Geldanamycin derivative inhibition of HGF/SF-mediated Met tyrosine kinase receptor-dependent urokinase-plasminogen activation

Article abstract of DOI:10.1016/j.bmc.2005.05.038

Ansamycins, including geldanamycin and the derivative 17-allylamino-17- demethoxygeldanamycin, and radicicol are known for their ability to tightly bind to the ATP-binding site of the amino-terminal domain region of heat shock protein 90. We have found that geldanamycin and some of its derivatives can inhibit hepatocyte growth factor/scatter factor-mediated Met tyrosine kinase receptor-dependent urokinase-plasminogen activation at femtomolar levels. Assessment is made of structural requirements for such an activity and evidence is given that distinguishes the target of such an activity from that of heat shock protein 90.

Full text of DOI:10.1016/j.bmc.2005.05.038

Bioorganic & Medicinal Chemistry 13 (2005) 4960–4971
Geldanamycin derivative inhibition of HGF/SF-mediated
Met tyrosine kinase receptor-dependent
urokinase-plasminogen activation
Yuehai Shen,a, Qian Xie,b, Monica Norberg,a Edward Sausville,c
George Vande Woudeb and David Wenkerta,*
aDepartment of Physiology, Michigan State University, East Lansing, MI 48864, USA
bVan Andel Research Institute, Grand Rapids, MI 49503, USA
cDivision of Cancer Treatment and Diagnosis, National Cancer Institute, Bethesda, MD 20892, USA
Received 17 March 2005; revised 24 May 2005; accepted 24 May 2005
Available online 22 June 2005
Abstract—Ansamycins, including geldanamycin and the derivative 17-allylamino-17-demethoxygeldanamycin, and radicicol are
known for their ability to tightly bind to the ATP-binding site of the amino-terminal domain region of heat shock protein 90.
We have found that geldanamycin and some of its derivatives can inhibit hepatocyte growth factor/scatter factor-mediated Met
tyrosine kinase receptor-dependent urokinase-plasminogen activation at femtomolar levels. Assessment is made of structural
requirements for such an activity and evidence is given that distinguishes the target of such an activity from that of heat shock
protein 90.
ꢀ 2005 Elsevier Ltd. All rights reserved.
1. Introduction
Recent work has shown that Met signaling pathway is a
potential therapeutic target for cancer therapy. Studies
have shown the efficacy of Met-directed ribozyme and
antisense strategies for reduction of Met and HGF/SF
expression, tumor growth, and metastatic tumor poten-
tial.5–8 NK4, a HGF/SF fragment possessing its amino-
terminal four-kringle domain, is a competitive HGF/SF
antagonist for the Met receptor9 and has been demon-
strated to inhibit tumor invasion and metastasis, as well
as tumor angiogenesis.10 Monoclonal antibodies direct-
ed to HGF/SF neutralize its activity with inhibition of
human xenograft tumor growth in athymic nu/nu
mice.11 The indole-based receptor tyrosine kinase inhib-
itors, K252a and PHA-665752, inhibit Met kinase
activity and Met-driven tumor growth and metastatic
potential.12,13
Signaling via the Met protooncogene, the tyrosine ki-
nase receptor for hepatocyte growth factor/scatter factor
(HGF/SF), is critical for normal physiological, as well as
pathological, processes. The Met signal transduction
pathway is involved in aspects of embryogenesis, prolif-
eration, motility, differentiation, and cellular survival.
Dysregulation of Met signaling has been found to
contribute to the development and growth of primary
tumors and their metastases. Evidence for this feature
is multifold.1 Inappropriate expression of Met is found
in many types of human cancers (www.vai.org/vari/met-
andcancer). Cell lines coexpressing Met and HGF/SF
become tumorigenic and metastatic in nude mice.2
Transgenic mice having HGF/SF inappropriately direct-
ed to various tissues developed diverse tumors.3 Impor-
tantly, activating mutations of the Met receptor have
been found in human carcinomas.1,4
A recent study screened for inhibitors of the Met recep-
tor signal transduction pathway.14 HGF/SF induces the
expression of the urokinase plasminogen activator
(uPA) and its receptor (uPAR), mediators of cell inva-
sion and metastasis. A cell-based assay utilizing the
induction of uPA and uPAR, and subsequent conver-
sion of plasminogen to plasmin allowed the screening
of compounds for inhibitory properties in MDCK-2
Keywords: Geldanamycin; Hepatocyte growth factor/scatter factor;
Urokinase-plasminogen activation; Met tyrosine kinase receptor.
*
Corresponding author. Tel.: +1 517 355 6475x1142; fax: +1 517 355
5125; e-mail: wenkert@msu.edu
Each contributed substantially to this work.
0968-0896/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2005.05.038

Y. Shen et al. / Bioorg. Med. Chem. 13 (2005) 4960–4971
4961
cells. Geldanamycin (1) and some derivatives thereof
were found to exhibit high inhibitory activity: at fem-
tomolar concentrations. More recently, this exquisite
inhibitory activity has been extended to include addi-
tional activities of the invasion complex, notably the in
vitro invasion of human tumor cells through three-
dimensional Matrigel.15 No loss of Met expression was
observed at lower than nanomolar concentrations, indi-
cating that the observed inhibitory activity was indepen-
dent of down-regulation of the Met receptor.
anamycin derivatives for an unknown target(s) and pro-
vide further evidence that distinguishes the femtomolar
target(s) from hsp90.
2. Results
2.1. Chemistry
Different classes of geldanamycin derivatives were synthe-
sized for this study. 17-Alkylamino-17-demethoxygel-
danamycin derivatives can bind hsp90 with high affinity
inhibiting its function, and numerous derivatives of this
class were prepared (Fig. 2). The 17-methoxy group of
geldanamycin is easily displaced by primary amines. 17-
Allylamino-17-demethoxygeldanamycin (4) and 17-(2-
dimethylaminoethyl)amino-17-demethoxygeldanamycin
(5), each of which is currently in clinical trials33,34 (NCI
protocol IDÕs are available at: http://cancer.gov/clinical-
trials), were prepared in near quantitative yield from gel-
danamycin by reaction with an appropriate amine.
Similarly prepared in high yield were the known 17-
amino-17-demethoxygeldanamycin (6),35 17-(2-chloro-
ethyl)amino-17-demethoxygeldanamycin (7), 17-(2-flu-
oroethyl)amino-17-demethoxygeldanamycin (8),36 as
well as the new derivatives 17-(2-acetylaminoethyl)ami-
no-17-demethoxygeldanamycin (9), 17-(6-acetylam-
inohexyl)amino-17-demethoxygeldanamycin (10), 17-
(6-biotinylaminohexyl)amino-17-demethoxygeldanamy-
cin (11), 17-(8-acetamido-3,6-dioxaoctylamino)-17-
demethoxygeldanamycin (12), and 17-carboxymethyl-
amino-17-demethoxygeldanamycin (13). The 17-meth-
oxy group of geldanamycin has been found to be displace-
able by the secondary amine azetidine and 17-N-
azetidinyl-17-demethoxygeldanamycin (14) was so pre-
pared.35 In a similar fashion, treatment of geldanamycin
Geldanamycin and 17-alkylamino-17-demethoxygel-
danamycin derivatives are best known for their ability
to bind to the ATP-binding site of the amino-terminal
domain region of heat shock protein 90 (hsp90).16–20
Hsp90 belongs to the structural protein family of
GHKL ATPases.21 This abundant protein helps in the
regulation of activity, turnover, and trafficking of vari-
ous critical proteins. It facilitates in the folding and
regulation of proteins in cellular signaling, such as
transcription factors, steroid receptors, and protein
kinases.22–25 The function of hsp90 is blocked by ansa-
mycin natural products, such as geldanamycin26 and
macbecin I (2),27,28 as well as radicicol (3)29,30 (Fig. 1).
It is the blockage of hsp90 function that has been as-
cribed to the antitumor effect of 17-allylamino-17-
demethoxygeldanamycin (4), a drug now used in clinical
trials.31,32
The target(s) for disruption of the Met signal transduc-
tion pathway at femtomolar levels in tumor cells by gel-
danamycin and its derivatives has remained unknown.
The above-described disruption of hsp90 function is an
effect known for this ansamycin class of drugs at higher
concentrations, that is, micromolar and greater. We
assess here the structure–activity relationship of geld-
Figure 1. Geldanamycin, macbecin I, and radicicol.

4962
Y. Shen et al. / Bioorg. Med. Chem. 13 (2005) 4960–4971
Figure 4. 17-N-Azetidinyl-17-demethoxygeldanamycin derivatives.
with acetic anhydride and 4-dimethylaminopyridine to
provide 11-O-acetyl-17-N-azetidinyl-17-demethoxygel-
danamycin (18).35 The 7-urethane group of compound
14 could be removed per slight modification of the liter-
ature procedure35 by treatment with potassium tert-
butoxide in tert-butanol (in lieu of the solvent dimethyl-
sulfoxide, which gave a lower product yield) to provide
17-N-azetidinyl-7-decarbamoyl-17-demethoxygeldana-
mycin (19).
Figure 2. 17-Amino-17-demethoxygeldanamycin derivatives.
with aziridine gave the new compound 17-N-aziridinyl-
17-demethoxygeldanamycin (15).
Finally, dihydrogeldanamycin (20), the hydroquinone
version of geldanamycin, and the geldanamycin-related
benzoquinone ansamycin macbecin I (2), were prepared
as per reported methods [sodium dithionite reduction of
geldanamycin and ferric chloride oxidation of macbecin
II (21), respectively] (Fig. 5).
Two derivatives of the geldanamycin derivative class,
geldanoxazinone, were synthesized (Fig. 3). Such deriv-
atives can be prepared by acid-catalyzed condensation
of geldanamycin with a 2-aminophenol.37 5-Bromo-2-
aminophenol and 5-iodo-2-aminophenol were thus used
to prepare adducts 16 and 17 in 60% and 44% yield,
respectively.
2.2. Biological evaluation
In an effort to investigate the effect of modification of
the ansa ring of geldanamycin on activity, an active
17-aminosubstituted-17-demethoxygeldanamycin deriv-
ative, 17-N-azetidinyl-17-demethoxygeldanamycin (14),
was used for making such changes (Fig. 4). The 11-hy-
droxyl group of the latter compound could be esterified
The aforementioned compounds were investigated for
their inhibitory effect on the HGF/SF-Met-induced uro-
kinase plasminogen activation (uPA) pathway, as per
a previously described method.14 The 17-alkyl- and
17-cycloalkyl-amino-17-demethoxygeldanamycin deriv-
atives 4–8, 14, and 15 were found to be active at low
concentrations for the inhibitory effect on the Met-uPA
pathway in the cell-based assay, whereas compounds 9–
13 were inactive. Table 1 shows the uPA inhibition index
for the compounds investigated. Dramatically, the com-
pounds fall into two groups: those with an index greater
than 12 and those less than 8. The former group displays
activity well into the femtomolar range.
The bromo-substituted and iodo-substituted geldanox-
azinones 16 and 17, respectively, were found to be inhib-
itory to the Met signaling pathway only at nanomolar
concentrations, that is, <8 IC50 (Table 1).
The modifications introduced to the ansa ring of a high-
ly active compound 14 (>15 IC50) led to the formation
of compounds 18 and 19 that exhibited only <8 IC50
Met-uPA signaling inhibitory activity (Table 1).
Testing of the inhibitory activity of the dihydroquinones
dihydrogeldanamycin (20) and macbecin II (21) was
Figure 3. Geldanoxazinones.

