Structure-activity relationship in a purine-scaffold compound series with selectivity for the endoplasmic reticulum Hsp90 paralog Grp94

Article

pubs.acs.org/jmc

Structure−Activity Relationship in a Purine-Scaﬀold Compound

Series with Selectivity for the Endoplasmic Reticulum Hsp90 Paralog

Grp94

Hardik J. Patel,^†Pallav D. Patel,^†,‡Stefan O. Ochiana,^†Pengrong Yan,^†Weilin Sun,^†Maulik R. Patel,^†

Smit K. Shah,^†Elisa Tramentozzi,^§James Brooks,^∥Alexander Bolaender,^†Liza Shrestha,^†Ralph Stephani,^‡

Paola Finotti,^§Cynthia Leifer,^∥Zihai Li,^⊥Daniel T. Gewirth,^#Tony Taldone,*^,†and Gabriela Chiosis*^,†

^†Molecular Pharmacology and Chemistry Program, Sloan-Kettering Institute, New York, New York 10021, United States

^‡Department of Pharmaceutical Sciences, College of Pharmacy and Allied Health Professions, St. John’s University, Jamaica, New

York 11439, United States

^§Department of Pharmacology and Anesthesiology, University of Padua, Largo E. Meneghetti 2, 35131, Padua, Italy

^∥Department of Microbiology and Immunology, Cornell University, Ithaca, New York 14850, United States

^⊥Department of Microbiology and Immunology, Medical University of South Carolina, Charleston, South Carolina United States

^#Hauptman-Woodward Medical Research Institute, Buﬀalo, New York 14203, United States

S

* Supporting Information

ABSTRACT: Grp94 is involved in the regulation of a restricted number of

proteins and represents a potential target in a host of diseases, including cancer,

septic shock, autoimmune diseases, chronic inﬂammatory conditions, diabetes,

coronary thrombosis, and stroke. We have recently identiﬁed a novel allosteric

pocket located in the Grp94 N-terminal binding site that can be used to design

ligands with a 2-log selectivity over the other Hsp90 paralogs. Here we perform

extensive SAR investigations in this ligand series and rationalize the aﬃnity and

paralog selectivity of choice derivatives by molecular modeling. We then use this to

design 18c, a derivative with good potency for Grp94 (IC₅₀= 0.22 μM) and

selectivity over other paralogs (>100- and 33-fold for Hsp90α/β and Trap-1,

respectively). The paralog selectivity and target-mediated activity of 18c was

conﬁrmed in cells through several functional readouts. Compound 18c was also

inert when tested against a large panel of kinases. We show that 18c has biological

activity in several cellular models of inﬂammation and cancer and also present here

for the ﬁrst time the in vivo proﬁle of a Grp94 inhibitor.

Grp94 has recently also emerged as an attractive target in

cancer. Clinically, Grp94 expression correlates with advanced

stage and poor survival in a variety of cancers and is closely

linked to cancer growth and metastasis in melanoma, breast

cancer, ovarian cancer, multiple myeloma, lung cancer, and

inﬂammation-associated colon cancer.^2,7,8It also is linked to

tumor resistance, where Grp94 expression in tumors decreases

sensitivity to X-rays and its suppression sensitizes cells to

chemotherapy.^2,7While the exact role for Grp94 in cancer

remains elusive, its function as an essential chaperone for TLRs,

integrins, GARP, and Wnt coreceptor LRP5/6 is suggestive of a

potential link between Grp94 and aberrant signaling via these

receptors. Evidence also points to Grp94’s involvement at the

ER level in regulating a number of tyrosine kinases, such as

EGFR and HER2, and in regulating the secretion of IGF-I and

IGF-II.^2,7,8Finally, the recent discovery that tumor-speciﬁc

INTRODUCTION

■

Glucose-regulated protein 94 (Grp94),¹also known as gp96,

ERp99, and endoplasmin, is the endoplasmic reticulum (ER)

paralog of the Hsp90 family of molecular chaperones. In

addition to Grp94, the members of this family include the

cytosolic Hsp90α, Hsp90β, and the mitochondrial tumor

necrosis factor receptor-associated protein-1 (Trap-1). While

highly homologous, with over 50% sequence identity in their

N-terminal regulatory domains and analogous ligand binding

cavities, it is now recognized that each paralog has a unique

biochemical role and is responsible for chaperoning a distinct

set of client proteins.²⁻⁴

Sustained medical interest in cytoplasmic Hsp90 has come

from the discovery that these chaperones play important roles

in maintaining the functional conformation of a large number

of malignancy- and neurodegeneration-driving proteins, such as

kinases, transcription factors, and antiapoptotic proteins.³In

addition, Trap-1 along with Hsp90 maintains mitochondrial

integrity by protecting against oxidative stress and apoptosis.^5,6

Received: February 3, 2015

Published: April 22, 2015

3922

DOI: 10.1021/acs.jmedchem.5b00197

J. Med. Chem. 2015, 58, 3922−3943

Journal of Medicinal Chemistry

Article

Figure 1. A schematic depiction of (a) 1, (b) 2 and 3, and the amino acids present in site 3 of Grp94 (adapted from refs 11, 14, and17); (c) 4 and

the amino acids present in the allosteric site 2 of Grp94 (adapted from ref 10); and (d) 5 (pan-Hsp90 inhibitor) and the amino acids found in site 1

of Grp94. The conserved adenine binding pocket is shown in orange. The adenine moiety is shown in blue, the quinone in green, and the resorcinol

core in red.

translocation of Grp94 to the plasma membrane regulates the

altered expression and function of membrane-associated cancer

proteins, such as HER2, and its role in sustaining the

transforming ability of these proteins point to a third avenue

for Grp94’s involvement in cancer.^9,10

conformation due to a 5-amino acid (QEDGQ) insertion in its

helix 1,4,5-subdomain. This conformation exposes a cavity

adjacent to the adenine binding pocket that accommodates the

5′-N-ethylcarboxamido moiety of NECA (depicted as site 3 in

Figure 1a). In yeast Hsp90, the analogous cavity is blocked by

the backbone carbonyl oxygen of Gly121, which suggests that

this is the origin of NECA binding selectivity.¹¹This cavity is

not utilized by bound nucleotides, as modeling shows that both

α- and β-phosphates of ADP experience severe steric clashes

with the carbonyl oxygen of Gly196 in Grp94. Instead, binding

of nucleotides induces large conformational changes in helices

1,4,5 in Grp94 that enlarge the ATP binding cavity and allow

for nucleotide binding.¹³Unfortunately, attempts to develop

more potent analogs of NECA have been unsuccessful.¹⁴While

NECA has served an important role in substantiating the

feasibility of developing Grp94-selective inhibitors, its use as a

probe to investigate Grp94 function suﬀers from its potent

agonist activity against several cellular adenosine receptors.¹⁵

Radamide (RDA, 2) (Figure 1b) is a chimeric molecule

composed of elements from both geldanamycin and radicicol

and is the second ligand reported to bind into the small binding

pocket unique to Grp94 (site 3, Figure 1b).^14,16X-ray crystal

structures of RDA bound to the N-terminal domain of either

canine Grp94 (PDB ID: 2GFD) or yeast Hsp90 (PDB ID:

2FXS) reveal that it adopts distinct poses when bound to either

of the two paralogs. While the resorcinol moiety binds similarly

into the adenine binding cavity of each paralog, the quinone

ring is oriented diﬀerently. In Grp94, the quinone ring is

capable of binding in two distinct poses, one of which orients

the quinone ring into the same binding pocket that

accommodates the 5′-N-ethylcarboxamido moiety of NECA

Because Grp94 is involved in the regulation of a restricted

number of proteins involved in channeling immune and

inﬂammatory signals, it has been proposed that, in addition

to cancer, numerous other diseases, including septic shock,

autoimmune diseases, chronic inﬂammatory conditions, dia-

betes, coronary thrombosis, and stroke, would beneﬁt from

targeting Grp94 with small-molecule inhibitors.^2,7,8Most

important from a therapeutic perspective, the chaperoning

activity of Grp94 is ATPase-dependent and uncoupled from its

ability to bind peptides, a role of Grp94 linked to the

immunogenic activity of Grp94/peptide complexes. Together,

these ﬁndings highlight a potential therapeutic role for small

molecules that bind selectively to Grp94 and inhibit its

chaperoning function.

The rational development of selective Grp94 inhibitors has

been challenging due to a high degree of sequence conservation

in the ATP binding pockets of the four Hsp90 paralogs (Figure

S1, Supporting Information). Nevertheless, screening identiﬁed

5′-N-ethylcarboxamidoadenosine (NECA, 1) (Figure 1a) as a

high-aﬃnity ligand with selectivity for Grp94.^11,12NECA has

sub-micromolar (K_d= 200 nM) binding aﬃnity for Grp94,

while having no aﬃnity for cytoplasmic Hsp90s.^11,12This

serendipitous discovery was initially puzzling given the fact that

the amino acids surrounding the adenine binding cavity in

Grp94 and Hsp90 are almost entirely conserved. However,

further examination revealed that Grp94 adopts a distinct

3923

DOI: 10.1021/acs.jmedchem.5b00197

J. Med. Chem. 2015, 58, 3922−3943

Journal of Medicinal Chemistry

Article

Figure 2. (a) Schematic depiction of the pose adapted by ligands when accessing the two binding sites, site 2 being speciﬁc to Grp94 and site 1 being

common to Grp94 and Hsp90. (b) Pose adapted by 4 when bound to Grp94 showing the hydrophobic stabilization of the 8-aryl group with the

hydrophobic residues of site 2 and the pose adapted by 5 when bound to Hsp90. (c) Surface area of the Grp94-selective ligands reported in ref 10

(left) and of the pan-Hsp90 ligand 5 (right) (surface area was generated with Macromodel/Molecular Surface). (d) Proposed modiﬁcation sites on

the purine scaﬀold performed to understand the structure−activity relationship in this series (as shown in blue).

that is unique to Grp94 (site 3, Figure 1a,b) and another that

orients it toward the solvent. In Hsp90, the quinone ring is

blocked from binding into the analogous site 3 pocket by Lys98

and instead adopts the solvent-oriented pose. Though RDA

interacts with the two paralogs in a very distinct manner and is

capable of binding into the unique pocket of Grp94, it is not

very selective for Grp94 (K_d= 0.52 μM) over Hsp90 (K_d= 0.87

μM).¹⁴Since the quinone moiety of RDA is joined to the

resorcinol via an amide linkage, upon binding to Grp94 it can

adopt either the cis-amide conformation to bind into the

hydrophobic pocket unique to Grp94 or the trans-amide

conformation to orient toward the outside of the pocket, similar

to that adopted when bound to Hsp90. Realizing that this

conformational ﬂexibility was the likely cause for the lack of

selectivity of RDA, Duerfeldt et al.¹⁷rationalized that

compounds of greater selectivity for Grp94 might be obtained

by restricting the orientation of RDA to the cis-conformation.

Furthermore, observing that the pocket was lined by Phe199,

Tyr200, and Trp223 (Figure 1b) made them seek to enhance

its hydrophobicity and potential for π−π interactions. Utilizing

imidazole as a cis-bioisostere to direct a more hydrophobic

benzyl moiety into the pocket provided among others

compound BnIm (3, Figure 1b). While BnIm was shown to

have an IC₅₀= 1.15 μM for Grp94 in a ﬂuorescent polarization

(FP) competitive binding assay using FITC-GM as a probe,¹⁸

similar competition experiments for each of the other isoforms

were not performed.

