Discovery of novel 17-phenylethylaminegeldanamycin derivatives as potent hsp90 inhibitors

Article abstract of DOI:10.1111/cbdd.12371

Twenty-six 17-phenylethylamine-modified geldanamycin derivatives were synthesized and evaluated for antiproliferation activity in human cancer cell lines, LNCaP and MDA-MB-231. Five derivatives (2j, 2q, 2v, 2x, and 2y) showed excellent in vitro antitumor activities. Among them, compound 2y was the most potent lead, with IC₅₀ values of 0.27 ± 0.11 and 0.86 ± 0.23 μm for LNCaP and MDA-MB-231, respectively. In particular, compound 2y was more active than its precursor geldanamycin against LNCap cells. Liver injury test in mice demonstrated that 2y group showed no significant difference for serum alanine aminotransferase (ALT) activity versus vehicle control, indicating that 2y was a promising antitumor candidate. Preliminary structure-activity relationship (SAR) and molecular dynamics (MD) simulations of this new series of geldanamycin derivatives were also investigated, suggesting a theoretical model of 17-phenylethylaminegeldanamycins binding to Hsp90. Compound 2y was revealed to be a promising antitumor agent through the synthesis, structure-activity relationship analysis and hepatotoxicity evaluation of 26 novel 17-phenylethylaminegeldanamycin derivatives, which showed the most potent antitumor activity against LNCap cells and the least hepatotoxicity. The binding model of compound 2y to Hsp90 was simulated by Amber12 and shown by PyMoL.

Full text of DOI:10.1111/cbdd.12371

Received Date : 18-Mar-2014
Revised Date : 18-May-2014
Accepted Date : 22-May-2014
Article type : Research Article
Discovery of novel 17-phenylethylaminegeldanamycin
derivatives as potent Hsp90 inhibitors
Zhenyu Li,a Lejiao Jia,b Jifeng Wang,c,d Xingkang Wu,a Guowei Shi,c,d Chunhua Lu,a and
Yuemao Shena,∗
aKey Laboratory of Chemical Biology (Ministry of Education), School of Pharmaceutical
Sciences, Shandong University, No. 44 West Wenhua Road, Jinan 250012, Shandong, P. R.
China
bDepartment of Pharmacy, Shandong University Qilu Hospital, No. 107 West Wenhua Road,
Jinan 250012, Shandong, P. R. China
cDepartment of Urology, the Fifth People's Hospital of Shanghai, Fudan University, No. 801
Heqing Road, Shanghai 200240, P. R. China
dUrology Research Center, Fudan University, No. 801 Heqing Road, Shanghai 200240, P. R.
China
∗Corresponding author. Tel./Fax: +86 531 88382108; e-mail: yshen@sdu.edu.cn.
This article has been accepted for publication and undergone full peer review but has
not been through the copyediting, typesetting, pagination and proofreading process
which may lead to differences between this version and the Version of Record. Please
cite this article as an 'Accepted Article', doi: 10.1111/cbdd.12371
This article is protected by copyright. All rights reserved.

Twenty-six 17-phenylethylamine-modified geldanamycin derivatives were
synthesized and evaluated for anti-proliferation activity in human cancer cell
lines, LNCaP and MDA-MB-231. Five derivatives (2j, 2q, 2v, 2x and 2y) showed
excellent in vitro antitumor activities. Among them, compound 2y was the most
potent lead, with IC50 values of 0.27 0.11 and 0.86 0.23 µM for LNCaP and
MDA-MB-231, respectively. In particular, compound 2y was more active than its
precursor geldanamycin against LNCap cells. Liver injury test in mice
demonstrated that 2y group showed no significant difference for serum alanine
aminotransferase (ALT) activity versus vehicle control, indicating that 2y was a
promising antitumor candidate. Preliminary structure–activity relationship
(SAR) and molecular dynamics (MD) simulations of this new series of
geldanamycin derivatives were also investigated, suggesting a theoretical model of
17-phenylethylaminegeldanamycins binding to Hsp90.
Keywords: Geldanamycin, synthesis, Hsp90, antitumor activities, hepatotoxicity
Geldanamycin (GA), a benzoquinone ansamycin (BQA) polyketide, was isolated
from Streptomyces hygroscopicus var. geldanus in 1970 (1). GA was initially thought
to be a nonspecific kinase inhibitor, but then found to target the heat shock protein 90
(Hsp90) and to be the first natural product inhibitor of Hsp90 (1-2). GA binds to the
N-terminal domain ATP binding site of Hsp90, inhibiting the chaperone activity of
the protein. This inhibition leads to the disruption of the Hsp90-client protein complex
(3-4). The unchaperoned client proteins are subsequently degraded by the proteasome,
resulting in inhibition of cell proliferation and induction of apoptosis which endows
the inhibitors of Hsp90 to be promising potent antitumor agents (5).
GA was a potent antitumor agent (6-7), but showed significant hepatotoxicity in
animals (8), which was possibly a result of reaction with biological nucleophiles at the
This article is protected by copyright. All rights reserved.

