4-Pyridylanilinothiazoles that selectively target von Hippel - Lindau deficient renal cell carcinoma cells by inducing autophagic cell death

Article abstract of DOI:10.1021/jm901457w

Renal cell carcinomas (RCC) are refractory to standard therapy with advanced RCC having a poor prognosis; consequently treatment of advanced RCC represents an unmet clinical need. The von Hippel-Lindau (VHL) tumor suppressor gene is mutated or inactivated in a majority of RCCs. We recently identified a 4-pyridyl-2-anilinothiazole (PAT) with selective cytotoxicity against VHL-deficient renal cells mediated by induction of autophagy and increased acidification of autolysosomes. We report exploration of structure-activity relationships (SAR) around this PAT lead. Analogues with substituents on each of the three rings, and various linkers between rings, were synthesized and tested in vitro using paired RCC4 cell lines. A contour map describing the relative spatial contributions of different chemical features to potency illustrates a region, adjacent to the pyridyl ring, with potential for further development. Examples probing this domain validated this approach and may provide the opportunity to develop this novel chemotype as a targeted approach to the treatment of RCC. 2009 American Chemical Society.

Full text of DOI:10.1021/jm901457w

J. Med. Chem. 2010, 53, 787–797 787
DOI: 10.1021/jm901457w
4-Pyridylanilinothiazoles That Selectively Target von Hippel-Lindau Deficient Renal Cell Carcinoma
Cells by Inducing Autophagic Cell Death
Michael P. Hay,*,† Sandra Turcotte,‡ Jack U. Flanagan,† Muriel Bonnet,† Denise A. Chan,‡ Patrick D. Sutphin,‡
Phuong Nguyen,‡ Amato J. Giaccia,‡ and William A. Denny†
†Auckland Cancer Society Research Centre, Faculty of Medical and Health Sciences, The University of Auckland, Private Bag 92019, Auckland,
New Zealand and ‡Department of Radiation Oncology, Stanford University School of Medicine, Stanford, California 94305
Received October 1, 2009
Renal cell carcinomas (RCC) are refractory to standard therapy with advanced RCC having a poor
prognosis; consequently treatment of advanced RCC represents an unmet clinical need. The von
Hippel-Lindau (VHL) tumor suppressor gene is mutated or inactivated in a majority of RCCs. We
recently identified a 4-pyridyl-2-anilinothiazole (PAT) with selective cytotoxicity against VHL-deficient
renal cells mediated by induction of autophagy and increased acidification of autolysosomes. We
report exploration of structure-activity relationships (SAR) around this PAT lead. Analogues
with substituents on each of the three rings, and various linkers between rings, were synthesized
and tested in vitro using paired RCC4 cell lines. A contour map describing the relative spatial
contributions of different chemical features to potency illustrates a region, adjacent to the pyridyl ring,
with potential for further development. Examples probing this domain validated this approach and may
provide the opportunity to develop this novel chemotype as a targeted approach to the treatment of
RCC.
Introduction
We recently developed a synthetic lethal18,19 screen to
identify small molecules that are selectively cytotoxic to
cells that have lost the VHL tumor suppressor gene.20
Subsequently, we have screened several diverse chemical
libraries and identified a number of “hits”. Hit compounds
were validated in a growth inhibition assay, and the most
promising compounds were further characterized by clono-
genic survival assays. We demonstrated the pyridylanili-
nothiazole (PAT) chemotype (Figure 1) to be active in this
context and found that compounds of this chemotype
induced autophagy in both cell lines but led to cell death
in a VHL-dependent fashion.21 The lead molecule, STF-
62247 (11) (Table 1), reduced tumor growth in vivo in VHL-
deficient RCC compared to genetically matched cells that
express wild-type VHL.21
Renal cell carcinomas (RCCsa) represent a small but
significant fraction of adult malignancies1 and are refractory
to standard chemotherapy and radiotherapy. Up to 45% of
patients present locally advanced or metastatic disease,2 and
these patients have an extremely poor prognosis.3 The current
standard of care, immunotherapy using interferon or inter-
leukin-2, has modest success with responses in a small fraction
of patients with metastatic RCC.4 Although recent antian-
giogenic therapies such as sunitinib (Sutent) and sorafenib
(Nexavar) haveprovided modest responses, itremainsunclear
which patients will have durable responses. This, combined
with the observation of stable or increasing patient mortality,5
defines the treatment of advanced RCC as an unmet clinical
need.
In this study, we have synthesized a range of analogues
(1-106) with substituents on each ring of the PAT chemotype
as well as modifications to the linker units between the three
domains. We have examined their in vitro activity as VHL-
selectivecytotoxinsusing pairedrenalcellcarcinoma celllines:
the VHL-null cell line RCC4 and the matched line RCC4-
VHL, which has VHL function restored. These well charac-
terized cell lines are isogenic with the exception of VHL
status.8 The selectivity of the analogues was determined as
the ratio (RCC4-VHL GI50)/(RCC4 GI50). We have exam-
ined the structure-activity relationships (SAR) for VHL-
selective cytotoxicity. A quantitative analysis of these SAR
was used todevelop a commonmolecular alignment for VHL-
selective members of the chemotype, and this alignment was
used to predict cytotoxicity and assessed as a tool to guide
further analogue synthesis. Initial attempts to use this tool to
direct synthesis are reported.
The von Hippel-Lindau (VHL) tumor suppressor gene is
inactivated in ∼75% of RCCs.6,7 This gene encodes for an E3
ubiquitin ligase that specifically regulates the protein stability
of a family of transcription factors called hypoxia inducible
factors (HIF).8 Loss of VHL results in constitutive expression
of HIF leading to increased transcription of genes involved in
metabolism, angiogenesis, invasion and metastasis, and pro-
liferation leading to an aggressive phenotype.9-13 In addition
to its role in HIF regulation, pVHL has been implicated in a
variety of HIF-independent processes.14-17
*To whom correspondence should be addressed. Phone: 64-9-
9236598. Fax: 64-9-3737502. E-mail: m.hay@auckland.ac.nz.
a Abbreviations: RCC, renal cell carcinoma; VHL, von Hippel-
Lindau; HIF, hypoxia-inducible factor; PAT, pyridylanilinothiazole;
SAR, structure-activity relationship; CoMFA, comparative molecular
field analysis; PLS, partial least-squares.
r
2009 American Chemical Society
Published on Web 12/08/2009
pubs.acs.org/jmc

788 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 2
Hay et al.