3 (i.e., site 3, Figure 1b).¹⁷In addition to its distinct spatial

orientation, site 2 is signiﬁcantly more hydrophobic compared

to site 1. Small-molecule purine-based ligands that utilize this

novel pocket exhibited greater than 100-fold preference for

Grp94 over Hsp90α/β and a 10−100-fold preference over

Trap-1.¹⁰The crystal structure of Grp94 in complex with the

selective ligand 4 (Figure 1c) showed that while the purine

moiety of 4 occupied the site ﬁlled by the adenine ring of ATP

and of the pan-Hsp90 inhibitor 5 (PU-H71; Figure 1d), the 8-

aryl group of Grp94-selective compounds inserted into site 2

when bound to Grp94 (Figure 1c).¹⁰Selectivity for Grp94 over

Hsp90 is achieved because the analogous site 2 is blocked in

Hsp90. Not all ligands of this purine-scaﬀold class can access

the site 2 pocket (Figure 2a−c). Those ligands that can access

site 2 must be capable of adopting a “backward bent”

conformation, whereas other ligands such as 5 adopt a “forward

bent” conformation and bind to site 1 with no particular

selectivity for the paralogs.¹⁰

LIGAND DESIGN, COMPUTATIONAL ANALYSES OF

GRP94−LIGAND INTERACTION, AND

BIOCHEMICAL TESTING

■

While our earlier work provided the rationale for the observed

paralog selectivity, the detailed SAR of these purine-based

inhibitors has not been described. Here, we discuss the design

and synthesis of purine-based ligands that preferentially bind to

site 2, which is unique to Grp94. Speciﬁcally, we report a

tractable SAR of purine-based ligands that bind selectively and

with varied degrees of aﬃnity to the novel allosteric site 2

found only in Grp94. We focused our synthetic eﬀorts on the

exploration of various substituents on the C8-aryl ring (Figure

2d, substituents 2′−5′), the linker between the purine and the

aryl moiety (Figure 2d, L), as well as at the N9/N3 position of

the adenine core. We also explore the possibility of exchanging

the N3 with a carbon atom, a modiﬁcation claimed to improve

Recently, we have reported the discovery of a novel,

therapeutically viable allosteric pocket in Grp94 (i.e., site 2,

Figure 1c) whose occupancy led to the inhibition of Grp94’s

chaperoning activity.¹⁰The newly discovered site 2, which is

composed of an almost completely hydrophobic cleft lined by

Leu104, Leu163, Phe199, Ala202, Phe203, Val211, Ile247, and

Leu249, is distinct from the pocket previously reported for 1

(i.e., site 3, Figure 1a)¹¹and the resorcinolic N-benzylimidazole

3924

DOI: 10.1021/acs.jmedchem.5b00197

J. Med. Chem. 2015, 58, 3922−3943

Journal of Medicinal Chemistry

Article

a

Scheme 1

a

Reagents and conditions: (a) neocuproine, CuI, NaOt-Bu or Na₃PO₄or K₂CO₃, DMF, 110 °C; (b) pent-4-yn-1-yl 4-methylbenzenesulfonate,

Cs₂CO₃, DMF, 80 °C; (c) Cl(CH₂)_nCCH or Cl(CH₂)_nCN or Br(CH₂)_nCF₃or Br(CH₂)_nC₆H₅, Cs₂CO₃, DMF, rt or 80 °C; (d) pent-4-yn-2-yl 4-

methylbenzenesulfonate, Cs₂CO₃, DMF, 80 °C; (e) ROH, PPh₃, DBAD, CH₂Cl₂, toluene, rt.

a

Scheme 2

a

Reagents and conditions: (a) m-CPBA, 30 min, rt; (b) P(OPh)₃, pyridine, microwave 220 °C, 30 min; (c) 5-chloropent-1-yne, Cs₂CO₃, DMF, 80

°C; (d) 2 equiv of Cs₂CO₃, DMF, 80 °C.

aﬃnity for Hsp90 in the pan-Hsp90 inhibitor series.^19,20The

design of these ligands is based on the resolved crystal structure

of PU-H54 (4) with Grp94 (Figure 2b) and on the determined

volume available for ligand binding inside site 2 (Figure 2c).

The aﬃnities of the synthesized compounds were determined

using competitive binding assays to recombinant Grp94

protein.^21,22To determine the selectivity proﬁle of our

compounds, we also tested them for binding against puriﬁed

Hsp90α/β and Trap-1, using a published assay protocol.²²

Further, to rationalize the aﬃnity trends noted for these

molecules and understand how the nature and size of

substituents aﬀect Grp94 binding, we used the reported X-ray

crystal structures of PU-H54 (4) bound to the N-domain of

Grp94 (PDB ID: 3O2F)¹⁰to dock compounds into the

allosteric binding site (site 2).

method previously described by us to generate key

intermediates 8a−r,t−w (Scheme 1).²³Brieﬂy, 8-mercaptoa-

denine (6) was reacted with various aryl iodides (7) using CuI/

neocuproine as catalyst and NaOt-Bu (or Na₃PO₄or K₂CO₃)

as base in anhydrous DMF at 110 °C to generate 8-arylsulfanyl

adenine intermediates (8a−r,t−w). Alkylation of 8a−r,t−w

with pent-4-yn-1-yl 4-methylbenzenesulfonate in the presence

of Cs₂CO₃yielded a mixture of N9-isomers (9a−r,t−w) and

N3-isomers (1 and 9a′−d′,g′,j′−n′,q′,v′−w′)²⁴in roughly 2:1

ratio, which were separated by preparative TLC. The synthesis

of 9s was carried out through a route that was previously

described²⁵from the reaction of 8-bromo-9-(pent-4-yn-1-yl)-

9H-purin-6-amine²⁵with 5-mercaptoisophthalonitrile²⁶in the

presence of t-BuOK in DMF at 130 °C.

The synthesis of compounds 13a−s was carried out by

alkylation of intermediate 8g or 8n with the appropriate alkyl

alcohols under Mitsunobu conditions or alkyl halides or

tosylates in the presence of Cs₂CO₃to yield the desired N9-

isomers after chromatographic separation (Scheme 1).

CHEMISTRY

■

8-Arylsulfanyl derivatives containing the N9-pent-4-yne chain

were synthesized by utilizing a copper-catalyzed cross-coupling

3925

DOI: 10.1021/acs.jmedchem.5b00197

J. Med. Chem. 2015, 58, 3922−3943

Journal of Medicinal Chemistry

Article

a

Scheme 3

a

Reagents and conditions: (a) 3-(tert-butoxycarbonyl-isopropyl-amino)propyl tosylate, Cs₂CO₃, DMF, rt; (b) TFA, 0 °C; (c) ArI (7), neocuproine,

CuI, NaOt-Bu or Na₃PO₄or K₂CO₃, DMF, 110 °C; (d) TFA, 80 °C, 3 h; (e) 1,3-dibromopropane, Cs₂CO₃, DMF, 1.5−3h; (f) isopropylamine,

DMF, rt.

Table 1

IC₅₀(μM)

compd

R₁

Hsp90α

>100

Hsp90β

>100

Grp94

Trap-1

9a

9b

9c

9d

9e

9f

2′-Cl

0.35 0.05

6.2 2.5

29.4 9.1

1.4 0.2

1.13 0.2

1.5 0.2

0.12 0.04

0.39

ND

2′-OCH₃

2′-CH₂OH

2′-OCF₃

>100

>300

>50

>100

ND

4′-OCH₃

4′-Cl

9g

9h

9i

2′,4′-di-Cl

>300

ND

145.4 13.4

ND

2′-Cl,4′-Br

2′,4′-di-CH₃

2′,4′-di-OCH₃

2′,5′-di-Cl

2′,5′-di-CH₃

2′-Cl,5′-CF₃

3′,5′-di-Cl

3′-Cl,5′-Br

3′,5′-di-Br

3′,5′-di-CH₃

2′,6′-di-Cl-pyridine

3′,5′-di-CN

3′,4′-di-Cl

>300

>50

>300

>50

1.63 0.47

34.5 3.5

0.53 0.10

1.14 0.14

0.5 0.2

0.26 0.01

0.24

13.8 1.8

ND

9j

9k

9l

60.4 1.6

>300

>100

>500

>100

>50

>100

>300

>100

>500

>100

ND

5.1 1.8

203.7 11.0

ND

9m

9n

9o

9p

9q

9r

9s

9t

83.6 11.8

ND

0.29

ND

>300

8.3

>300

ND

11.45 0.92

1.9

157.9 9.7

9.6

>50

ND

7.73

ND

1.96

ND

0.695

ND

9u

9v

9w

5

2′,3′,4′-tri-Cl

2′,4′,5′-tri-Cl

2′,4′,6′-tri-CH₃

15.97

>200

>500

0.043 0.006

39.47

>200

>500

0.042 0.003

0.430

3.60

0.17 0.11

2.1 0.56

0.026 0.004

78.4 8.9

65.5 1.3

0.205 0.006

3926

DOI: 10.1021/acs.jmedchem.5b00197

J. Med. Chem. 2015, 58, 3922−3943

Journal of Medicinal Chemistry

Article

Figure 3. Predicted binding poses for 9g, 9g′, and 9i when bound to the N-terminal domain of Grp94 (PDB ID: 3O2F).

The synthesis of analogs of 9g with varying linker groups is

shown in Scheme 2. Treatment of 9g with 2 equiv of m-CPBA

at rt yielded a mixture of sulfoxide 14a and sulfone 14b, in

which each were readily isolated in 23% and 34% yield,

respectively. The O-linked derivative 14c was obtained from

reaction of 8-bromo-9-(pent-4-yn-1-yl)-9H-purin-6-amine²⁵

with 2,4-dichlorophenol and KOt-Bu in 51% yield. Micro-

wave-assisted cyclocondensation of 4,5,6-triaminopyrimidine

(15) with 2,4-dichlorophenylacetic acid (16) yielded 8-

benzyladenine 17 in 79% yield.²⁷Heating a mixture of 17

with 1-chloropent-4-yne and 1.2 equiv of Cs₂CO₃at 80 °C

yielded 14d, which could be converted to carbonyl derivative

14e in 50% yield by heating at 80 °C with additional Cs₂CO₃.

Compounds having an ionizable 3-isopropylamino-propyl

chain were synthesized by alkylation of 8-arylsulfanyl adenine

intermediates 8g,i,n with 3-(tert-butoxycarbonyl-isopropyl-

amino)propyl tosylate in the presence of Cs₂CO₃, resulting in

Boc-protected N9- and N3-regioisomers after separation.²⁸

Boc-deprotection of the N9- and N3-regioisomers in the

presence of TFA yielded desired derivatives 18a,c and 18b′,c′

(Scheme 3), respectively. Related imidazopyridines 18d−f

(Scheme 3) were prepared using modiﬁcations to a synthetic

route described earlier.²⁰PMB-protected intermediates 20d−f

were obtained from CuI/neocuproine-catalyzed coupling of

19²⁰with various aryl iodides (7), which were then deprotected

with TFA to yield 21d−f. Alkylation of 21d−f with 1,3-

dibromopropane in the presence of Cs₂CO₃followed by

reaction with isopropylamine yielded imidazopyridines 18d−f.

Interestingly, our SAR shows that the chlorine functionality is

generally the preferred substituent in any combination as

compared to other substituents explored, such as −OCH₃,

−OCF₃, −CF₃, −CN, or −CH₃(e.g., see 9a vs 9b and 9d, or

9g vs 9i, or 9n vs 9q and 9s), the only exception being

bromine, which we ﬁnd to be tolerable (see 9g vs 9h, or 9n vs

9o or 9p). Although subtle electronic eﬀects could play a role,

we surmise that the size and lipophilicity of the chlorine atom

are ideal to satisfy binding into the hydrophobic pocket site 2.

In order to obtain a better understanding of this preference, as

well as to provide further insight toward the design of more

potent ligands, we next pursued docking studies.

Indeed, in our docking analysis, chlorine at 2′-position is

oriented nicely to form hydrophobic interactions with Leu163,

Ile247, and Val211 (Figure 3a,b), which explains the higher

binding aﬃnity of 9a for Grp94 over other substituents (i.e.,

9b−d). More polar substituents, as in 9b (OCH₃), 9c

(CH₂OH), and 9d (OCF₃), are less favorably oriented toward

these hydrophobic residues and may account for the decrease in

binding aﬃnity. At the para-position, binding aﬃnity appears to

be less aﬀected by substituent changes, as both −OCH₃(9e)

and −Cl (9f) at this position result in compounds with similar

aﬃnity (Table 1). Our ﬁndings suggest that while select purine

analogs are indeed able to adopt the “backward bent”

conformation and bind into site 2, the space available in the

pocket is somewhat restricted at 2′-position, where smaller

substituents are preferred, as evidenced by the drastic reduction

in aﬃnity observed for larger substituents. Our analysis of the

eﬀects of di- and trisubstitution on Grp94 aﬃnity and selectivity

further validated chlorine as the best substituent choice for the

8-aryl moiety. Brieﬂy, dichloro substitutions at 2′,4′-, 2′,5′-, and

3′,5′-positions on the 8-phenyl ring retained high aﬃnity and

selectivity but also indicated that 2′,4′-substitution (i.e., 9g) is

preferred for higher aﬃnity, followed by 3′,5′ (9n) and 2′,5′

(9k), as shown in Table 1. Concurrently, the same positions on

the aryl ring substituted with methyl (9i, 9q, and 9l) reveal a

signiﬁcant loss of activity in the low micromolar range

compared to the nanomolar aﬃnity observed for the chlorine

isomers. Additionally, the SAR also shows that 3′,4′-

disubstitution with chlorine (9t) is not as well-tolerated as

2′,4′ (9g), 3′,5′ (9n), and 2′,5′ (9k) given the observed loss in

aﬃnity and selectivity. Substitution of one of the chlorines with

a bromine as in 2′-Cl,4′-Br (9h) and 3′-Cl,5′-Br (9o) as well as

STRUCTURE−ACTIVITY RELATIONSHIP IN THE

GRP94 INHIBITOR SERIES

■

Modiﬁcations on the 8-Aryl Ring. We began our SAR

exploration by focusing ﬁrst on substitutions on the 8-aryl ring

(Figure 2d, substituents 2′−5′) given that this region of the

molecule can potentially occupy the hydrophobic site 2. Thus,

with an interest in better understanding the polarity and

volume of this pocket, a varied set of compounds was

synthesized having substitutions on the 8-aryl ring focused

primarily on the 2′-, 4′-, 2′,4′-, 2′,5′-, and 3′,5′-positions,

though other substitutions were also explored (Tables 1 and 4).