Y. Shen et al. / Bioorg. Med. Chem. 13 (2005) 4960–4971
4963
Figure 5. Dihydroquinone ansamycins.
Table 1. uPA-plasmin inhibition index of compounds
Compound
Chemical name
uPA-plasmin inhibition indexa
8
4
17-(2-Fluoroethyl)amino-17-demethoxygeldanamycin
17-Allylamino-17-demethoxygeldanamycin
17-N-Aziridinyl-17-demethoxygeldanamycin
17-Amino-17-demethoxygeldanamycin
17-N-Azetidinyl-17-demethoxygeldanamycin
17-(2-Dimethylaminoethyl)amino-17-demethoxygeldanamycin
Geldanamycin
19
18.0
15.7
15.3
15
15
6
14
5
14.9
14.3
14.0
12.7
7.9
1
7
17-(2-Chloroethyl)amino-17-demethoxygeldanamycin
Dihydrogeldanamycin
20
18
3
11-O-Acetyl-17-N-azetidinyl-17-demethoxygeldanamycin
Radicicol
Macbecin II
7.9
6.5
21
2
Macbecin I
6.4
6.3
13
9
17-Carboxymethylamino-17-demethoxygeldanamycin
17-(2-Acetylaminoethyl)amino-17-demethoxygeldanamycin
50-Iodogeldanoxazone
5.8
5.8
17
12
19
11
10
16
17-(8-Acetamido-3,6-dioxaoctylamino)-17-demethoxygeldanamycin
17-N-Azetidinyl-7-decarbamyl-17-demethoxygeldanamycin
17-(6-Biotinylaminohexyl)amino-17-demethoxygeldanamycin
17-(6-Acetylaminohexyl)amino-17-demethoxygeldanamycin
50-Bromogeldanoxazone
5.8
5.7
5.5
5.3
5.3
a uPA-plasmin inhibition index or IC50 is the negative log of the drug concentration at which 50% inhibition of uPA occurs when MDCK cells are
treated with HGF/SF.14 Compounds with IC50 higher than 12 are referred to fM-GAi, while compounds with index lower than 8 belong to nM-
GAi.
done and compared with those of their benzoquinone
analogs, geldanamycin and macbecin I (3). Additionally,
testing was done of radicicol (3). Though radicicol29 and
macbecins I27 and II15 are known for their affinity for
hsp90, each of these compounds exhibited poor activity
in our HGF/SF-induced uPA-plasmin assays. However,
dihydrogeldanamycin (20) was found to be >12 IC50 and
active, as was seen also with geldanamycin (Table 1).
ical investigations of 17-N-allylamino-17-demethoxygel-
danamycin (4) for cancer treatment.32,33,39 Preliminary
reports of clinical trials have indicated that the drug
has efficacy as an anticancer therapeutic. It has been
reported that hepatic toxicity is dose-limiting. Non-
dose-limiting toxicities have been reported to include
anemia, anorexia, nausea, emesis, and diarrhea.32,33
Our results indicate that various geldanamycin deriva-
tives are inhibitors of HGF/SF-induced Met-dependent
urokinase activation in MDCK cells and in cancer cell
lines in a fashion that is distinct from the traditional
hsp90 binding and inhibition paradigm. Our lines of evi-
dence for this characteristic are as follows.
3. Discussion
Investigations into the therapeutic potential of geldana-
mycin and its derivatives have been focusing primarily
on biological processes where hsp90 plays a critical
role.33,38,39 Multiple proteins critical in cancer cell
survivability and proliferation are dependent on this
chaperone protein.32,40 The ability of geldanamycin
derivatives to block the function of hsp90 has led to clin-
Geldanamycin derivatives can act as inhibitors of the
Met-urokinase-plasminogen activation pathway at con-
centrations far below that needed for inhibition of hsp90
function. Although currently we cannot exclude the pos-

4964
Y. Shen et al. / Bioorg. Med. Chem. 13 (2005) 4960–4971
sibility that some isoforms of hsp90 or a specific com-
plex of drug and hsp90 are contributing to this exquisite
sensitivity, our highly active compounds exhibit this
activity at femtomolar levels, well below the nanomolar
concentrations that have been ascribed to hsp90 inhibi-
tion. Indeed, our most active compound, 17-(2-fluoro-
ethyl)amino-17-demethoxygeldanamycin (8), exhibited
activity to 10ꢀ19 M concentrations. Of the 21 com-
pounds tested for their inhibitory effect on the HGF/
SF-Met-induced urokinase plasminogen activation
(uPA) pathway in our cell-based assays, nine of these
had inhibitory activity in the 10ꢀ12 to 10ꢀ19 M range.
The remaining 12 compounds had inhibitory activity
in a distinct 10ꢀ5 to 10ꢀ8 M range.
mycin in the macbecin II-treated cells was the same as
that in the parental MDCK cells. HGF/SF could still
significantly upregulate uPA activity and this could also
be inhibited by geldanamycin at femtomolar levels.
These studies have further confirmed the fact that
geldanamycin inhibits HGF/SF-induced uPA activity
through non-hsp90 target(s).15
Although the unknown target(s) of our active geldana-
mycin derivatives remain to be identified, our data
suggest some structure–activity relationships. Some 17-
N-aminoderivatized-17-demethoxygeldanamycin com-
pounds were found to be active in our cell-based assays,
but others were not. The latter includes those with long-
er 17-N-aminosubstitutions, for example, 9–12, in addi-
tion to the carboxylate derivative 13.
Additionally, we have found that the activity seen does
not always correlate with affinity to human a-hsp90.
The known hsp90 binding compounds, geldanamycin
(1),26,16,41,19,42 17-allylamino-17-demethoxygeldanamy-
cin (4),18,42,43 and 17-(2-dimethylaminoethyl)amino-17-
demethoxygeldanamycin (5),20 were found to be highly
active compounds in our cell-based assays. However,
the known hsp90-binding compounds, macbecin I
(2),27 macbecin II (21),15 and, in particular, the hsp90
high affinity compound radicicol (3),19,42 were found
to have low inhibitory activity in our assays of the
HGF/SF-Met-induced uPA pathway.
In regard to ansa ring modifications, when the 7-ure-
thane group was removed from the active geldanamycin
derivative 14, the resulting decarbamoylated compound
19 was found to be inactive. The urethane functionality
is known to undergo hydrogen bonding interactions with
several aminoacid residues in the crystal structures of gel-
danamycin and geldanamycin derivative 5 complexed
with the amino-terminal domain of hsp90.16,20 Addition-
ally, it has been reported by Schnur et al. that the
7-urethane was needed for anti-erbB-2 activity.35 The
7-urethane of geldanamycin derivatives is buried deep
in the ATP-binding site of hsp90. Thus, it is possible that
the binding site for geldanamycin of our unknown tar-
get(s) for Met function may share similarities with this
binding area of hsp90. We additionally modified the ac-
tive geldanamycin derivative 14 with acetylation of the
11-hydroxyl group to give 18, which was inactive in
our cell-based assays.
Further evidence distinguishing the mode of action of
high activity compounds from acting via hsp90 comes
from the observed effect of our compounds on hsp90
and Met expression. The reported cellular effect of gel-
danamycin is such that hsp90 is usually up-regulated
and that of Met expression is down-regulated in vi-
tro.44,45,15 We have observed this effect of geldanamycin
on hsp90 and Met expression levels only at higher con-
centrations (<8 IC50). At subnanomolar concentrations
(>12 IC50), where uPA activity remains inhibited, we
did not detect any change of either hsp90 or Met
expression.15
As mentioned, the activity we observed showed differ-
ences with the known relative affinity of compounds
with hsp90. Although the binding site on the target(s)
of our cell-based uPA assays remains unknown, the site
may possibly also be an ATP-binding site, albeit with
some differences. It has recently been reported by Kamal
et al. that a high-affinity conformation of hsp90 in tu-
mor cells accounts for the tumor selectivity of 17-N-all-
ylamino-17-demethoxygeldanamycin (4) and radicicol
(3).43 The hsp90 of the tumor cells forms multichaper-
one complexes, whereas normal tissue hsp90 is not so
complexed. It is not possible to speculate if such tar-
get(s) is so complexed changing the conformation of
the geldanamycin-binding site.
Finally, evidence for a non-hsp90 direct effect of our ac-
tive compounds comes from the effect of tested com-
pounds on tumor cell invasion and proliferation. High
activity geldanamycin derivatives not only inhibit uPA
activity at femtomolar level, they also inhibit tumor cell
invasion in vitro.15 However, proliferation is inhibited
only at nanomolar levels, the same concentration levels
of low activity geldanamycin derivatives.14 This suggests
that geldanamycin inhibits proliferation and invasion by
several mechanisms. Thus, while proliferation may be
affected via inhibition of hsp90 function, invasion may
be affected by geldanamycin interaction with unknown
target(s). In support of this hypothesis, the MDCK cells
were intentionally cultured in the presence of macbecin
II (21). This ansamycin inhibits both invasion and pro-
liferation activity at nanomolar level. MDCK cells were
maintained at the highest non-toxic concentrations of
macbecin II (21) (3 lM) for several months. Under these
conditions, both Met and hsp90 returned to parental
levels and Met responsiveness to HGF/SF was restored,
but hsp90 appeared to remain in complex with macbe-
cin. Strikingly, the uPA-plasmin sensitivity to geldana-
That the Met uPA-plasmin and invasion pathway is so
exquisitely sensitive to our active compounds is also sug-
gestive that such compounds may play a catalytic role in
the disruption of the pathway. Dihydrogeldanamycin
(20) was found to be active in our assays, albeit slightly
less so than geldanamycin itself. The former has been
reported to be air-oxidizable to the latter,36 and thus this
cannot be discounted as a possible contributing factor to
dihydrogeldanamycinÕs observed activity. However, the
related ansamycins macbecin I (2) and its reduction
product macbecin II (21) were both observed to be inac-
tive. Both of these latter compounds bind hsp90.27,15