Table 2. GI50 Values and Selectivity Ratios for Substituted 4-PATs
Figure 1
compd
R1
2-Fl
R2 RCC4 GI50,a μM
ratiob
Table 1. GI50 Values and Selectivity Ratios for Substituted 4-PATs
33
34
35
36
37
6-Fl
>40
>40
6.5
ndc
ndc
6.4
>9
10
2-Me
3-Me
3-Me
3-CH2-
6-Me
4-Me
5-Me
4.5
2.0
CH2CH2-4
38
39
40
41
42
43
3-OH
3-OH
3-Me
4-Me
4-OMe
4-OH
12
4.6
ntd
2.3
1.9
ndc
ndc
ntd
70
compd
R1
RCC4 GI50,a μM
ratiob
1
H
7
>10
>3.1
>1.2
ndc
>2.8
>1.4
ndc
ndc
>3
17
3-OMe
3-OMe
3-OCH2O-4
4-OMe
5-OMe
24
2
2-OH
2-OMe
2-Me
2-Fl
12.8
33.4
>40
14.3
27.6
>40
>40
28.5
3
>40
>40
3
4
5
44
45
3-Cl
3,4,5-OMe3
4-Cl
13.5
>40
>3
ndc
6
2-Cl
7
2-CF3
2-NO2
3-OH
3-OMe
3-Me
3-Et
a Single determinations in triplicate. b Ratio = (RCC4-VHL GI50)/
(RCC4 GI50). c nd, not determined because of low solubility. d nt, not
tested.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
2.1
2.5
19
9.6
Scheme 1a
3-iPr
3-tBu
3-F
17
1.4
>40
>40
3
ndc
ndc
5
3-Cl
3-Br
12.4
>40
>40
>40
5.3
>3.2
ndc
ndc
ndc
8
3-CN
3-CF3
3-NO2
4-OH
4-OMe
14
60
3.6
1.3
a Reagents: (i) Br2, HOAc/HBr; (ii) NH4SCN, aqueous HCl; (iii)
PhCOCl, NH4SCN, acetone, then aqueous NaOH; (iv) EtOH. $ =
commercially available.
4-OCH2CH2NMe2
4-Me
2.5
20
16
4-CH2CH2OH
4-CH2CH2NMe2
4-F
5
50
1.2
6
38
>7
>2
ndc
ndc
ndc
ntd
Scheme 2a
4-Cl
4-CN
>40
>40
>40
ntd
4-CF3
4-NO2
4-NH2
a Single determinations in triplicate. b Ratio = (RCC4-VHL GI50)/
(RCC4 GI50). c nd, not determined because of low solubility. d nt, not
tested.
Chemistry
a Reagents: (i) MsCl, i-Pr2NEt, DCM; (ii) Me2NH, DMF; (iii) H2,
Pd/C, EtOH.
Synthesis. Initially, we synthesized a set of analogues
(1-45) with substituents on the aniline ring (Tables 1 and
2). These analogues were conveniently prepared using the
Hantzsch thiazole synthesis, condensing pyridylbromoke-
tones with phenylthioureas (Scheme 1). 4-Pyridyl-2-bromo-
ethanone 109 was readily prepared from 4-acetylpyridine
using Br2 in HBr/HOAc. Thioureas 110-152 were commer-
cially available or were prepared from the corresponding
aniline. Direct reaction of the appropriate aniline and
NH4SCN in acidic solution readily provides thioureas, albeit
in low to moderate yields. For less accessible (or more
expensive) anilines, reaction of benzoyl isocyanate and ani-
line to give corresponding benzoyl thioureas followed by
hydrolysis provided higher yields.22 Elaboration of the pri-
mary alcohol 25 to the dimethylamine 26, and reduction of
the nitroaniline 31 to aminoaniline 32, was accomplished
using standard chemistries (Scheme 2).
Modification of the linker between the thiazole and aniline
rings was examined. Alkylation at the 2-amino position of 11

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 2 789
Scheme 3a
a Reagents: (i) NaH, RI, DMF; (ii) NH2CSNH2, EtOH; (iii) benzoic acid, (COCl)2, DMF, DCM; (iv) 3-methylbenzaldehyde, THF; then NaBH4, H2O.
Scheme 4a
a Reagents: (i) Br2, HOAc/HBr; (ii) EtOH; (iii) HCHO, Et3N, THF. $ = commercially available.
Scheme 5a
a Reagents: (i) MnO2, benzene; (ii) Ph3PdCHCO2Me, DCM; (iii) H2, Pd/C, EtOH,EtOAc; (iv) NaOH, EtOH; (v) LiAlH4, THF; (vi) MsCl, i-Pr2NEt,
DCM; (vii) morpholine, DMF.
gave the 2-alkylaminothiazoles 46-48 (Scheme 3). Reaction
of bromoketone 109 with thiourea gave the 2-aminothiazole
153, which was reacted with a series of acid chlorides to
give the amides 49-52. Reductive amination of 153 with
3-methylbenzaldehyde gave the secondary amine 53.
Bromination of 4-pyridylpropanone gave the bromide
154, which was coupled with thioureas 120 and 144 to give
5-methylthiazoles 54 and 55, respectively (Scheme 4). For-
mylation of pyridylthiazoles (1, 11, 24, and 37) gave the
corresponding 5-hydroxymethyl analogues 56-59. Oxida-
tion of hydroxymethylthiazole 57 gave aldehyde 60, which
underwent Wittig reaction to give a mixture of E and Z
isomers of the unsaturated ester 61 (Scheme 5). Reduction
of 61 gave the ester 62, which was hydrolyzed to the acid 63
and then reduced to the propanol 64. Mesylation and reaction
with morpholine gave the morpholide 65. The hydroxy-
methylthiazole 57 was further elaborated to a series of solu-
bilized thiazole analogues 66-69 (Scheme 6). Bromination of
ethyl isonicotinoylacetate gave the bromoketoester 155 which
was condensed with thioureas 120 and 144 to give 5-ethyles-
ters 70 and 71, respectively (Scheme 7). Hydrolysis of ester 70
and coupling with a variety of amines gave the amides 73-76.