A deleterious eﬀect on activity was observed for more polar

functionalities, as evidenced by compounds 9c and 9r, which

further conﬁrmed the hydrophobic proﬁle of the pocket.

3927

DOI: 10.1021/acs.jmedchem.5b00197

J. Med. Chem. 2015, 58, 3922−3943

Journal of Medicinal Chemistry

Article

Table 2

IC₅₀(μM)

Grp94

compd

R₁

Hsp90α

Hsp90β

Trap-1

9n

−(CH₂)₃CCH

−(CH₂)₂CCH

−(CH₂)₄CCH

−CH₂CCCH₃

−(CH₂)₂CF₃

−(CH₂)₃CF₃

−(CH₂)₄CF₃

−(CH₂)₅CF₃

−(CH₂)₃CN

−CH₂C₆H₅

−(CH₂)₂C₆H₅

−CH(CH₃)(CH₂)₂CH₃

−CH(CH₃)CH₂CCH

−CH(C₂H₅)CH₂CCH

−(CH₂)₃CH₃

−CH(CH₃)(CH₂)₂CH₃

−(CH₂)₂CCH

>500

>50

>500

>100

ND

0.26 0.01

0.287

0.665

5.1

83.6 11.8

ND

13a

13b

13c

13d

13e

13f

>50

ND

>100

>50

ND

0.56

0.502

2.74

1.648

0.429

>5

ND

>100

>50

ND

13g

13h

13i

ND

>50

4.58

ND

>50

13j

>50

>2.5

14.3

5.8

>100

ND

13l

>50

13m

13n

13o

13p

13q

13r

13s

>50

ND

>100

>50

ND

>5

ND

0.8

ND

>100

>50

ND

2.48

0.39

5.0

>50

ND

7.23

ND

−CH(CH₃)CH₂CCH

−CH(C₂H₅)CH₂CCH

>50

ND

>50

ND

0.92

>50

Table 3

IC₅₀(μM)

compd

X

Hsp90α

Hsp90β

Grp94

Trap-1

9g

S

>300

24.9

>300

ND

0.12 0.04

145.4 13.4

ND

14a

14b

14c

14d

14e

SO

OSO

O

5.4

4.7

0.5

0.7

0.34

>25

ND

>100

ND

CH₂

ND

CO

4.98

the dibrominated analog 3′,5′-di-Br (9p) shows that bromine is

well-tolerated at these positions. Combined, the available data

proposed 3′,5′ and 2′,4′ as the optimal sites for substitution on

the 8-aryl ring, yielding compounds with excellent selectivity

and potency for Grp94.

In accordance with our previous ﬁndings, these compounds

were each selective for Grp94. As mentioned above, the

substitution pattern in these compounds was biased toward

those that were known to favorably position the compound to

attain a backward bent conformation. This should not be taken

to mean that these compounds cannot attain the forward bent

conformation and potentially bind into site 1 present in Hsp90

(or Grp94 for that matter). These compounds are capable of

attaining a forward bent conformation; however, the sub-

stituents chosen are known to not be ideal for binding into site

1 through previous SAR studies with Hsp90.²⁴It can be

envisioned that for any ligand there potentially exists an

equilibrium whereby the 8-aryl group can adopt either a

3928

DOI: 10.1021/acs.jmedchem.5b00197

J. Med. Chem. 2015, 58, 3922−3943

Journal of Medicinal Chemistry

Article

Figure 4. Design of 18c and its proposed interaction with the Hsp90 paralogs. (a) Interactions of 9n and 18c with the amino acids lining the Grp94

allosteric site 2. (b) Docking of compound 18c in the allosteric site 2 of Grp94. (c) Overlay of 18c on to sites 1 and 2 of the four Hsp90 paralogs.

Residues that constrain either the availability of the site 2 pocket or the interaction of 18c with the protein are shown in red.

backward or forward bent conformation and thus occupy either

site 2 or site 1, respectively. Energy minimization calculations

show that certain ligands prefer a backward bent conformation

and are thus favored entropically, but the diﬀerence in energy

between these conformations is not enough to explain the

observed selectivity. There is also an enthalpic component

whereby stabilizing binding interactions that favor binding into

one pocket over another are also necessary. The enhanced

potency of ligands bearing more hydrophobic substituents is a

direct consequence of the greater hydrophobic character of site

2 in Grp94. Combined, these two factors explain the high

aﬃnity and selectivity these ligands display for Grp94.

Modiﬁcations of the Linker between the Purine Core

and the 8-Aryl Moiety. We next focused our eﬀorts on

examining the eﬀect on potency and selectivity of the linker

joining the purine core to the 8-aryl ring (Figure 2d). The

adenine core binds in an essentially identical manner to each

isoform, whereby a key interaction is formed between the

exocyclic −NH₂and N1 of the adenine with a conserved

Asp149 of Grp94 (Asp93 of Hsp90α/β, Asp158 of Trap1) and

a network of water molecules at the bottom of the ATP-binding

pocket.²²Since the adenine core is essentially static in binding

to each of the isoforms, it is the linker atom that is a critical

element to the ligand attaining the conformation necessary for

selective Grp94 inhibition through binding into site 2. Naturally

then, the linker group can be expected to have a signiﬁcant

impact on Grp94 binding/selectivity by impacting the dihedral

angle, which is indeed the case (Figure S2, Supporting

Information). Substitution of the S in 2′,4′-di-Cl analog 9g

with O (14c), CH₂(14d), CO (14e), SO (14a), and O

SO (14b) was investigated by the synthesis of the

corresponding analogs (Scheme 2). Substitution with O or

CH₂was tolerated, giving only a 4−6-fold decrease in activity;

however, this was not the case for the sulfoxide and sulfone

derivatives, which showed a signiﬁcant loss of activity (∼40-

fold, Table 3). Overall, the sulfur linker was found to be the

best choice, as it provided the most potent analog.

Modiﬁcations of the N3/N9-Substituents. We next

examined the eﬀect of diﬀerent N9-substituents on Grp94

binding aﬃnity and selectivity (Table 2). The appendage at the

N9-position accommodated various chain lengths up to ﬁve

carbons with a further increase resulting in a decrease in

potency. Increasing the chain length from ﬁve to six carbons

(9n vs 13b) at the N9-position resulted in approximately 3-fold

decrease in potency for Grp94; however, a decrease in chain

length to four carbons did not aﬀect binding aﬃnity (9n vs

13a). As seen in Figure 4a, the alkyne moiety of pent-4-ynyl

(9n) and but-3-ynyl (13a) orients closer to Phe195 and makes

critical π−π interactions, whereas hex-5-ynyl (13b) is unable to

form such interactions because of its orientation away from this

amino acid. As predicted by docking studies, moving the alkyne

internally led to a decrease in Grp94 binding aﬃnity (13a vs

13c, Table 2) which can be attributed to the inability of this

compound to form π−π interactions with Phe195. Replace-

ment of the alkyne with a triﬂuoromethyl group was also

extensively investigated; however, this did not oﬀer any

improvement (see 13d−g, Table 2). Similarly, isosteric

replacement of the alkyne with the more polar cyano group

(13h, Table 2), while tolerated, does not oﬀer any improve-

ment over alkyne 9n. Introduction of bulkier substituents, such

as benzyl (13i) and phenethyl (13j), result in a drastic decrease

in aﬃnity, caused most likely by steric clash with Phe195.

We had previously shown that C1′-substitution on the N9-

substituent with a methyl group was a positive factor toward

enhancing the selectivity of compounds for Grp94.¹⁰In fact,

this was one subtype of compound that was identiﬁed as

3929

DOI: 10.1021/acs.jmedchem.5b00197

J. Med. Chem. 2015, 58, 3922−3943

Journal of Medicinal Chemistry

Article

Table 4

IC₅₀(μM)

compd

R₁

Hsp90α

>100

Hsp90β

>100

Grp94

Trap-1

Hsp90α/ Grp94

9a′

9b′

9c′

9d′

9g′

4

2′-Cl

100 10

>100

ND

−

>210

>21

−

>48

>5

4

2′-OCH₃

2′-CH₂OH

2′-OCF₃

>100

>500

>250

>100

ND

>100

2′,4′-di-Cl

>500

2.37 0.79

11.77 3.24

>100

9.6 0.8

54.1 4.6

ND

2′,4′-di-CH₃

2′,4′-di-OCH₃

2′,5′-di-Cl

2′,5′-di-CH₃

2′-Cl,5′-CF₃

3′,5′-di-Cl

3′,5′-di-CH₃

2′,4′,5′-tri-Cl

2′,4′,6′-tri-CH₃

2′,4′-di-CH₃

3′,5′-di-Cl

>250

9j′

9k′

9l′

>100

35.3

>100

100

5

ND

>100

ND

9m′

9n′

9q′

9v′

9w′

18b′

18c′

5

>100

ND

>100

8

ND

>100

2.08

20

ND

>100

19.5 0.2

ND

>100

5

>100

>25

2.43

ND

>100

0.042 0.003

>100

>39

1

0.043 0.006

0.026 0.004

0.205 0.006

Table 5

IC₅₀(μM)

compd

X

R₁

Hsp90α

Hsp90β

ND

Grp94

Trap-1

18a

18c

18d

18e

18f

18g

5

N

C

N

−

2′,4′-di-Cl

3′,5′-di-Cl

2′,4′-di-Cl

3′,5′-di-Cl

I

20.3 1.2

27.3 3.5

>50

0.41 0.1

ND

41.8 1.3

ND

0.22 0.08

3.44

7.3 0.5

14.1

>50

ND

1.08

7.2

0.028

ND

0.101

ND

N(CH₃)₂

I

0.100

ND

0.265

ND

0.043 0.006

a

0.042 0.003

0.026 0.004

0.205 0.006

a

22

−

33

38

190

1586

a

Values were obtained from ref 22.

Grp94-selective with greater than 10-fold selectivity over

Hsp90α, Hsp90β and Trap1. Modeling indicated that the

presence of a methyl group at this position favorably disposed

the 8-aryl ring into the backward bent conformation. Because

the substituents used to decorate the 8-aryl moiety were not

particularly favorable for binding, these ligands were not potent

for Grp94. In contrast to Grp94 in this region, Hsp90 is

characterized by a more limited volume, and indeed, previously

reported SAR showed that C1′-substitution completely ablated

Hsp90 binding.²⁹Using this information, we had hoped to

further enhance the selectivity for Grp94 through substitution

at the C1′-position while at the same time incorporating the

3930

DOI: 10.1021/acs.jmedchem.5b00197

J. Med. Chem. 2015, 58, 3922−3943

Journal of Medicinal Chemistry

Article

favorable substituents that we show above to enhance potency

(i.e., 2′,4′-di-Cl and 3′,5′-di-Cl). Unfortunately, it was not

possible to ascertain the eﬀects of such modiﬁcation on

selectivity, since at the highest concentration tested (maximal

solubility) for these analogs (13l−n,p,r,s, Table 2) there was

essentially no inhibition of Hsp90α. However, the results do

clearly show for both the butyl and butyne series that

substitution with methyl (i.e., 13l,m,p,r) or ethyl (i.e., 13n,s)

results in a decrease in potency for Grp94 relative to the

unsubstituted analogs 13a,o,q.

To rationalize the binding aﬃnity, we docked compound 18c

into the binding site of Grp94 (Figure 4b, PDB ID: 3O2F).