Y. Shen et al. / Bioorg. Med. Chem. 13 (2005) 4960–4971
4965
It may be that the active ansamycin derivatives partici-
pate in a catalytic electron-transfer redox process and
that the oxidation–reduction potential between dihydro-
geldanamycin (20) and geldanamycin is critical for it to
be able to do so. The potential between the two macbec-
ins may be sufficiently different not to allow this. It has
been reported that geldanamycin can participate in the
generation of superoxide via a redox process.46
the mixture was washed with brine, dried over anhy-
drous sodium sulfate, and concentrated. Separation by
flash chromatography on silica gel (1:2 hexane/ethyl ace-
tate) gave the product as a purple solid (5.3 mg, 99%).
IR (KBr) (cmꢀ1) 3464, 3333, 2958, 2929, 2825, 1728,
1691, 1652, 1571, 1485, 1372, 1323, 1189, 1101, 1057;
UV (95% EtOH) (nm) 332 (e = 2.0 · 104); 1H NMR
(CDCl3, 500 MHz) d 9.14 (s, 1H), 7.28 (s, 1H), 6.93
(br d, J = 11.5 Hz, 1H), 6.56 (br dd, J = 11.5, 11.0 Hz,
1H), 6.38 (br t, J = 6.0 Hz, 1H), 5.94–5.81 (m, 3H),
5.30–5.24 (m, 2H), 5.17 (s, 1H), 4.82 (br s, 2H), 4.29
(br d, J = 10.0 Hz, 1H), 4.21 (br s, 1H), 4.18–4.08 (m,
2H), 3.55 (ddd, J = 9.0, 6.5, 2.0 Hz, 1H), 3.43 (ddd,
J = 9.0, 3.0, 3.0 Hz, 1H), 3.34 (s, 3H), 3.25 (s, 3H),
2.72 (dqd, J = 9.5, 7.0, 2.0 Hz, 1H), 2.63 (d,
J = 14.0 Hz, 1H), 2.34 (dd, J = 14.0, 11.0 Hz, 1H), 2.00
(br s, 3H), 1.78 (d, J = 1.0 Hz, 3H), 1.78–1.74 (m, 2H),
1.74–1.67 (m, 1H), 0.99–0.95 (m, 6H); 13C NMR
(CDCl3, 125 MHz, assignment of protonated carbons
aided by DEPT) d 183.8 (18-C), 180.9 (21-C), 168.4 (1-
C), 156.0 (7-O2CNH2), 144.6 (17-C), 141.2 (20-C),
135.8 (5-C), 134.9 (2-C), 133.7 (9-C), 132.7 (8-C),
132.5 (20-C), 126.9 (4-C), 126.5 (3-C), 118.5 (30-C),
108.8 (19-C), 108.7 (16-C), 81.6 (7-C), 81.4 (12-C),
81.2 (6-C), 72.6 (11-C), 57.1 (6- or 12-OCH3), 56.7 (6-
or 12-OCH3), 47.8 (10-C), 35.0 (13-C), 34.3 (15-C),
32.3 (10-C), 28.4 (14-C), 22.9 (14-CH3), 12.8 (8-CH3),
12.6 (2-CH3), 12.3 (10-CH3); HRMS (FAB) [M+H]+
calcd for C31H44N3O8, 586.3129; found 586.3120.
The low concentrations we have found that are needed
to arrest the Met signaling responsible for the invasive
and metastatic behavior of solid tumors makes the
active compounds of our cell-based assays attractive as
potential drugs. The low concentrations needed may
eliminate the observed dose-dependent toxicities of gel-
danamycin derivatives. Successful identification and iso-
lation of the unknown targets of such derivatives may
allow a better design of compounds that may be effica-
cious in blocking this Met signaling pathway.
4. Conclusions
Geldanamycin and some 17-amino-17-demethoxygel-
danamycin derivatives were found to block hepatocyte
growth factor/scatter factor-mediated Met tyrosine
kinase receptor-dependent urokinase-plasmin activation
and scattering at femtomolar levels. In contrast, other
ansamycins (macbecins I and II), geldanamycin deriva-
tives, and radicicol required concentrations several logs
higher (greater than nanomolar) to effect such inhibi-
tion. Additionally, the ability of tested compounds to
effect this inhibition was discordant with the known
ability to bind hsp90. These results are indicative of a
novel target(s) for hepatocyte growth factor/scatter fac-
tor-mediated urokinase-plasmin activation.
5.3. 17-(2-Dimethylaminoethyl)amino-17-demethoxygel-
danamycin (5)48
N,N-Dimethylethylenediamine (6.0 ll, 0.055 mmol) was
added to a solution of (+)-geldanamycin (4.3 mg,
7.7 lmol) in chloroform (1.0 ml). The mixture was stir-
red at room temperature. Upon complete conversion of
geldanamycin shown by thin-layer chromatography
(4 h), the mixture was washed with 0.5% aqueous sodium
hydroxide solution and brine, dried over anhydrous sodi-
um sulfate, and concentrated. Separation by flash chro-
matography on silica gel (9:1 ethyl acetate/methanol)
gave the product as a purple solid (4.5 mg, 95%). IR
(KBr) (cmꢀ1) 3462, 3329, 2932, 2871, 2824, 2774, 1733,
1690, 1653, 1565, 1485, 1373, 1321, 1253, 1188, 1100,
1055; UV (95% EtOH) (nm) 332 (e = 1.7 · 104); 1H
NMR (CDCl3, 500 MHz) d 9.18 (s, 1H), 7.24 (s, 1H),
7.04 (br t, J = 5.0 Hz, 1H), 6.94 (br d, J = 11.5 Hz,
1H), 6.57 (br dd, J = 11.5, 11.0 Hz, 1H), 5.90 (br d,
J = 9.5 Hz, 1H), 5.84 (dd, J = 11.0, 10.0 Hz, 1H), 5.17
(s, 1H), 4.75 (br s, 2H), 4.42 (br s, 1H), 4.29 (br d,
J = 10.0 Hz, 1H), 3.70–3.42 (m, 3H), 3.57 (br dd,
J = 9.0, 6.5 Hz, 1H), 3.34 (s, 3H), 3.25 (s, 3H), 2.72
(dqd, J = 9.5, 7.0, 2.0 Hz, 1H), 2.67 (d, J = 14.0 Hz,
1H), 2.55 (t, J = 5.5 Hz, 2H), 2.38 (dd, J = 14.0,
11.0 Hz, 1H), 2.25 (s, 6H), 2.01 (br s, 3H), 1.83–1.68
(m, 3H), 1.78 (br s, 3H), 0.98 (d, J = 7.0 Hz, 3H), 0.95
(d, J = 6.5 Hz, 3H); MS (FAB) [M+H]+ found 617.
5. Experimental
5.1. General methods
Melting points are uncorrected. Infrared spectra were
recorded on a Matton Galaxy Series FTIR 3000 spectro-
photometer. Ultraviolet–visible spectra were recorded on
a Hitachi U-4001 spectrophotometer. 1H and 13C NMR
spectra were recorded on Varian Inova-600, UnityPlus-
500, VRX-500, or VRX-300 spectrometers. The number-
ing used in assignments is based on the geldanamycin
ring system unless otherwise indicated. Mass spectra
were performed by the MSU Mass Spectrometry Facili-
ty. Geldanamycin and macbecin II were provided by the
National Cancer Institutes. Macbecin I was synthesized
from macbecin II as per published procedure.47 Radici-
col was obtained commercially (Sigma-Aldrich). Anhy-
drous solvents were purified as per standard methods.
5.2. 17-Allylamino-17-demethoxygeldanamycin (4)35,36
(+)-Geldanamycin (5.1 mg, 9.0 lmol) was stirred with
allylamine (10.0 ll, 0.13 mmol) in chloroform (1.5 ml)
at room temperature. Upon complete conversion of gel-
danamycin shown by thin-layer chromatography (18 h),
5.4. 17-Amino-17-demethoxygeldanamycin (6)36,49,50
Concentrated aqueous solution of ammonia (28%,
0.70 ml, 0.010 mol) was added to a solution of (+)-gel-