790 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 2
Hay et al.
Isomeric 2- and 3-pyridyl analogues 77-82 were readily
prepared by reaction of the bromides 156 and 157 with
thioureas 110, 120, and 130 (Scheme 8). Modification of
the linker between the pyridine and thiazole rings was
examined. The bromides 160 and 163 could be conveniently
prepared from the 2- and 3-pyridylethanones by an extension
of a regioselective bromination/debromination strategy.23
Conversion of 3- or 4-pyridineacetic acid to their Weinreb
amides, 158 and 161, and then to the corresponding methyl
ketones, 159 and 162, with subsequent bromination and
condensation with thiourea 120 gave pyridylmethylthiazoles
83 and 84 with extended linker chains.
Finally, we sought to elaborate the pyridine ring to explore
the SAR for this domain. Thus, reaction of amine 11 with
MCPBA gave the N-oxide 85 (Scheme 9). In most instances
preparation of the Weinreb amides of 4-pyridinecarboxylic
acids and Grignard reaction with MeMgBr gave 4-etha-
nones,24,25 which were brominated and condensed with 120
to give the corresponding PATs (Schemes 9 and 10). The
2-methoxypyridine acid was converted to the acid chloride
and treated with TMSCH2N2, and the intermediate diazo-
ketone was treated with HBr to give the bromoketone 164
(Scheme 9). The 2-Me analogue 167 was obtained by reaction
of 2-picoline N-oxide with Me2SO4 and KCN to give the
nitrile 165. Grignard reaction gave the ketone 166 which was
brominated to give 167. Elaboration of the 2-Br PAT 90
using Stille chemistry [Pd(PPh3)4, allyltributyltin] proved to
be unsuccessful, but the application of Sonogashira chem-
istry was effective. Thus, reaction of bromide 90 with TMS-
acetylene in the presences of PdCl2(PPh3)2 and CuI, followed
by removal of the TMS protecting group with TBAF, gave
the 2-alkyne 92 in moderate yield. Reduction of the acetylene
group gave the 2-ethyl analogue 93. A similar sequence with
propargyl alcohol gave the 2-substituted propynol 94 and the
corresponding propanol 95. A series of 3-substituted pyri-
dines 96-100 were synthesized via Weinreb amide chemistry
(Scheme 10). The 3-nitro PAT 100 was reduced to the
3-amino PAT 101 with subsequent acetylation giving acet-
amide 102. Sonogashira chemistry was used to elaborate the
3-bromo PAT 99 to a series of 3-alkynyl and 3-alkyl PATs
103-106.
Scheme 6a
In preliminary experiments to explore modeling predic-
tions, we synthesized two additional analogues. Songashira
reaction of bromide 99 gave the methyl ether 107, while
Suzuki reaction of bromide 99 gave the phenyl ketone 108.
QSAR Modeling
Alignment Generation. A minimum energy conformation
for 11 was obtained using the OMEGA2 package (Openeyes),
a Reagents: (i) MsCl, i-Pr2NEt, DCM; (ii) amine, DMF.
Scheme 7a
a Reagents: (i) Br2, CHCl3; (ii) EtOH; (iii) NaOH, EtOH; (iv) (COCl)2, DMF, DCM; (v) amine, DCM. $ = commercially available.
Scheme 8a
a Reagents: (i) Br2, HOAc/HBr; (ii) arylthiourea, EtOH; (iii) MeNHOMe, EDCI, HOBT, Et3N, DCM; (iv) MeMgBr, THF. $ = commercially
available.

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 2 791
Scheme 9a
a Reagents: (i) MCPBA, DCM; (ii) (COCl)2, DMF, DCM; (iii) TMSCH2N2, THF, then HBr; (iv) arylthiourea, EtOH; (v) Me2SO4, then KCN, aq
EtOH; (vi) MeMgBr, THF; (vii) Br2, HOAc/HBr; (viii) Me(OMe)NH HCl, EDCI, HOBT, Et3N, DCM; (ix) PdCl2(PPh3)2, CuI, TMSCtCH, Et3N,
3
DMF, then TBAF; (x) H2, Pd/C, EtOH; (xi) PdCl2(PPh3)2, CuI, HCtCCH2OH, Et3N, DMF. $ = commercially available.
Scheme 10a
a Reagents: (i) Me(OMe)NH HCl, EDCI, HOBT, Et3N, DCM; (ii) MeMgBr, THF; (iii) Br2, HOAc/HBr; (iv) 3-methylphenylthiourea 118, EtOH;
3
(v) H2, Pd/C, EtOH; (vi) Ac2O, dioxane; (vii) PdCl2(PPh3)2, CuI, TMSCtCH, Et3N, DMF, then TBAF; (ix) PdCl2(PPh3)2, CuI, HCtCCH2OH, Et3N,
DMF; (x) PdCl2(PPh3)2, CuI, HCtCCH2OMe, Et3N, DMF; (xi) PdCl2dppf, K2CO3, CH3COPhB(OH)2, toluene, EtOH, H2O, DMF. $ =
commercially available.
with default settings. The mmff94s force field was used for
both construction and torsion driving, while the rms torsion
driving parameter was set at 0.6. All other structures were
generated from this conformation using the SKETCHER
module in SYBYL8.0 (Tripos) and minimized using MAX-
MIN2, with the Tripos force field and Gasteiger-Huckel
charges. Minimization was performed using 1000 steps of step
descents followed by conjugate gradients until convergence at
˚
the 0.05 kcal/(mol A) level. A distance dependent dielectric
3
function was used with a dielectric constant of 80. Where the
compound could have more than one substitution pattern, all
conformers were constructed. Molecules were superimposed
on the basis of overlap of molecular volume using the ROCS
alignment package (Openeyes) at default settings.

792 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 2
Hay et al.
