Synthesis and biological activity of simplified denoviose-coumarins related to novobiocin as potent inhibitors of heat-shock protein 90 (hsp90)

Article abstract of DOI:10.1016/j.bmcl.2008.01.128

A new series of coumarin inhibitors of hsp90 lacking the noviose moiety as well as substituents on C-7 and C-8 positions of the aromatic ring was synthesised and their hsp90 inhibitory activity has been delineated: for example, their capacity to induce th

Full text of DOI:10.1016/j.bmcl.2008.01.128

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 2495–2498
Synthesis and biological activity of simplified
denoviose-coumarins related to novobiocin as potent inhibitors
of heat-shock protein 90 (hsp90)
b
a
a
b
´
Christine Radanyi, Gaelle Le Bras, Samir Messaoudi, Celine Bouclier,
¨
a
a
b
´
Jean-Franc¸ois Peyrat, Jean-Daniel Brion, Veronique Marsaud,
b
a,
*
ˆ
Jack-Michel Renoir and Mouad Alami
aUniv Paris-Sud, CNRS, BioCIS-UMR 8076, Laboratoire de Chimie The´rapeutique, Faculte´ de Pharmacie,
ˆ
IFR 141, 5 rue J.-B. Cle´ment, Chatenay-Malabry F-92296, France
bUniv Paris-Sud, CNRS, UMR 8612, Laboratoire de Pharmacologie Cellulaire et Mole´culaire des Anticance´reux,
ˆ
Faculte´ de Pharmacie, IFR 141, 5 rue J.-B. Cle´ment, Chatenay-Malabry F-92296, France
Received 10 January 2008; revised 30 January 2008; accepted 31 January 2008
Available online 14 February 2008
Abstract—A new series of coumarin inhibitors of hsp90 lacking the noviose moiety as well as substituents on C-7 and C-8 positions
of the aromatic ring was synthesised and their hsp90 inhibitory activity has been delineated: for example, their capacity to induce the
degradation of client proteins and to inhibit estradiol-induced transcription in human breast cancer cells. In cell proliferation assay,
the most active compound 5g was ꢀ8 times more potent than the parent novobiocin natural compound.
Ó 2008 Elsevier Ltd. All rights reserved.
The aminocoumarins novobiocin and clorobiocin are
well-established antibiotics agents that act as inhibitors
of the bacterial ATP binding gyrase B, a type II DNA
topoisomerase.1 The natural product novobiocin (Nvb)
has attracted renewed attention because of its antitu-
mour activity.2 The mechanism of action of novobiocin
was found to inhibit the heat-shock proteins 90 (hsp90)
via interaction with a previously unrecognized C-termi-
nal ATP binding site.3 This ATP-dependent molecular
chaperone is well known to be involved in the folding
of many client proteins including, among others, kinases
(Akt, Raf-1, Her-2, cdk-4), transcription factors like ste-
roid receptors and p53, that are directly associated with
all six hallmarks of cancer.4 Inhibition of hsp90 by Nvb
induced the degradation of the hsp90 client proteins via
the ubiquitin–proteasome pathway.2 Consequently,
hsp90 has evolved as an exciting new target in cancer
drug discovery.5
Unfortunately, the ability of Nvb to induce degradation
of hsp90 client protein (e.g., ErbB2 in SkBr3 breast can-
cer)2 is relatively weak (700 lM) and requires further
investigation to provide more potent compounds with
improved pharmaceutical properties. Among the most
active analogues, Blagg et al.6 highlighted the crucial
role of the noviose moiety at the 7-position of the cou-
marin ring for the biological activity, whereas the 4-hy-
droxy and 8-methyl groups in Nvb are not
indispensable. In more recent studies,7 30-descarba-
moyl-4-deshydroxynovobiocin DHN2 proved to be a
more effective and selective hsp90 inhibitor (Fig. 1).
As part of our research devoted to nonsugar coumarin
analogues that target hsp90, we recently showed that
the removal of the noviose moiety together with the
introduction of a tosyl substituent at C-4 coumarins
provides 4-tosylcyclonovobiocic acid (4TCNA) as a lead
compound having a C-7 free-phenolic function and a
C-8 methyl group.8 Biological evaluation of 4TCNA
revealed higher potency than Nvb itself to induce
client-protein degradation. In our continuing struc-
ture–activity relationship (SAR) study, our aim was to
determine what modifications are necessary to selec-
tively provide more potent nonsugar coumarin com-
Keywords: hsp90; Novobiocin; Coumarin inhibitors; Anti-tumour;
Steroid receptors.
*
Corresponding author. Tel.: +33 146 83 58 87; fax: +33 146 83 58
28; e-mail: mouad.alami@u-psud.fr
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.01.128

