Radanamycin, a macrocyclic chimera of radicicol and geldanamycin

Article abstract of DOI:10.1016/j.bmcl.2006.01.086

Radicicol and geldanamycin are potent inhibitors of the Hsp90 protein folding machinery, which is an emerging target for the development of cancer chemotherapeutics. However, radicicol is inactive in vivo and geldanamycin suffers from its redox-active beh

Full text of DOI:10.1016/j.bmcl.2006.01.086

Bioorganic & Medicinal Chemistry Letters 16 (2006) 2459–2462
Radanamycin, a macrocyclic chimera of radicicol and geldanamycin
Mingwen Wang, Gang Shen and Brian S. J. Blagg*
Department of Medicinal Chemistry, The University of Kansas, 1251 Wescoe Hall Drive, Malott 4070,
Lawrence, KS 66045-7563, USA
Received 17 January 2006; revised 20 January 2006; accepted 23 January 2006
Available online 7 February 2006
Abstract—Radicicol and geldanamycin are potent inhibitors of the Hsp90 protein folding machinery, which is an emerging target
for the development of cancer chemotherapeutics. However, radicicol is inactive in vivo and geldanamycin suffers from its redox-
active behavior that produces toxicity unrelated to Hsp90 inhibition. It was proposed that a chimeric molecule containing the res-
orcinol ring of radicicol and the quinone of geldanamycin could provide an opportunity to elucidate structure–activity relationships
for both natural products and serve as a starting point for the development of more potent inhibitors. Synthesis of the macrocyclic
chimera named radanamycin is reported along with the biological activity exhibited by this compound in MCF-7 breast cancer cells.
ꢀ 2006 Elsevier Ltd. All rights reserved.
The 90 kDa heat shock proteins (Hsp90) represent a
promising target for the development of cancer chemo-
therapeutics as a result of their ability to fold nascent
oncogenic polypeptides into biologically active pro-
teins.1–3 Not surprisingly, the Hsp90 protein folding
process is dependent upon ATP-hydrolysis to provide
energy for the conformational maturation of nascent
polypeptides.4 However, unlike most ATP-binding pro-
teins, Hsp90 binds ATP in a bent conformation that is
very unique and only eukaryotic enzymes MutL and his-
tidine kinase have been shown to bind nucleotides in a
similar manner.5 Moreover, two natural products have
been identified as potent inhibitors of the Hsp90-medi-
ated protein folding process and there are currently
more than 20 clinical trials in progress based on
Hsp90-targeted drugs. All of these trials utilize deriva-
tives of geldanamycin (GDA, IC50 = 49 nM, Fig. 1),
which has been shown to produce toxicity unrelated to
Hsp90 inhibition.6 Another natural product inhibitor
of Hsp90 is radicicol (RDC, IC50 = 23 nM), which pro-
duces the most potent inhibitory activity yet identified,
however, it has no activity in vivo because it is rapidly
metabolized into inactive compounds that have little
or no affinity for Hsp90.7
developed small molecule antagonists of the Hsp90
molecular chaperone that are either based upon the pur-
ine skeleton (PU24FCl, IC50 = 2 lM),8 dihydroxyphenyl
pyrazoles (CCT018159, IC50 = 4.1 lM),9 or more
recently novobiocin (IC50 ꢀ 700 lM).10 In an effort to
develop new Hsp90 inhibitors and to understand struc-
ture–activity relationships for the natural products, we
superimposed the co-crystal structures of GDA and
RDC bound to Hsp90.5 As can be seen in Figure 2,
the resorcinol moiety of RDC occupies the same binding
region as the carbamate of GDA. Both of these moieties
provide similar interactions as those observed between
the adenine ring of ADP and Hsp90. In contrast, the
quinone ring of GDA extends toward the protein sur-
face in analogy to the macrocyclic ring and the epoxide
of RDC. Although the entire three-dimensional crystal
structure of Hsp90 has not been determined, experimen-
tal data and individual structures of the N- and middle
domains have supported contact between these two re-
gions, suggesting the region that binds to the quinone
and epoxide may not be solvent exposed.11–13 To in-
crease interactions of each inhibitor with Hsp90, we pro-
posed that a molecule containing portions of each
natural product should be capable of producing effective
Hsp90 inhibitors that could be modified for increased
potency.
To circumvent problems associated with the natural
product inhibitors of Hsp90, several institutions have
Originally, our proposed inhibitor was designed by com-
paring the co-crystal structure of each natural product
bound to Hsp90. As can be seen in Figure 2, both mol-
ecules adopt a bent conformation upon Hsp90 binding,
which is similar to that of the natural substrate, ADP.5
Keywords: Hsp90; Inhibitors; Geldanamycin; Radicicol; Cancer.
*
Corresponding author. Tel.: +1 785 864 2288; fax: +1 785 864
5326; e-mail: bblagg@ku.edu
0960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.01.086

