Design strategies to target crystallographic waters applied to the Hsp90 molecular chaperone

Article abstract of DOI:10.1016/j.bmcl.2011.04.130

A series of novel and potent small molecule Hsp90 inhibitors was optimized using X-ray crystal structures. These compounds bind in a deep pocket of the Hsp90 enzyme that is partially comprised by residues Asn51 and Ser52. Displacement of several water mol

Full text of DOI:10.1016/j.bmcl.2011.04.130

Bioorganic & Medicinal Chemistry Letters 21 (2011) 3557–3562
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design strategies to target crystallographic waters applied to the Hsp90
molecular chaperone
Pei-Pei Kung a, , Piet-Jan Sinnema b, Paul Richardson a, Michael J. Hickey a, Ketan S. Gajiwala a, Fen Wang a,
⇑
Buwen Huang a, Guy McClellan a, Jeff Wang a, Karen Maegley a, Simon Bergqvist a, Pramod P. Mehta a,
Robert Kania a
a Pfizer Worldwide Research and Development, La Jolla Laboratories, 10770, Science Center Dr., San Diego, CA 92121, USA
b Syncom B. V., Kadijk 3, 9747AT Groningen, The Netherlands
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of novel and potent small molecule Hsp90 inhibitors was optimized using X-ray crystal struc-
tures. These compounds bind in a deep pocket of the Hsp90 enzyme that is partially comprised by resi-
dues Asn51 and Ser52. Displacement of several water molecules observed crystallographically in this
pocket using rule-based strategies led to significant improvements in inhibitor potency. An optimized
inhibitor (compound 17) exhibited potent Hsp90 inhibition in ITC, biochemical, and cell-based assays
Received 11 March 2011
Revised 25 April 2011
Accepted 27 April 2011
Available online 5 May 2011
(Kd = 1.3 nM, Ki = 15 nM, and cellular IC50 = 0.5 lM).
Keywords:
Ó 2011 Elsevier Ltd. All rights reserved.
Hsp90 inhibitor
Structure-based design technique
Pyrrolopyrimidine
Heat shock protein (Hsp90) is an important target in the devel-
opment of cancer therapy due to its role at the crossroads of multi-
ple signaling pathways associated with cell proliferation and cell
viability.1 It is an ATPase-dependent protein folding molecular
chaperone.2 It plays a key role in the stability and function of many
proteins that are important to cancer cells, such as Bcr-Abl, Erb-B2,
EGFR, BRaf, Akt, Met, VEGF, and Flt3.3 Inhibition of Hsp90 leads to
simultaneous blockage of multiple signal transduction pathways,
thus making Hsp90 an attractive anticancer target.
The natural product Hsp90 inhibitor geldanamycin (GA), its
semi-synthetic analogs (17-AAG and 17-DMAG, Fig. 1),4 and sev-
eral non-quinone small molecule inhibitors have entered clinical
trials.5–9 Many other examples of non-quinone small molecule
HSP90 inhibitors are also known.10
via improving the binding affinity to the enzyme using structure-
based design techniques.
We initially suspected that the large difference between bio-
chemical and cellular potencies of 1 might be due to the presence
of reactive and/or unstable functional groups in the compound’s
chemical structure (e.g., the 2-chloro-pyridine moiety and/or the
boc-substituted pyrrole). Accordingly, we explored the replace-
ment of these potential structural liabilities. Compound 2 (Fig. 3),
which incorporates a methyl group in place of the chloro substitu-
ent of 1, exhibited 10-fold weaker biochemical inhibitory potency
relative to the latter molecule. Further replacement of the boc-pyr-
role present in 2 with a phenyl moiety gave compound 3 and re-
sulted in the loss of another 10-fold in biochemical potency
relative to 2 (Table 1). Unfortunately, neither compound 2 nor 3
Our Hsp90 research program began with a high throughput
screening campaign using a competitive binding assay in which
the ability of test compounds to displace a tritium-labeled resor-
cinol amide ligand11,12 was examined (Fig. 2). This effort led to
the discovery of the 4-chloro-2-aminopyrimidine-containing com-
pound 1 (Fig. 3) which displayed a Ki of 261 nM and a correspond-
ing ligand efficiency (LE)13 of 0.47. Unfortunately, the compound
(cell IC50s >10
l
M) displayed measurable improvement in cell
O
17
R
O
N
H
exhibited an IC50 of about 45 lM in a cell-based assay which exam-
O
GA: R=OCH3
17-AAG: R=-NHCH2CH=CH2
17-DMAG: R=-NHCH2CH2NMe2
ined the degradation of the Hsp90 client protein Akt.14 Below we
describe our efforts to enhance the inhibitor’s cellular potency
O
O
7
NH2
O
OH
O
⇑
Corresponding author.