reactive C-17 methoxy group (9) or 19-position of the quinone ring (10). To date,
great efforts have been made to modify the C-17 position of GA, generating a number
of 17-fatty amine substituted GA analogs (9, 11-14). Indeed, introducing substituents
at C-17 position of GA yielded some less toxic and more soluble derivatives such as
17-allylamino-17-demethoxygeldanamycin (tanespimycin, 17-AAG) and
17-[2-(dimethylamino)ethyl]amino-17-demethoxygeldanamycin
(alvespimycin,17-DMAG) (Figure 1) (15-16). Although 17-AAG showed some
promise in Phase II trials for the treatment of breast cancer (17), the clinic trial has
been terminated, indicating that the GA derivatives with less hepatotoxicity are
urgently in demand.
In this study, we incorporated phenylethylamine scaffold into the C-17 position of GA
through application of the structure-based bioisosterism approach (18-20) in order to
not only increase the binding of 17-phenylethylaminegeldanamycins to Hsp90, but
also decrease the hepatotoxicity (Figure 2). The effects of the different substituents
(halogens, methoxy group, methyl group, nitro group and trifluoromethyl group) on
the phenyl group were also investigated. Especially, introduction of fluorine enhanced
the binding interactions, metabolic stability and physicochemical properties (21). The
antitumor activities of twenty-six geldanamycin derivatives against LNCaP and
MDA-MB-231 cell lines as well as their in vivo hepatotoxicity were evaluated.
Experimental section
Chemistry
All melting points were determined on a micromelting point apparatus and were
uncorrected. 1H-NMR and 13C-NMR spectra were recorded on a Bruker Avance (600
MHz) spectrometer. Chemical shifts (δ) are in parts per million (ppm) downfield from
TMS (δ); multiplicity; observed coupling constant (J) in hertz (Hz); proton count;
This article is protected by copyright. All rights reserved.

assignment. Multiplicities are recorded as singlet (s), doublet (d), doublet of doublet
(dd), doublet of triplet (dt), triplet (t), quarter (q), multiplet (m) and broad singlet (bs)
where appropriate. Mass spectra were obtained on an electrospray ionization mass
spectrometer as the value m/z. Thin-layer chromatography (TLC) was self-made silica
gel (GF254) sheets. Flash chromatography was performed using 200-300 mesh silica
gel.
General procedure for preparation of compounds 2a–z
To geldanamycin (34 mg, 0.06 mmol, 1.0 equiv.) in dichloromethane (8 mL) was
added substitutional phenylethylamine (0.60 mmol, 10.0 equiv.). The reaction mixture
was stirred overnight at room temperature, and the color changed from yellow to
purple. TLC was used to monitor the progress of the reaction. After the reaction
complete, the mixture was diluted with 20 mL of dichloromethane and washed with
1.5 N HCl (2 × 20 mL) followed by brine (3 × 30 mL). The organic layer was dried
over anhydrous sodium sulfate, filtered, and evaporated to dryness. The residue was
purified by flash column chromatography (silica gel, ethyl acetate/petroleum ether =
1:1) to afford a purple solid.
17-(3-trifluoromethylphenylamino)-17-demethoxygeldanamycin (2y). Purple solid
(30.2 mg, 84.1%); mp 120-122 ºC; 1H NMR (600 MHz, CDCl3) δ 9.13 (s, 1H,
1-CONH), 7.55 (d, J = 7.5 Hz, 1H, 6’-H), 7.49 – 7.46 (m, 2H, 8’-H and 5’-H), 7.41
(d, J = 7.4 Hz, 1H, 4’-H), 7.27 (s, 1H, 19-H), 6.95 (d, J = 11.3 Hz, 1H, 3-H), 6.60 (t, J
= 11.3 Hz, 3-NH), 6.27 (s, 1H, 17-NH), 5.90 (d, J = 9.2 Hz, 1H, 5-H), 5.86 (d, J =
10.6 Hz, 1H, 9-H), 5.19 (s, 1H, 7-H), 4.87 (bs, 2H, 7-CONH2), 4.31 (d, J = 9.8 Hz,
1H, 6-H), 4.21 (bs, 1H, 11-OH), 3.86 – 3.82 (m, 1H, 1’-H), 3.78 – 3.74 (m, 1H, 1’-H),
3.57 – 3.54 (m, 1H, 11-H), 3.45 – 3.43 (m, 1H, 12-H), 3.35 (s, 3H, 12-OMe), 3.26 (s,
3H, 6-OMe), 3.04 – 3.00 (m, 2H, 2’-H), 2.75 – 2.72 (m, 1H, 10-H), 2.70 (d, J = 14.2
This article is protected by copyright. All rights reserved.