4966
Y. Shen et al. / Bioorg. Med. Chem. 13 (2005) 4960–4971
danamycin (5.0 mg, 9.0 lmol) in acetonitrile (5.0 ml) at
room temperature. The yellow solution turned slowly
dark red. Upon complete conversion of geldanamycin
shown by thin-layer chromatography (5 h), the mixture
was partitioned between ethyl acetate and brine. The
organic phase was washed with brine, dried over anhy-
drous sodium sulfate, and concentrated. Separation of
the solid residue by flash chromatography on silica gel
(1:2 hexane/ethyl acetate) gave the product as a dark
red solid (4.6 mg, 95%). IR (KBr) (cmꢀ1) 3452, 3339,
2957, 2931, 2825, 1721, 1692, 1617, 1591, 1495, 1374,
1323, 1250, 1190, 1133, 1101, 1055; UV (95% EtOH)
5.6. 17-(2-Fluoroethyl)amino-17-demethoxygeldanamycin
(8)36
Sodium hydroxide aqueous solution (1.10 M, 0.53 ml,
0.58 mmol) was added to a mixture of (+)-geldanamycin
(5.5 mg, 9.8 lmol) and 2-fluoroethylamine hydrochlo-
ride (65 mg, 0.59 mmol) in acetonitrile (1.0 ml). The
mixture was stirred at room temperature. Upon com-
plete conversion of geldanamycin shown by thin-layer
chromatography (12 h), the mixture was partitioned be-
tween ethyl acetate and brine. The organic phase was
washed with brine, dried over anhydrous sodium sulfate,
and concentrated. Separation by flash chromatography
on silica gel (1:2 hexane/ethyl acetate) gave the product
as a purple solid (5.7 mg, 98%). IR (KBr) (cmꢀ1) 3465,
3330, 2954, 2927, 2873, 1728, 1691, 1653, 1576, 1487,
1375, 1323, 1255, 1190, 1103, 1051; UV (95% EtOH)
1
(nm) 328 (e = 2.0 · 104); H NMR (CDCl3, 500 MHz)
d 9.08 (s, 1H), 7.26 (s, 1H), 6.95 (br d, J = 11.5 Hz,
1H), 6.56 (br dd, J = 11.5, 11.0 Hz, 1H), 5.89–5.82 (m,
2H), 5.37 (br s, 2H), 5.17 (s, 1H), 4.73 (br s, 2H), 4.29
(br d, J = 10.0 Hz, 1H), 3.98 (br s, 1H), 3.59 (ddd,
J = 9.0, 6.5, 2.0 Hz, 1H), 3.42 (ddd, J = 9.0, 3.0,
3.0 Hz, 1H), 3.34 (s, 3H), 3.25 (s, 3H), 2.75 (dqd,
J = 9.5, 7.0, 2.0 Hz, 1H), 2.65 (d, J = 14.0 Hz, 1H),
2.01 (br s, 3H), 1.97–1.75 (m, 4H), 1.79 (d, J = 1.0 Hz,
3H), 0.99–0.97 (m, 6H); 13C NMR (CDCl3, 125 MHz)
d 183.1, 180.4, 167.9, 156.1, 146.0, 140.4, 135.8, 135.0,
134.0, 133.0, 126.9, 126.6, 110.3, 108.6, 81.9, 81.2,
81.1, 72.2, 57.1, 56.8, 35.0, 34.7, 32.2, 28.7, 23.8, 12.8,
12.5, 12.2; HRMS (FAB) [M+H]+ calcd for
C28H40N3O8, 546.2816; found 546.2818.
1
(nm) 332 (e = 1.7 · 104); H NMR (CDCl3, 500 MHz)
d 9.10 (s, 1H), 7.29 (s, 1H), 6.94 (br d, J = 11.5 Hz,
1H), 6.57 (br dd, J = 11.5, 11.0 Hz, 1H), 6.36 (br t,
J = 5.0 Hz, 1H), 5.88–5.83 (m, 2H), 5.18 (s, 1H), 4.75
(br s, 2H), 4.69–4.57 (m, 2H), 4.30 (br d, J = 10.0 Hz,
1H), 3.94–3.76 (m, 2H), 3.56 (br d, J = 9.0 Hz, 1H),
3.43 (ddd, J = 9.0, 3.0, 3.0 Hz, 1H), 3.35 (s, 3H), 3.26
(s, 3H), 2.73 (dqd, J = 9.5, 7.0, 2.0 Hz, 1H), 2.70 (d,
J = 14.0 Hz, 1H), 2.30 (dd, J = 14.0, 11.0 Hz, 1H), 2.01
(br s, 3H), 1.80–1.76 (m, 2H), 1.78 (d, J = 1.0 Hz, 3H),
1.75–1.68 (m, 1H), 0.99 (d, J = 7.0 Hz, 3H), 0.97 (d,
J = 6.5 Hz, 3H); 13C NMR (CDCl3, 125 MHz) d 183.8,
181.2, 168.4, 156.0, 144.9, 140.9, 135.9, 135.0, 133.6,
132.8, 127.0, 126.5, 109.7, 109.1, 81.6, 81.5 (d,
J = 170 Hz), 81.4, 81.2, 72.6, 57.2, 56.7, 46.0 (d,
J = 20 Hz), 35.1, 34.3, 32.4, 28.8, 23.0, 12.8, 12.6, 12.5;
HRMS (FAB) [M]+ calcd for C30H42FN3O8, 591.2956;
found 591.2952.
5.5. 17-(2-Chloroethyl)amino-17-demethoxygeldanamycin
(7)50
Sodium hydroxide aqueous solution (2.80 M, 0.75 ml,
2.1 mmol) was added to a mixture of (+)-geldanamycin
(11.7 mg, 0.021 mmol) and 2-chloroethylamine hydro-
chloride (242 mg, 2.1 mmol) in acetonitrile (3.0 ml).
The mixture was stirred at room temperature. Upon
complete conversion of geldanamycin shown by thin-
layer chromatography (20 h), the mixture was parti-
tioned between ethyl acetate and brine. The organic
phase was washed with brine, dried over anhydrous
sodium sulfate, and concentrated. Separation by flash
chromatography on silica gel (1:2 hexane/ethyl acetate)
gave the product as a purple solid (12.0 mg, 95%). IR
(KBr) (cmꢀ1) 3334, 2938, 2874, 2822, 1733, 1696,
1653, 1577, 1489, 1375, 1325, 1274, 1190, 1136, 1101,
1060; UV (95% EtOH) (nm) 332 (e = 1.9 · 104); 1H
NMR (CDCl3, 500 MHz) d 9.09 (s, 1H), 7.29 (s, 1H),
6.94 (br d, J = 11.5 Hz, 1H), 6.56 (ddd, J = 11.5, 11.0,
1.0 Hz, 1H), 6.35 (br t, J = 5.0 Hz, 1H), 5.87 (br d,
J = 9.5 Hz, 1H), 5.85 (br dd, J = 11.0, 10.0 Hz, 1H),
5.18 (s, 1H), 4.72 (br s, 2H), 4.30 (br d, J = 10.0 Hz,
1H), 4.03 (br s, 1H), 3.94–3.83 (m, 2H), 3.75–3.67 (m,
2H), 3.56 (ddd, J = 9.0, 6.5, 2.0 Hz, 1H), 3.43 (ddd,
J = 9.0, 3.0, 3.0 Hz, 1H), 3.35 (s, 3H), 3.26 (s, 3H),
2.73 (dqd, J = 9.5, 7.0, 2.0 Hz, 1H), 2.70 (d,
J = 14.0 Hz, 1H), 2.24 (dd, J = 14.0, 11.0 Hz, 1H), 2.01
(br s, 3H), 1.78 (d, J = 1.0 Hz, 3H), 1.80–1.75 (m, 2H),
1.75–1.68 (m, 1H), 1.00–0.96 (m, 6H); 13C NMR
(CDCl3, 125 MHz) d 183.8, 181.2, 168.3, 155.9, 144.7,
140.8, 135.9, 135.0, 133.6, 132.9, 127.0, 126.5, 110.0,
109.1, 81.6, 81.4, 81.2, 72.7, 57.1, 56.7, 46.9, 42.7, 35.1,
34.4, 32.4, 28.8, 23.0, 12.8, 12.6, 12.5; MS (FAB)
[M+H]+ found 608.
5.7. 17-(2-Acetylaminoethyl)amino-17-demethoxygel-
danamycin (9)36
N-Acetylethylenediamine (90%, 10.0 ll, 0.094 mmol)
was added to a solution of (+)-geldanamycin (5.0 mg,
8.9 lmol) in chloroform (1.0 ml) at room temperature.
Upon complete conversion of geldanamycin shown by
thin-layer chromatography (10 h), the mixture was
washed with distilled water, dried over anhydrous sodi-
um sulfate, and concentrated. Separation by flash chro-
matography on silica gel (ethyl acetate) gave the product
as a purple solid (4.5 mg, 80%). IR (KBr) (cmꢀ1) 3449,
3338, 2932, 2881, 2824, 1718, 1685, 1654, 1569, 1487,
1
1374, 1323, 1269, 1189, 1102, 1057; H NMR (CDCl3,
500 MHz) d 9.12 (s, 1H), 7.23 (s, 1H), 6.94 (br d,
J = 11.5 Hz, 1H), 6.63 (br t, J = 5.0 Hz, 1H), 6.56 (br
dd, J = 11.5, 11.0 Hz, 1H), 5.88 (br d, J = 9.5 Hz, 1H),
5.84 (dd, J = 11.0, 10.0 Hz, 1H), 5.80 (br t, J = 6.0 Hz,
1H), 5.17 (s, 1H), 4.72 (br s, 2H), 4.29 (br d,
J = 10.0 Hz, 1H), 4.17 (br s, 1H), 3.77–3.62 (m,
2H), 3.58–3.46 (m, 3H), 3.42 (ddd, J = 9.0, 3.0, 3.0 Hz,
1H), 3.34 (s, 3H), 3.25 (s, 3H), 2.73 (dqd, J = 9.5,
7.0, 2.0 Hz, 1H), 2.64 (d, J = 14.0 Hz, 1H), 2.33
(dd, J = 14.0, 11.0 Hz, 1H), 2.01 (s, 3H), 2.00 (d,
J = 1.0 Hz, 3H), 1.80–1.76 (m, 2H), 1.78 (d, J =
1.0 Hz, 3H), 1.74–1.67 (m, 1H), 0.98 (d, J = 7.0 Hz,
3H), 0.95 (d, J = 6.5 Hz, 3H); HRMS (FAB)