3D-QSAR Analysis. A comparative molecular field ana-
lysis (CoMFA)26 3D-QSAR method was used to determine a
relationship between the compounds tested and in vitro
cytotoxicity. Briefly, CoMFA was performed using a cubic
lattice extending 4 A around the aligned molecules with a
˚
point spacing of 2 A. A distance dependent dielectric along
with cutoffs of 30 kcal/mol and smooth transition were used
for both steric and electrostatic energy calculations. A C.3
atom with charge þ1 was used as the probe. The partial least-
squares (PLS) regression analysis was performed using the
QSAR module, with the optimum number of components
extracted from the standard error of predictions as deter-
mined using the SAMPLS protocol with leave-one-
out cross-validation. The optimum number of components
was used in the conventional nonvalidated PLS ana-
lysis. Only compounds that had both GI50 and cytoto-
xicity ratios and substitutions up to one rotatable bond
were used (41 compounds; see Supporting Information for
set).
˚
Biological Assays. Growth inhibition was determined by
GI50 assays, using a 4-day drug exposure of RCC4 and
RCC4/VHL cells in 96-well plates as described previously.21
The selectivity ratio was calculated as the intraexperiment
ratio RCC4-VHL/RCC4 (Tables 1-5). Selected compounds
(11, 37, 98, and 105) were tested in a clonogenic survival
assay to assess their ability to kill RCC4 and RCC4/VHL
cells in vitro. Three hundred cells were plated in 60 mm
dishes and allowed to attach overnight. The next day, cells
were treated with different concentrations of drug and
colonies were stained 10 days later. These compounds were
also examined for their ability to induce formation of
intracytoplasmic vacuoles; thus, cells were incubated with
drug (5 μM) for 24 h before phase contrast images were
taken.
Table 3. GI50 Values and Selectivity Ratios for Linkers
compd
R1
Y
RCC4 GI50,a μM
ratiob
46
47
48
49
50
51
52
53
3-Me
3-Me
3-Me
H
NMe
NEt
28
40
0.6
0.6
ndc
ndc
ndc
ndc
ndc
ndc
NCH2CH2OH
NHCO
NHCO
>40
>40
>40
>40
>40
>40
3-Me
4-Me
4-OMe
3-Me
NHCO
NHCO
NHCH2
Results and Discussion
a Single determinations in triplicate. b Ratio = (RCC4-VHL GI50)/
(RCC4 GI50). c nd, not determined because of low solubility.
Wehave prepared a set of PAT analogues (1-106) designed
to explore the SAR for each of the domains of the chemotype.
Table 4. GI50 Values and Selectivity Ratios for Thiazole Substituents
compd
R1
Me
R2
X
RCC4 GI50,a μM
ratiob
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
H
Me
Me
10
12
>4
1.5
3-CH2CH2CH2-4
H
Me
H
H
CH2OH
CH2OH
CH2OH
CH2OH
CHO
30
>1.3
2.4
1.2
7
14
H
CH2CH2CH2
4-Me
6.8
15
1.4
Me
Me
H
H
H
H
H
H
H
H
H
H
H
ndc
2.8
CHdCHCO2Me
CH2CH2CO2Me
CH2CH2CO2H
CH2CH2CH2OH
CH2HC2CH2Nmorph
CH2NMe2
2.5
>40
ntd
3.6
5.8
16
Me
Me
ndc
ntd
Me
Me
3.7
1.6
Me
Me
2.3
1.4
CH2NEt2
13
13
Me
Me
CH2N-piperidine
CH2N-morpholine
CO2Et
1.9
2.5
>40
>40
ntd
19.2
16
4.8
Me
ndc
ndc
ntd
CH2CH2CH2
Me
Me
CO2Et
CO2H
H
H
H
H
H
CONH2
CONMe2
>2
>2.5
ndc
ndc
Me
Me
CONHCH2CH2NMe2
CONHCH2CH2CH2N-morph
>40
>40
Me
a Single determinations in triplicate. b Ratio = (RCC4-VHL GI50)/(RCC4 GI50). c nd, not determined because of low solubility. d nt, not
tested.

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 2 793
Table 5. GI50 Values and Selectivity Ratios for A Ring Substituents
compd
ring
2-pyridyl
R1
R2
RCC4 GI50,a μM
ratiob
77
78
H
H
35.1
37
0.9
0.9
2-pyridyl
2-pyridyl
3-pyridyl
3-pyridyl
3-pyridyl
3-pyridyl-CH2
4-pyridyl-CH2
4-pyridyl-N-oxide
4-pyridyl
4-pyridyl
4-pyridyl
4-pyridyl
4-pyridyl
4-pyridyl
4-pyridyl
4-pyridyl
4-pyridyl
4-pyridyl
4-pyridyl
4-pyridyl
4-pyridyl
4-pyridyl
4-pyridyl
4-pyridyl
4-pyridyl
4-pyridyl
4-pyridyl
4-pyridyl
4-pyridyl
4-pyridyl
4-pyridyl
H
3-Me
4-OH
H
79
80
H
36
0.7
ndc
H
>40
>40
>40
>40
20.3
>40
28.7
>40
4.5
81
82
H
3-Me
4-OH
3-Me
3-Me
3-Me
3-Me
3-Me
3-Me
3-Me
3-Me
3-Me
3-Me
3-Me
3-Me
3-Me
3-Me
3-Me
3-Me
3-Me
3-Me
3-Me
3-Me
3-Me
3-Me
3-Me
3-Me
3-Me
3-Me
ndc
H
ndc
83
84
H
ndc
H
24.4
ndc
85
86
H
2-OMe
2-Me
2-F
2-Cl
2-Br
2,6-Cl2
1.2
ndc
87
88
>8.9
3.8
4.3
89
90
2.2
2.0
91
92
15
2.3
2-CtCH
2-Et
17.8
6.4
>2.2
>6.3
1.6
93
94
2-CtC-CH2OH
2-CH2CH2CH2OH
3-Me
9.3
5.2
95
96
>7.7
>7.7
>20
>13.8
>10
4.1
5.2
97
98
3-F
3-Cl
2
2.9
99
3-Br
3-NO2
3.9
8.2
100
101
102
103
104
105
106
107
108
3-NH2
3-NHAc
>40
33.3
1.35
6.2
ndc
1.0
4.4
3-CtCH
3-Et
2.8
3-CtC-CH2OH
3-CH2CH2CH2OH
3-CtC-CH2OMe
3-Ph-4-Ac
0.24
6.8
11.9
>5.9
45
0.62
3.9
9
a Single determinations in triplicate. b Ratio = (RCC4-VHL GI50)/(RCC4 GI50). c nd, not determined because of low solubility.