Hz, 1H, 15-H), 2.40 (t, J = 11.9 Hz, 1H, 15-H), 2.02 (s, 3H, 2-Me), 1.79 (s, 3H,
8-Me), 1.78 – 1.76 (m, 2H, 13-H), 1.70 – 1.68 (m, 1H, 14-H), 1.00 (d, J = 6.5 Hz, 3H,
10-Me), 0.94 (d, J = 6.0 Hz, 3H, 14-Me); 13C NMR (150 MHz, CDCl3) δ 183.7,
180.9, 168.4, 156.1, 144.5, 141.2, 138.2 (3’-C), 135.9, 134.9, 133.7, 132.8, 132.1
(8’-C), 131.3 (d, J = 32.2 Hz, 5’-C), 129.5 (7’-C), 127.0, 126.6, 125.4 (d, J = 3.5 Hz,
4’-C), 124.1 (d, J = 3.8 Hz, 6’-C), 123.9 (d, J = 272.5 Hz, 5’-CF3), 109.0, 108.8, 81.6,
81.4, 81.2, 72.6, 57.2, 56.7, 46.6 (1’-C), 35.8 (2’-C), 35.0, 34.4, 32.3, 28.6, 22.9, 12.8,
12.6, 12.4; ESI-MS: m/z 740.4 [M + Na]+ C37H46F3N3NaO8 calcd. 740.3.
In vitro antitumor activity assays
The cytotoxicity was measured by the MTT assay as described in the literature.
(22-23) Briefly, MDA-MB-231 cells were seeded in 96-well plates at the density of
3×103 cells/well/100 μL. After 24 h of cultivation, the cells were treated in triplicate
with various concentrations of compounds for 72 h in 5% CO2 incubator at 37°C. Cell
viability was measured using MTT
[3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] assay. In short, 20 μL
of MTT (5 mg/ml) solution were added to each well. The plate was incubated for an
additional 4 h, and then the medium was removed. Then, 200 μL DMSO was added to
each well to dissolve any purple formazan crystals formed. The plates were
vigorously shaken before taking measurement of relative color intensity. The
absorbance of each well was measured by a microplate reader (Bio-Rad680, USA) at
a test wavelength of 570 nm. The cell inhibitory rate was calculated with the
following equation:
OD control well − OD treated well
OD control well − OD blank well
Inhibition rate =
×100%
The cytotoxicities of compounds were expressed as IC50 that was defined as the drug
concentration required inhibiting growth by 50% relative to controls.
This article is protected by copyright. All rights reserved.

In vivo toxicity assays
Twenty-seven male Kunming mice (weighing 18-22 g) were obtained from the
Experimental Animal Center of Shandong University (Jinan, China). The experiments
were conducted according to the Guide for the Care and Use of Laboratory Animals
and approved by the Animal Care and Use Committee of Shandong University. All
efforts were made to minimize the number of animals used and their suffering.
The mice were group-housed in stainless steel cages (10 mice per cage) in an
air-conditioned room with temperature maintained at 20 2ºC and 12/12 h light/dark
cycle for one week before the experiment. The mice were allowed standard mice
chow diet and drinking water ad libitum throughout the study. The doses for the
compounds were selected on the basis of previously published study (24). Totally
twenty-seven mice were randomly divided into nine groups of three mice each and
were treated as the following details: Group I blank control mice treated with 5%
glucose injection; Group II vehicle control mice; Group III-IX treated with GA,
17-AAG, 2j, 2q, 2v, 2x and 2y, respectively (10 mg/kg).
All the Kunming mice were administered intravenously via tail veins and free
accessed to standard mice chow diet and drinking water. Every day injected one time,
continuously for three days. After that, blood samples were collected for measuring
liver enzymes. The blood serum samples were obtained from centrifugalization under
8000 rpm for 4 mins. To assess damage to the hepatic parenchyma, levels of serum
alanine aminotransferase (ALT) and aspartate aminotransferase (AST) were
determined using commercial available ALT Assay Kit and AST Assay Kit (Shanghai
Rongsheng Biological Pharmaceutical Co., Ltd., Shanghai, China) based on a
spectrophotometric method, respectively. The blood serum samples were treated
according to the manufacturer’s instruction. The ALT and AST activity were
expressed as U/L. All data were expressed as mean standard deviation (SD).
This article is protected by copyright. All rights reserved.

Fluorescence polarization assay (FPA)
FPA was used to evaluate for Hsp90 inhibitors based on displacement of a
fluorescent compound GA-FITC that binds specifically to the Hsp90 ATP-binding
site and was carried out as described (25-26). Briefly, reaction mixes (100 µL)
containing 20 mM HEPES (pH7.3), 50 mM KCl, 2 mM DTT, 5 mM MgCl2, 20 mM
Na2MoO4, 0.01% Triton X-100 with 0.1 mg/mL BSA, 50 nM of recombinant Hsp90α
protein (ab80369, abcam®, USA), 5 nM of GA-FITC (ab141589, abcam®, USA), and
varying concentrations of tested compounds or GA were added in a low binding black
96-well plates (Corning Costar #3650). After incubation for 5h, plates were read on
Omega POLARstar Multi-Mode Microplate Reader (BMG Labtech, Germany) with
excitation at 485 nm and emission at 520 nm. Polarization values (mP) were used to
calculate competitive effects, which were represented as percentage of control
(%control) and calculated following the formula: %control = 100 × (mPc −
mPf)/(mPb − mPf), where mPc, mPb, and mPf are recorded mP values from wells
containing tested compounds, control wells containing both GA-FITC and Hsp90, and
wells containing only the fluorescent probe GA-FITC, respectively. IC50 values were
determined as the competitor concentrations at which 50% of bound GA-FITC was
displaced.
Western blot assay
LNCaP cells were treated without or with different concentrations of compounds for
24 h, and then harvested and lysed in ice-cold lysis buffer (20 mM Tris-HCl, pH 7.4,
150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton, 2.5 mM sodium
pyrophosphate, 1 mM β-glycerolphosphate, 1 mM sodium orthovanadate, 1 mg/ml
leupeptin, 1 mM phenylmethylsulfonyl fluoride). The lysate was mixed with equal
volume of 2× loading buffer (4% SDS, 10% 2-mercaptoethanol, 20% glycerol, and
0.2 mg/ml bromophenol blue in 0.1 M Tris-HCl, pH 6.8), and boiled for 10 min
immediately. The boiled lysates were subjected to 8-12% SDS-PAGE, and then
This article is protected by copyright. All rights reserved.