2496
C. Radanyi et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2495–2498
OH
OH
H
N
OH
H
4
N
3
O
7
O
O
O
O
O
O
O
8
MeO
O
R1
MeO
R2
O
HO
DHN2
O
OH
Novobiocin (Nvb): R1 = Me, R2 = NH2
OH
O
Clorobiocin: R1 = Cl, R2 =
N
H
O
S
O
O
R2
O
O
H
N
H
N
R1
R1 , R2 = alkyle, aryle
O
O
O
HO
O
O
4TCNA
A
Figure 1. Hsp90 inhibitors and general structure A of the synthesised compounds.
pounds that target hsp90. Herein we report the synthesis
of a simplified scaffold 3-(N-substituted) aminocouma-
rin of type A (Fig. 1), which includes two centres for
introduction of diversity into coumarin molecule and
disclose the impact of the absence of the 7-hydroxy moi-
ety and the 8-methyl groups on the hsp90 inhibitory
activity. The potencies of newly synthesised coumarins
were evaluated for their capacity to inhibit estradiol
(E2)-induced transcription in human breast cancer cells,
to induce the proteasome-mediated degradation of sev-
eral known hsp90 client proteins, including Raf-1, the
estrogen receptor a (ERa) and the progesterone recep-
tors A and B (PRA, PRB).
dimethoxybenzamide to provide directly 4a using vari-
ous combinations of Pd/L/base mixtures (e.g.,
Pd(OAc)2, Pd2(dba)3/Xantphos, BINAP, Xphos, dppf/
Cs2CO3, K3PO4, K2CO3, etc.) resulted unfortunately
in unsuccessful results, presumably for steric consider-
ations. Therefore, the requisite aminocoumarins of gen-
eral structure A were prepared by the route shown in
Scheme 1. The synthesis was initiated with the reduc-
tion of the nitro function of commercially available
3-nitro-4-hydroxycoumarin
2 under Pd/C-catalysed
hydrogenation to afford the corresponding 3-amino-4-
hydroxycoumarin 3 in 84% yield. Further coupling
with a series of aryl and alkyl carboxylic acids in the
presence of PyBOP and DIEA in DMF at room tem-
perature gave 3-acylaminocoumarin derivatives 4a–l
in moderate to good yields (44–91%). Because of the
strong insolubility derivatives 4k and 4l in the usual or-
ganic solvents, their biological activity could not be
delineated.
Initially, the synthesis of 3-aminocoumarin derivatives
4 was examined by the palladium-catalysed C–N bond
coupling reaction of known 3-bromo-4-hydroxy-cou-
marin 1 with a series of amide nucleophiles as we pre-
viously described.9 All our attempts to react 1 with 2,4-
OH
O
OH
O
OH
O
OTs
O
Ar
H
N
H
N
H
N
X
NH2
O
R
R
R
(c)
(d)
(b)
(a)
O
O
O
O
O
O
O
O
3
4
5
6a-f
1
2
X = Br
OMe
O
Me
Me
X = NO2
OMe
OMe
OMe
OMe
OMe
OMe
6a: Ar =
6b: Ar =
R =
a:
OMe
f:
R =
R =
MeO
MeO
O
Me
Me
g:
Me
10
R =
R =
b:
c:
R =
R =
Me
h:
i:
R =
R =
R =
Me
10
MeO
OMe
OMe
Cl
6c: Ar =
6d: Ar =
6e: Ar =
R =
R =
R =
Me
10
MeO
OMe
j:
Me
OMe
OMe
O
O
OMe
OMe
Me
10
OMe
OMe
d:
e:
k:
l:
R =
R =
R =
R =
OMe
OMe
OMe
O
O
O
Me
6f: Ar =
R =
OMe
Scheme 1. Reagents and conditions: (a) H2, Pd/C, MeOH, HCl (1%), 3 h, rt, 84%; (b) RCO2H (1.2 equiv), PyBOP (1 equiv), DIEA (3 equiv), DMF,
18 h, rt; (c) TsCl (2 equiv), pyridine, 16 h, rt; (d) ArB(OH)2 (1.3 equiv), K3PO4 (3 equiv), Bu4NBr (10 mol %), PdCl2dppf (5 mol %), MeCN, 80 °C,
overnight.