E-mail address: peipei.kung@pfizer.com (P.-P. Kung).
Figure 1. Structures of benzoquinone ansamycins.
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.04.130

3558
P.-P. Kung et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3557–3562
correlated well with the corresponding Ki data generated in the
T
displacement assay (Table 1) with the latter values consistently
being 2- to 3-fold higher.
Encouragingly, one of these inhibitors displayed enhanced
activity in the cell-based assay as compared to the original screen-
T
NH
Cl
O
HO
ing hit (compare 6 with 1, cellular IC50s of 10 and 45 lM, respec-
N
tively). However, the micromolar IC50 values of both compounds
indicated further improvements in the inhibitor series were
required.
Compound 7 displayed improved cellular potency relative to
compounds 3–5 (Table 1) and improved aqueous solubility when
OH
O
Figure 2. Tritium-labeled resorcinol amide ligand.
compared with 6 (65 lM compared with 12 lM for compound
6). This latter improvement is probably due to the presence of
the methoxy group in compound 7). A 1.7 Å co-crystal structure
of 7 bound to the Hsp90 enzyme was obtained, illuminating a un-
ique opportunity to apply structure-based design strategies that
target crystallographic water molecules (Fig. 4 and Fig. 5, PDB code
3RLP).
O
N
O
N
R
N
NH2
Compound 7 bound to the ATP site and interacted with Asp93
and a conserved water molecule (W1). Similar interactions have
been reported by others and a comparison of various ligand bind-
ing modes was also recently published.16 Further analysis of this
co-crystal structure identified a small unfilled pocket formed by
Hsp90 residues Asn51, Leu48, Ile91, Val186, and Ser52 near the
aminopyrimidine ring of 7. The co-crystal structure also revealed
several water molecules (W2, W3, and W4) that either formed
hydrogen bonds to the inhibitor or comprised a network in a deep
buried binding pocket that compound 7 was not occupying (Fig. 5).
This water network was also observed in a co-crystal structure of
the PU317 inhibitor bound to the Hsp90 protein.18 However, the
importance of the interactions between these water molecules
and the corresponding inhibitors has not yet been completely
explored.
1, R=Cl, Ki= 0.3 µM, cell IC50= 45 µM.
2, R=CH3, Ki= 3.7 µM, cell IC50= 0% inhib. @ 10 µM.
Figure 3. HTS hit (compound 1) and hit follow-up (compound 2).
potency, suggesting that further modification would be required to
achieve our goals.
From the analysis of the minimized conformation of compound
1, the preferred Hsp90 binding mode likely required that the pyr-
role and aminopyrimidine rings reside in a non-planar orientation
to relieve torsional strain. A methoxy group was therefore intro-
duced at the ortho-position of the phenyl moiety present in 3 to en-
force such a conformation. The resulting molecule (compound 4,
Table 1) improved biochemical Hsp90 inhibition potency relative
to 3. To further optimize the potency, a small library of compounds
was then prepared by coupling various 2-substituted phenyl boro-
nic acids/esters to the 4-methyl-2-aminopyrimidine core. This
activity afforded three molecules which exhibited further im-
proved Hsp90 inhibition activity relative to 4 (compounds 5, 6,
and 7, Table 1). Since the binding assay employed a competitive
method to measure affinity relative to pre-bound, tritiated ligand
( Fig. 2), we periodically used isothermal titration calorimetry
(ITC)15 experiments to measure the absolute binding affinity of a
selected subset of inhibitors. The dissociation constants (Kd)
Deeper analysis of the small water pocket just internal to the
aminopyrimidine ring of 7 indicated that it was formed between
a b-sheet (cyan, above) and
a-helix (red, below), as shown in
Figure 4. Despite the secondary structure of the flanking protein,
which tied up backbone donors and acceptors into satisfied
H-bonds, the pocket was filled with the three water molecules
(W2–W4). Each water molecule had some hydrogen bonds satisfied
by either compound 7, polar side chains of the protein (Asn51,
Ser52, and Asp93) or one of the other water molecules within this
buried network (Fig. 5).