Unlike the hydrophobic chain of 9n, the polar chain at N9

orients away from Phe195 (Figure 4a,b) and is predicted to

form hydrogen bonds with surface amino acids such as Glu158

in Grp94, which seems to compensate for the loss of

hydrophobic and π−π interactions with Phe195, resulting in

a similar inhibition proﬁle (18c vs 9n, Tables 5 and 1). To

determine the reason for ligand selectivity, sites were generated

by SiteMap using a previously described procedure¹⁰on each

protein, and the sites were overlaid onto the structure of 18c

(Figure 4c). In Grp94, there are two sites available for the right

side phenyl ring to orient into (sites 1 and 2). The phenyl ring

of 18c orients into energetically favorable site 2 (backward

bent) and is stabilized by hydrophobic interactions with

Leu104, Leu163, Phe199, Ala202, Phe203, Val211, Ile247,

and Leu249. Orientation of 18c into site 1 of Grp94 was less

preferred due to energy constrains and unfavorable interactions

with Trp223 (NH). The ability of compound 18c to access site

2 and favorably bind in this pocket results in the observed high

aﬃnity of this ligand for Grp94. Only site 1 is available for the

phenyl ring to orient into in Hsp90α, Hsp90β, and Trap-1

(Figure 4c) since a steric clash of the phenyl ring with Phe138

(Hsp90α/β) or Phe205 (Trap-1) would preclude binding into

site 2. This restriction as well as unfavorable interactions of the

hydrophobic substituents of 18c (3,5-di-Cl) with polar amino

acids of site 1 [i.e., Trp162 (NH) and Tyr139 (OH) in

Hsp90α/β and Tyr206 (OH) in Trap-1] results in low-aﬃnity

binding to Hsp90α/β and Trap-1 and the resulting selectivity

for Grp94.

Modiﬁcation of the Purine CoreSubstitution of N3

with C. In an attempt to explore further structural

modiﬁcations that might improve pharmacological properties,

we sought to modify the 3-position nitrogen of the purine to a

carbon, as shown in Scheme 3. Previously, such a modiﬁcation

was utilized by researchers at Curis in the development of

Hsp90 inhibitors, since they had rationalized that the N3

nitrogen lacks direct contact with Hsp90, although it does make

water-mediated hydrogen bonds with Gly97 and Asp102.³⁰The

resulting imidazopyridines are reported to bind to Hsp90 with

high aﬃnity,^19,20and one of these analogs, CUDC-305 (22;

Table 5), is in clinical trials for cancer. Since such a change

would be expected to alter the physiochemical properties, we

sought to make a similar modiﬁcation to 18a and 18c.

Somewhat surprisingly, we found that such a change

substantially diminished potency (5−8-fold for 18d and 18e)

for Grp94, indicating that the N3 is signiﬁcant for binding of

18a and 18c to Grp94. Examination of the crystal structure for

PU-H54 (4) bound to Grp94 was not much help for us in

understanding the role of N3, since this analog is N3-

substituted and would alter the H-bond capabilities of the

nitrogen by diminishing electron density as well as sterically

aﬀecting it. Therefore, we relied upon the crystal structure of

NECA (1) bound to Grp94, which shows that N3 forms water-

mediated H-bonds with Asn162 and Lys114 (Figure S3,

Supporting Information) in a manner highly similar to that

observed for PU-H71 (5) bound to Hsp90.³⁰Therefore, on the

basis of this as well as our binding results between the two

series (i.e., 18a vs 18d, and 18c vs 18e, Table 5), water-

mediated hydrogen bonds made with the ligand through N3 are

signiﬁcant for aﬃnity (Figure S3, Supporting Information).

Previously, when we had compared the binding mode of

some pan-inhibitors to the various isoforms of Hsp90, we found

In addition to the N9-substituted compounds described

above, we also examined the potency and selectivity of their

N3-substituted analogs (Table 4), which were conveniently

isolated as minor components in the syntheses of N9-

compounds. We ﬁnd that compared to the N9-series, N3-

substituted compounds were considerably less potent (Table 4

vs Tables 1 and 5). In fact, because of solubility limitations an

accurate IC₅₀could not be determined for a number of the

compounds. The compounds that we were able to evaluate,

such as 9a′, 9g′, 4, 9k′, 9n′, 9v′, 9w′, and 18c′ (Table 4),

displayed a 10−400-fold decrease in aﬃnity for Grp94.

Importantly, the SAR of the N3-isomers correlated to that of

the N9-alkylated series, since the 2′,4′- and 2′,5′-dichloro

analogs were more potent than their dimethyl analogs (9g′ vs 4

and 9k′ vs 9l′, Table 4). Molecular docking studies suggested a

distinct orientation and interaction of the N3- and N9-alkyl

chain with Grp94. While the pent-4-ynyl moiety in the N9-

position results in a bend that orients the chain closer to the

surface of the protein to form hydrophobic interactions with

Phe195 and Leu163 and π−π stacking with Phe195 (2.2 Å) of

Grp94 (e.g., 9g, Figure 3a), this is not the case for the N3-

analogs, where the pent-4-ynyl moiety is positioned further

away and orients instead to polar residues such as Lys114,

Gln194, and Asp110, accounting for the decrease in binding

aﬃnity (9g vs 9g′, Figure 3a,b).

As described above, our initial SAR focused on modiﬁcation

of the 8-phenyl ring and resulted in the identiﬁcation of 2′,4′-

and 3′,5′-dichloro-substituted analogs 9g and 9n (Table 1),

respectively, as potent and selective Grp94 inhibitors. However,

we found that these compounds had limited aqueous solubility

and were less than optimal for use in cell-based studies and,

moreover, diﬃcult to formulate for in vivo studies. Therefore,

in an attempt to improve aqueous solubility, we aimed to

modify the N9-position based on our prior knowledge of the

purine scaﬀold.²⁸Additionally, the X-ray crystal structure of

PU-H54 (4) with Grp94 reveals the presence of polar amino

acids near the exit of the pocket that could potentially form H-

bonds with a hydrophilic N9-substituent. On the basis of our

earlier experience in the development of the pan-Hsp90

inhibitor PU-H71 (5), where the polar N9-amine projects out

into the solvent, we decided to introduce the ionizable 3-

isopropylamino-propyl group.²⁸Gratifyingly for us, replace-

ment of the pent-4-ynyl chain of 9g and 9n with the ionizable

3-isopropylamino-propyl chain provided derivatives 18a and

18c, which maintained the previously observed binding aﬃnity

for Grp94, though the 3,5-dichloro analog 18c was slightly

better than the 2,4- compound 18a. As such, 18c shows good

potency for Grp94 (IC₅₀= 0.22 μM) and selectivity over other

paralogs (>100- and 33-fold for Hsp90α/β and Trap-1,

respectively, Table 5). Additionally, 18c can be formulated as

either a mesylate or hydrochloride salt, which increases its

water solubility (i.e., >10 mg/mL).

3931

DOI: 10.1021/acs.jmedchem.5b00197

J. Med. Chem. 2015, 58, 3922−3943

Journal of Medicinal Chemistry

Article

Figure 5. Grp94-speciﬁc inhibitor 18c dose-dependently inhibits several Grp94-mediated cellular eﬀects, such as (a) TLR9 expression, (b) TNFα

production in cells stimulated with CpG, (c) IGF-II production by diﬀerentiated myoblasts, and (d) HER2 degradation in HER2-overexpressing

breast cancer cells. Selectivity of 18c for Grp94 over the cytosolic Hsp90 is demonstrated by the lack of activity in cells stimulated by LPS (b) and no

Hsp70 induction in SKBr3 cells (e), both eﬀects mediated by Hsp90. PU-H71 (5) and GM are pan-Hsp90 inhibitors (i.e., bind equally well to both

Grp94 and the cytosolic Hsp90s). Data points are mean error, n = 4.

that the aﬃnity of CUDC-305 (22) for Grp94 (IC₅₀= 190 nM)

was ∼6-fold lower than for Hsp90α (IC₅₀= 33 nM) or Hsp90β

(IC₅₀= 38 nM).²²This was attributed by modeling to be due to

steric clashes that occur between the dimethylamine group with

Phe195 and Leu163 in the Grp94 binding pocket.²²The

signiﬁcance of a nitrogen or carbon at the 3-position of the

adenine ring was not considered at the time. In order to further

elucidate the eﬀects of such substitution we prepared both the

dimethylamine analog (18g) and the C3 analog (18f) of the

pan-inhibitor PU-H71 (5). We chose to make analogs of PU-

H71 (5) because of the similar aﬃnity this compound displays

for Grp94 and Hsp90α/β, which would allow us to more

carefully gauge such eﬀects. Substitution of the iodine with

dimethylamine (18g) resulted in a similar decrease (∼3-fold) in

relative aﬃnity for Grp94 that is observed with CUDC-305

(22), and when compared to PU-H71 (5), substitution results

in a 10-fold loss of aﬃnity for Grp94, supporting the notion of a

steric clash that was earlier suggested by modeling. Similarly,

the C3 analog of 5, 18f, displayed ∼4-fold lower aﬃnity for

Grp94 compared to Hsp90α and when compared to PU-H71

(5), further supporting the importance of N3 for binding to

Grp94.

Supporting Information). This panel contains a set of kinases

covering AGC, CAMK, CMGC, CK1, STE, TK, TKL, lipid,

and atypical kinase families, plus important mutant forms.

Developed by Ambit Biosciences, it employs proprietary active-

site-dependent competition binding assays to determine how

compounds bind to kinases.³¹It is based on a competition

binding assay that quantitatively measures the ability of a

compound to compete with an immobilized, active-site-directed

ligand and can be used in detection of multiple inhibitor types

(e.g., type I, type II, and non-ATP competitive). Compound

18c was inert against the kinase panel, while only two minor

interactions were identiﬁed for compound 9g (MKNK2 and

ROCK2, 34% and 32% inhibition, respectively) (Figure S4,

Supporting Information).

Biological Evaluation of Grp94 Selectivity in Cells. We

next addressed the paralog speciﬁcity of these compounds in a

cellular system because the biochemical aﬃnity and selectivity

of these compounds measured in vitro with recombinant

protein may diﬀer from their eﬀects on Hsp90 paralogs within

cells. There are multiple factors that could contribute to such a

proﬁle, including diﬀerences in the cellular abundance and

distribution of the Hsp90 paralogs. Additionally, post-transla-

tional modiﬁcations (PTMs) and complex formation of

intracellular Grp94 could potentially render a distinct inhibition

proﬁle to that observed compared to recombinant protein.³²

Thus, to conﬁrm that these compounds act in cells through

paralog-speciﬁc inhibition, we tested the selective target

modulation of compound 18c by several functional readouts.

TLRs are a family of pattern recognition receptors for

pathogens that have critical roles in innate immunity against

infection.³³Thirteen TLRs exist in mammals and they localize

BIOCHEMICAL AND BIOLOGICAL EVALUATION OF

THE GRP94 LIGANDS

■

Oﬀ-Target Selectivity Testing. Because these ligands are

based on a purine scaﬀold, there is the misconception that they

tend to nonspeciﬁcally interact with ATP-binding pockets, such

as those of kinases. Both derivative 18c and 9g were tested at

10 μM against the scanEDGE 97 kinase panel (Figure S4,

3932

DOI: 10.1021/acs.jmedchem.5b00197

J. Med. Chem. 2015, 58, 3922−3943

Journal of Medicinal Chemistry

Article

Figure 6. (a) Relative binding aﬃnity of Grp94 inhibitors for Hsp90α and Grp94 was measured using a ﬂuorescence polarization assay as previously

reported.²²(b) The HER2-overexpressing breast cancer cells SKBr3 were treated for 24 h with the indicated concentrations of Grp94 inhibitors, and

proteins were analyzed by Western blot. Representative blots of triplicate experiments are shown. (c) Viability of cancer cells treated for 72 h with

the indicated Grp94 inhibitor concentrations was measured using the CellTiter-Glo cell viability assay. Data points are mean error, n = 4.

either on the cell surface or in the endolysosome compartment.