Thus, a range of substituents on each of the aniline, thiazole,
and pyridyl rings were synthesized, as well as analogues with
modified linkers between the three rings. The ability of these
analogues to inhibit cell growth in a VHL-dependent manner
was investigated by determining the GI50 values against a pair
of renal cell carcinoma cell lines: the VHL-null cell line RCC4
and the matched line RCC4-VHL, which has VHL function
restored. The selectivity of the analogues was determined as
the ratio (RCC4-VHL GI50)/(RCC4 GI50). In many instances
poor aqueous solubility prevented the determination of a GI50
value against the RCC4-VHL cell line, leading to a greater
value being expressed for the selectivity ratio. In the instances
where inequalities were obtained for both cell lines the ratio
was not determined.
The parent compound 1 was modestly potent with a GI50 of
7 μM and selectivity for the VHL-negative RCC4 cell line of
>10-fold (Table 1). Addition of substituents to the 2-position
of the aniline ring (2-8) was associated with a loss of potency
and selectivity. Compounds with small, lipophilic substituents
at the 3-position, such as the initial lead 11 with a 3-Me
substituent, were active against the RCC4 cell line with
low micromolar cytotoxicity and 10- to 20-fold selectivity.
Increasing steric bulk in the 3-position (e.g., 12-14) was
associated with a loss of potency and selectivity for RCC4.
Polar substituents (e.g., 18-20) resulted in inactive com-
pounds. A similar theme was observed at the 4-position, where
compounds with small lipophilic substituents such as 4-Me 24
and 4-F 27 had low micromolar potency, but increasing
polarity at this position was associated with a loss of activity
(e.g., 29-31). Although an extended neutral side chain was
tolerated in compound 25, more polar analogues with basic
amine side chains (23, 26) were less potent and not selective.
The SAR for multiply substituted anilines (33-45)
mirrored that seen for single substituents (Table 2). 2,
6-Disubstitution was not tolerated, whereas dialkyl substi-
tuted compounds spanning the 3-, 4-, and 5-positions (35-37)
had potencies similar to those of their monosubstituted coun-
terparts but displayed lower selectivity. More polar combina-
tions (38-45) resulted in diminished cytotoxic potency and a
loss of selectivity.
Replacement of the aniline NH with an N-alkyl substituent
(46-48) or conversion to an amide (49-52) or benzylamine
(53) consistently abolished activity, suggesting a key binding
interaction for the aniline NH (Table 3).
We next focused on substituents at the 5-position of the
thiazole ring (Table 4). Although small alkyl substituents

794 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 2
Hay et al.
(54-59) gave compounds with only slightly reduced cytotoxic
potency, they suffered a complete loss of selectivity. Longer
chain alkyl substituents with terminal polar groups (61-69)
generally allowed retention of cytotoxic potency but resulted
in low selectivity, suggesting off-target activity. Analogues
with more polar ester or amide substituents (71, 73-76) were
generally neither potent nor selective. Thus, substitution on
the thiazole ring was not a fruitful endeavor.
model development were tested. The data for the best model
are presented in Table 6, and the putative bioactive con-
formation is shown in Figure 2 along with steric (Figure 2a)
and electrostatic (Figure 2b) field contours. Although the
data are limited in the GI50 range they cover, and the low q2
value indicates that the predictive value of this model is
also limited, the data do provide some support for a relation-
ship between this conformation of the PAT core and cyto-
toxicity and indicate that steric fields account for 65.6% of
the variation in the data while electrostatic fields explain
34.4%.
Figure 2 shows that within the model development com-
pound set, the contouring around the aniline ring is more
complex than around the pyridine ring. A large blue contour
around the 3-position of the aniline ring indicates that
electrostatic effect around this site is favored while it is
disfavored around the 4- and 5-positions. Only a region
close to the 5-position of this ring reported a positive effect
on GI50 for steric bulk, while additional steric bulk around
the 4-position is likely to have a negative effect. Compounds
23, 25, and 26 were not included in the QSAR analysis
because of their flexibility; however, they probe this region
with both steric bulk and electrostatic properties. The GI50
values for the basic compounds increase, a result in part
consistent with the disfavored electrostatic contour around
this site; by contrast the smaller alkylhydroxyl side chain is
accepted at this position with only a modest loss of activity,
also supporting the yellow steric contour. Strikingly, the
contour maps report the positive effect of steric bulk at the
3-position of the pyridine ring, while electrostatic effects
around this region are likely to have a negative effect on
GI50.
Maintaining or increasing the selective cytotoxicity for
RCC4 is important to the development of this series of
compounds; thus, we overlaid those compounds with ratios
of 10 or more (including 12, ratio 9.6) (Figure 2). Features
common to compounds with high selectivity include the
positive steric contour, green volume, at the pyridine ring
3-position while avoiding interaction with an adjacent dis-
favored electrostatic red region. In addition, selective com-
pounds also had increased steric bulk around the 4- and
5-positions of the aniline ring while avoiding an interaction
with the disfavored electrostatic region at the aniline 3- and
4-positions. Substituents on the pyridine ring provide the
greatest potential for the definition of a new ligand domain.
We then tested the ability of the alignment to guide
synthesis by exploring the steric pocket adjacent to the
3-position of the pyridyl ring. Two simple analogues of
the existing series were prepared. Thus, synthesis of the
Regioselectivity in the pyridine ring was explored with a
series of 2- and 3-pyridyl analogues (77-82) (Table 5). These
were all inactive as were 3- and 4-pyridyl analogues (83, 84)
with a methylene spacer to the thiazole ring. Masking the
4-pyridyl moiety as the N-oxide (85) also ablated activity.
Collectively, these data point to a key binding interaction
for the 4-pyridyl moiety. Halide substituents at the 2-position
of the pyridyl ring (88-91) gave potent analogues with
low selectivity. Although the 2-methyl analogue 86 was
inactive, higher analogues (93, 95) were potent and selective
compounds. Incorporation of the acetylene unit (e.g., 92
and 94, respectively) resulted in compounds with lower
potency and a loss of selectivity. The 3-halo substituted
analogues (97-99) were potent and selective analogues,
whereas more polar substituents (100-102) resulted in lower
potency and a loss of selectivity. The 3-alkyl substitutions
(96, 104, and 106) generated good potency and modest
selectivity, and interestingly, the inclusion of an acetylene unit
(103, 105) resulted in a clear increase in cytotoxic potency and
selectivity.