Table 1
R2
R3
R1
N
N
NH2
Compd#
R1
R2
R3
Enzyme Kia
(l
M)
Kd
(l
M)
Cell IC50 (lM)
3
4
5
6
7
H
H
H
Cl
Cl
Cl
H
H
H
H
22
7.9
2.7
0.35
0.14
ND
3.5 0.36
ND
0.104 0.001
0.084 0.0006
>10b
>10b
27
10.5
10
OCH3
OCH3
Cl
Cl
OCH3
a
Assay error is less than 20%.
No measurable activity seen at 10 lM.
b

P.-P. Kung et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3557–3562
3559
features to form a fourth interaction to the water if needed,
(4) if a water had three or four good H-bonds, one with a H-
bond partner on the original ligand, then maintain a similar
H-bond feature in future designs.
Analysis of the Hsp90 network revealed that W1 was an exam-
ple of the last point made above. It formed four H-bonds with very
good distances and angles, one of which was to the pyrimidine
nitrogen of 7. The design strategy dictates that we leave this water
and maintain the H-bond acceptor of 7 in future inhibitors.
Analysis of the W2–W4 network was less straightforward. The
structure showed three good H-bonds to W2 and four to W3, but
the geometry around W3 was far from ideal. Only two H-bonds
to W4 were counted because (1) the longer distance to Ser52, (2)
the OH of Ser52 already had donor and acceptor interactions satis-
fied and, (3) because of the sharp angles of such a H-bond, which
would bisect the two H-bonds to W4 that are counted above to
give angles of 55° and 61°. For W4, with its two H-bonds and
semi-lipophilic environment due to proximity to Ile91 and
Val186 (not visible in Fig. 5), the above strategy dictated that it
should be targeted for displacement by a lipophilic group. With
W4 targeted for displacement, its H-bond to W3 was lost, leaving
only three H-bonds to W3 of poor geometry. One of the three
remaining H-bonds came from the ligand, where new designs
could easily remove that partner. Targeting W3 for displacement
also left W2 with two remaining H-bonds because they formed
H-bonds with each other. An additional consideration for W2
was that one of the remaining two H-bonds was with the ligand,
which can be engineered to accommodate. Taken together, all
three waters should be targeted for displacement as a group while
the ligand H-bond partners to these waters are simultaneously
eliminated to remove the desolvation penalty of binding a water-
less pocket. Specifically, the design strategy was to fill the deep
pocket as completely as possible while maintaining other core
interactions of aminopyrimidine series. Displacing one or two
waters was insufficient according to the analysis. Such an incom-
plete approach would leave a high energy environment for the
remaining waters or a high energy cavity and would insufficiently
test the hypothesis.
Figure 4. A ribbon diagram of Hsp90 shows that the protein surrounds 7 and three
buried waters with secondary structures, -helix from below and b-sheet from
above. Therefore, all backbone amide H-Bond donors and acceptors are satisfied
adjacent to the waters.
a
Lastly, the protein should be considered in designs so that the
H-bond donor of Asn51 can be matched with designed structural
feature that present an H-bond acceptor.
Inspired by the above analysis, compound 8 was prepared as a
negative challenge to the hypothesis and to test the importance
of the ligand H-bond interactions to the interior water network.
In this case, a ligand H-bond to W2 was removed by migrating
the pyrimidine nitrogen to the exterior, where it could continue
to H-bond, but with bulk waters. This regioisomer was also seen
as a necessary modification to construct more complex designs
that target the full spirit of the above hypothesis. Compound 8 lost
10-fold inhibitory potency relative to 7, suggesting that the forma-
tion of a frank hydrogen bond between the latter molecule and W2
is a significant contribution to potent Hsp90 binding through sta-
bilization of the whole internal water network. Indeed, a crystal
structure of 8 was obtained and showed that the network re-
mained, but W2 was now 3.3 Å from 8 compared with the 3.0 Å
distance to 7.
Figure 5. X-ray structure of 7 bound to the ATP site of human Hsp90. Four buried
water molecules (W1–W4) form a hydrogen bond interactions, three (W2–W4) in a
network with the inhibitor, each other and to the binding pocket region of the
protein.