Y. Shen et al. / Bioorg. Med. Chem. 13 (2005) 4960–4971
4967
[M+H]+ calcd for C32H47N4O9, 631.3343; found
631.3344.
2H), 2.66 (d, J = 14.0 Hz, 1H), 2.38 (dd, J = 14.0,
11.0 Hz, 1H), 2.21–2.15 (m, 2H), 2.01 (br s, 3H), 1.78
(d, J = 1.0 Hz, 3H), 1.78–1.32 (m, 17H), 0.98 (d,
J = 7.0 Hz, 3H), 0.95 (d, J = 6.5 Hz, 3H); HRMS
(FAB) [M+H]+, calcd for C44H67N6O10S, 871.4592;
found 871.4619.
5.8. 17-(6-Acetylamino-1-hexyl)amino-17-demethoxygel-
danamycin (10)
A solution of (+)-geldanamycin (5.7 mg, 0.010 mmol)
and N-(6-aminohexyl)acetamide51 (5.5 mg, 0.035 mmol)
in chloroform was stirred at room temperature. Upon
complete conversion of geldanamycin shown by thin-
layer chromatography (20 h), the mixture was washed
with distilled water, dried over anhydrous sodium sul-
fate, and concentrated. Separation by flash chromato-
graphy on silica gel (ethyl acetate) gave the product as
a purple solid (5.7 mg, 82%). IR (KBr) (cmꢀ1) 3445,
3323, 3202, 2931, 2865, 2824, 1723, 1687, 1653, 1562,
1486, 1371, 1322, 1256, 1188, 1135, 1106; UV (95%
EtOH) (nm) 333 (e = 1.2 · 104); 1H NMR (CDCl3,
500 MHz) d 9.17 (br s, 1H), 7.26 (s, 1H), 6.94 (br d,
J = 11.5 Hz, 1H), 6.57 (br dd, J = 11.5, 11.0 Hz, 1H),
6.26 (br t, J = 5.0 Hz, 1H), 5.89 (br d, J = 9.5 Hz, 1H),
5.85 (dd, J = 11.0, 10.0 Hz, 1H), 5.42 (br s, 1H), 5.18
(s, 1H), 4.73 (br s, 2H), 4.31 (br s, 1H), 4.29 (br d,
J = 10.0 Hz, 1H), 3.59–3.39 (m, 4H), 3.35 (s, 3H),
3.27–3.19 (m, 2H), 3.25 (s, 3H), 2.74 (dqd, J = 9.5, 7.0,
2.0 Hz, 1H), 2.66 (d, J = 14.0 Hz, 1H), 2.39 (dd, J
=14.0, 11.0 Hz, 1H), 2.01 (br s, 3H), 1.96 (s, 3H),
1.80–1.75 (m, 2H), 1.78 (d, J = 1.0 Hz, 3H), 1.73–1.62
(m, 3H), 1.55–1.47 (m, 2H), 1.46–1.33 (m, 4H), 0.99
(d, J = 7.0 Hz, 3H), 0.95 (d, J = 6.5 Hz, 3H); 13C
NMR (CDCl3, 125 MHz) d 183.9, 180.7, 170.0, 168.4,
156.0, 144.9, 141.5, 135.9, 135.0, 133.8, 132.8, 127.0,
126.6, 108.7, 108.4, 81.7, 81.5, 81.2, 72.7, 57.2, 56.7,
45.8, 39.4, 35.1, 34.4, 32.4, 29.7, 29.6, 28.6, 26.5, 26.4,
23.4, 22.9, 12.8, 12.6, 12.4; HRMS (FAB) [M+H]+ calcd
for C36H55N4O9, 687.3969; found 687.3967.
5.10. 17-[2-[2-(2-Acetylaminoethoxy)ethoxy]ethyl]amino-
17-demethoxygeldanamycin (12)
A
mixture of 2,20-(ethylenedioxy)bis(ethylamine)
(56.0 ll, 0.38 mmol), acetic anhydride (46.0 ll,
0.48 mmol), and triethylamine (73.2 ll, 0.52 mmol) in
chloroform (1.0 ml) was stirred for 1 h at room tempera-
ture and then concentrated to dryness under high vacu-
um. The colorless solid residue was then stirred with
(+)-geldanamycin (4.0 mg, 7.1 lmol) in chloroform
(1.0 ml). Upon complete conversion of geldanamycin
shown by thin-layer chromatography (20 h), the mixture
was washed with distilled water, dried over anhydrous
sodium sulfate, and concentrated. Separation by flash
chromatography on silica gel (5:1 ethyl acetate/metha-
nol) gave the desired product as a purple solid (1.1 mg,
21%). IR (KBr) (cmꢀ1) 3446, 3336, 2960, 2929, 2877,
1727, 1689, 1655, 1566, 1487, 1375, 1325, 1261, 1190,
1103, 1057; 1H NMR (CDCl3, 300 MHz) d 9.17 (s,
1H), 7.25 (s, 1H), 6.94 (br d, J = 11.5 Hz, 1H), 6.78 (br
t, J = 5.0 Hz, 1H), 6.57 (br dd, J = 11.5, 11.0 Hz, 1H),
6.36 (br s, 1H), 5.89 (br d, J = 9.5 Hz, 1H), 5.85 (dd,
J = 11.0, 10.0 Hz, 1H), 5.18 (s, 1H), 4.74 (br s, 2H),
4.29 (br d, J = 10.0 Hz, 1H), 4.26 (br s, 1H), 3.78–3.40
(m, 14H), 3.35 (s, 3H), 3.25 (s, 3H), 2.78–2.64 (m, 2H),
2.39 (dd, J = 14.0, 11.0 Hz, 1H), 2.01 (br s, 3H), 1.98
(s, 3H), 1.78–1.67 (m, 3H), 1.78 (d, J = 1.0 Hz, 3H),
0.99–0.94 (m, 6H); HRMS (FAB) [M+H]+ calcd for
C36H55N4O11, 719.3867; found 719.3864.
5.9. (+)-Biotin 17-(6-aminohexyl)amino-17-demethoxy-
geldanamycin amide (11)
5.11. 17-Carboxymethylamino-17-demethoxygeldana-
mycin (13)
1,6-Diaminohexane (10.0 mg, 0.086 mmol) was added to
a solution of (+)-geldanamycin (5.0 mg, 8.9 lmol) in
chloroform (1.0 ml) at room temperature. Upon com-
plete conversion of geldanamycin shown by thin-layer
chromatography (20 h), the mixture was washed with
0.5% aqueous sodium hydroxide solution and brine,
dried over potassium carbonate, and concentrated.
The resulted dark purple solid was then stirred over-
night with (+)-biotin N-hydroxysuccinimide ester
(3.0 mg, 8.8 lmol) in DMF (1.0 ml). Removal of the sol-
vent and separation by flash chromatography on silica
gel (9:1 ethyl acetate/methanol) gave the product as a
purple solid (6.5 mg, 85%). IR (KBr) (cmꢀ1) 3327,
2931, 2864, 1709, 1651, 1562, 1485, 1371, 1325, 1255,
(+)-Geldanamycin (3.1 mg, 5.5 lmol) was stirred at
room temperature with glycine sodium salt (10.7 mg,
0.11 mmol) in a mixture of ethanol (1.2 ml) and water
(0.3 ml). Upon complete conversion of geldanamycin
shown by thin-layer chromatography (3 h), the purple
mixture was acidified with diluted hydrochloric acid
and partitioned between chloroform and distilled water.
The organic phase was dried over anhydrous sodium
sulfate and concentrated. Separation by flash chroma-
tography on silica gel (1:1 ethyl acetate/methanol) gave
the product as a purple solid (3.2 mg, 96%). IR (KBr)
(cmꢀ1) 3446, 3305, 2929, 2875, 1734, 1693, 1655, 1618,
1574, 1485, 1394, 1319, 1267, 1139, 1072; 1H NMR
(CDCl3, 500 MHz) d 8.91 (s, 1H), 7.25 (s, 1H), 6.83
(br s, 1H), 6.80 (br d, J = 11.5 Hz, 1H), 6.60 (br dd,
J = 11.5, 11.0 Hz, 1H), 5.86–5.80 (m, 2H), 5.16 (s, 1H),
4.95 (br s, 2H), 4.33–4.21 (m, 2H), 4.27 (br d,
J = 10.0 Hz, 1H), 3.54 (dd, J = 9.0, 2.0 Hz, 1H), 3.42
(ddd, J = 9.0, 3.0, 3.0 Hz, 1H), 3.33 (s, 3H), 3.25 (s,
3H), 2.70 (dqd, J = 9.5, 7.0, 2.0 Hz, 1H), 2.59 (d,
J = 14.0 Hz, 1H), 2.27 (dd, J = 14.0, 11.0 Hz, 1H), 2.07
(br s, 3H), 1.80–1.75 (m, 2H), 1.62–1.54 (m, 1H), 1.77
(br s, 3H), 0.98 (d, J = 7.0 Hz, 3H), 0.91 (d,
1
1099, 731; H NMR (CDCl3, 500 MHz) d 9.19 (s, 1H),
7.24 (s, 1H), 6.94 (br d, J = 11.5 Hz, 1H), 6.56 (br dd,
J = 11.5, 11.0 Hz, 1H), 6.28 (br t, J = 5.0 Hz, 1H), 5.87
(br d, J = 9.5 Hz, 1H), 5.84 (dd, J = 11.0, 10.0 Hz,
1H), 5.88–5.77 (m, 2H), 5.17 (s, 1H), 5.15 (br s, 1H),
4.87 (br s, 2H), 4.50 (dd, J = 7.5, 5.0 Hz, 1H), 4.32–
4.29 (m, 2H), 4.23 (br s, 1H), 3.58–3.41 (m, 4H), 3.34
(s, 3H), 3.26 (s, 3H), 3.24–3.20 (m, 2H), 3.17–3.12 (m,
1H), 2.91 (dd, J = 13.0, 5.0 Hz, 1H), 2.75–2.69 (m,

4968
Y. Shen et al. / Bioorg. Med. Chem. 13 (2005) 4960–4971
J = 6.5 Hz, 3H); HRMS (FAB) [M+H]+ calcd for
C30H42N3O10, 604.2870; found 604.2867.
carbons aided by DEPT) d 183.8 (18-C), 183.2 (21-
C), 168.3 (1-C), 156.0 (7-O2CNH2), 152.7 (17-C),
138.8 (20-C), 136.1 (5-C), 134.9 (2-C), 133.3 (9-C),
133.1 (8-C), 127.0 (4-C), 126.4 (3-C), 125.4 (16-C),
111.6 (19-C), 81.6 (7-C), 81.1 (12-C), 81.1 (6-C), 72.7
(11-C), 57.2 (6- or 12-OCH3), 56.7 (6- or 12-OCH3),
35.1 (13-C), 33.6 (15-C), 32.3 (10-C), 29.2 (17-NCH2),
28.9 (14-C), 23.3 (14-CH3), 12.9 (8-CH3), 12.5 (2-
CH3), 12.4 (10-CH3); HRMS (FAB) [M+H]+ calcd
for C30H42N3O8, 572.2926; found 572.2968.
5.12. 17-(1-Azetidinyl)-17-demethoxygeldanamycin (14)52
Azetidine (4.0 ll, 0.059 mmol) was added to a solution
of (+)-geldanamycin (7.5 mg, 0.013 mmol) in dichlo-
romethane (1.5 ml) with stirring. Upon complete
conversion of geldanamycin shown by thin-layer chro-
matography (40 min), the mixture was washed with
brine, dried over anhydrous sodium sulfate, and concen-
trated. Separation by flash chromatography on silica gel
(1:2 hexane/ethyl acetate) gave the product as a deep
5.14. 3-Bromo-4-nitrophenol and 3-bromo-6-nitrophenol
purple solid (7.7 mg, 98%). IR (in CH2Cl2) (cmꢀ1
)
Fuming nitric acid (3.8 ml, 89 mmol) in glacial acetic
acid (12 ml) was added for over 35 min to a solution
of 3-bromophenol (15.2 g, 87.9 mmol) in glacial acetic
acid (60 ml) in a flask with a surrounding ice bath.
The reaction mixture was stirred at room temperature
for an additional 30 min, poured on ice, and then con-
centrated in vacuo. Medium pressure chromatography
on silica gel (1:2 ethyl acetate/hexanes) allowed separa-
tion of products 3-bromo-4-nitrophenol (3.47 g,
15.9 mmol, 18% yield); mp 130–131 ꢁC, following
recrystallization from ether/hexanes (reported mp 130–
131 ꢁC54 and 131 ꢁC55); 1H NMR (DMSO-d6,
500 MHz) d 7.99 (d, 1H, J = 9 Hz), 7.18 (d, 1H,
J = 3 Hz), 6.91 (dd, 1H, J = 9, 3 Hz); and 3-bromo-6-
nitrophenol (1.94 g, 8.90 mmol, 10% yield, following
recrystallization from ether/hexanes); mp 41.5–42.5 ꢁC
(reported mp 42–45 ꢁC56 and 42 ꢁC55); 1H NMR
(CDCl3, 500 MHz) d 10.60 (s, 1H), 7.95 (d, 1H,
J = 9 Hz), 7.35 (d, 1H, J = 2 Hz), 7.11 (dd, 1H, J = 9,
2 Hz); 13C NMR (CDCl3; assignments aided by HMQC)
d 122.9 (C-2), 123.8 (C-4), 126.0 (C-5), 132.2 (C-3), 132.7
(C-6), 155.2 (C-1); IR (KBr) (cmꢀ1) 3450 (broad), 1612,
1578, 1527, 1475, 1311, 1235, 1186, 900.
3422, 3075, 3049, 2986, 1733, 1686, 1651, 1605, 1540,
1486, 1420, 1375, 1283, 1260, 1103, 1047; 1H NMR
(CDCl3, 500 MHz) d 9.16 (s, 1H), 7.10 (s, 1H), 6.92
(br d, J = 11.5 Hz, 1H), 6.56 (br dd, J = 11.5, 11.0 Hz,
1H), 5.92 (br d, J = 9.5 Hz, 1H), 5.82 (dd, J = 11.0,
10.0 Hz, 1H), 5.15 (s, 1H), 4.79 (br s, 2H), 4.72–4.58
(m, 4H), 4.28 (br d, J = 10.0 Hz, 1H), 3.54 (br d,
J = 9.0 Hz, 1H), 3.43 (ddd, J = 9.0, 3.0, 3.0 Hz, 1H),
3.34 (s, 3H), 3.24 (s, 3H), 2.71 (dqd, J = 9.5, 7.0,
2.0 Hz, 1H), 2.59 (d, J = 14.0 Hz, 1H), 2.42 (quintet,
J = 8.0 Hz, 2H), 2.23 (dd, J = 14.0, 11.0 Hz, 1H), 2.00
(br s, 3H), 1.78 (br s, 3H), 1.77–1.73 (m, 2H), 1.69–
1.62 (m, 1H), 0.97 (d, J = 7.0 Hz, 3H), 0.94 (d,
J = 6.5 Hz, 3H); 13C NMR (CDCl3, 125 MHz, assign-
ment of protonated carbons aided by DEPT and
HMQC) d 185.8 (18-C), 178.4 (21-C), 168.4 (1-C),
156.0 (7-O2CNH2), 145.9 (17-C), 140.5 (20-C), 135.5
(5-C), 135.1 (2-C), 134.0 (9-C), 132.6 (8-C), 126.7 (4-
C), 126.6 (3-C), 109.6 (19-C), 109.2 (16-C), 81.8 (7-C),
81.6 (12-C), 81.3 (6-C), 72.5 (11-C), 58.9 (10- and 30-
C), 57.1 (6- or 12-OCH3), 56.7 (6- or 12-OCH3), 35.1
(13-C), 34.1 (15-C), 32.3 (10-C), 28.1 (14-C), 22.9 (14-
CH3), 18.4 (20-C), 12.7 (8-CH3), 12.6 (2-CH3), 12.2
(10-CH3); MS (FAB) [M+H]+ found 586.
5.15. 2-Amino-5-bromophenol
5.13. 17-(1-Aziridinyl)-17-demethoxygeldanamycin (15)
3-Bromo-6-nitrophenol (0.292 g, 1.34 mmol) was stirred
in a 0.5% aqueous sodium hydroxide solution (30 ml).
Sodium hydrosulfite (2.00 g of 85%, 9.76 mmol) was
added to the reaction flask and this solution was stirred
at room temperature for 15 min. The reaction flask was
then acidified with diluted hydrochloric acid until a pH
of 5 was obtained. The reaction mixture was then
extracted with diethyl ether (3· 40 ml), the combined
organic layers were dried over anhydrous sodium sulfate
and concentrated to provide crude 2-amino-5-bromo-
phenol, which was recrystallized from ethyl ether/hex-
anes to provide the pure product (0.151 g, 0.80 mmol,
60% yield); mp 125–127 ꢁC (dec.) (reported mp 149.5–
Aziridine53 (0.30 ml, 5.80 mmol) was added to a solu-
tion of (+)-geldanamycin (5.8 mg, 0.010 mmol) in
dichloromethane (2.0 ml). The mixture was stirred at
room temperature. Upon complete conversion of gel-
danamycin shown by thin-layer chromatography
(25 min), the mixture was washed with brine, dried
over anhydrous sodium sulfate, and concentrated. Sep-
aration by flash chromatography on silica gel (1:2 hex-
ane/ethyl acetate) gave the product as an orange solid
(5.6 mg, 95%). IR (KBr) (cmꢀ1) 3438, 3338, 3192,
2925, 2827, 1736, 1701, 1687, 1644, 1585, 1517, 1457,
1
1367, 1272, 1192, 1112; H NMR (CDCl3, 500 MHz)
1
150.5 ꢁC57); H NMR (CD3CN, 500 MHz) d 7.08 (br s,
d 8.77 (s, 1H) 7.27 (s, 1H), 6.91 (br d, J = 11.5 Hz,
1H), 6.55 (br dd, J = 11.5, 11.0 Hz, 1H), 5.86–5.80
(m, 2H), 5.17 (s, 1H), 4.80 (br s, 2H), 4.30 (br d,
J = 10.0 Hz, 1H), 3.52 (ddd, J = 9.0, 6.5, 2.0 Hz, 1H),
3.42–3.37 (m, 2H), 3.34 (s, 3H), 3.27 (s, 3H), 2.73
(dqd, J = 9.5, 7.0, 2.0 Hz, 1H), 2.57 (d, J = 14.0 Hz,
1H), 2.50 (dd, J = 14.0, 11.0 Hz, 1H), 2.44–2.33 (m,
4H), 2.00 (br s, 3H), 1.80–1.76 (m, 2H), 1.77 (br s,
3H), 1.75–1.69 (m, 1H), 0.99–0.96 (m, 6H); 13C
NMR (CDCl3, 125 MHz, assignment of protonated
1H), 6.82 (d, 1H, J = 2 Hz), 6.78 (dd, 1H, J = 8, 2 Hz),
6.56 (d, 1H, J = 8 Hz), 4.03 (br s, 2H); IR (KBr)
(cmꢀ1) 3496 (broad), 3377, 3298, 1598, 1502, 1431,
1269, 1210, 916, 877.
5.16. 50-Bromogeldanoxazinone (16)37,14
A mixture of (+)-geldanamycin (21.8 mg, 0.039 mmol)
and 2-amino-5-bromophenol (14.6 mg, 0.078 mmol) in