Grp94 is one of several accessory proteins that regulate

intracellular traﬃcking of TLR9. In the absence of Grp94 or

when Grp94 chaperoning function is inhibited pharmacologi-

cally, TLR9 fails to exit the endoplasmic reticulum, and

therefore, Grp94-deﬁcient macrophages fail to recognize

unmethylated CpG DNA and induce TNF-α production.³⁴

Impairment of Grp94 function increases the proteolytic

sensitivity of TLR9, possibly because, in the absence of

Grp94, TLR9 is not folded properly and is turned over in the

ER or secretory pathways. Indeed, 18c dose-dependently

inhibited the expression of TLR9 in the post-ER compartment

(Figure 5a) and the CpG induced TNF-α production (Figure

5b). 18c also dose-dependently inhibited IGF-II secretion by

diﬀerentiated myoblasts (Figure 5c) and reduced the steady-

state levels of HER2 kinase in HER2 overexpressing breast

cancer cells (Figure 5d), both Grp94-mediated cellular

processes.^10,35

While these functional readouts conﬁrm Grp94 inhibition in

cells by 18c, they do not exclude potential interference with

Hsp90, the cytosolic paralog. Therefore, we also investigated

the ability of 18c to modulate Hsp90-speciﬁc functions (Figure

5e). One such function is related to Hsp90’s ability to maintain

the transcription factor HSF-1 sequestered to the cytosol,

where it directly binds to it and prevents its activation.³⁶When

Hsp90 is inhibited, HSF-1 is released from Hsp90, trimerizes to

an active transcription factor, and translocates to the nucleus to

bind to several promoters of heat-shock proteins, resulting in

increased transcription and synthesis of these proteins, such as

Hsp70. Indeed, inhibition of Hsp90, by either the Hsp90-

selective inhibitor PU29F¹⁰or by the pan-Hsp90 inhibitors

GM^37,38or PU-H71,^28,39led to upregulated levels of Hsp70. As

a conﬁrmation of its retained paralog speciﬁcity in cells, 18c did

not modulate this Hsp90-related event at concentrations that

resulted in Grp94 inhibition in cells.

target the N-terminal domain from those that target the C-

terminus of Hsp90.^40,41There is even considerable diﬀerence

observed in vitro and in vivo between those inhibitors that

target the N-terminal nucleotide binding pocket.⁴²⁻⁴⁴Despite

sharing a similar binding site, these inhibitors can display

signiﬁcantly diﬀerent properties, and it seems that no two

inhibitors are the same as a result of their ability to sample

diﬀerent conformations and chaperone complexes.³²

In this context, we were intrigued by the previous

observation that BnIm (3) lacked antiproliferative eﬀects

against HER2-overexpressing SKBr3 cells.¹⁷This contrasted

with our ﬁndings that inhibition of Grp94 by 18c or knock-

down of Grp94 by siRNA led to growth impairment in these

cells.¹⁰In order to directly compare the eﬀects of these two

classes of compounds, we synthesized 3 according to published

procedures.^16,17As noted above, 3 binds into site 3 (Figure 1b),

in contrast to 18c, which binds into site 2. While the IC₅₀of 3

for Grp94 is reported,¹⁸its speciﬁcity for the other Hsp90s was

not speciﬁed, so the potency and selectivity against

recombinant protein could not be clearly assessed. Although

in silico docking predicted for 3 a binding mode that oriented

the benzyl ring into the pocket of Grp94, no crystal structure

was provided to conﬁrm this hypothesis. The selectivity of 3 for

Grp94 was hypothesized on the basis of cell-based functional

assays that were interpreted to be readouts of speciﬁc Grp94 or

Hsp90α/β function, and the precise paralog selectivity/potency

proﬁle of 3 remains unclear.

Thus, we ﬁrst chose to determine both the potency and

selectivity of 3 against Grp94 and Hsp90α. We obtained an

IC₅₀of 1.2 and 14 μM for Grp94 and Hsp90α, respectively

(∼12-fold selective) (Figure 6a). The binding aﬃnity of 3 for

Grp94 is similar to that recorded for this compound (IC₅₀= 1.5

μM).¹⁸Next, we tested the eﬀect of BnIm on HER2, a protein

that we have recently shown to be dependent on Grp94 in

SKBr3 cells.¹⁰Speciﬁcally, we demonstrated that the stability

and function of HER2 at the plasma membrane of tumor cells

characterized by high HER2 expression are critically regulated

by Grp94.¹⁰Consistent with previous results,¹⁷we observed no

eﬀect of 3 in SKBr3 cells up to 10 μM, a concentration

previously reported to inhibit both TLR9 traﬃcking and IGF-II

secretion.¹⁷We increased the concentration of BnIm to 100

μM, a concentration recently shown to inhibit the chaperoning

of myocilin by Grp94.¹⁸At 100 μM, a slight eﬀect could be

seen for 3 on SKBr3 cells, as noted by minor PARP cleavage

and possibly a slight reduction in HER2 (Figure 6b) and by

cellular toxicity (Figure 6c); this, however, we could not

Distinct Grp94 Inhibitors May Result in Nonoverlap-

ping Phenotypes. Molecular chaperones are atypical drug

targets because they can exist as ﬂuid complexes. They can exist

in cells as components of complexes bound to nucleotide,

cochaperone, or client proteins and are subject to PTMs.³²

Combined these two factors help to regulate chaperone

function and enable the cell to mount a stress response as is

needed under diverse cellular conditions. These can also serve

to explain the diﬀerences observed with diﬀerent inhibitors. For

instance, there is a signiﬁcant diﬀerence observed both at the

biochemical and at the cellular level for Hsp90 inhibitors that

3933

DOI: 10.1021/acs.jmedchem.5b00197

J. Med. Chem. 2015, 58, 3922−3943

Journal of Medicinal Chemistry

Article

Figure 7. In vivo proﬁle of 18c. (a) 18c (75 mg/kg, ip) was administered to mice (n = 4 per experimental point) bearing breast tumors, and mice

were sacriﬁced at 2 and 24 h postadministration. Inhibitor levels in the indicated tissues and plasma were determined by LC−MS/MS following

administration of 18c. Each data point represents the mean error. (b) The concentrations of 18c in tumor to brain, heart, and plasma of mice as in

panel a were determined by LC−MS/MS and the ratios plotted. (c) 18c was injected as a single dose iv or oral at 10 mg/kg to B6D2F1 mice (n = 2

male mice per experimental point, 20 mice total). The agent was formulated in 30% captisol, 5 mM citrate buﬀer at pH 4.2. Agent levels (μg/g or

μg/mL) were determined by LC−MS/MS and converted to molar concentrations with the approximation that 1g = 1 mL. Concentrations were

graphed against the time of sacriﬁce, which represents the time after agent administration. Each point represents mean error.

speciﬁcally attribute to Grp94 inhibition, because at this

concentration Hsp70 induction (indicating cytosolic Hsp90

inhibition) was also noted (Figure 6b).

increased to 200:1 and the tumor vs heart to 100:1, indicating

speciﬁc retention of 18c in the tumor mass (Figure 7b). We

further validated our ﬁndings in B6D2F1 mice, in which we

conducted a “classical” pilot PK study (Figure 7c). The fraction

of agent that reached systemic circulation (F) after oral

administration was 80%, which is suggestive of good absorption

and limited ﬁrst-pass metabolism (Figure 7c). This study re-

enforces our ﬁndings in the tumor-bearing mice; speciﬁcally, we

see a high C_max, AUC, and %F (po) in the plasma in the time

interval of 0−4 h, which is in accord with the rapid and good

biodistribution we determined above (Figure 7a). These studies

are encouraging, as they indicate that we have a very good

pharmacophore and lead compound in 18c. They also suggest

that speciﬁc tumor targeting with rapid clearance from normal

tissue is also possible in the Grp94 purine-scaﬀold series,

similarly to that observed for the purine-scaﬀold pan-Hsp90

inhibitors.⁴⁶⁻⁴⁹

Through these data we cannot conclusively attribute the lack

of eﬀect for 3 in SKBr3 cancer cells on its distinct binding

interaction with Grp94; while we favor an explanation that

diverse binding modes may enable each compound to sample a

diverse range of conformations, resulting in diﬀerent

phenotypes, other mechanisms such as an inability of 3 to

access Grp94 at the plasma membrane (such as Grp94 being

translocated in SKBr3 to chaperone HER2¹⁰) might also

account for the phenotypic diﬀerences. Nonetheless, these

results demonstrate that the eﬀect of small-molecule inhibition

of Grp94 is context-dependent and suggests that future eﬀorts

in the development of Grp94 inhibitors should implement, in

addition to biochemical tests, relevant phenotypic and cell-

speciﬁc target modulation assays.

In Vivo Pharmacokinetics. So far there is no report on the

in vivo proﬁle of a Grp94 inhibitor. Studies conducted with

pan-Hsp90 inhibitors indicate a tumor-targeting and retention

proﬁle for such compounds.^32,39This was rationalized by the

existence of Hsp90 species in tumors distinct from those

expressed in normal cells.^32,42,45Speciﬁcally, the presence of

aberrant proteins, such as those present in cancer cells, may

shunt Hsp90 to a highly cochaperone bound and dependent

complex that exhibits high aﬃnity not only for its endogenous

ligand, ATP, but also for ligands that occupy the ATP pocket.

This stress-speciﬁc Hsp90 complex we termed the oncogenic

Hsp90 species.⁴²In contrast, in normal cells, Hsp90 exists in

complexes of dynamic nature that regulate normal cell

physiology, also termed housekeeping Hsp90 species.

To address the question whether a tumor-speciﬁc retention

proﬁle may also extend to the allosteric Grp94 inhibitors

described here, we performed a pilot PK study for 18c (Figure

7). No acute toxicity was observed upon intraperitoneal

administration of the agent to mice at doses as high as 150

mg/kg (higher was not tested due to solubility limitations). In

this PK study we incorporated two time points, a 2 and 24 h

time of sacriﬁce postadministration (Figure 7a). The early time

point is incorporated to test the biodistribution of the agent to

the site of its action, the tumor; 18c was readily distributed to

tumor mass with ∼850 μM noted in tumor at 2 h vs 100 μM in

the plasma. At 24 h, the ratio of agent in tumor vs plasma

CONCLUSIONS

■

This report is the ﬁrst to describe a comprehensive SAR of

Grp94 inhibitors that interact with the allosteric site 2 and to

oﬀer valuable insight into ways to overcome the high structural

similarity among Hsp90s in the ligand-binding pocket. Whereas

previous eﬀorts in the ﬁeld provided useful information on the

feasibility of achieving selectivity for this Hsp90 isoform, the

current work provides detailed knowledge on the ligand

characteristics required for selective Grp94 inhibition both in

vitro and in cells. It also presents several context-dependent

biological assays that probe binding to Grp94 in cellular models

of disease and thus provides a blueprint for future medicinal

chemistry eﬀorts. Not last, it provides preliminary in vivo

testing of a Grp94 inhibitor to suggest speciﬁc and eﬀective

targeting of tumors by a Grp94 ligand.

In this paper we describe the design and synthesis of purine-

based ligands that preferentially bind to site 2, which is unique

to Grp94, and confers selectivity over other paralogs.

Speciﬁcally, this work resulted in compounds such as 18c,

which shows good binding aﬃnity for Grp94 and selectivity

over the other Hsp90 paralogs. To come to this conclusion, we

synthesized, characterized, and tested approximately 70 ligands

that together encompass nine points of variation on the

pharmacophore. As such, we extensively probed substitutions

on the 8-aryl, the nature of the linker between the purine ring

3934

DOI: 10.1021/acs.jmedchem.5b00197

J. Med. Chem. 2015, 58, 3922−3943

Journal of Medicinal Chemistry

Article

and the 8-aryl, and the nature and length of linkers attached at

either N3 or N9 of the purine, and last, we modiﬁed the purine

core itself by replacing N3 with a carbon.

In summary, our ﬁndings suggest Grp94 as a valid target for

therapeutic intervention and position the Grp94 inhibitors such

as 18c as therapeutic leads in the development of drugs for the

treatment of immune-related disorders and certain cancers.

We also provide here a rationale for the preferred chemical

spaces that elicit a “backward bent” that is directed toward the

unique binding site of Grp94 and optimally complements

volume and polarity characteristics of the site 2 binding pocket

to result in high aﬃnity binding. The conformation observed

for the binding of these ligands to Grp94 is distinct from all

previous structures (apo and ligand bound) reported for Grp94

and it is important to note that this novel pocket (site 2)

becomes exposed and subsequently available for binding only

after ligand-induced conformational changes to Grp94. Upon

interaction, compounds such as those described here are

capable of inducing a conformational change in Grp94 that

results in the exposure of a unique binding pocket that can then

be ﬁlled by the ligand.

We further validated that 18c acts in cells through Grp94-

speciﬁc inhibition using a number of functional readouts.