We employed simple, rule-based drug design strategies for
the crystallographic waters. First, all the H-bonds for a water
molecule of interest were identified. These could include H-
bonds to ligand, protein, or other water molecules, provided
each partner was not already engaged in an H-bond (e.g., the
a-helix and b-sheet backbone amides were excluded as dis-
cussed earlier). Good H-bonds were those with distances be-
tween heavy atoms of 2.6–3.2 Å, near-tetrahedral geometry
with angles between H-bonds of ꢀ80° to 140°, and appropriate
geometry considerations for the H-bond partner (e.g., 120° and
in plane for amides). From this, appropriate strategies were
determined as follows: (1) displace waters that were weakly
bound by less than three good H-bonds with newly designed li-
gand structural features that present lipophilic groups in lipo-
philic pockets, or polar groups to remake key H-bonds that
would otherwise be left unmet, (2) leave waters that were
tightly bound with at least three good H-bonds that anchor it
to the protein, (3) incorporate newly designed ligand structural
In order to test the hypothesis that replacing W2 could gain affin-
ity to the Hsp90 enzyme, we further investigated whether a bicyclic
ring system derived from 8 could be accommodated in the Hsp90
binding site. The resulting pyrrolopyrimidine inhibitor (9,Fig. 6) dis-
played improved affinity to the Hsp90 enzyme compared to that
exhibited by 8 (Ki of 1.7 lM compared with 10 lM). The added
atoms of the fused ring compensated for the loss of the H-bond to
the water network and displaced W2. Importantly, the new bicyclic
core provideda vector from which interactionswith the side chain of
Asn51 could be introduced.

3560
P.-P. Kung et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3557–3562
Cl
Cl
O
O
Cl
Cl
N
N
N
N
H
H2N
N
8
9
Enzyme Ki= 1.7 µM
Cell IC50= NI @ 10 µM
Enzyme Ki= 10 µM
Cell IC50 = NI @ 10 µM
Figure 6. Design of 9 from 8.
Table 2
Figure 7b. Fo–Fc difference electron density map at 2.5r contour level for
compound 13.
Cl
N
O
improved Hsp90 inhibition activity relative to compound 9 (10–13,
Table 2). Inhibitor 13, bearing a cyano group at the R1 position, was
especially potent. The nitrile moiety present in the compound was
designed to fill space occupied by W2 and directly form a H-bond
with the Hsp90 Asn51 side chain. A 1.9 Å co-crystal structure of 13
and Hsp90 was subsequently obtained that confirmed this hypoth-
esis (Figs. 7a and b, PDB code 3RLQ). Remarkably, W3 remained,
although it was displaced 1.2 Å from its original position in the
7-Hsp90 co-crystal structure. The edge of the bicyclic aryl ring of
13 is 3.2 Å from W3 with the highly polarized CH at R2 pointing to-
ward the water, suggesting that a polarized aryl C–H H-bond is
able to provide a stabilizing interaction.
Design from the 13-Hsp90 co-crystal structure suggested that
the pocket could be filled via displacement of the remaining water
molecules (W3 and W4) with hydrophobic groups (Fig. 7). Several
pyrrolopyrimidines bearing small alkyl groups at the R2 position
were synthesized (14–17, Table 2). Compared to the corresponding
molecules containing a hydrogen atom at the R2 position, these
compounds displayed small additional improvements or no
improvements in potency when evaluated in biochemical and
cell-based assays (Table 2). The potency gains for compounds 16
and 17, realized through beneficial hydrophobic displacement of
the remaining two waters, was attenuated by the exchange of
favorable water interactions with the polar side chain of Ser52
with repulsion due to R2 sterics.
Cl
R1
N
R2
N
H
b
Compd#
R1
R2
Enzyme Kia
(
lM)
Cell IC50 (lM)
9
H
H
H
H
H
2
NI at 10
ND
l
M
10
11
12
13
14
15
16
17
COCH3
F
Cl
CN
H
Cl
0.2
0.2
0.2
0.04
2
0.1
0.03
0.015
ND
2
1
H
CH3
CH3
CH3
NI at 10
ND
l
M
CN
CN
0.7
0.5
CH2CH3
a,b
Assay error for both assays are 10–15%.