2460
M. Wang et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2459–2462
syntheses and inhibitory activities of the open chain chi-
O
O
meras of RDC and GDA (radester and radamide)14,15
and based on their activities proposed that the conform-
ationally constrained macrocyclic analogue would be
more effective as entropic penalties would be minimized.
In this letter, we report the synthesis and evaluation of
radanamycin (1), a macrocyclic chimera of radicicol
and geldanamycin.
OMe
O
O
O
N
H
HO
O
O
O
Cl
O
O
NH2
HO
OH
Synthesis of radanamycin began by modification of the
procedure used to prepare the seco-derivatives. The
TBS-protected resorcinolic methyl ester 3 was synthe-
sized from ester 2 by previously reported methods
(Scheme 1).14 The silyl groups were removed and the es-
ter dealkylated with the lithium anion of n-propylthiol
to furnish the corresponding acid, 4. DCC-mediated
coupling of the free acid with the MOM-protected pre-
cursor to the quinone ring (5)15 provided the ester prod-
uct 6, in good yield. Subsequent oxidation of the olefin
and conversion to the acid, gave 7 in high yield.
O
Radicicol
Geldanamycin
O
O
OMe
MeO
MeO
F
N
N
NH
N
N
N
Cl
PU24FCl
NH2
HO
CCT018159
OH
OH
The nitro substituent was reduced without competitive
removal of the protecting groups by brief exposure to
palladium on carbon under hydrogen gas for a limited
time (Scheme 2).16 Prolonged exposure to these condi-
tions resulted in the accumulation of several undesired
products.
OH
H
N
O
O
O
O
O
MeO
O
Novobiocin
OH
Macrolactamization was problematic at the onset, but
was finally achieved by treatment of the crude amino
acid 8, with HATU and HOBt to furnish the desired
macrocycle 9, in excellent yield. 1H NMR of 9 provided
evidence for a bent conformation similar to that ob-
served for the natural products bound to Hsp90 by vir-
tue of the upshield shift and anisotropy of several
aliphatic protons. In analogy to the method used for
the construction of the seco-agents, radester and rada-
mide,14,15 the MOM-protecting groups were removed
O
H2N
OH
OMe
OH
OMe
HN
O
HN
O
HO
HO
O
Cl
CO2Me
Cl
O
HO
OH
HO
OH
Radamide hydroquinone
Radester hydroquinone
Figure 1. Hsp90 inhibitors.
OMe
O
OMe
O
1. LiOH
Cl
2. LiSnPr
A
Ref. 15
83% over
2 steps
HO
OH
TBSO
OTBS
2
3
NH
O
OH
Cl
OMOM
OMe
MeO
HO
O
OH
OH
DCC,
DMAP
76%
O2N
Cl
HO
NH2
+
OH
O
B
MOMO
HN
OH
HO
O
5
cl
4
1
OMOM
OMe
OMOM
OMe
O2N
O2N
O
1. OsO4,
NaIO4
Figure 2. Conformations of geldanamycin and radicicol bound to
yeast Hsp90 as well as the chimeric compound, radanamycin (1).
OH
R
R
2. NaClO2
81% over
2 steps
O
O
Cl
HO
Cl
O
O
A simple disconnection across the apex of each natural
product, followed by hybridization of the RDC resor-
cinol moiety with the GDA quinone, afforded the
macrocyclic compound. Previously, we reported the
R = MOM
HO
OH
OH
7
6
Scheme 1. Synthesis of 7.