3D-QSAR Model for Cell Cytotoxicity for the PAT Series.
Collectively, the SAR data provided inferences for further
analogue development, but in the absence of information on
the molecular target of the chemotype we sought to generate
a more complete description of a possible bioactive confor-
mation for the PAT chemotype by using CoMFA to assess
the relationship between structure and differential cell killing
ability using VHL selectivity.
Alignments that reflected the conformer possibilities for
the substitutions present in the set of compounds used for
Table 6. Statistical Analysis of the QSAR Model
SAMPLS statistics
LOO cross-validated q2 (CV correlation coefficient)
N (no. of components)
0.344
3
SEP (standard error of prediction)
non-cross-validated r2 (correlation coefficient)
SEE (standard error of estimate)
F (F ratio)
0.434
0.825
0.225
57.991
0.656
0.344
steric
electrostatic
Figure 2. (a) 3D-QSAR conformer model relating PAT conformation to cytotoxicity. Electrostatic fields are contoured at 80% favored (blue)
and 20% disfavored (red). (b) Steric fields are contoured at 80% favored (green) and 20% disfavored (yellow).

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 2 795
Table 7. Induction of Autophagy by PATs in RCC4 Cells
3-propargyl methyl ether 107 was achieved using a Sonoga-
shira reaction with bromide 99 (Scheme 10). Similarly,
Suzuki coupling of bromide 99 with 4-acetylphenylboronic
acid gave the ketone 108. The ether 107 retained VHL
cytotoxic potency [GI50(RCC4) = 0.62 μM] and was more
selective (45-fold) when compared to the propargyl alcohol
105 (11-fold), confirming the potential use of the steric
pocket to access improved PAT analogues. The use of a
phenyl ring to probe the steric space was less successful, with
108 displaying low micromolar potency [GI50(RCC4) = 3.9
μM] and modest selectivity (9-fold). The presence of this
steric pocket may be useful when designing affinity chroma-
tography reagents suitable for protein purification and
identification studies.
We evaluated several of the more active compounds from
the SAR set to confirm that their mechanism of action was
the same as the initial lead 11. Compounds 37, 98, and 105
were tested for their ability to suppress clonogenic survival
and were shown to selectively kill RCC4 cells compared to
RCC4/VHL (see Supporting Information for survival
curves). Compounds 37, 98, 105,and 107 were also examined
for their ability to induce autophagy (Table 7). Treatment
with the compounds (5 μM) for 24 h initiates the formation
of intracytoplasmic vesicles in a similar manner to 11, which
in combination with the clonogenic data confirms the mode
of action for these analogues as VHL-selective induction of
autophagy resulting in cell death.
to further explore the chemotype and develop a predictive
model. CoMFA analysis suggested the alignment had low
predictive power; however, the alignment did consistently
interpret all the data. The alignment model also suggested
the presence of a steric pocket adjacent to the 3-position of the
pyridine ring that might provide a new domain for analogue
development. Preliminary efforts to explore this pocket were
successful with two analogues displaying increased VHL-
selective cytotoxicity. The activity of these additional analo-
gues demonstrates the potential of the alignment to identify
new domains to explore in the search for more potent and
selective VHL-dependent cytotoxins. Further studies are cur-
rently underway to utilize the QSAR around this chemotype
to improve potency and selectivity, as well as to develop tools
to explore the novel mechanism of action of the class. This will
enable further development of this novel targeted approach to
the treatment of renal cell carcinoma.
Experimental Section
Chemistry. General experimental details are described in the
Supporting Information. All final products were analyzed by
reverse-phase HPLC (Altima C18 5 μm column, 150 mm ꢀ
3.2 mm; Alltech Associated, Inc., Deerfield, IL) using an Agilent
HP1100 equipped with a diode-array detector. Mobile phases
were gradients of 80% acetonitrile/20% H2O (v/v) in 45 mM
ammonium formate at pH 3.5 and 0.5 mL/min. Purity was
determined by monitoring at 330 ( 50 nm and was >95% in all
cases.
Example of Synthetic Methods. See Supporting Information
for full experimental details.
Conclusions
We have synthesized an exploratory SAR compound set to
investigate the chemical space around the PAT chemotype.
Empirical observation identified several key features neces-
sary for selective cytotoxicity against VHL-negative renal
carcinoma cells. The aniline-NH and the 4-pyridyl-N are
likely to be involved in hydrogen bonding to the as yet
undetermined molecular target, while an unsubstituted thia-
zole is required for VHL selectivity. The activities of
compounds with small lipophilic substituents at the 3- and
4-positions of the aniline ring are suggestive of the presence of
a lipophilic pocket in the molecular target. A subset of VHL-
selective compounds was used to create a molecular alignment
Method A. Preparation of Pyridyl-2-bromoethanones. Br2 (80
mmol) was added dropwise to a stirred a solution of acetylpyri-
dine (80 mmol) in 30% HBr/HOAc (100 mL) at 15 °C. The
mixture was stirred at 40 °C for 1 h and then at 75 °C for 1 h. The
mixture was cooled to 20 °C, diluted with Et2O (400 mL), and
stirred for 30 min. The precipitate was filtered, washed with
Et2O (25 mL), and dried under vacuum to give the bromoetha-
none as the hydrobromide salt.
Method B. Preparation of Thioureas. NH4SCN (50 mmol) was
added to a stirred solution of m-toluidine (50 mmol) in 1 M HCl
(50 mL) at 100 °C and the solution stirred at 100 °C for 16 h. The
solution was diluted with water (60 mL) and stood at 5 °C for
2 h. The precipitate was filtered, washed with water (5 mL),

796 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 2
Hay et al.
washed with ether (5 mL), and dried. The precipitate was
purified by column chromatography, eluting with an appropri-
ate blend (30-100%) of EtOAc/petroleum ether, to give the
thiourea.