Accordingly, we synthesized several pyrrolopyrimidines con-
taining various R1 substituents and observed that many displayed
This latter point can be visualized in the solved 16-Hsp90 co-
crystal structure ( Fig. 8, PDB code 3RLR). The oxygen of Ser52
was moved 0.5 Å from its location when 13 was bound to Hsp90.
Even with the movement, the distance between the Ser52 oxygen
and the methyl group of 16 was only 2.9 Å, suggesting a clash. The
net result was a negligible potency gain for 16 compared to 13
(Ki = 30 and 40 nM, respectively). The ethyl derivative 17, which
now more completely filled the deep pocket, exhibited a Ki potency
of 10 nM, with a measured Kd of 1.3 nM, and sub-lM cellular
potency. This corresponded to high binding efficiency, with an
LE = 0.52 and LipE19 = 4.1.
Compounds 3–8 were synthesized following the reported pro-
cedure in the literature.20 Compound 9–17 were synthesized in
an analogous manner. Specifically, the synthesis of compound 17
is depicted in Scheme 1. The 2-ethyl 4-chloro-6-dimethyl-7H-pyr-
rolo[2,3-d]pyrimidine (18) was prepared from cyano-acetic acid
methyl ester follow the procedure reported by Hutchison et al.21
Subsequent alkylation with 2-(trimethylsilyl)-ethoxymethyl chlo-
ride gave 19 in 87% yield. Bromination of the C-5 position of 19
using N-bromosuccinimide gave 20 in 98% yield. Compound 20
Figure 7a. Binding interactions of 13 to Hsp90. The cyano moiety of pyrrolopyr-
imidine core replaced water molecule W2 to form a direct hydrogen bond with the
side chain of Asn51. Oxygen atoms of structural water molecules are depicted in
cyan. Oxygen atoms of the Hsp90 protein and bound inhibitor are colored red.

P.-P. Kung et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3557–3562
3561
novel lead for further lead optimization. In comparison with the
modified HTS hit (compound 7), it displayed similar LE, LipE (17,
LE = 0.52 and LipE = 4.1; 7, LE = 0.55 and LipE = 3.7, LE and LipE
were calculated by the Kd), and improved the cell phenotypic effect
by at least 20-fold.
Acknowledgment
We thank Dr. Bas Dros at Syncom for helpful discussions and
preparation of the manuscript.
Supplementary data
Supplementary data (synthetic procedures and analytical data)
associated with this article can be found, in the online version, at
doi:10.1016/j.bmcl.2011.04.130.
Figure 8. Ser52 movement away from the methyl group of 16 is only enough to
give separation of 2.9 Å.
References and notes
1. Workman, P.; Burrows, F.; Neckers, L.; Rosen, N. Ann. N.Y. Acad. Sci. 2007, 1113,
202.
2. Pearl, L. H.; Prodromou, C. Curr. Opin. Struct. Biol. 2000, 10, 46.
3. Solit, D. B.; Osman, I.; Polsky, D.; Panageas, K. S.; Daud, A.; Goydos, J. S.;
Teitcher, J.; Wolchok, J. D.; Germino, F. J.; Krown, S. E.; Coit, D.; Rosen, N.;
Chapman, P. B. Clin. Cancer Res. 2008, 14, 8302.
4. Whitesell, L. et al Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 8324.
5. Kasibhatla, S. R.; Hong, K.; Biamonte, M. A.; Busch, D. J.; Karjian, P. L.;
Sensintaffar, J. L.; Kamal, A.; Lough, R. E.; Brekken, J.; Lundgren, K.; Grecko, R.;
Timony, G. A.; Ran, Y.; Mansfield, R.; Fritz, L. C.; Ulm, E.; Burrows, F. J.; Boehm,
M. F. J. Med. Chem. 2007, 50, 2767.
was then carboxylated with CO2 in the presence of the
Turbo-Grignard reagent22 (isopropylmagnesium chloride, lithium
chloride complex) to give compound 21 in 96% yield. Reaction of
21 with oxalyl chloride and ammonia afforded carboxylic-acid–
amide compound 22. Dehydration of 22 then gave compound 23
containing the C3-cyano group in 77% yield over two steps. Initial
attempts at Suzuki coupling of 23 with 2523 resulted in a messy
mixture. Deprotection of 23 with BF3ÁOEt2 gave 24 in quantitative
yield. The subsequent Suzuki coupling reaction of 24 with 25 then
gave a much cleaner reaction mixture and provided compound
1724 in 45% yield. In summary, compound 17 represents a new
6. Nakashima, T.; Ishii, T.; Tagaya, H.; Seike, T.; Nakagawa, H.; Kanda, Y.; Akinaga,
S.; Soga, S.; Shiotsu, Y. Clin. Cancer Res. 2010, 16, 2792.