M. Wang et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2459–2462
2461
OMOM
OMe
120
100
80
60
40
20
0
H2N
OH
HATU, HOBt,
iPr2NEt
O
Pd(C), H2
7
R
76% over
2 steps
R = MOM
O
Cl
HO
O
OH
8
0.01
0.1
1
10
100
OMOM
Figure 4. Anti-proliferative activity of radanamycin in MCF-7 breast
cancer cells. Concentrations are reported on the x axis in lM.
HN
O
OMe
TMSCl,
NaI
R
1
O
81%
and further derivatization of these molecules is likely
to afford analogues with increased activity and perhaps,
useful alternatives to the geldanamycin derivatives in
clinical trials. Additional studies with these chimeric spe-
cies are in progress and structure–activity relationships
for these agents will be reported in due course.
Cl
O
HO
OH
9
Scheme 2. Synthesis of radanamycin, 1.
GDA
0
.01 0.1
0.5 1.0 2.0
5.0
10.0
Acknowledgments
Her2
Akt
The authors gratefully acknowledge support of this pro-
ject by the NIH COBRE RR017708 and The University
of Kansas Center for Research. G.S. is the recipient of a
Susan G. Komen Dissertation Award.
Hsp90
Actin
References and notes
Figure 3. Western blot analyses of Her2, Akt, Hsp90, and actin.
Concentrations of radanamycin are listed above each lane in lM.
GDA was used as a positive control at 0.5 lM.
1. Goetz, M. P.; Toft, D. O.; Ames, M. M.; Erlichman, C.
Ann. Oncol. 2003, 14, 1169.
2. Chiosis, G.; Vilenchik, M.; Kim, J.; Solit, D. Drug Discov.
Today 2004, 9, 881.
3. Neckers, L.; Neckers, K. Exp. Opin. Emerg. Drugs 2005,
10, 137.
with in situ generated trimethylsilyl iodide to give hydro-
quinone 1.17
4. Avila, C.; Kornilayev, B.; Blagg, B. S. J. Bioorg. Med.
Chem. 2006, 14, 1134.
5. Roe, S. M.; Prodromou, C.; O’Brien, R.; Ladbury, J. E.;
Piper, P. W.; Pearl, L. H. J. Med. Chem. 1999, 42, 260.
6. Dikalov, S.; Landmesser, U.; Harrison, D. G. J. Biol.
Chem. 2002, 277, 25480.
7. Geng, X.; Yang, Z.-Q.; Danishefsky, S. J. Synlett 2004, 8,
1325.
8. Chiosis, G.; Lucas, B.; Shtil, A.; Huezo, H.; Rosen, N.
Bioorg. Med. Chem. 2002, 10, 3555.
9. Cheung, K.-M.; Matthews, T. P.; James, K.; Rowlands,
M. G.; Boxall, K. J.; Sharp, S. Y.; Maloney, A.; Roe, S.
M.; Prodroou, C.; Pearl, L. H.; Aherne, W.; McDonald,
E.; Workman, P. Bioorg. Med. Chem. Lett. 2005, 15, 3338.
10. 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.
11. Scheibel, T.; Siegmund, H. I.; Jaenicke, R.; Ganz, P.; Lilie,
H.; Buchner, J. Proc. Natl. Acad. Sci. U.S.A. 1999, 96,
1297.
With radanamycin in hand, we sought to determine its
ability to inhibit Hsp90 as determined by the degrada-
tion of Hsp90-dependent client proteins. When 1 was
incubated with MCF-7 breast cancer cells for 24 h, a
dramatic effect on Hsp90-dependent client proteins
(Her2 and Akt) was observed (Fig. 3). These Hsp90-de-
pendent substrates were readily degraded in the presence
of 1–5 lM of radanamycin, which correlated inversely
with Hsp90 levels. Actin is not an Hsp90-dependent sub-
strate and remains unaffected by Hsp90 inhibition,
thereby allowing a suitable control for comparing rela-
tive protein concentrations.
Consistent with the degradation of Hsp90-dependent
substrates, radanamycin exhibited anti-proliferative
activity with an IC50 = 1.2 0.1 lM (Fig. 4), thereby
linking Hsp90 inhibition directly to the inhibition of cell
growth. The values obtained from these studies compare
well to those of previously identified inhibitors of the
Hsp90 protein folding machinery as noted earlier in this
letter.8–10,13,14
12. Huai, Q.; Wang, H.; Liu, Y.; Kim, H.-Y.; Toft, D.; Ke, H.
Structure 2005, 13, 579.
13. Soti, C.; Racz, A.; Csermely, P. J. Biol. Chem. 2002, 277,
7066.
14. Clevenger, R. C.; Blagg, B. S. J. Org. Lett. 2004, 6, 4459.
15. Shen, G.; Blagg, B. S. J. Org. Lett. 2005, 7, 2157.
16. Intermediate 7 (69 mg, 0.12 mmol) was dissolved in EtOH
(2.0 mL) before 10% palladium on carbon (15 mg) was
Results from these studies indicate that chimeric
compounds composed of RDC’s resorcinol ring and
GDA’s quinone ring produce potent Hsp90 inhibitors