transferred to PVDF membranes (Millipore), which was then blocked with 5% BSA
in phosphate-buffered saline with 0.1% Tween-20 for 1 h, and incubated overnight
with the corresponding primary antibodies (Anti-β-actin (sc-47778), anti-Hsp70
(sc-24), and anti-EGFR (sc-03) were purchased from Santa Cruz Biotech (USA);
Anti-Raf-1 (#9422), anti-AKt (#9272), anti-HER2 (29D8) (#2165), were purchased
from Cell Signaling Technology (USA)) in the blocking solution at 4°C. Primary
antibodies were detected using either a Peroxidase-conjugated ImmunoPure goat
anti-rabbit IgG (H+L) or Peroxidase-conjugated ImmunoPure Goat Anti-Mouse IgG
(H+L) secondary antibody and enhanced chemiluminescence.
Molecular modeling
Compound 2y was ultimately converted to the PDBQT format using AutoDock Tools
[http://mgltools.scripps.edu], which is required for AutoDock Vina (27). The
3-dimensional (3D) structure of Hsp90 was downloaded from the Protein Data Bank
(PDB ID: 1YET). The molecular docking was conducted using AutoDock Vina,
which uses a unique algorithm that implements a machine learning approach to its
scoring function. Using AutoDock Tools, the PDB (1YET) structure was converted
from a pdb file to a pdbqt file and the search grid was identified as center_x: 40.695,
center_y: -46.782 and center_z: 65.693 with dimensions size_x: 16, size_y: 16, and
size_z: 16. Then the compound 2y was docked into the GA binding site of Hsp90. For
Vina docking, the default parameters were used if it was not mentioned. Then a MD
study was performed to revise the docking result.
The Amber 12.0 (28) program was used for MD simulations of the selected
docked pose. Compound 2y was first prepared by ACPYPE (29), a tool based on
ANTECHAMBER (30-31) for generating automatic topologies and parameters in
different formats for different molecular mechanics programs, including calculation of
partial charges. Then, the forcefield “leaprc.gaff” (generalized amber forcefield) was
used to prepare the ligand, while “leaprc.ff12SB” was used for the receptor. The
This article is protected by copyright. All rights reserved.

system was placed in a rectangular box (with a 10.0 Å boundry) of TIP3P water using
the “SolvateOct” command with the minimum distance between any solute atoms.
Equilibration of the solvated complex was done by carrying out a short minimization
(500 steps of each steepest descent and conjugate gradient method), 500 ps of heating,
and 50 ps of density equilibration with weak restraints using the GPU (NVIDIA®
Tesla K20c) accelerated PMEMD (Particle Mesh Ewald Molecular Dynamics)
module. At last, 8 ns of MD simulations were carried out. All the molecular dynamics
were performed on Dell Precision T5500 workstation.
Results and discussion
Synthesis of 2a-z
The title compounds (2a-z) were prepared as shown in Scheme 1 according to the
similar method as reported in the literatures (11-14, 32). In total, twenty-six
compounds were synthesized by substituting GA (1) with a variety of substitutional
phenylethylamines at the C-17 position (Table 1). Their structures (2a-z) were
characterized by physicochemical, MS, 1H-NMR and 13C-NMR spectral data.
In vitro antiproliferative activity
All twenty-six 17-phenylethylaminegeldanamycin derivatives were evaluated by
MTT assay for their inhibitory activities against both human prostatic cancer cell line
LNCaP and human breast cancer cell line MDA-MB-231. GA and 17-AAG were used
as the positive control drugs (Table 1, blue shaded). According to the types of
substituents, twenty-six geldanamycin derivatives were divided into four small series
(Table 1).
Halogen-substituted derivatives (2a-i) exhibited moderate activities against both
LNCaP and MDA-MB-231 (Table 1, yellow shaded). The para-chlorine substituted
This article is protected by copyright. All rights reserved.