Speciﬁcally, we show that 18c inhibits both the post-ER

expression of TLR9 as well as CpG-induced/TLR9-mediated

TNF-α production. 18c also inhibits IGF-II secretion by

diﬀerentiated myoblasts and reduces the steady-state levels of

HER2 kinase in HER2-overexpressing breast cancer cells. At

these Grp94-modulatory concentrations, we show that 18c has

no eﬀects on Hsp90-mediated functions in cells, providing

proof for its in cell selectivity.

We also present data comparing the activity of two Grp94

classes and show that the phenotypes of the two compounds do

not necessarily overlap. As the ﬁeld is coming to appreciate,

chaperones regulate proteins by co-opting activating cochaper-

ones and adapter proteins. In disease, these complexes are

context-dependent, in that they form to buﬀer the speciﬁc

alterations in the proteome. The eﬃcacy of an agent targeting a

chaperone in these cells is therefore determined by its ability to

sample and engage such complexes, stressing the need to test

chaperone inhibitors in the disease context they are therapeuti-

cally developed for.

EXPERIMENTAL SECTION

■

1

Chemistry. H and ¹³C NMR spectra were recorded on a Bruker

AV-III 400, 500, or 600 MHz NMR spectrometer. Chemical shifts are

reported in δ values in ppm downﬁeld from TMS as the internal

standard. ¹H data are reported as follows: chemical shift, multiplicity (s

= singlet, d = doublet, t = triplet, q = quartet, br = broad, m =

multiplet), coupling constant (Hz), integration. ¹³C chemical shifts are

reported in δ values in ppm downﬁeld from TMS as the internal

standard. High-resolution mass spectra were recorded on a Waters

LCT Premier system. Low-resolution mass spectra were obtained on

Waters Acquity Ultra Performance LC with electrospray ionization

and SQ detector. The purities of title compounds were determined by

analytical HPLC, performed on a Waters Autopuriﬁcation system with

PDA, MicroMass ZQ and ELSD detector, and a reverse-phase column

(Waters X-Bridge C18, 4.6 × 150 mm, 5 μM) using method A [(a)

H₂O + 0.1% TFA and (b) CH₃CN + 0.1% TFA, 5 to 95% b over 10

min at 1.2 mL/min] or method B [(a) H₂O + 0.1% TFA and (b)

CH₃CN + 0.1% TFA, 20 to 90% b over 16 min at 1.0 mL/min]. Title

compounds used were >95% pure. Analytical thin-layer chromatog-

raphy was performed on 250 μM silica gel F₂₅₄plates. Preparative thin

layer chromatography was performed on 1000 μM silica gel F₂₅₄plates.

Flash column chromatography was performed by employing 230−400

mesh silica gel. Solvents were HPLC grade. PU-H71 (5) was

synthesized as previously reported.²⁸The synthesis of 18g is described

in the Supporting Information. CUDC-305 (22) was purchased from

ChemieTek.

General Procedure for Synthesis of 8-Arylsulfanylpurines (8). 8-

Mercaptoadenine (6); corresponding aryl iodide 7 (300 mol %);

neocuproine hydrate (10 mol %); CuI (10 mol %); NaOt-Bu, K₂CO₃,

or Na₃PO₄(200 mol %); and DMF (2 mL) were charged in a dry

vessel under nitrogen atmosphere. The reaction vessel was sealed and

placed in an oil bath (110 °C) and stirred for 24 h. The reaction

mixture was then cooled to room temperature and DMF was removed

in vacuo. The crude material was puriﬁed by silica gel ﬂash

c h r o m a t o g r a p h y e l u t i n g w i t h

a

g r a d i e n t o f

CH₂Cl₂:CH₃OH:CH₃COOH at 60:1:0.5 to 20:1:0.5 to aﬀord the

desired products.

We also provide for the ﬁrst time information regarding the

in vivo proﬁle of a selective Grp94 inhibitor. When

administered to tumor-bearing mice, 18c was cleared relatively

rapidly from normal tissue while being selectively retained in

tumor; this is remarkable considering that Grp94 is similarly

expressed in most tissues and organs. Such phenomenon was

previously observed with pan-Hsp90 inhibitors.^32,39,40It is

believed that the tumor-speciﬁc eﬀects of HSP inhibitors are

due to an activation of the chaperones in tumors by factors such

as proteotoxic stress. Because of these features, and like the

Hsp90 inhibitors, it is not envisioned that Grp94 inhibitors will

be targeted systemically but, rather, will show preferential

localization in tumors. We have conﬁrmed these concepts with

our pharmacokinetic data.

8-((3-Bromo-5-chlorophenyl)thio)-9H-purin-6-amine (8o). Yield:

42%. MS (ESI): m/z 356.1/358.1 [M + H]⁺.

8-((3,5-Dibromophenyl)thio)-9H-purin-6-amine (8p). Yield: 40%.

MS (ESI): m/z 401.9 [M + H]⁺.

8-((2,6-Dichloropyridin-4-yl)thio)-9H-purin-6-amine (8r). Yield:

50%. MS (ESI): m/z 312.8 [M + H]⁺.

8-((3,4-Dichlorophenyl)thio)-9H-purin-6-amine (8t). Yield: 69%.

MS (ESI): m/z 312.0 [M + H]⁺.

8-((2,3,4-Trichlorophenyl)thio)-9H-purin-6-amine (8u). Yield:

43%. MS (ESI): m/z 347.7 [M + H]⁺.

8-(Mesitylthio)-9H-purin-6-amine (8w). Yield: 53%. ¹H NMR (400

MHz, DMSO): δ 13.1 (br s, 1H), 8.02 (s, 1H), 6.93−7.04 (m, 5H),

2.33 (s, 6H), 2.25 (s, 3H). MS (ESI): m/z 286.1 [M + H]⁺.

Synthesis of 8a−n, 8q, and 8v followed a similar protocol. The

synthesis and characterization of these coupling products have been

described elsewhere.²⁴

It is yet unclear whether such a tumor-retention feature

extends to all Grp94 pharmacophores. For Hsp90 inhibitors it

was determined that not all inhibitors exhibit such preference

for tumor species, possibly because not all inhibitors interact

with the protein in an identical fashion.⁴²Each inhibitor may

preferentially sample speciﬁc conformations and complexes of

the HSP, and it is expected that those that sample preferentially

the tumor, active complexes would be those with higher

therapeutic index.³²This remains to be seen as more Grp94

inhibitor classes move to in vivo studies.

General Procedure for the Synthesis of 9 and 9′. A mixture of 8-

arylsulfanyladenine 8 (100 mmol), Cs₂CO₃(100 mmol), and pent-4-

ynyl 4-methylbenzenesulfonate (120 mmol) in DMF (1.3 mL) under

nitrogen atmosphere was heated at 80 °C for 30 min. Following

solvent removal, the crude material was puriﬁed by preparatory TLC

with CHCl₃:MeOH:NH₄OH at 10:1:0.5 or CHCl₃:MeOH:AcOH at

10:1:0.5 to provide the corresponding 9-alkyl-8-arylsulfanyladenine

derivatives 9 and 3-alkyl-8-arylsulfanyladenine derivatives 9′.

8-(2-Chlorophenylthio)-3-(pent-4-ynyl)-3H-purin-6-amine (9a′).

Yield: 10%. ¹H NMR (400 MHz, CDCl₃/MeOH-d₄): δ 8.02 (s,

1H), 7.48−7.51 (m, 1H), 7.40−7.42 (m, 1H), 7.14−7.17 (m, 2H),

3935

DOI: 10.1021/acs.jmedchem.5b00197

J. Med. Chem. 2015, 58, 3922−3943

Journal of Medicinal Chemistry

Article

4.49−4.51 (m, 2H), 2.17−2.22 (m, 4H), 2.05−2.06 (m, 1H). MS

(ESI): m/z 344.0 [M + H]⁺. HRMS (ESI): m/z [M + H]⁺calcd for

C₁₆H₁₅N₅ClS 344.0737, found 344.0720. HPLC: (b) t_R= 7.84 min,

99.6%.

8-(2-Methoxyphenylthio)-3-(pent-4-ynyl)-3H-purin-6-amine

(9b′). Yield: 25%. ¹H NMR (400 MHz, CDCl₃/MeOH-d₄): δ 7.97 (s,

1H), 7.45 (d, J = 7.7 Hz, 1H), 7.22 (d, J = 7.8 Hz, 1H), 6.86−6.90 (m,

2H), 4.47 (t, J = 6.6 Hz, 2H), 3.87 (s, 3H), 2.21−2.26 (m, 4H), 2.05−

2.07 (m, 1H). ¹³C NMR (100 MHz, CDCl₃/MeOH-d₄) δ 157.7,

152.6, 141.8, 132.5, 128.3, 121.1, 111.1, 81.6, 70.5, 55.9, 49.0, 27.0,

15.3. MS (ESI): m/z 340.0 [M + H]⁺. HRMS (ESI): m/z [M + H]⁺

calcd for C₁₇H₁₈N₅OS 340.1232, found 340.1218. HPLC: (a) t_R= 7.70

min, 98.8%; (b) t_R= 6.44 min, 99.1%.

(2-(6-Amino-3-(pent-4-ynyl)-3H-purin-8-ylthio)phenyl)methanol

(9c′). Yield: 21%. ¹H NMR (400 MHz, CDCl₃/MeOH-d₄): δ 7.96 (s,

1H), 7.74 (d, J = 7.6 Hz, 1H), 7.60 (d, J = 7.5 Hz, 1H), 7.44 (t, J = 7.4

Hz, 1H), 7.29 (t, J = 7.5 Hz, 1H), 4.92 (s, 2H), 4.40 (t, J = 6.4 Hz,

2H), 3.47 (br s, 1H), 2.11−2.20 (m, 4H), 2.03−2.06 (m, 1H). ¹³C

NMR (100 MHz, CDCl₃/MeOH-d₄): δ 146.2, 142.6, 137.3, 131.5,

130.9, 129.1, 82.1, 71.2, 64.9, 49.6, 27.1, 15.6. MS (ESI): m/z 340.1

[M + H]⁺. HRMS (ESI): m/z [M + H]⁺calcd for C₁₇H₁₈N₅OS

340.1232, found 340.1242. HPLC: (a) t_R= 7.07 min, 98.9%; (b) t_R=

5.40 min, 99.0%.

4.45 (t, J = 6.3 Hz, 2H), 2.43 (s, 3H), 2.37 (s, 3H), 2.17−2.27 (m,

3H), 2.04 (m, 2H). ¹³C NMR (100 MHz, CDCl₃/MeOH-d₄): δ 152.0,

151.2, 141.7, 137.7, 135.9, 134.5, 130.9, 130.2, 129.2, 81.8, 70.5, 48.9,

27.0, 20.8, 20.4, 15.3. HRMS (ESI): m/z [M + H]⁺calcd for

C₁₈H₂₀N₅S 338.1439, found 338.1435. HPLC: (a) t_R= 8.67 min,

95.7%.

8-(2-Chloro-5-(triﬂuoromethyl)phenylthio)-3-(pent-4-ynyl)-3H-

purin-6-amine (9m′). Yield: 18%. ¹H NMR (400 MHz, CDCl₃/

MeOH-d₄): δ 8.08 (s, 1H), 7.84 (s, 1H), 7.57 (d, J = 8.3 Hz, 1H), 7.47

(d, J = 8.4 Hz, 1H), 4.48 (t, J = 6.5 Hz, 2H), 2.16−2.26 (m, 4H),

2.09−2.10 (m, 1H). ¹³C NMR (100 MHz, CDCl₃/MeOH-d₄): δ

157.4, 153.2, 150.4, 142.8, 138.2, 134.5, 130.1, 129.4, 128.9, 124.8,

121.9, 81.5, 70.4, 28.3, 26.7, 15.0. MS (ESI): m/z 411.8 [M + H]⁺.

HPLC: (a) t_R= 9.38 min, 95.0%; (b) t_R= 7.66 min, 97.5%.

8-(3,5-Dichlorophenylthio)-3-(pent-4-ynyl)-3H-purin-6-amine

(9n′). Yield: 14%. ¹H NMR (400 MHz, CDCl₃/MeOH-d₄): δ 8.03 (s,

1H), 7.46 (s, 2H), 7.30 (s, 1H), 4.50−4.52 (m, 2H), 2.20−2.24 (m,

3H), 2.04−2.07 (m, 2H). ¹³C NMR (100 MHz, CDCl₃/MeOH-d₄): δ

157.6, 152.6, 150.8, 142.9, 136.3, 134.8, 128.7, 127.1, 120.4, 81.3, 70.1,

43.8, 26.7, 14.7. MS (ESI): m/z 377.8 [M + H]⁺. HRMS (ESI): m/z

[M + H]⁺calcd for C₁₄H₁₅N₅Cl₂S 378.0347, found 378.0340. HPLC:

(a) t_R= 9.50 min, 99.3%; (b) t_R= 7.57 min, 99.5%.