7. Woodhead, A. J.; Angove, H.; Carr, M. G.; Chessari, G.; Congreve, M.; Coyle, J. E.;
Cosme, J.; Graham, B.; Day, P. J.; Downham, R.; Fazal, L.; Feltell, R.; Figueroa, E.;
Cl
Br
Cl
Cl
N
N
N
N
N
b
a
N
N
N
H
N
O
O
20
18
N
19
Si
c
Si
OH
Cl
N
NH2
O
Cl
O
Cl
N
N
e
N
d
N
N
N
N
N
O
O
O
21
Cl
Si
22
23
Si
Si
f
Cl
O
O
N
Cl
N
Cl
Cl
B
N
O
O
N
N
H
N
N
H
N
25
24
17
Scheme 1. Reagents and conditions: (a) (CH3)3Si(CH2)2OCH2Cl, NaH, DMF, 0–23 °C, 3 h, 87%; (b) NBS, DMF, 4–23 °C, 2 h, 98%; (c) i-PrMgCl/LiCl, THF, 50 °C, 3 h, CO2, À78 °C,
30 min, 96%, (d) COCl2, DCM, ammonia, 97%; (e) CF3CO2H, DCM, 0–23 °C, 1 h, 77%; (f) BF3ÁOEt2, quant.; (g) Pd(PPh3)4, Na2CO3(aq), compound 25, 1,4-dioxane, microwave,25
100 °C, 45 min, 45%.

3562
P.-P. Kung et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3557–3562
were added to each well starting at 10
Frederikson, M.; Lewis, J.; McMenamin, R.; Murray, C. W.; O’Brien, M. A.; Parra,
l
M with 3-fold serial dilution of 11
L.; Patel, S.; Philips, T.; Rees, D. C.; Rich, S.; Smith, D.-M.; Trewartha, G.;
Vinkovic, M.; Williams, B.; Woolford, A. J.-A. J. Med. Chem. 2010, 53, 5956.
8. Huang, K. H.; Veal, J. M.; Fadden, R. P.; Rice, J. W.; Eaves, J.; Strachan, J.-P.;
Barabasz, A. F.; Foley, B. E.; Barta, T. E.; Ma, W.; Silinski, M. A.; Hu, M.; Partridge,
J. M.; Scott, A.; DuBois, L. G.; Freed, T.; Steed, P. M.; Ommen, A. J.; Smith, E. D.;
Hughes, P. F.; Woodward, A. R.; Hanson, G. J.; McCall, W. S.; Markworth, C. J.;
Hinkley, L.; Jenks, M.; Geng, L.; Lewis, M.; Otto, J.; Pronk, B.; Verleysen, K.; Hall,
S. E. J. Med. Chem. 2009, 52, 4288.
points of various concentrations. 24 h post adding compounds, cell lysates
were prepared and subjected to Luminex 100 system using the Beadlyte Akt/
PKB kit–Upstate #46-605.
15. Wiseman, T.; Williston, S.; Brandts, J. F.; Lin, L. N. Anal. Biochem. 1989, 179, 131.
16. Biamonte, M. A.; Van de Water, R.; Arndt, J. W.; Scannevin, R. H.; Perret, D.; Lee,
W.-C. J. Med. Chem. 2010, 53, 3.
17. Chiosis, G.; Timaul, M. N.; Lucas, B.; Munster, P. N.; Zheng, F. F.; Sepp-
Lorenzino, L.; Rosen, N. Chem. Biol. 2001, 8, 289.
9. Brough, P. A.; Aherne, W.; Barril, X.; Borgognoni, J.; Boxall, K.; Cansfield, J. E.;
Cheung, K.-M. J.; Collins, I.; Davies, N. G. M.; Drysdale, M. J.; Dymock, B.; Eccles,
S. A.; Finch, H.; Fink, A.; Hayes, A.; Howes, R.; Hubbard, R. E.; James, K.; Jordan,
A. M.; Lockie, A.; Martins, V.; Massey, A.; Mathews, T. P.; McDonald, E.;
Northfiel, C. J.; Pearl, L. H.; Prodromou, C.; Ray, S.; Raynaud, F. I.; Roughley, S.