Y. Shen et al. / Bioorg. Med. Chem. 13 (2005) 4960–4971
4969
glacial acetic acid (2.0 ml) was stirred at 78 ꢁC under
5.18. 2-Amino-5-iodophenol
nitrogen for 19 h, then cooled, and concentrated. Sepa-
ration of the deep orange residue by flash chromato-
graphy on silica gel (hexane:ethyl acetate) gave a crude
product contaminated with unreacted (+)-geldanamy-
cin. This solution was then dissolved in chloroform
and subjected to preparative HPLC separation (Waters
Nova-Pak Silica 6 lm 7.8 · 300 mm column, 2.0 ml/
min, 2:3 CHCl3/EtOAc) to afford the product as a bright
orange powder (16.4 mg, 60% yield); mp 274–278 ꢁC
(dec.) (lit.37 mp 275–278 ꢁC). IR (KBr) (cmꢀ1) 3442,
3342, 3209, 2954, 2926, 2878, 1734, 1700, 1615, 1583,
1507, 1384, 1314, 1192, 1111, 1061 (lit.37 1605, 1585,
3-Iodo-6-nitrophenol (0.993 g, 3.75 mmol) was stirred in
an aqueous sodium hydroxide solution (0.06 M, 100 ml
water). Sodium hydrosulfite (4.62 g of 85%, 22.6 mmol)
was added to the reaction flask and this solution was
stirred at room temperature for 40 min. The reaction
flask was then cooled with a surrounding ice bath and
acetic acid was added until a pH of 5–6 was obtained.
The reaction mixture was then extracted with methylene
chloride (3· 200 ml), the combined organic layers were
dried over anhydrous magnesium sulfate and concen-
trated to provide crude 6-amino-3-iodophenol (0.533 g,
mp 99.5–100.5 ꢁC), which was recrystallized from ethyl
ether/hexanes to provide the pure product (0.463 g,
1.97 mmol, 53% yield); mp 126–128 ꢁC (dec.) (reported
mp 141 ꢁC59); 1H NMR (CD3CN, 500 MHz) d 7.04
(br s, 1H), 6.97 (d, 1H, J = 2 Hz), 6.95 (dd, 1H, J = 8,
2 Hz), 6.45 (d, 1H, J = 8 Hz), 4.05 (br s, 2H); IR
(KBr) (cmꢀ1) 3455 (broad), 3380, 3305, 1714, 1504,
1430, 1365, 1279, 1257, 1223, 890; Anal. Calcd for
C6H6INO: C, 30.66; H, 2.57; N, 5.96. Found: C,
30.65; H, 2.42; N, 5.92.
1
1505); H NMR (CDCl3, 500 MHz) d 9.13 (br s, 1H),
8.33 (s, 1H), 7.73 (d, J = 8.5 Hz, 1H), 7.60 (d,
J = 2.0 Hz, 1H), 7.53 (dd, J = 8.5, 2.0 Hz, 1H), 7.03
(br d, J = 11.5 Hz, 1H), 6.60 (br dd, J = 11.5, 11.0 Hz,
1H), 5.96 (br d, J = 9.5 Hz, 1H), 5.86 (dd, J = 11.0,
10.0 Hz, 1H), 5.21 (s, 1H), 4.72 (br s, 2H), 4.35 (br d,
J = 10.0 Hz, 1H), 4.25 (br s, 1H), 3.64 (br dd, J = 9.0,
6.5 Hz, 1H), 3.46 (ddd, J = 9.0, 3.0, 3.0 Hz, 1H), 3.37
(s, 3H), 3.27 (s, 3H), 2.82–2.71 (m, 3H), 2.08 (br s,
3H), 1.98–1.86 (m, 2H), 1.85–1.77 (m, 1H), 1.81 (d,
J = 1.0 Hz, 3H), 1.01 (d, J = 7.0 Hz, 3H), 0.99 (d,
J = 6.5 Hz, 3H); 13C NMR (CDCl3, 125 MHz) d 180.7,
168.4, 156.0, 148.5, 145.0, 143.5, 136.8, 135.5, 135.3,
133.9, 133.0, 132.9, 130.9, 129.1, 126.7, 125.2, 119.3,
117.5, 112.9, 81.9, 81.3, 81.2, 72.2, 57.1, 56.8, 35.3,
33.1, 32.2, 29.4, 27.6, 23.3, 12.8, 12.7, 12.1; HRMS
(FAB) [M+H]+ calcd for C34H41BrN3O8, 698.2077;
found 698.2080.
5.19. 50-Iodogeldanoxazinone (17)
A mixture of (+)-geldanamycin (4.8 mg, 8.6 lmol) and
2-amino-5-iodophenol (4.0 mg, 0.017 mmol) in glacial
acetic acid (1.0 ml) was stirred at 78 ꢁC under nitrogen
for 20 h, then cooled, and concentrated. Separation of
the deep orange residue by flash chromatography on sil-
ica gel (hexane/ethyl acetate) gave a crude product con-
taminated with unreacted (+)-geldanamycin. This
mixture was then dissolved in chloroform and subjected
to preparative HPLC separation (Waters Nova-Pak Sil-
ica 6 lm 7.8 · 300 mm column, 2.0 ml/min, 2:3 CHCl3/
EtOAc) to afford the product as a bright orange powder
(2.8 mg, 44%). IR (in CH2Cl2) (cmꢀ1) 3139, 3076, 3048,
2995, 2967, 1733, 1684, 1599, 1580, 1496, 1447, 1423,
5.17. 3-Iodo-4-nitrophenol and 3-iodo-6-nitrophenol
Fuming nitric acid (3.0 ml, 75 mmol) in glacial acetic
acid (12 ml) was added for over 25 min to a solution
of 3-iodophenol (15.03 g, 68.3 mmol) in glacial acetic
acid (60 ml) in a flask with a surrounding ice bath.
The reaction mixture was stirred at room temperature
for an additional 30 min, poured on ice and then con-
centrated in vacuo, taken up with water (150 ml), and
extracted with methylene chloride (2· 300 ml), and the
combined methylene chloride layers were dried over
anhydrous magnesium sulfate and evaporated to give
17 g organic residue. Medium pressure chromatography
on silica gel (1:2 ethyl acetate/hexanes) allowed the sep-
aration of products 3-iodo-4-nitrophenol (6.93 g,
26.1 mmol, 38% yield); mp 121–123 ꢁC; 1H NMR
(CDCl3, 300 MHz) d 7.98 (d, 1H, J = 9 Hz), 7.54 (d,
1H, J = 3 Hz), 6.92 (dd, 1H, J = 9, 3 Hz), 5.54 (br s,
1H); IR (KBr) (cmꢀ1) 3150 (broad), 1600, 1580, 1512,
1404, 1336, 1298, 1212, 1121, 1023, 870; and 3-iodo-6-ni-
trophenol (3.07 g, 11.6 mmol, 17% yield); mp 92–94 ꢁC,
following recrystallization from methylene chloride/hex-
anes (reported mp 96 ꢁC58); 1H NMR (CDCl3,
300 MHz) d 10.53 (s, 1H), 7.76 (d, 1H, J = 9.0 Hz),
7.59 (d, 1H, J = 2.0 Hz), 7.33 (dd, 1H, J = 9.0, 2.0 Hz);
13C NMR (CDCl3; assignments aided by HMQC) d
105.2 (C-3), 125.6 (C-5), 129.2 (C-2), 129.7 (C-4),
133.4 (C-6), 154.6 (C-1); IR (KBr) (cmꢀ1) 3430 (broad),
1604, 1571, 1518, 1463, 1317, 1225, 1172, 1055, 888;
Anal. Calcd for C6H4INO3: C, 27.19; H, 1.52; N, 5.29.
Found: C, 27.36; H, 1.57; N, 5.15.
1
1260, 1098; H NMR (CDCl3, 500 MHz) d 9.12 (br s,
1H), 8.30 (s, 1H), 7.79 (d, J = 2.0 Hz, 1H), 7.71 (dd,
J = 8.5, 2.0 Hz, 1H), 7.55 (d, J = 8.5 Hz, 1H), 7.01 (br
d, J = 11.5 Hz, 1H), 6.59 (br dd, J = 11.5, 11.0 Hz,
1H), 5.94 (br d, J = 9.5 Hz, 1H), 5.84 (dd, J = 11.0,
10.0 Hz, 1H), 5.20 (s, 1H), 4.71 (br s, 2H), 4.33 (br d,
J = 10.0 Hz, 1H), 4.24 (br s, 1H), 3.63 (ddd, J = 9.0,
6.5, 2.0 Hz, 1H), 3.45 (ddd, J = 9.0, 3.0, 3.0 Hz, 1H),
3.36 (s, 3H), 3.26 (s, 3H), 2.79–2.70 (m, 3H), 2.06 (br
s, 3H), 1.97–1.84 (m, 2H), 1.82–1.74 (m, 1H), 1.80 (d,
J = 1.0 Hz, 3H), 0.99 (d, J = 7.0 Hz, 3H), 0.97 (d,
J = 6.5 Hz, 3H); 13C NMR (CDCl3, 125 MHz) d 180.8,
168.4, 156.0, 148.7, 144.9, 143.3, 136.9, 135.6, 135.3,
135.0, 133.9, 133.6, 133.0, 131.0, 126.8, 126.6, 125.2,
117.5, 112.9, 96.6, 81.9, 81.4, 81.3, 72.2, 57.1, 56.8,
35.4, 33.0, 32.3, 27.7, 23.3, 12.8, 12.6, 12.2; HRMS
(FAB) [M+H]+ calcd for C34H41IN3O8, 746.1938; found
746.1937.
5.20. 11-O-Acetyl-17-(1-azetidinyl)-17-demethoxygel-
danamycin (18)35
17-(1-Azetidinyl)-17-demethoxygeldanamycin (3.2 mg,
5.5 lmol) was stirred with acetic anhydride (5.2 ll,