Acknowledgment. The authors thank Dr. Maruta Boyd
and Sisira Kumara for technical assistance. This work was
supported by the U.S. National Cancer Institute under Grant
CA82566 (M.P.H., P.D.S.) and Grant CA123823 (D.A.C.),
by Canadian Institutes of Health Research (S.T.), and by the
Auckland Division of the Cancer Society of New Zealand
(W.A.D.).
Method C. Alternative Preparation of Thioureas. Benzoyl
chloride (65 mmol) was added dropwise to a solution of
NH4SCN (69 mmol) in dry acetone (50 mL). The mixture was
stirred at reflux temperature for 15 min. The heating was
removed, and m-toluidine (46 mmol) was added dropwise
keeping the solution at reflux. The mixture was stirred at reflux
temperature for a further 30 min, cooled to 20 °C, and poured
onto ice. The mixture was stirred for 30 min, and then the
precipitate was filtered and washed with water (15 mL) and
dried. The solid was added to a solution of NaOH (10% w/v, 100
mL) at 80 °C and stirred at 80 °C for 30 min, then cooled to 20 °C
and poured onto ice/HCl (6 M, 50 mL). The pH of the mixture
was adjusted to 10 with concentrated NH3 solution and stirred
for 30 min. The precipitate was filtered and washed with water
(20 mL) and dried. The precipitate was purified by column
chromatography, eluting with an appropriate blend (30-100%)
of EtOAc/petroleum ether, to give the thiourea.
Method D. Preparation of Thiazoles. A mixture of bromoke-
tone hydrobromide (3 mmol) and phenylthiourea (3 mmol) in
EtOH (20 mL) was stirred at reflux temperature for 1 h. The
mixture was cooled to 20 °C and diluted with water (50 mL), the
pH adjusted to ∼8 with aqueous NH3, and the mixture stirred
at 20 °C for 2 h. The precipitate was filtered, washed with water
(5 mL), and dried. The solid was purified by column chroma-
tography, eluting with an appropriate blend (30-100%) of
EtOAc/petroleum ether, to give the thiazole.
Supporting Information Available: Experimental details and
characterization data for synthetic intermediates 109-190 and
PATs 1-108. This material is available free of charge via the
Internet at http://pubs.acs.org.
References
(1) Jemal, A.; Siegel, R.; Ward, E.; Hao, Y.; Xu, J.; Thun, M. J. Cancer
statistics, 2009. Ca;Cancer J. Clin. 2009, 59, 225–249.
(2) Chow, W. H.; Devesa, S. S.; Warren, J. L.; Fraumeni, J. F., Jr.
Rising incidence of renal cell cancer in the United States. JAMA, J.
Am. Med. Assoc. 1999, 281, 1628–1631.
(3) Motzer, R. J.; Mazumdar, M.; Bacik, J.; Berg, W.; Amsterdam, A.;
Ferrara, J. Survival and prognostic stratification of 670 patients
with advanced renal cell carcinoma. J. Clin. Oncol. 1999, 17, 2530–
2540.
(4) Nelson, E. C.; Evans, C. P.; Lara, P. N., Jr. Renal cell carcinoma:
current status and emerging therapies. Cancer Treat. Rev. 2007, 33,
299–313.
(5) Murai, M.; Oya, M. Renal cell carcinoma: etiology, incidence and
epidemiology. Curr. Opin. Urol. 2004, 14, 229–233.
(6) Gnarra, J. R.; Tory, K.; Weng, Y.; Schmidt, L.; Wei, M. H.; Li, H.;
Latif, F.; Liu, S.; Chen, F.; Duh, F.-M.; Lubensky, I.; Duan,
D. R.; Florence, C.; Pozzatti, R.; Walther, M. M.; Bander,
N. H.; Grossman, H. B.; Brauch, H.; Pomer, S.; Brooks, J. D.;
Isaacs, W. B.; Lerman, M. I.; Zbar, B.; Linehan, W. M. Mutations
of the VHL tumour suppressor gene in renal carcinoma. Nat.
Genet. 1994, 7, 85–90.
Method E. Preparation of Pyridylamides. Et3N (6.0 mmol)
was added to a stirred suspension of 4-pyridinylacetic acid (1.5
mmol), EDCI (1.7 mmol), HOBT H2O (1.7 mmol), and MeN-
3
HOMe HCl (2.3 mmol) in dry DCM (50 mL), and the mixture
3
(7) Kim, W. Y.; Kaelin, W. G. Role of VHL gene mutation in human
was stirred at 20 °C for 16 h. The resulting solution was diluted
with DCM (100 mL), washed with water (2 ꢀ 30 mL), washed
with brine (30 mL), and dried, and the solvent was evaporated.
The residue was purified by column chromatography, eluting
with a gradient of an appropriate blend (0-10%) of MeOH/
EtOAc, to give the amide.
cancer. J. Clin. Oncol. 2004, 22, 4991–5004.
(8) Maxwell, P. H.; Wiesener, M. S.; Chang, G. W.; Clifford, S. C.;
Vaux, E. C.; Cockman, M. E.; Wykoff, C. C.; Pugh, C. W.; Maher,
E. R.; Ratcliffe, P. J. The tumour suppressor protein VHL targets
hypoxia-inducible factors for oxygen-dependent proteolysis. Nat-
ure 1999, 399, 271–275.
(9) Gnarra, J. R.; Zhou, S.; Merrill, M. J.; Wagner, J. R.; Krumm, A.;
Papavassiliou, E.; Oldfield, E. H.; Klausner, R. D.; Linehan, W. M.
Post-transcriptional regulation of vascular endothelial growth
factor mRNA by the product of the VHL tumor suppressor gene.
Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 10589–10594.
(10) Iliopoulos, O.; Levy, A. P.; Jiang, C.; Kaelin, W. G., Jr.; Goldberg,
M. A. Negative regulation of hypoxia-inducible genes by the von
Hippel-Lindau protein. Proc. Natl. Acad. Sci. U.S.A. 1996, 93,
10595–10599.
(11) Knebelmann, B.; Ananth, S.; Cohen, H. T.; Sukhatme, V. P.
Transforming growth factor alpha is a target for the von Hippel-
Lindau tumor suppressor. Cancer Res. 1998, 58, 226–231.