8-((3-Bromo-5-chlorophenyl)thio)-9-(pent-4-yn-1-yl)-9H-purin-6-

amine (9o). Yield: 37%. ¹H NMR (400 MHz, CDCl₃): δ 8.36 (1H, s),

7.42−7.44 (2H, m), 7.33 (1H, s), 6.19 (2H, br s), 4.33 (2H, t, J = 7.3

Hz), 2.27 (2H, td, J = 6.6, 2.4 Hz), 1.99−2.03 (3H, m). ¹³C NMR

(CDCl₃): δ 154.9, 153.5, 151.5, 143.5, 135.9, 134.7, 131.2, 131.0,

128.7, 123.4, 120.3, 82.3, 69.7, 43.0, 28.3, 16.0. HRMS (ESI): m/z [M

+ H]⁺calcd for C₁₆H₁₄N₅SClBr 423.9821, found 423.9822.

3-(Pent-4-ynyl)-8-(2-(triﬂuoromethoxy)phenylthio)-3H-purin-6-

1

amine (9d′). Yield: 19%. H NMR (400 MHz, CDCl₃/MeOH-d₄): δ

8.01 (s, 1H), 7.50 (d, J = 7.9 Hz, 1H), 7.21−7.27 (m, 2H), 7.13 (t, J =

7.1 Hz, 1H), 4.48 (t, J = 6.1 Hz, 2H), 2.17−2.20 (m, 4H), 2.04−2.05

(m, 1H). ¹³C NMR (100 MHz, CDCl₃/MeOH-d₄): δ 158.3, 153.1,

142.2, 132.1, 128.1, 127.9, 126.9, 122.5, 120.7, 81.8, 70.5, 49.0, 26.9,

15.2. MS (ESI): m/z 393.7 [M + H]⁺. HRMS (ESI): m/z [M + H]⁺

calcd for C₁₇H₁₅N₅F₃OS 394.0949, found 394.0946. HPLC: (a) t_R=

8.82 min, 99.7%; (b) t_R= 7.19 min, 99.4%.

8-((3,5-Dibromophenyl)thio)-9-(pent-4-yn-1-yl)-9H-purin-6-

amine (9p). Yield: 37%. ¹H NMR (400 MHz, CDCl₃): δ 8.31 (1H, s),

7.62 (1H, t, J = 1.6 Hz), 7.50 (2H, d, J = 1.7 Hz), 6.22 (2H, br s), 4.33

(2H, t, J = 7.5 Hz), 2.27 (2H, td, J = 6.8, 2.6 Hz), 1.99−2.05 (3H, m).

¹³C NMR (CDCl₃): δ 154.2, 152.9, 151.6, 145.8, 134.1, 131.8, 128.2,

8-(2,4-Dichlorophenylthio)-3-(pent-4-ynyl)-3H-purin-6-amine

(9g′). Yield: 22%. ¹H NMR (400 MHz, CDCl₃/MeOH-d₄): δ 8.10 (s,

1H), 7.50 (s, 1H), 7.42 (d, J = 8.4 Hz, 1H), 7.23−7.25 (m, 1H), 4.46

(t, J = 6.5 Hz, 2H), 2.24−2.27 (m, 2H), 2.14−2.19 (m, 3H). ¹³C NMR

(100 MHz, CDCl₃/MeOH-d₄): δ 154.8, 152.9, 150.0, 142.8, 133.9,

132.3, 130.6, 129.3, 128.8, 127.2, 120.4, 81.3, 70.0, 49.8, 26.7, 14.7. MS

(ESI): m/z 377.7 [M + H]⁺. HRMS (ESI): m/z [M + H]⁺calcd for

C₁₆H₁₄N₅Cl₂S 378.0347, found 378.0335. HPLC: (a) t_R= 9.18 min,

98.4%; (b) t_R= 7.35 min, 99.0%.

123.6, 114.5, 82.4, 69.7, 43.0, 28.2, 16.0. HRMS (ESI): m/z [M + H]⁺

calcd for C₁₆H₁₄N₅SBr₂465.9337, found 465.9329.

8-(3,5-Dimethylphenylthio)-3-(pent-4-ynyl)-3H-purin-6-amine

(9q′). Yield: 16%. ¹H NMR (400 MHz, CDCl₃/MeOH-d₄): δ 8.01 (s,

1H), 7.23 (s, 2H), 7.00 (s, 1H), 4.45 (t, J = 7.5 Hz, 2H), 2.26 (s, 6H),

2.21−2.25 (m, 4H), 1.94−1.99 (m, 1H). MS (ESI): m/z 338.1 [M +

H]⁺. HRMS (ESI): m/z [M + H]⁺calcd for C₁₈H₂₀N₅S 338.1439,

found 338.1426. HPLC: (a) t_R= 8.80 min, 98.9%; (b) t_R= 7.36 min,

99.2%.

8-(2,4-Dimethylphenylthio)-3-(pent-4-ynyl)-3H-purin-6-amine

1

(1). Yield: 27%. H NMR (400 MHz, CDCl₃/MeOH-d₄): δ 7.94 (s,

1H), 7.48 (d, J = 7.7 Hz, 1H), 7.08 (s, 1H), 6.97 (d, J = 7.7 Hz, 1H),

4.36 (t, J = 6.1 Hz, 2H), 2.51 (s, 3H), 2.39 (s, 3H), 2.15−2.19 (m,

4H), 2.03−2.05 (m, 1H). ¹³C NMR (100 MHz, CDCl₃/MeOH-d₄): δ

162.6, 151.9, 151.1, 141.5, 141.1, 138.6, 134.5, 131.3, 127.3, 127.2,

121.7, 81.8, 70.5, 48.8, 26.9, 21.1, 20.8, 15.2. MS (ESI): m/z 338.1 [M

+ H]⁺. HRMS (ESI): m/z [M + H]⁺calcd for C₁₈H₂₀N₅S 338.1439,

found 338.1427. HPLC: (a) t_R= 8.24 min, 99.4%; (b) t_R= 7.19 min,

99.6%.

8-((2,6-Dichloropyridin-4-yl)thio)-9-(pent-4-yn-1-yl)-9H-purin-6-

1

amine (9r). Yield: 22%. H NMR (600 MHz, CDCl₃/MeOH-d₄): δ

8.32 (s, 1H), 7.47 (s, 1H), 7.18 (s, 1H), 4.38 (t, J = 7.2 Hz, 2H), 2.26−

2.29 (m, 2H), 2.03−2.05 (m, 3H). ¹³C NMR (150 MHz, CDCl₃/

MeOH-d₄): δ 159.3, 157.6, 155.0, 154.9, 153.9, 124.4, 124.3, 85.8,

73.6, 47.3, 41.7, 32.3, 19.7. HRMS (ESI): m/z [M + H]⁺calcd for

C₁₅H₁₃N₆SCl₂379.0299, found 379.0312.

8-((3,4-Dichlorophenyl)thio)-9-(pent-4-yn-1-yl)-9H-purin-6-

amine (9t). Yield: 54%. ¹H NMR (400 MHz, CDCl₃): δ 8.25 (1H, s),

7.58 (1H, d, J = 2.2 Hz), 7.40 (1H, d, J = 8.4 Hz), 7.25 (1H, dd, J =

8.3, 2.1 Hz), 6.97 (2H, br s), 4.32 (2H, t, J = 7.4 Hz), 2.28 (2H, td, J =

6.8, 2.6 Hz), 1.97−2.02 (3H, m). ¹³C NMR (CDCl₃): δ 154.8, 153.3,

151.5, 144.6, 133.6, 133.0, 132.5, 131.3, 130.6, 130.2, 120.1, 82.3, 69.6,

42.9, 28.2, 16.0. MS (ESI): m/z 378.9 [M + H]⁺. HPLC: t_R= 8.50

min. HRMS (ESI): m/z [M + H]⁺calcd for C₁₆H₁₄N₅SCl₂378.0347,

found 378.0353.

9-(Pent-4-yn-1-yl)-8-((2,3,4-trichlorophenyl)thio)-9H-purin-6-

amine (9u). Yield: 39%. ¹H NMR (400 MHz, CDCl₃): δ 8.35 (1H, s),

7.29 (1H, d, J = 8.6 Hz), 6.97 (1H, d, J = 8.6 Hz), 6.18 (2H, br s), 4.34

(2H, t, J = 7.3 Hz), 2.26 (2H, td, J = 6.8, 2.4 Hz), 2.03 (2H, p, J = 6.9

Hz), 1.98 (1H, t, J = 2.5 Hz). ¹³C NMR (CDCl₃): δ 154.8, 153.6,

151.5, 143.2, 133.5, 133.4, 133.0, 132.4, 128.8, 128.4, 120.5, 82.1, 69.7,

43.1, 28.4, 16.0. MS (ESI): m/z 412.0 [M + H]⁺. HPLC: t_R= 8.78

min. HRMS (ESI): m/z [M + H]⁺calcd for C₁₆H₁₃N₅SCl₃411.9957,

found 411.9967.

8-(2,4-Dimethoxyphenylthio)-3-(pent-4-ynyl)-3H-purin-6-amine

1

(9j′). Yield: 7%. H NMR (400 MHz, CDCl₃/MeOH-d₄): δ 7.99 (s,

1H), 7.53 (d, J = 8.1 Hz, 1H), 6.56−6.59 (m, 2H), 4.43 (t, J = 6.8 Hz,

2H), 3.86 (s, 3H), 3.82 (s, 3H), 2.12−2.26 (m, 5H). ¹³C NMR (100

MHz, CDCl₃/MeOH-d₄): δ 162.2, 160.6, 151.8, 141.6, 137.3, 135.2,

108.8, 105.3, 99.1, 81.5, 70.1, 55.6, 55.1, 26.7, 14.8. MS (ESI): m/z

370.0 [M + H]⁺. HRMS (ESI): m/z [M + H]⁺calcd for C₁₈H₂₀N₅O₂S

370.1338, found 370.1350. HPLC: (a) t_R= 7.79 min, 97.9%; (b) t_R=

6.73 min, 98.7%.

8-(2,5-Dichlorophenylthio)-3-(pent-4-ynyl)-3H-purin-6-amine

(9k′). Yield: 21%. ¹H NMR (400 MHz, CDCl₃/MeOH-d₄): δ 8.06 (s,

1H), 7.43 (s, 1H), 7.26−7.28 (m, 1H), 7.06−7.08 (m, 1H), 4.50−4.52

(m, 2H), 2.21−2.24 (m, 4H), 2.05−2.06 (m, 1H). ¹³C NMR (100

MHz, CDCl₃/MeOH-d₄): δ 153.6, 142.6, 132.7, 130.3, 129.7, 127.3,

81.7, 70.6, 49.2, 26.9, 15.2. MS (ESI): m/z 377.8 [M + H]⁺. HRMS

(ESI): m/z [M + H]⁺calcd for C₁₆H₁₄N₅Cl₂S 378.0347, found

378.0362. HPLC: (a) t_R= 9.07 min, 95.6%; (b) t_R= 7.30 min, 97.8%.

8-(2,5-Dimethylphenylthio)-3-(pent-4-ynyl)-3H-purin-6-amine

3-(Pent-4-ynyl)-8-(2,4,5-trichlorophenylthio)-3H-purin-6-amine

(9v′). Yield: 23%. ¹H NMR (400 MHz, CDCl₃/MeOH-d₄): δ 8.04 (s,

1H), 7.73 (s, 1H), 7.55 (s, 1H), 4.50 (t, J = 6.4 Hz, 2H), 2.08−2.29

1

(9l′). Yield: 20%. H NMR (400 MHz, CDCl₃/MeOH-d₄): δ 7.96 (s,

1H), 7.41 (s, 1H), 7.13 (d, J = 7.7 Hz, 1H), 7.04 (d, J = 7.7 Hz, 1H),

3936

DOI: 10.1021/acs.jmedchem.5b00197

J. Med. Chem. 2015, 58, 3922−3943

Journal of Medicinal Chemistry

Article

(m, 4H), 2.06−2.07 (m, 1H). ¹³C NMR (100 MHz, CDCl₃/MeOH-

d₄): δ 159.6, 154.7, 152.4, 144.3, 135.3, 134.9, 134.7, 133.5, 132.9,

132.4, 128.8, 83.4, 72.2, 48.6, 28.6, 16.8. MS (ESI): m/z 411.9 [M +

H]⁺. HRMS (ESI): m/z [M + H]⁺calcd for C₁₆H₁₃N₅Cl₃S 411.9957,

found 411.9947. HPLC: (a) t_R= 9.84 min, 99.5%; (b) t_R= 7.96 min,

99.4%.