derivative 2c (4-Cl) was more potent than its homologs 2a (4-F) and 2i (4-Br) in both
two cell lines. However, introducing of chlorine to both meta- and para-position of
phenyl group obtained 2f (3,4-2Cl) resulting in a complete loss of inhibitory activities.
Moreover, para-bromine-substituted derivative 2i (4-Br) and
ortho-bromine-substituted derivative 2g (2-Br) showed less active than its congener
2h (3-Br) against both LNCaP and MDA-MB-231.
In monomethoxyl substituted series (2j-l), the activity order was 2j (2-OMe) >
2k (3-OMe) > 2l (4-OMe), and the IC50 value of most potent compound 2j was 0.41
0.12 μM against MDA-MB-231 (Table 1, gray shaded). Dimethoxyl-substituted
derivatives (2m-s) showed less potent to both cell lines, indicating that introduction of
methoxyl group to the compound except ortho-postion will decrease the potency.
Interestingly, the activity order of the monomethyl-substituted derivatives (2t-v)
was not following the rule of monomethoxyl substituted series. Introducing methyl
group to the para-position of the phenyl ring showed more potent activity than the
other two positions (ortho- and meta-position), while the compound 2w (2,5-2Me)
bearing two methyl groups nearly lost its activity (Table 1, pink shaded).
Nitro- and trifluoromethyl-substituted derivatives (2x-z) produced moderate to
excellent anti-proliferative potencies against both LNCaP and MDA-MB-231 (Table
1, green shaded). For instance, 2x (2-CF3) and 2z (4-NO2) did not present remarkable
activities against both cell lines. However, compound 2y containing a trifluoromethyl
group at meta-position of phenyl group was the most active one with the IC50 values
of 0.27 0.11 μM against LNCaP and 0.86 0.23 μM against MDA-MB-231,
respectively. The activity of the compound 2y was more potent than its precursor GA
(IC50 = 0.43 0.15 μM) toward LNCap cells. The promising compounds 2j, 2q, 2v,
2x and 2y were chosen as the representative ones of this class to undertake a study of
their hepatotoxicity evaluation in mice.
This article is protected by copyright. All rights reserved.

In vivo hepatotoxicity evaluation
The AST and ALT are the most commonly used biomarkers for assessment of
hepatotoxicity, since the damaged hepatocytes could leak large amounts of these
enzymes into the bloodstream, leading to the increase of serum ALT and AST (33).
ALT is more specific for hepatic injury than AST because ALT is present mainly in
the cytosol of the liver and in low concentrations elsewhere, while AST is found in
more types of cell (e.g. heart, intestine, muscle) (34-35). As shown in Table 2, the
AST and ALT activities of GA treated group were 317.2 39.1 U/L and 273.0 25.3
U/L, respectively, which were both significantly higher than that of vehicle control
group (P < 0.001). However, positive control drug 17-AAG did not show remarkable
hepatotoxicity. The ALT level of 2q treated mice slightly increased, while 2j, 2v, 2x
and 2y treated groups showed no significant differences versus vehicle control mice
(P > 0.05). These results suggested that compound 2y was a promising antitumor
agent with low hepatotoxicity, which can be further developed for the treatment of
prostate cancer.
Hsp90 inhibitory activity
To further confirm the mechanism of action of the newly synthesized compounds
(2a–2z), the most active compound 2y was selected to carry out an anti-Hsp90 assay
using fluorescence polarization assay (FPA) with GA and 17-AAG as the reference
drugs. The results revealed (Table 3) that compound 2y displayed inhibitory activity
with an IC50 value of 2.29 µM, which is comparable to the reference drug 17-AAG
(IC50 = 0.78 µM). This result indicated that the 17-phenylethylaminegeldanamycin
derivatives can be classified as Hsp90 inhibitors.
Inhibition of Hsp90 in cellular system
Inhibition of Hsp90 leads to the proteasomal degradation of a subset of signaling
proteins that require Hsp90 chaperone activity for their conformational maturation
This article is protected by copyright. All rights reserved.

(14). A decrease in Hsp90 client proteins (including Raf, Akt, EGFR and Her2) and a
compensatory increase in Hsp70 are the two most common biomarkers used for
inhibition of Hsp90 in cells (10). As shown in Figure 3, western blot analysis
indicated that compensatory induction of Hsp70 by the inhibitors was in a
concentration-dependent manner. The levels of Hsp70 were all markedly increased
when treated with 2y, GA and 17AAG at the concentration of 5 µM (Figure 3).
Moreover, compared with the control band (DMSO), the Raf, Akt, EGFR and Her2 of
2y-treated cells had markedly decreased at the concentration of 5 µM. These results
indicated that compound 2y inhibits the activity of Hsp90. When considered together
with its markedly decreased hepatotoxicity in mice, this suggests that 2y may have a
greater therapeutic window than its parent GA and hence considerable potential for
application in the therapy of cancer.
Molecular modeling
To better elucidate the antitumor potencies of 17-phenylethylaminegeldanamycins at
a molecular level and to understand the structural basis of their binding mode, a
docking study of compound 2y was first performed using AutoDock Vina (27) and
then a molecular dynamics (MD) simulation was performed to refine the docking
result by means of Amber 12. The X-ray crystal structure of Hsp90 with GA taken
from PDB (1YET) was used as the input structure. Hsp90-2y complex was
equilibrated after 8 ns MD simulation, and the plot of RMSD (in ångstrom) of the
complex was shown in Figure 4.
The theoretical binding mode of 2y to the ATP/ADP binding site of Hsp90 is
shown in Figure 5. The result suggested that this class of compounds shares a similar
binding mode with GA (36). Compound 2y adopts a compact conformation to bind
inside the pocket of ATP/ADP binding domain of Hsp90 (Figure 5). The
benzoquinone group of 2y binds near the entrance of the pocket, while the ansamycin
ring having dimensions similar to that of a polypeptide in a turn conformation (36). A
This article is protected by copyright. All rights reserved.