D.; Sharp, S. Y.; Surgenor, A.; Walmsley, D. L.; Webb, P.; Wood, M.; Workman,
P.; Wright, L. J. Med. Chem. 2008, 51, 196.
18. PDB code: 1uym Wright, L.; Barril, X.; Dymock, B.; Sheridan, L.; Surgenor, A.;
Beswick, M.; Drysdale, M.; Collier, A.; Massey, A.; Davies, N.; Fink, A.; Fromont,
C.; Aherne, W.; Boxall, K.; Sharp, S.; Workman, P.; Hubbard, R. E. Chem. Biol.
2004, 11, 775.
19. (a) Lesson, P. D.; Springthorpe, B. Nat. Rev. Drug Disc. 2007, 6, 881; (b)
Ryckmans, T.; Edwards, M. P.; Horne, V. A.; Correia, A. M.; Owen, D. R.;
Thompson, L. R.; Tran, I.; Tutt, M. F.; Young, T. Bioorg. Med. Chem. Lett. 2009, 19,
4406.
10. Janin, Y. L. Drug Discovery Today 2010, 15, 342.
11. Compounds were evaluated for potency against Hsp90 using
a
SPA
20. Chessari, G.; Congreve, M. S.; Callaghan, O.; Cowan, S. R.; Murray, C. W.;
Woolford, A. J.-A.; O’Brien, M. A.; Woodhead, A. J. WO 2006 123165 A2.
21. Hutchison, D. L.; Martinelli, M. J.; Wilson, T. M. WO 2000 00201 A1.
22. Krasovskiy, A.; Knochel, P. Angew. Chem. 2004, 116, 3396.
23. Compound 25 was synthesized following an analogous procedure from Kung,
P.-P., Meng, J. J. WO 2008 059368 A2.
(scintillation proximity assay) competition binding assay. Briefly, either full
length or N-terminal HSP-90 that contains a 6-His tag on its C-terminus were
bound to copper on Yttrium-silicate scintillant beads. Test compounds which
bind to Hsp90 competed with tritiated propyl-geldanamycin (3H-pGA) at the
ATP-binding site on Hsp90. For Ki determinations, % inhibition values are
plotted versus log inhibitor concentration and fit to four parameter logistic
equation. The IC50 is converted to competitive Ki using Cheng–Prusoff equation.
24. Compound 17 was characterized by 1H NMR and elemental analysis to be >90%
purity.
Where Ki is the inhibition constant,
L
is the total concentration of
1H NMR (400 MHz, DMSO-d6) d ppm 1.31 (t, J = 7.7 Hz, 3 H) 2.71 (s, 3 H) 2.90 (q,
J = 7.6 Hz, 2 H) 3.89 (s, 3 H) 7.31 (s, 1 H) 7.77 (s, 1 H) 13.15 (br s, 1H). Anal.
Calcd for C17H14Cl2N4O: C, 54.70; H, 4.53; N, 14.37. Found: C, 55.06; H, 4.11, N,
14.25.
pGA = 200 nM, and Kd is the dissociation constant for pGA = 240 nM.
12. Kung, P.-P.; Huang, B.; Zhang, G.; Zhou, J. Z.; Wang, J.; Digits, J. A.; Skaptason, J.;
Yamazaki, S.; Neul, D.; Zientek, M.; Elleraas, J.; Mehta, P.; Yin, M.-J.; Hickey, M.
J.; Gajiwala, K. S.; Rodgers, C.; Davies, J. F.; Gehring, M. R. J. Med. Chem. 2010, 53,
499.
13. Hopkins, L. A.; Groome, C. R.; Alex, A. Drug Discovery Today 2004, 9, 430.
14. Akt lum assay: non small cell lung carcinoma cell line H1299 was cultured in
DMEM + 10%FBS, 10,000cells/well were cultured in 96-well plate. Compounds
25. Biotage Initiator™ Sixty. Microwave reaction was performed in a sealed tube.
Irradiation was automatically adjusted by the microwave software to maintain
a constant 100 °C reaction temperature.