C. Radanyi et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2495–2498
2497
In order to evaluate the importance of the free 4-hydro-
xyl group in derivatives 4, we planned to introduce an
aryl substituent by metal-catalysed cross coupling reac-
tion. For this purpose, the use of aryl tosylates as elec-
trophiles for the Suzuki-type coupling is a very
attractive procedure.10 The synthesis of the required 4-
tosylcoumarins 5 was achieved from coumarins 4 using
tosyl chloride in the presence of pyridine as a solvent
at room temperature. Compounds 5 were then submit-
ted to cross-coupling with various arylboronic acids un-
der our previously best conditions (K3PO4, PdCl2(dppf),
Bu4NBr, MeCN, 80 °C).8 As expected, the correspond-
ing coupling products 6 were obtained in good yields
(52–96%).
ies 4–6, compounds 4b, 4c, 4j, 4k, 5e, 5g, 5i, 5h and 6c
were able to decrease the levels of ERa and Raf-1,
whereas compounds 4f and 5d were able to induce the
loss of only ERa without affecting Raf-1 (Fig. 2A).
The activity of all these compounds was inhibited by
the proteasome-inhibitor MG-132, except that of 6c
(not shown). Other derivatives proved to be inactive.
These data suggest that different analogues may affect
selectively various hsp-90/client protein complexes lead-
ing to depletion of different key signalling proteins. As
shown in Figure 2A, compound 6c induced the degrada-
tion of other proteins reflecting a strong toxicity likely
unrelated to hsp90 inhibition and was not further eval-
uated. In dose–response analysis (Fig. 2B) compound
5g revealed to be the strongest inhibitor since processing
of ERa, Raf-1 and both progesterone receptors (PRB
and PRA) occurred over a range of 40–100 lM .
We had previously established that 4TCNA inhibits
estradiol (E2)-induced transcription in MELN cells,
and targets several hsp90-client proteins including ERa
to proteasome-mediated degradation. In addition,
4TCNA promotes apoptosis through activation of casp-
ases 7 and 8 in ER-positive MCF-7 human breast cancer
cells. With Nvb and 4TCNA in hands as reference com-
pounds, in a large screening experiment, we found that
among all the novobiocin analogues from the three ser-
We further assessed the potency of these novobiocin
analogues to affect transcription in MELN cells. As
shown in Figure 3, compound 4g enhances both the ba-
sal and the E2-induced transcription, whereas 4h, 5d, 5c,
5i, 5e, 5h and 5g inhibit the E2-induced Luciferase
(LUC) expression. Surprisingly, almost all compounds
Figure 2. Stability of hsp90 client proteins in the presence of various novobiocin analogues. (A) MCF-7 breast cancer cells were exposed to 100 lM
Nvb or analogues for 16 h and cell lysates were analyzed by Western blotting with regard to the levels of ERa and Raf-1. (B) In addition to detection
of these two hsp90 client proteins, PRA and PRB fate was analyzed in cell lysates following exposure of MCF-7 cells to increasing amounts of 5g.
Primary antibodies were: ERa (HC20), Raf-1 (C12) from Santa Cruz; PR (Ab-8) from Labvision Corp., all used at 1 lg/mL. Complexes were
detected with peroxidase-labelled secondary antibody. C, control cell lysate from cell exposed to DMSO. The visualization of a nonspecific (NS) band
served as constant protein loading.
14
-E2
+ E2
12
10
8
6
4
2
0
Compounds
Figure 3. Effects of novobiocin analogues on ER-mediated transcription in MELN breast cancer cells. Cells were grown and treated for 18 h as
described before8 with 200 lM Nvb and analogues in the presence (+) or not (ꢁ) of 0.1 nM E2. LUC activity was measured and expressed as fold
induction relative to untreated control cells.