high density of van der Waals contacts are formed because of the extensive surface
complementarity between 2y and the pocket of Hsp90. Importantly, six hydrogen
bond contacts are observed between 2y and Hsp90, and one pair from the carbamate
group of 2y can be reasonably described as one of the most important intermolecular
interactions in the complex. It is worthy to point out that the phenylethylamine at the
C-17 position of 2y forms a cation-π interaction with residue Lys103, which is the
main difference between 2y and GA.
Taken together, our molecular simulation allowed us to rationalize the activity
profile of the 17-phenylethylaminegeldanamycins against Hsp90, which provided
valuable information for further design of novel effective Hsp90 inhibitors.
Conclusions
In this study, we synthesized twenty-six novel 17-phenylethylaminegeldanamycin
derivatives, which were structurally confirmed by 1H-NMR, 13C-NMR and MS
spectral analysis and evaluated for their inhibitory activities against cancer cells,
LNCaP and MDA-MB-231. The results showed that five of the derivatives (2j, 2q,
2v, 2x and 2y) exhibited evident antitumor activities. In particular, 2y showed the
most potent activity with an IC50 value 0.27 0.11 μM against LNCaP cells, which
was more active than its precursor GA. Liver injury test in mice demonstrated that 2y
group showed no significant difference for ALT activity versus vehicle control (P <
0.05), indicating that 2y was a promising antitumor candidate. Docking and MD
refinement of the Hsp90-2y complex give us an explanation of theoretical binding
model of 17-phenylethylaminegeldanamycins at molecular level and some hints about
the structural modifications of GA. Further study on this family of Hsp90 inhibitors
and antitumor agents are undergoing in our laboratory.
This article is protected by copyright. All rights reserved.

Acknowledgments
This work was supported by a grant from the 973 Program (2010CB833802), NSFC
Projects (81273384, 90913024), the Independent Innovation Foundation of Shandong
University to C.H. Lu (No. 2010TB016), the Post-graduate Independent Innovation
Fund of Shandong University to Z.Y. Li (YZC12095) and Shanghai Nature Science
Foundation of Shanghai Science and Technology Committee, China (13ZR1432700).
Conflict of interest
We declare that we have no conflict of interest.
Supporting Information
Additional Supporting Information about physicochemical, MS, 1H-NMR and
13C-NMR spectral data of 2a-x and 2z, and 1H-NMR and 13C-NMR spectra of 2a-z
can be found in the online version of this article.
References
1. DeBoer C., Meulman P.A., Wnuk R.J., Peterson D.H. (1970) Geldanamycin, a new antibiotic. J
Antibiot (Tokyo);23:442-447.
2. Stebbins C.E., Russo A.A., Schneider C., Rosen N., Hartl F.U., Pavletich N.P. (1997) Crystal
structure of an Hsp90-geldanamycin complex: targeting of a protein chaperone by an antitumor agent.
Cell;89:239-250.
3. Roe S.M., Prodromou C., O'Brien R., Ladbury J.E., Piper P.W., Pearl L.H. (1999) Structural basis for
inhibition of the Hsp90 molecular chaperone by the antitumor antibiotics radicicol and geldanamycin. J
Med Chem;42:260-266.
This article is protected by copyright. All rights reserved.

4. Stebbins C.E., Russo A.A., Schneider C., Rosen N., Hartl F.U., Pavletich N.P. (1997) Crystal
structure of an Hsp90-geldanamycin complex: targeting of a protein chaperone by an antitumor agent.
Cell;89:239-250.
5. Tian Z.-Q., Wang Z., MacMillan K.S., Zhou Y., Carreras C.W., Mueller T., Myles D.C., Liu Y.
(2009) Potent cytotoxic C-11 modified geldanamycin analogues. J Med Chem;52:3265-3273.
6. An W.G., Schnur R.C., Neckers L., Blagosklonny M.V. (1997) Depletion of p185erbB2, Raf-1 and
mutant p53 proteins by geldanamycin derivatives correlates with antiproliferative activity. Cancer
Chemother Pharmacol;40:60-64.
7. Neckers L., Schulte T.W., Mimnaugh E. (1999) Geldanamycin as a potential anti-cancer agent: its
molecular target and biochemical activity. Invest New Drugs;17:361-373.
8. Supko J.G., Hickman R.L., Grever M.R., Malspeis L. (1995) Preclinical pharmacologic evaluation of
geldanamycin as an antitumor agent. Cancer Chemother Pharmacol;36:305-315.
9. Tian Z.Q., Liu Y., Zhang D., Wang Z., Dong S.D., Carreras C.W., Zhou Y., Rastelli G., Santi D.V.,
Myles D.C. (2004) Synthesis and biological activities of novel 17-aminogeldanamycin derivatives.
Bioorg Med Chem;12:5317-5329.
10. Kitson R.R., Chang C.H., Xiong R., Williams H.E., Davis A.L., Lewis W., Dehn D.L., Siegel D.,
Roe S.M., Prodromou C., Ross D., Moody C.J. (2013) Synthesis of 19-substituted geldanamycins with
altered conformations and their binding to heat shock protein Hsp90. Nat Chem;5:307-314.
11. 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., et al. (1995) Inhibition of the oncogene product p185erbB-2 in
vitro and in vivo by geldanamycin and dihydrogeldanamycin derivatives. J Med Chem;38:3806-3812.
12. 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., et al. (1995) erbB-2 oncogene inhibition by geldanamycin
derivatives: synthesis, mechanism of action, and structure-activity relationships. J Med
Chem;38:3813-3820.
13. Kasuya Y., Lu Z., Kopeckova P., Kopecek J. (2001) Improved synthesis and evaluation of
17-substituted aminoalkylgeldanamycin derivatives applicable to drug delivery systems. Bioorg Med
Chem Lett;11:2089-2091.
14. Lee K., Ryu J.S., Jin Y., Kim W., Kaur N., Chung S.J., Jeon Y.J., Park J.T., Bang J.S., Lee H.S.,
Kim T.Y., Lee J.J., Hong Y.S. (2008) Synthesis and anticancer activity of geldanamycin derivatives
derived from biosynthetically generated metabolites. Org Biomol Chem;6:340-348.
This article is protected by copyright. All rights reserved.