4970
Y. Shen et al. / Bioorg. Med. Chem. 13 (2005) 4960–4971
0.055 mmol) and DMAP (7.3 mg, 0.060 mmol). Upon
complete conversion of starting material shown by
thin-layer chromatography (40 h), the mixture was
washed with brine, dried over anhydrous sodium sulfate,
and concentrated. Separation by flash chromatography
on silica gel (1:2 hexane/ethyl acetate) gave the product
was added. The mixture was stirred at room tempera-
ture. Upon complete conversion of geldanamycin shown
by thin-layer chromatography (1 h), the organic layer
was separated, washed with brine, dried over anhydrous
sodium sulfate, and concentrated. Separation of the solid
residue by flash chromatography on silica gel (3:1 hex-
ane/ethyl acetate) afforded a pale yellow solid (3.3 mg,
as a purple solid (3.2 mg, 93%). IR (in CH2Cl2) (cmꢀ1
)
1
3686, 3536, 3420, 3069, 3052, 2930, 1734, 1689, 1649,
1601, 1585, 1549, 1486, 1435, 1374, 1273, 1102; 1H
NMR (CDCl3, 500 MHz) d 9.37 (s, 1H), 7.13 (br s,
1H), 6.94 (s, 1H), 6.50 (ddd, J = 11.5, 11.0, 1.0 Hz,
1H), 5.81 (dd, J = 11.0, 7.5 Hz, 1H), 5.45 (br s, 1H),
5.28 (br d, J = 10.0 Hz, 1H), 5.04 (dd, J = 8.0, 3.5 Hz,
1H), 4.64–4.54 (m, 4H), 4.48 (br d, J = 7.5 Hz, 1H),
3.63 (br s, 1H), 3.33 (s, 3H), 3.31 (s, 3H), 2.85–2.77
(m, 1H), 2.71 (br d, J = 10.0 Hz, 1H), 2.38 (quintet,
J = 8.0 Hz, 2H), 2.06–2.00 (m, 1H), 1.98 (br s, 3H),
1.97 (s, 3H), 1.71–1.56 (m, 2H), 1.68 (br s, 3H), 1.28–
1.18 (m, 1H), 0.96–0.93 (m, 6H); 13C NMR (CDCl3,
125 MHz) d 186.2, 178.0, 170.6, 169.2, 155.7, 145.6,
140.4, 135.6, 134.8, 132.9, 128.3, 126.2, 109.2, 108.6,
80.0, 79.2, 78.4, 75.1, 58.5, 57.6, 56.1, 35.8, 33.0, 30.1,
29.7, 21.6, 20.9, 18.5, 15.6, 14.1, 12.3; HRMS (FAB)
[M+H]+ calcd for C33H46IN3O9, 628.3234; found
628.3237.
94%). H NMR (CDCl3, 500 MHz) d 8.34 (s, 1H), 8.08
(s, 1H), 8.02 (br s, 1H), 6.76 (br d, J = 11.5 Hz, 1H),
6.37 (br dd, J = 11.5, 11.0 Hz, 1H), 5.94 (br d,
J = 9.5 Hz, 1H), 5.64 (dd, J = 11.0, 10.0 Hz, 1H), 5.04
(br s, 1H), 4.95 (s, 1H), 4.65 (br s, 2H), 4.29 (br d,
J = 10.0 Hz, 1H), 3.81 (s, 3H), 3.61 (br d, J = 9.0 Hz,
1H), 3.43 (br d, J = 9.0 Hz, 1H), 3.33 (s, 3H), 3.21 (s,
3H), 2.79–2.74 (m, 2H), 2.35 (br d, J = 14.0 Hz, 1H),
1.82–1.65 (m, 3H), 1.76 (br s, 6H), 0.92 (d, J = 6.5 Hz,
3H), 0.86 (d, J = 7.0 Hz, 3H); HRMS (FAB) [M]+ calcd
for C29H42N2O9, 562.2890; found 562.2886.
5.23. HGF/SF-Met-uPA-plasmin cell-based assay
Cells were seeded in 96-well plates at 1500 cells/well
(with the exception of SK-LMS-1 cells, which were seed-
ed at 5000 cells/well) to detect color intensity, either with
MTS (Promega) for cell growth determination or via
Chromozyme PL (Boehringer–Mannheim) for uPA-
plasmin activity measurement. Cells were grown over-
night in DMEM/10% FBS, as previously described.14
Drugs were dissolved in DMSO, serially diluted from
stock concentrations into DMEM/10% FBS medium,
and added to appropriate wells. Immediately after the
addition of drugs or reagents, HGF/SF (60 ng/ml) was
added to all wells (with the exception of wells used as
controls to calculate basal growth and uPA-plasmin
activity levels). Twenty-four hours after the addition of
drugs and HGF/SF, plates were processed for the deter-
mination of uPA-plasmin activity as follows: Wells were
washed twice with DMEM (without phenol red; Life
Technologies, Inc.), and 200 ll of reaction buffer [50%
(v/v) 0.05 U/ml plasminogen in DMEM (without phenol
red), 40% (v/v) 50 mM Tris buffer (pH 8.2), and 10% (v/
v) of 3 mM Chromozyme PL (Boehringer–Mannheim)
in 100 mM glycine solution] was added to each well.
The plates were then incubated at 37 ꢁC, 5% CO2 for
4 h, at which time the absorbances generated were read
on an automated spectrophotometric plate reader at a
single wavelength of 405 nm. uPA-plasmin inhibition
index or IC50 is the negative log10 of the concentration
at which uPA-plasmin activity is inhibited by 50%.14
5.21. 17-(1-Azetidinyl)-7-decarbamyl-17-demethoxygel-
danamycin (19)52,35
Potassium tert-butoxide (5.3 mg, 0.045 mmol) was added
to a solution of 17-(1-azetidinyl)-17-demethoxygeldana-
mycin (5.0 mg, 8.5 lmol) in tert-butanol (4.0 ml) under
nitrogen atmosphere. The reaction mixture was stirred
at room temperature for 1 h and then quenched by parti-
tioning between ethyl acetate and brine. The organic layer
was washed with brine, dried over anhydrous sodium sul-
fate, and concentrated. Separation by flash chromato-
graphy on silica gel (hexane/ethyl acetate) gave the prod-
uct as a purple solid (4.4 mg, 95 %). IR (KBr) (cmꢀ1
)
3461, 3330, 2955, 2927, 2871, 2826, 1685, 1652, 1539,
1489, 1404, 1381, 1287, 1255, 1191, 1136, 1106;
1H NMR (CDCl3, 500 MHz) d 9.16 (s, 1H), 7.09 (s,
1H), 6.90 (br d, J = 11.5 Hz, 1H), 6.54 (br dd, J = 11.5,
11.0 Hz, 1H), 5.98 (dd, J = 11.0, 10.0 Hz, 1H), 5.70 (br
d, J = 9.5 Hz, 1H), 4.72–4.59 (m, 4H), 4.16 (br d,
J = 10.0 Hz, 1H), 3.98 (s, 1H), 3.52 (dd, J = 9.0, 2.0 Hz,
1H), 3.41 (ddd, J = 9.0, 3.0, 3.0 Hz, 1H), 3.34 (s, 3H),
3.23 (s, 3H), 2.73 (dqd, J = 9.5, 7.0, 2.0 Hz, 1H), 2.57
(d, J = 14.0 Hz, 1H), 2.42 (quintet, J = 8.0 Hz, 2H),
2.23 (dd, J = 14.0, 11.0 Hz, 1H), 2.01 (d, J = 1.0 Hz,
3H), 1.77–1.71 (m, 2H), 1.74 (d, J = 1.0 Hz, 3H), 1.70–
1.62 (m, 1H), 0.97 (d, J = 7.0 Hz, 3H), 0.94 (d,
J = 6.5 Hz, 3H); 13C NMR (CDCl3, 125 MHz) d 185.8,
178.4, 168.6, 145.8, 140.5, 137.2, 136.1, 134.8, 132.0,
126.9, 125.9, 109.5, 109.2, 81.8, 80.5, 80.3, 72.9, 58.9,
56.7, 56.3, 34.9, 34.2, 32.2, 28.2, 22.9, 18.4, 12.6, 12.4,
11.8; MS (FAB) [M+H]+ found 543.
Acknowledgments
This work was supported by a grant from the Michigan
Life Sciences Corridor (G.V.W. principle investigator)
and a Robert J. Lynch/American Brain Tumor Associa-
tion grant (to Q.X.).
5.22. 17,21-Dihydrogeldanamycin (20)36
References and notes
(+)-Geldanamycin (3.5 mg, 6.2 lmol) was dissolved in
ethyl acetate (2.5 ml) and then aqueous solution
(2.5 ml) of sodium dithionite (ꢁ85%, 0.50 g, 2.4 mmol)
1. Birchmeier, C.; Birchmeier, W.; Gherardi, E.; Vande
Woude, G. F. Nat. Rev. Mol. Cell Biol. 2003, 4, 915–925.