(12) Staller, P.; Sulitkova, J.; Lisztwan, J.; Moch, H.; Oakeley, E. J.;
Krek, W. Chemokine receptor CXCR4 downregulated by von
Hippel-Lindau tumour suppressor pVHL. Nature 2003, 425,
307–311.
(13) Erler, J. T.; Bennewith, K. L.; Nicolau, M.; Dornhofer, N.; Kong,
C.; Le, Q. T.; Chi, J. T.; Jeffrey, S. S.; Giaccia, A. J. Lysyl oxidase is
essential for hypoxia-induced metastasis. Nature 2006, 440, 1222–
1226.
(14) Ohh, M.; Yauch, R. L.; Lonergan, K. M.; Whaley, J. M.; Stemmer-
Rachamimov, A. O.; Louis, D. N.; Gavin, B. J.; Kley, N.; Kaelin,
W. G., Jr.; Iliopoulos, O. The von Hippel-Lindau tumor suppres-
sor protein is required for proper assembly of an extracellular
fibronectin matrix. Mol. Cell 1998, 1, 959–968.
Method F. Preparation of Pyridyl Ketones. A solution of
MeMgBr (3 M in hexanes, 7.5 mmol) was added slowly to a
stirred solution of amide (5 mmol) in dry THF (50 mL) at 0 °C,
and the mixture was stirred at 0 °C for 8 h. The reaction mixture
was quenched with saturated aqueous NH4Cl solution (20 mL)
and the mixture extracted with EtOAc (3 ꢀ 50 mL). The com-
bined organic fraction was washed with water (40 mL), washed
with brine (40 mL), and dried, and the solvent was evaporated.
The residue was purified by column chromatography, eluting
with 30% EtOAc/petroleum ether, to give ketone.
GI50 Assays. GI50 values for compounds were determined by
XTT assay. For 2,3-bis[2-methoxy-4-nitro-5-sulfophenyl]-2H-
tetrazolium-5-carboxanilide inner salt (XTT) assays, 5000 cells
were plated in 96-well plates. The next day, vehicle or drug was
added to each well and incubated for 4 days. The medium was
aspirated, and phenol red-free medium with 0.3 mg/mL XTT
(Sigma Chemical Co) in phenol red-free medium, 20% FCS,
and 2.65 μg/mL N-methyldibenzopyrazine methyl sulfate was
added. The cells were incubated at 37 °C for 1-2 h, and
absorbance was read at 450 nm. GI50 values for each compound
were calculated using linear interpolation.
(15) Hergovich, A.; Lisztwan, J.; Barry, R.; Ballschmieter, P.; Krek, W.
Regulation of microtubule stability by the von Hippel-Lindau
tumour suppressor protein pVHL. Nat. Cell Biol. 2003, 5, 64–70.
(16) Okuda, H.; Saitoh, K.; Hirai, S.; Iwai, K.; Takaki, Y.; Baba, M.;
Minato, N.; Ohno, S.; Shuin, T. The von Hippel-Lindau tumor
suppressor protein mediates ubiquitination of activated atypical
protein kinase C. J. Biol. Chem. 2001, 276, 43611–43617.
(17) Na, X.; Duan, H. O.; Messing, E. M.; Schoen, S. R.; Ryan, C. K.;
di Sant’Agnese, P. A.; Golemis, E. A.; Wu, G. Identification of the
RNA polymerase II subunit hsRPB7 as a novel target of the von
Hippel-Lindau protein. EMBO J. 2003, 22, 4249–4259.
Clonogenic Assays. Three hundred cells were plated into
60 mm tissue culture dishes in DMEM. The next day, cells were
treated with vehicle or drug and were further incubated for an
additional 10 days. After 10 days, the medium was removed and
colonies were fixed and stained in 95% ethanol and 0.1% crystal
violet for 15 min. The stain was removed, and plates were
washed in deionized water. Colonies were quantified. All con-
ditions were measured in triplicate, and all experiments were
performed in triplicate.

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 2 797
(18) Sutphin, P. D.; Chan, D. A.; Giaccia, A. J. Dead cells don’t
form tumors: HIF-dependent cytotoxins. Cell Cycle 2004, 3,
160–163.
(19) Kaelin, W. G., Jr. The concept of synthetic lethality in the context
of anticancer therapy. Nat. Rev. Cancer 2005, 5, 689–689.
(20) Sutphin, P. D.; Chan, D. A.; Li, J. M.; Turcotte, S.; Krieg, A. J.;
Giaccia, A. J. Targeting the loss of the von Hippel-Lindau tumor
suppressor gene in renal cell carcinoma cells. Cancer Res. 2007, 67,
5896–5905.
(21) Turcotte, S.; Sutphin, P. D.; Chan, D. A.; Hay, M. P.; Denny,
W. A.; Giaccia, A. J. A novel molecule targeting VHL-deficient
renal cell carcinoma that induces autophagic cell death. Cancer Cell
2008, 14, 90–102.
M. J.; Powell, E. T.; Molinari, A. J. Improved procedures for the
preparation of cycloalkyl-, arylalkyl-, and arylthioureas. Synthesis
1988, 456–459.
(23) Choi, H. Y.; Chi, D. Y. Non-selective bromination-selective deb-
romination strategy: selective bromination of unsymmetrical ke-
tones on singly activated carbon against doubly activated carbon.
Org. Lett. 2003, 5, 411–414.
(24) Nahm, S.; Weinreb, S. M. N-Methoxy-N-methylamides as effective
acylating agents. Tetrahedron Lett. 1981, 39, 3815–3818.
(25) Singh, J.; Satamurthi, N.; Aidhen, I. S. The growing synthetic
utility of Weinreb’s amide. J. Prakt. Chem. 2000, 342, 340–347.
(26) Cramer, R. D.; Patterson, D. E.; Bunce, J. D. Comparative
molecular field analysis (CoMFA): effect of shape on binding of
steroids to carrier proteins. J. Am. Chem. Soc. 1988, 110, 5959–
5967.
(22) Rasmussen, C. R.; Villani, F. J., Jr.; Weaner, L. E.; Reynolds, B. E.;
Hood, A. R.; Hecker, L. R.; Nortey, S. O.; Hanslin, A.; Costanzo,