2462
M. Wang et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2459–2462
added. The mixture was purged with argon before H2 gas
1597, 1487, 1235, 1129, 1070 cmÀ1; HRMS (TOF-ES+)
found 626.1475 (M+H+), calcd 626.1480 for C24H29Cl
NO10.
was added. The heterogeneous mixture was stirred for 3 h
and then filtered through a plug of Celite. The eluent was
concentrated and the residue was redissolved in DMF
(10 mL) before slow addition to a mixture of HATU
(137 mg, 0.36 mmol), HOBt (49 mg, 0.36 mmol), and
diisopropyl ethyl amine (0.189 mL, 1.08 mmol) in DMF
(120 mL) over 4 h. Upon addition, the solution was heated
at 70 ꢁC for 15 h. The solvent was removed by distillation
at reduced pressure. EtOAC (20 mL) was added to the
residue and the resulting solution was washed with
saturated aqueous NH4Cl (3· 5 mL), H2O (5 mL), and
saturated aqueous NaCl (5 mL), dried (Na2SO4), filtered,
and concentrated. The residue was purified via chroma-
tography (SiO2, 20% hexanes in EtOAc) to afford 9
(48 mg, 76% yield, 2 steps) as a white solid: 1H NMR
(acetone-d6, 400 MHz) d 9.73 (s, 1H, –OH), 9.15 (s, 1H,
–OH), 8.22 (s, 1H), 6.97 (s, 1H), 5.52 (m, 1H), 5.18 (dd,
J = 6.6, 24.1 Hz, 2H), 4.90 (dd, J = 6.6, 24.1 Hz, 2H), 4.06
(s, 3H), 3.53 (s, 3H), 3.48 (s, 3H), 3.19 (dt, J = 4.8,13.3 Hz,
1H,), 3.19 (dd, J = 11.5, 13.2 Hz, 1H,), 3.06 (dd, J = 3.1,
13.2 Hz, 1H), 2.68 (dt, J = 4.0,13.3 Hz, 1H), 2.40 (dt,
J = 3.9,13.1 Hz, 1H), 2.04 (m, 1H), 1.45 (d, J = 6.3 Hz,
2H); 13C NMR (acetone-d6, 125 MHz) d 174.3, 168.7,
157.3, 156.0, 148.9, 147.6, 147.1, 139.6, 126.9, 126.8, 115.6,
113.4, 112.7, 102.4, 100.2, 95.8, 73.0, 60.8, 57.4, 56.0, 36.2,
31.8, 27.3, 20.9; IR (film) mmax 3301, 3119, 2929, 1651,
17. MOM-protected radanamycin 9 (76 mg, 0.144 mmol) was
dissolved in CH3CN (1.2 mL) and CH2Cl2 (1.2 mL) before
NaI (216 mg, 1.44 mmol) and TMSCl (0.182 mL,
1.44 mmol) were added at rt. Upon addition, the solution
turned cloudy and yellow. After 20 min, a saturated
aqueous solution of Na2S2O4 (2 mL) was added to the
mixture and stirred for 10 min. The aqueous phase was
extracted with EtOAc (3· 5 mL) and the combined
organic layers were washed with H2O (2· 5 mL), saturated
aqueous NaCl (5 mL), dried (Na2SO4), filtered, and
concentrated. The residue was purified via chromatogra-
phy (SiO2, 10% MeOH in CH2Cl2) to afford 1 (53 mg, 84%
1
yield) as a white solid: H NMR (acetone-d6, 400 MHz) d
7.82 (br s, 2H), 6.59 (s, 1H), 6.41 (s, 1H), 5.55 (m, 1H),
3.97 (s, 3H), 3.54 (dt, J = 5.4, 13.3 Hz, 1H), 3.27 (dd,
J = 11.6, 13.5 Hz, 1H), 2.96 (dd, J = 3.2, 13.5 Hz, 1H),
2.77 (dt, J = 3.6, 13.2 Hz, 1H), 2.18 (m, 1H), 1.97 (dt,
3
J = 5.4, 11.8 Hz, 1H), 1.41 (d, J = 6.3 Hz, 3H); C NMR
(acetone-d6, 125 MHz) d 174.1, 168.8, 158.0, 156.0, 146.8,
145.2, 143.0, 139.8, 120.4, 120.0, 114.5, 112.7, 111.9, 102.0,
73.2, 59.9, 35.7, 29.3, 27.1, 20.5; IR (film) mmax 3323, 3021,
1643, 1603, 1445, 1380, 1299, 1236, 1170 cmÀ1; HRMS
(TOF-ES+) found 438.0949 (M+H+), calcd 438.0956 for
C20H21Cl NO8.