Y. Shen et al. / Bioorg. Med. Chem. 13 (2005) 4960–4971
4971
2. Rong, S.; Segal, S.; Anver, M.; Resau, J. H.; Woude, G. F.
V. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 4731–4735.
3. Takayama, H.; LaRochelle, W. J.; Sharp, R.; Otsuka,
T.; Kriebel, P.; Anver, M.; Aaronson, S. A.; Merlino,
G. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 701–706.
4. Danilkovitch-Miagkova, A.; Zbar, B. J. Clin. Invest. 2002,
109, 863–867.
5. Abounader, R.; Ranganathan, S.; Lal, B.; Fielding, K.;
Book, A.; Dietz, H.; Burger, P.; Laterra, J. J. Natl. Cancer
Inst. 1999, 91, 1548–1556.
6. Jiang, W. G.; Grimshaw, D.; Lane, J.; Martin, T. A.;
Abounder, R.; Laterra, J.; Mansel, R. E. Clin. Cancer Res.
2001, 7, 2555–2562.
7. Abounader, R.; Lal, B.; Luddy, C.; Koe, G.; Davidson,
B.; Rosen, E. M.; Laterra, J. FASEB J. 2002, 16, 108–110.
8. Stabile, L. P.; Lyker, J. S.; Huang, L.; Siegfried, J. M.
Gene Ther. 2004, 11, 325–335.
31. Maloney, A.; Workman, P. Expert Opin. Biol. Ther. 2002,
2, 3–24.
32. Neckers, L.; Ivy, S. P. Curr. Opin. Oncol. 2003, 15, 419–
424.
33. Sausville, E. A.; Tomaszewski, J. E.; Ivy, P. Curr. Cancer
Drug Targets 2003, 3, 377–383.
34. Ivy, P. S.; Schoenfeldt, M. Oncology (Huntingt) 2004, 18,
610, 615, 619, 620.
35. Schnur, R. C.; Corman, M. L.; Gallaschun, R. J.; Cooper,
B. A.; Dee, M. F.; Doty, J. L.; Muzzi, M. L.; DiOrio, C. I.;
Barbacci, E. G.; Miller, P. E.; Pollack, V. A.; Savage, D.
M.; Sloan, D. E.; Pustilnik, L. R.; Moyer, J. D.; Moyer,
M. P. J. Med. Chem. 1995, 38, 3813–3820.
36. Schnur, R. C.; Corman, M. L.; Gallaschun, R. J.; Cooper,
B. A.; Dee, M. F.; Doty, J. L.; Muzzi, M. L.; Moyer, J. D.;
DiOrio, C. I.; Barbacci, E. G.; Miller, P. E.; OÕBrien, A.
T.; Morin, M. J.; Foster, B. A.; Pollack, V. A.; Savage, D.
M.; Sloan, D. E.; Pustilnik, L. R.; Moyer, M. P. J. Med.
Chem. 1995, 38, 3806–3812.
9. Date, K.; Matsumoto, K.; Shimura, H.; Tanaka, M.;
Nakamura, T. FEBS Lett. 1997, 420, 1–6.
10. Matsumoto, K.; Nakamura, T. Cancer Sci. 2003, 94, 321–
327.
37. Rinehart, K. L., Jr.; McMillan, M. W.; Witty, T. R.;
Tipton, C. D.; Shield, L. S. Bioorg. Chem. 1977, 6, 353–
369.
38. Workman, P. Curr. Cancer Drug Targets 2003, 3, 297–300.
39. Banerji, U.; Judson, I.; Workman, P. Curr. Cancer Drug
Targets 2003, 3, 385–390.
40. Maloney, A.; Clarke, P. A.; Workman, P. Curr. Cancer
Drug Targets 2003, 3, 331–341.
41. Xu, W.; Mimnaugh, E.; Rosser, M.; Nicchitta, C.; Marcu,
M.; Yarden, Y.; Neckers, L. J. Biol. Chem. 2001, 276,
3702–3708.
42. Carreras, C. W.; Schirmer, A.; Zhong, Z.; Santi, D. V.
Anal. Biochem. 2003, 317, 40–46.
43. Kamal, A.; Thao, L.; Sensintaffar, J.; Zhang, L.; Boehm,
M. F.; Fritz, L. C.; Burrows, F. J. Nature 2003, 425, 407–
410.
11. Cao, B.; Su, Y.; Oskarsson, M.; Zhao, P.; Kort, E. J.;
Fisher, R. J.; Wang, L. M.; Vande Woude, G. F. Proc.
Natl. Acad. Sci. U.S.A. 2001, 98, 7443–7448.
12. Morotti, A.; Mila, S.; Accornero, P.; Tagliabue, E.;
Ponzetto, C. Oncogene 2002, 21, 4885–4893.
13. Christensen, J. G.; Schreck, R.; Burrows, J.; Kuruganti,
P.; Chan, E.; Le, P.; Chen, J.; Wang, X.; Ruslim, L.;
Blake, R.; Lipson, K. E.; Ramphal, J.; Do, S.; Cui, J. J.;
Cherrington, J. M.; Mendel, D. B. Cancer Res. 2003, 63,
7345–7355.
14. Webb, C. P.; Hose, C. D.; Koochekpour, S.; Jeffers, M.;
Oskarsson, M.; Sausville, E.; Monks, A.; Vande Woude,
G. F. Cancer Res. 2000, 60, 342–349.
15. Xie, Q.; Gao, C.-F.; Shinomiya, N.; Sausville, E.; Hay, R.;
Gustafson, M.; Shen, Y.; Wenkert, D.; Vande Woude, G.
F. Oncogene 2005, 24, 3697–3707.
44. Nimmanapalli, R.; OÕBryan, E.; Bhalla, K. Cancer Res.
2001, 61, 1799–1804.
16. Stebbins, C. E.; Russo, A. A.; Schneider, C.; Rosen, N.;
Hartl, F. U.; Pavletich, N. P. Cell 1997, 89, 239–250.
17. Grenert, J. P.; Sullivan, W. P.; Fadden, P.; Haystead, T.
A.; Clark, J.; Mimnaugh, E.; Krutzsch, H.; Ochel, H. J.;
Schulte, T. W.; Sausville, E.; Neckers, L. M.; Toft, D. O.
J. Biol. Chem. 1997, 272, 23843–23850.
18. Schulte, T. W.; Neckers, L. M. Cancer Chemother.
Pharmacol. 1998, 42, 273–279.
19. Roe, S. M.; Prodromou, C.; OÕBrien, R.; Ladbury, J. E.;
Piper, P. W.; Pearl, L. H. J. Med. Chem. 1999, 42, 260–
266.
20. Jez, J. M.; Chen, J. C.; Rastelli, G.; Stroud, R. M.; Santi,
D. V. Chem. Biol. 2003, 10, 361–368.
21. Dutta, R.; Inouye, M. Trends Biochem. Sci. 2000, 25, 24–
28.
22. Fink, A. L. Physiol. Rev. 1999, 79, 425–449.
23. Richter, K.; Buchner, J. J. Cell. Physiol. 2001, 188, 281–
290.
45. Maulik, G.; Kijima, T.; Ma, P. C.; Ghosh, S. K.; Lin, J.;
Shapiro, G. I.; Schaefer, E.; Tibaldi, E.; Johnson, B. E.;
Salgia, R. Clin. Cancer Res. 2002, 8, 620–627.
46. Dikalov, S.; Landmesser, U.; Harrison, D. J. Biol. Chem.
2002, 277, 25480–25485.
47. Muroi, M.; Izawa, M.; Kosai, Y.; Asai, M. J. Antibiot.
1980, 33, 205–212.
48. Egorin, M. J.; Lagattuta, T. F.; Hamburger, D. R.; Covey,
J. M.; White, K. D.; Musser, S. M.; Eiseman, J. L. Cancer
Chemother. Pharmacol. 2002, 49, 7–19.
49. Li, L. H.; Clark, T. D.; Cowie, C. H.; Rinehart, K. L., Jr.
Cancer Treat. Rep. 1977, 61, 815–824.
50. Sasaki, K.; Yasuda, H.; Onodera, K. J. Antibiot. 1979, 32,
849–851.
51. Lee, Y.; Park, M.; Folk, J. J. Med. Chem. 1995, 38, 3053–
3061.
52. Schnur, R. C.; Corman, M. L. J. Org. Chem. 1994, 59,
2581–2584.
24. Picard, D. Cell. Mol. Life Sci. 2002, 59, 1640–1648.
25. Pratt, W. B.; Toft, D. O. Exp. Biol. Med. 2003, 228, 111–
133.
26. Whitesell, L.; Mimnaugh, E. G.; De Costa, B.; Myers,
C. E.; Neckers, L. M. Proc. Natl. Acad. Sci. U.S.A.
1994, 91, 8324–8328.
53. Allen, C. F. H.; Spangler, F. W.; Webster, E. R. Org.
Synth. Coll. 1963, 4, 433.
54. Wright, C.; Shulkind, M.; Jones, K.; Thompson, M.
Tetrahedron Lett. 1987, 28, 6389–6390.
55. Hodgson, H. H.; Moore, F. H. J. Chem. Soc. 1926, 155–
161.
27. Blagosklonny, M. V.; Toretsky, J.; Bohen, S.; Neckers, L.
Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 8379–8383.
28. Bohen, S. P. Mol. Cell. Biol. 1998, 18, 3330–3339.
29. Sharma, S. V.; Agatsuma, T.; Nakano, H. Oncogene 1998,
16, 2639–2645.
30. Schulte, T. W.; Akinaga, S.; Soga, S.; Sullivan, W.;
Stensgard, B.; Toft, D.; Neckers, L. M. Cell Stress
Chaperones 1998, 3, 100–108.
56. Hanzlik, R. P.; Weller, P. E.; Desai, J.; Zheng, J.; Hall,
L. R.; Slaughter, D. E. J. Org. Chem. 1990, 55, 2736–
2742.
57. Boyland, E.; Sims, P. J. Chem. Soc. 1954, 980–985.
58. Hodgson, H. H.; Moore, F. H. J. Chem. Soc. 1927, 630–
635.
59. Hodgson, H. H.; Kershaw, A. J. Chem. Soc. 1928, 2703–
2705.