15. Ronnen E.A., Kondagunta G.V., Ishill N., Sweeney S.M., DeLuca J.K., Schwartz L., Bacik J.,
Motzer R.J. (2006) A phase II trial of 17-(Allylamino)-17-demethoxygeldanamycin in patients with
papillary and clear cell renal cell carcinoma. Invest New Drugs;24:543-546.
16. Glaze E.R., Lambert A.L., Smith A.C., Page J.G., Johnson W.D., McCormick D.L., Brown A.P.,
Levine B.S., Covey J.M., Egorin M.J. (2005) Preclinical toxicity of a geldanamycin analog,
17-(dimethylaminoethylamino)-17-demethoxygeldanamycin (17-DMAG), in rats and dogs: potential
clinical relevance. Cancer Chemother Pharmacol;56:637-647.
17. Modi S., Stopeck A., Linden H., Solit D., Chandarlapaty S., Rosen N., D'Andrea G., Dickler M.,
Moynahan M.E., Sugarman S. (2011) HSP90 inhibition is effective in breast cancer: a phase II trial of
tanespimycin (17-AAG) plus trastuzumab in patients with HER2-positive metastatic breast cancer
progressing on trastuzumab. Clin Cancer Res;17:5132-5139.
18. Patani G.A., LaVoie E.J. (1996) Bioisosterism: a rational approach in drug design. Chem
Rev;96:3147-3176.
19. Olesen P.H. (2001) The use of bioisosteric groups in lead optimization. Curr Opin Drug Discovery
Dev;4:471-478.
20. Lima L.M., Barreiro E.J. (2005) Bioisosterism: a useful strategy for molecular modification and
drug design. Curr Med Chem;12:23-49.
21. Hagmann W.K. (2008) The many roles for fluorine in medicinal chemistry. J Med
Chem;51:4359-4369.
22. Carmichael J., DeGraff W.G., Gazdar A.F., Minna J.D., Mitchell J.B. (1987) Evaluation of a
tetrazolium-based semiautomated colorimetric assay: assessment of chemosensitivity testing. Cancer
Res.;47:936-942.
23. Huang Y., Wang J., Li G., Zheng Z., Su W. (2001) Antitumor and antifungal activities in endophytic
fungi isolated from pharmaceutical plants Taxus mairei, Cephalataxus fortunei and Torreya grandis.
FEMS Immunol. Med. Microbiol.;31:163-167.
24. Grem J.L., Morrison G., Guo X.-D., Agnew E., Takimoto C.H., Thomas R., Szabo E., Grochow L.,
Grollman F., Hamilton J.M. (2005) Phase I and pharmacologic study of
17-(allylamino)-17-demethoxygeldanamycin in adult patients with solid tumors. J Clin
Oncol;23:1885-1893.
25. Lu C., Liu D., Jin J., Deokar H., Zhang Y., Buolamwini J.K., Yu X., Yan C., Chen X. (2013)
Inhibition of gastric tumor growth by a novel Hsp90 inhibitor. Biochem Pharmacol;85:1246-1256.
This article is protected by copyright. All rights reserved.