2498
C. Radanyi et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2495–2498
Table 1. Anti-proliferative activity of the most active inhibitors
Acknowledgements
a
Compound
IC50 (lM)
The authors gratefully acknowledge support of this pro-
ject by the Ligue Nationale contre le Cancer (grants to
`
J.-M. R). C.B. is the recipient of the Ministere de l’Edu-
cation Nationale de la Recherche et de la Technologie.
´ ´ ´
GLB is a recipient of the Societe de Chimie Therapeu-
tique (SERVIER group).
5d
5g
40 ( 6)
35 ( 3)
50 ( 5)
70 ( 4)
75 ( 4)
45 ( 8)
260 ( 15)
5i
5e
5h
4TCNA
Nvb
a Values are means of three experiments, carried out in sextuplate and
standard deviation is given in parentheses.
References and notes
1. For reviews on structure and mode action of coumarins on
DNA gyrase, see: (a) Reece, R. J.; Maxwell, A. Crit. Rev.
Biochem. Mol. Biol. 1991, 26, 335; (b) Maxwell, A. Trends
Microbiol. 1997, 5, 102; (c) Lewis, R. J.; Singh, O. M. P.;
Smith, C. V.; Skarzynski, T.; Maxwell, A.; Wonacott, A.
J.; Wigley, D. B. EMBO J. 1996, 15, 1412.
of series 5 having a tosyl substituent on the C-4 position
of the coumarin nucleus, behave as stronger inhibitors
of ER-mediated transcription, except compound 5a, as
compared to Nvb.
2. Marcu, M. G.; Schulte, T. W.; Neckers, L. J. Natl. Cancer
Inst. 2000, 92, 242.
The in vitro activity of these derivatives was further
evaluated by their growth-inhibitory potency in MCF-
7 cells. The quantification of cell survival in MCF-7 cells
(initially 5000 cells/well of a 96-well plate, exposed after
24 h to serial dilution of the different analogues for 72 h)
was established by using the 3-(4,5-dimethylthiazol-2-
yl)-2,5-diphenyltetrazolium bromide assay (MTT, Sig-
ma). The experiment was conducted as described before8
and the IC50 values were measured as the drug concen-
tration that inhibits the cell growth by 50% compared
with growth of vehicle-treated cells. As shown in
Table 1, the IC50 of the most active analogues was in
the range of 35–75 lM not strongly different from that
of our lead compound 4TCNA. Furthermore, all the
analogues exhibited improved growth inhibition
potentials as compared to the parent compound Nvb.
Consistent with the degradation of hsp90 client proteins,
the most active inhibitors are those with the lowest
IC50.
3. (a) Marcu, M. G.; Chadli, A.; Bouhouche, I.; Catelli,
M.; Neckers, L. M. J. Biol. Chem. 2000, 275, 37181; (b)
Prodromou, C.; Siligardi, G.; O’Brien, R.; Woolfson, D.
N.; Regan, L.; Panaretou, B.; Ladbury, J. E.; Piper, P.
W.; Pearl, L. H. EMBO J. 1999, 18, 754; (c) Prodro-
mou, C.; Roe, S. M.; O’Brien, R.; Ladbury, J. E.; Piper,
P. W.; Pearl, L. H. Cell 1997, 90, 65; (d) Young, J. C.;
Obermann, W. M. J.; Hartl, F. U. J. Biol. Chem. 1998,
273, 18007; (e) Harris, S. F.; Shiau, A. K.; Agard, D. A.
Structure 2004, 12, 1087; (f) Allan, RK.; Mok, D.;
Ward, BK.; Ratajczak, T. J. Biol. Chem. 2006, 281,
7161.
4. (a) Hanahan, D.; Weinberg, R. A. Cell 2000, 100, 57;
(b) Richter, K.; Buchner, J. J. Cell. Physiol. 2001, 188,
281; (c) Neckers, L. Trends Mol. Med. 2002, 8, S55; (d)
Neckers, L.; Ivy, S. P. Curr. Opin. Oncol. 2003, 15,
419; (e) Zhang, H.; Burrows, F. J. Mol. Med. 2004, 82,
488.
5. (a) Chiosis, G.; Caldas Lopes, E.; Solit, D. Curr. Opin.
Invest. Drugs 2006, 7, 534; (b) Workman, P.; Burrows, F.;
Neckers, L.; Rosen, N. Ann. N.Y. Acad. Sci. 2007, 1113,
202.
In summary, results from these studies demonstrate that
the newly synthesised novobiocin analogues, based on a
simplified 3-aminocoumarin scaffold exhibited increased
inhibitory activity against the hsp90 protein folding pro-
cess. In this SARs study, we highlighted in denoviose
analogues 5, bearing a tosyl group on the 4-position,
that removal of C7/C8 substituents is not detrimental
for hsp90 inhibitory activity and strongly enhances the
capacity of the analogues to inhibit hsp90. Among these
analogues, compound 5g was identified to be the most
potent representative of the new family of simplified
coumarins. Our results suggest that 5g and 4TCNA
which exerted similar biological profile may be consid-
ered interesting compounds for the development of
more potent novobiocin analogues.
6. Yu, X. M.; Shen, G.; Neckers, L.; Blake, H.; Holzbeier-
lein, J.; Cronk, B.; Blagg, B. S. J. J. Am. Chem. Soc. 2005,
127, 12778.
7. Burlison, J. A.; Neckers, L.; Smith, A. B.; Maxwell, A.;
Blagg, B. S. J. J. Am. Chem. Soc. 2006, 128, 15529.
8. Le Bras, G.; Radanyi, C.; Peyrat, J.-F.; Brion, J.-D.;
Alami, M.; Marsaud, V.; Stella, B.; Renoir, J.-M. J. Med.
Chem. 2007, 50, 6189.
9. (a) Audisio, D.; Messaoudi, S.; Peyrat, J.-F.; Brion, J.-D.;
Alami, M. Tetrahedron Lett. 2007, 48, 6928; (b) Messao-
udi, S.; Audisio, D.; Brion, J.-D.; Alami, M. Tetrahedron
2007, 63, 10202.
10. (a) Wu, J.; Wang, L.; Fathi, R.; Yang, Z. Tetrahedron
Lett. 2002, 43, 4395; (b) Boland, G. M.; Donnelly, D. M.
X.; Finet, J. P.; Rea, M. D. J. Chem. Soc., Perkin Trans. 1
1996, 2591.