26. Llauger-Bufi L., Felts S.J., Huezo H., Rosen N., Chiosis G. (2003) Synthesis of novel fluorescent
probes for the molecular chaperone Hsp90. Bioorg Med Chem Lett;13:3975-3978.
27. Trott O., Olson A.J. (2010) AutoDock Vina: improving the speed and accuracy of docking with a
new scoring function, efficient optimization, and multithreading. J Comput Chem;31:455-461.
28. Pierce L.C., Salomon-Ferrer R., Augusto F. de Oliveira C., McCammon J.A., Walker R.C. (2012)
Routine access to millisecond time scale events with accelerated molecular dynamics. J Chem Theory
Comput;8:2997-3002.
29. da Silva A.W.S., Vranken W.F. (2012) ACPYPE-Antechamber python parser interface. BMC Res
Notes;5:367.
30. Wang J., Wolf R.M., Caldwell J.W., Kollman P.A., Case D.A. (2004) Development and testing of a
general amber force field. J Comput Chem;25:1157-1174.
31. Wang J., Wang W., Kollman P.A., Case D.A. (2006) Automatic atom type and bond type perception
in molecular mechanical calculations. J Mol Graph Model;25:247-260.
32. Wuest F., Bouvet V., Mai B., LaPointe P. (2012) Fluorine- and rhenium-containing geldanamycin
derivatives as leads for the development of molecular probes for imaging Hsp90. Org Biomol
Chem;10:6724-6731.
33. Thimmulappa R.K., Mai K.H., Srisuma S., Kensler T.W., Yamamoto M., Biswal S. (2002)
Identification of Nrf2-regulated genes induced by the chemopreventive agent sulforaphane by
oligonucleotide microarray. Cancer Res;62:5196-5203.
34. Giboney P.T. (2005) Mildly elevated liver transaminase levels in the asymptomatic patient. Am
Fam Physician;71:1105-1110.
35. Uno S., Dalton T.P., Derkenne S., Curran C.P., Miller M.L., Shertzer H.G., Nebert D.W. (2004) Oral
exposure to benzo[a]pyrene in the mouse: detoxication by inducible cytochrome P450 is more
important than metabolic activation. Mol Pharmacol;65:1225-1237.
36. Stebbins C.E., Russo A.A., Schneider C., Rosen N., Hartl F.U., Pavletich N.P. (1997) Crystal
structure of an Hsp90–geldanamycin complex: targeting of a protein chaperone by an antitumor agent.
Cell;89:239-250.
This article is protected by copyright. All rights reserved.

Table 1: In vitro cytotoxicity of compounds 2a-z against LNCaP and MDA-MB-231 cells
LNCaP
IC50 (μM)
> 10
MDA-MB-231
IC50 (μM)
4.62 0.33
5.17 1.26
2.03 0.36
7.00 0.67
6.38 1.27
> 10
Compounds
R
4-F
2-F, 6-Cl
4-Cl
2a
2b
2c
2d
2e
2f
2.19 0.12
2.86 0.45
> 10
2,3-2Cl
2,6-2Cl
3,4-2Cl
2-Br
2.46 0.33
> 10
5.08 1.21
1.47 0.87
> 10
6.86 2.13
3.23 0.56
5.51 0.77
0.41 0.12
2.48 0.23
3.64 0.73
2.26 0.45
9.32 2.87
4.37 0.54
2.56 0.67
2g
2h
2i
3-Br
4-Br
2j
2-OMe
3-OMe
4-OMe
2,3-2OMe
2,5-2OMe
3,4-2OMe
3,5-2OMe
5.28 1.35
> 10
2k
2l
> 10
2m
2n
2o
2p
4.40 2.09
1.34 0.23
> 10
6.85 0.45
This article is protected by copyright. All rights reserved.

4-OH,3-OMe
4-OH,3,5-2OMe
3,4-Methylenedioxyl
2-Me
0.81 0.19
3.24 0.78
> 10
3.26 0.15
4.28 1.08
5.75 1.56
7.57 2.43
> 10
2q
2r
2s
2t
7.33 2.15
> 10
3-Me
2u
4-Me
2.31 0.63
> 10
0.42 0.23
> 10
2v
2,5-2Me
2w
2-CF3
1.25 0.21
0.27 0.11
5.12 1.34
0.16 0.19
0.43 0.15
0.92 0.21
0.86 0.23
2.75 0.47
0.28 0.25
0.05 0.04
2x
3-CF3
2y
4-NO2
2z
17-AAG
GA
Table 2: In vivo hepatotoxicity of the active compounds in mice
Compounds
AST (U/L)
ALT (U/L)
266.0 17.2 60.1 5.2
220.0 31.2 99.6 4.8
220.2 21.8 57.7 6.7
2j
2q
2v
222.7 8.5
69.0 7.6
2x
2y
227.4 20.5 62.3 5.9
237.4 29.4 66.6 8.1
317.2 39.1 273.0 25.3
197.2 11.2 60.7 6.9
17-AAG
GA
Vehicle control
This article is protected by copyright. All rights reserved.

Table 3: IC50 values of 2y, 17-AAG and GA binding to Hsp90
Compounds IC50 (μM)
2.29
0.78
0.14
2y
17-AAG
GA
Figure 1: Structures of GA, 17-AAG and 17-DMAG.
This article is protected by copyright. All rights reserved.

Figure 2: Designed strategies of 17-phenylethylaminegeldanamycins.
Figure 3: Evaluation of 2y as the inhibitor of Hsp90 in the cellular system.
This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

O
O
H
N
O
O
O
MeO
MeO
O
R
NH2
N
H
R
N
H
MeO
CH2Cl2, r.t.
OH
MeO
OH
MeO
O
O
O
O
NH2
NH2
1
2a-z
Scheme 1: Synthesis route of 17-phenylethylaminegeldanamycins 2a-z.
This article is protected by copyright. All rights reserved.