8-(Mesitylthio)-9-(pent-4-yn-1-yl)-9H-purin-6-amine (9w). Yield:

32%. ¹H NMR (400 MHz, CDCl₃): δ 8.21 (s, 1H), 6.98 (s, 2H), 5.87

(br s, 2H), 4.31 (t, 2H), 2.40 (s, 6H), 2.29−2.34 (m, 2H), 2.23 (s,

3H), 2.12 (m, 2H), 2.01 (m, 1H). ¹³C NMR (100 MHz, CDCl₃): δ

153.7, 151.8, 151.7, 143.3, 140.2, 129.5, 123.1, 119.7, 82.5, 69.4, 49.9,

42.4, 28.2, 21.8, 21.0, 16.1. MS (ESI): m/z 352.1 [M + H]⁺. HRMS

(ESI): m/z [M + H]⁺calcd for C₁₉H₂₂N₅S 352.1596, found 352.1598.

HPLC: (a) t_R= 6.58 min, 99.6%.

HRMS (ESI): m/z [M + H]⁺calcd for C₁₅H₁₂N₅SCl₂364.0190, found

364.0194.

8-((3,5-Dichlorophenyl)thio)-9-(hex-5-yn-1-yl)-9H-purin-6-amine

(13b). Yield: 41%. ¹H NMR (400 MHz, CDCl₃): δ 8.29 (1H, s), 7.28

(3H, s), 6.71 (2H, br s), 4.26 (2H, t, J = 7.2 Hz), 2.22 (2H, td, J = 6.8,

2.7 Hz), 1.95 (1H, t, J = 2.6 Hz), 1.87−1.93 (2H, m), 1.53 (2H,

pentet, J = 6.8 Hz). ¹³C NMR (CDCl₃): δ 155.0, 151.2, 143.6, 135.9,

134.4, 129.0, 128.5, 128.2, 120.0, 83.3, 69.1, 43.5, 28.8, 25.3, 17.9.

HRMS (ESI): m/z [M + H]⁺calcd for C₁₇H₁₅N₅SCl₂392.0503, found

392.0493.

9-(But-2-yn-1-yl)-8-((3,5-dichlorophenyl)thio)-9H-purin-6-amine

(13c). Yield: 40%. ¹H NMR (500 MHz, CDCl₃): δ 8.33 (1H, s), 7.32

(2H, s), 7.28 (1H, s), 6.63 (2H, br s), 4.98−4.99 (2H, m), 1.70 (3H,

m). ¹³C NMR (CDCl₃): δ 155.1, 153.2, 150.8, 143.3, 135.7, 134.5,

128.5, 128.3, 120.0, 82.1, 71.7, 33.4, 3.5. HRMS (ESI): m/z [M + H]⁺

calcd for C₁₅H₁₂N₅SCl₂364.0190, found 364.0197.

8-(Mesitylthio)-3-(pent-4-ynyl)-3H-purin-6-amine (9w′). Yield:

1

15%. H NMR (400 MHz, CDCl₃/MeOH-d₄): δ 7.93 (s, 1H), 6.97

8-((3,5-Dichlorophenyl)thio)-9-(3,3,3-triﬂuoropropyl)-9H-purin-6-

(s, 2H), 4.38 (t, J = 6.5 Hz, 2H), 2.41 (s, 6H), 2.30 (s, 3H), 2.04−2.16

(m, 5H). ¹³C NMR (100 MHz, CDCl₃/MeOH-d₄): δ 157.4, 150.4,

143.0, 142.2, 139.2, 128.9, 124.8, 81.5, 70.1, 26.7, 21.6, 21.3, 20.5, 14.8.

MS (ESI): m/z 351.8 [M + H]⁺. HRMS (ESI): m/z [M + H]⁺calcd

for C₁₉H₂₂N₅S 352.1596, found 352.1594. HPLC: (a) t_R= 9.05 min,

99.6%; (b) t_R= 7.49 min, 99.7%.

Formation of 9a−n, 9q, and 9v followed a similar protocol.

Synthesis and characterization of these coupling products have been

described elsewhere.²⁴

8-Bromo-9H-purin-6-amine (11). Adenine (2.2 g, 16.3 mmol) was

added to a solution of bromine (6.0 mL, 117.7 mmol) in water (200

mL), and the resulting mixture was stirred overnight at room

temperature. The solvent was evaporated to dryness to yield 11, which

was used further without additional puriﬁcation. MS (ESI): m/z

213.5/215.6 [M + H]⁺.

1

amine (13d). Yield: 43%. H NMR (400 MHz, CDCl₃): δ 8.33 (1H,

s), 7.29 (3H, s), 6.44 (2H, br s), 4.49 (2H, t, J = 6.8 Hz), 2.64−2.68

(2H, m). ¹³C NMR (CDCl₃): δ 155.0, 153.4, 151.2, 143.5, 136.0,

133.8, 128.8, 128.4, 126.6, 120.2, 37.3, 33.3. HRMS (ESI): m/z [M +

H]⁺calcd for C₁₄H₁₁N₅SCl₂F₃408.0064, found 408.0074.

8-((3,5-Dichlorophenyl)thio)-9-(4,4,4-triﬂuorobutyl)-9H-purin-6-

1

amine (13e). Yield: 44%. H NMR (400 MHz, CDCl₃): δ 8.36 (1H,

s), 7.28 (3H, s), 6.16 (2H, br s), 4.30 (2H, t, J = 6.6 Hz), 2.10−2.14

(2H, m), 2.03−2.05 (2H, m). ¹³C NMR (CDCl₃): δ 154.9, 153.7,

151.5, 143.2, 135.9, 134.2, 128.6, 128.1, 125.1, 120.3, 42.5, 31.3, 22.5.

HRMS (ESI): m/z [M + H]⁺calcd for C₁₅H₁₃N₅SCl₂F₃422.0221,

found 422.0222.

8-((3,5-Dichlorophenyl)thio)-9-(5,5,5-triﬂuoropentyl)-9H-purin-6-

1

amine (13f). Yield: 45%. H NMR (400 MHz, CDCl₃): δ 8.28 (1H,

s), 7.28 (3H, s), 6.82 (2H, br s), 4.25 (2H, t, J = 7.2 Hz), 2.07−2.14

(2H, m), 1.85 (2H, pentet, J = 7.4 Hz), 1.57 (2H, pentet, J = 7.5 Hz).

¹³C NMR (CDCl₃): δ 155.1, 152.7, 151.2, 143.5, 135.9, 134.1, 128.6,

128.2, 125.4, 120.1, 43.3, 33.1, 28.8, 19.2. HRMS (ESI): m/z [M +

H]⁺calcd for C₁₆H₁₅N₅SF₃Cl₂436.0377, found 436.0363.

8-Bromo-9-(pent-4-yn-1-yl)-9H-purin-6-amine (12). A mixture of

8-bromoadenine (11; 0.1 g, 0.47 mmol), Cs₂CO₃(0.3 g, 0.94 mmol),

and 5-chloropent-1-yne (0.13g, 0.65 mmol) in DMF (1.5 mL) under

nitrogen protection was heated at 80 °C for 2 h. Following solvent

removal, the crude material was puriﬁed by preparatory TLC with

1

8-((3,5-Dichlorophenyl)thio)-9-(6,6,6-triﬂuorohexyl)-9H-purin-6-

CH₂Cl₂:MeOH:AcOH at 20:1:0.1 to provide 36 mg (30%) of 12. H

1

amine (13g). Yield: 47%. H NMR (400 MHz, CDCl₃): δ 8.35 (1H,

NMR (500 MHz, CDCl₃/MeOH-d₄): δ 8.29 (s, 1H), 4.33 (t, J = 7.2

Hz, 2H), 2.28−2.33 (m, 2H), 2.09 (q, J = 7.0 Hz, 2H), 2.01−2.03 (t,

1H). ¹³C NMR (125 MHz, CDCl₃/MeOH-d₄): δ 154.4, 153.1, 151.3,

127.4, 119.9, 82.4, 69.7, 43.8, 28.2, 16.1. MS (ESI): m/z 280.1/282.2

[M + H]⁺.

s), 7.29 (3H, s), 6.33 (2H, br s), 4.24 (2H, t, J = 7.2 Hz), 2.00−2.07

(2H, m), 1.79 (2H, pentet, J = 7.4 Hz), 1.57 (2H, pentet, J = 7.6 Hz),

1.37 (2H, pentet, J = 7.8 Hz). ¹³C NMR (CDCl₃): δ 155.1, 152.8,

151.3, 143.4, 135.9, 134.4, 128.5, 128.1, 125.8, 120.0, 43.6, 33.9, 29.4,

25.7, 21.4. HRMS (ESI): m/z [M + H]⁺calcd for C₁₇H₁₇N₅SCl₂F₃

450.0534, found 450.0549.

4-(6-Amino-8-((3,5-dichlorophenyl)thio)-9H-purin-9-yl)-

butanenitrile (13h). Yield: 41%. ¹H NMR (400 MHz, CDCl₃): δ 8.34

(1H, s), 7.31 (3H, s), 6.04 (2H, br s), 4.36 (2H, t, J = 7.0 Hz), 2.42

(2H, t, J = 7.1 Hz), 2.18 (2H, pentet, J = 7.2 Hz). ¹³C NMR (CDCl₃):

δ 154.8, 152.7, 151.5, 143.4, 136.0, 133.9, 129.2, 127.5, 120.2, 118.3,

42.4, 25.6, 14.9. HRMS (ESI): m/z [M + H]⁺calcd for C₁₅H₁₃N₆SCl₂

379.0299, found 379.0303.

9-Benzyl-8-((3,5-dichlorophenyl)thio)-9H-purin-6-amine (13i).

Yield: 48%. ¹H NMR (400 MHz, CDCl₃): δ 8.43 (1H, s), 7.23

(3H, s), 7.15−7.19 (3H, m), 7.04−7.05 (2H, m), 5.97 (2H, br s), 5.45

(2H, s). ¹³C NMR (CDCl₃): δ 154.9, 153.9, 151.7, 143.6, 135.6, 135.3,

134.5, 128.7, 128.2, 128.1, 127.9, 127.5, 120.8, 47.0. HRMS (ESI): m/z

[M + H]⁺calcd for C₁₈H₁₄N₅SCl₂402.0347, found 402.0335.

8-((3,5-Dichlorophenyl)thio)-9-phenethyl-9H-purin-6-amine

(13j). Yield: 45%. ¹H NMR (400 MHz, CDCl₃): δ 8.35 (1H, s), 7.22−

7.26 (4H, m), 7.19−7.20 (2H, m), 7.04−7.06 (2H, m), 6.45 (2H, br

s), 4.47 (2H, t, J = 7.2 Hz), 3.09 (2H, t, J = 7.3 Hz). ¹³C NMR

(CDCl₃): δ 153.8, 151.5, 151.2, 145.1, 136.8, 135.8, 134.0, 128.9,

128.8, 128.6, 128.5, 127.2, 120.1, 45.5, 35.7. HRMS (ESI): m/z [M +

H]⁺calcd for C₁₉H₁₆N₅SCl₂416.0503, found 416.0508.

5-((6-Amino-9-(pent-4-yn-1-yl)-9H-purin-8-yl)thio)-

isophthalonitrile (9s). A mixture of 5-mercaptoisophthalonitrile²⁶(11

mg, 0.069 mmol) and t-BuOK (7 mg, 0.069 mmol) in 1.5 mL of DMF

was stirred for 15 min at room temperature. Compound 12 (16 mg,

0.057 mmol) was then added and the reaction mixture was allowed to

stir at 80 °C for 2 h. Following solvent removal, the crude material was

puriﬁed by preparatory TLC with CH₂Cl₂:MeOH at 20:1 to provide

1

10.2 mg (51%) of the title compound. H NMR (600 MHz, CDCl₃/

MeOH-d₄): δ 8.38 (s, 1H), 7.98 (s, 2H), 7.85 (s, 1H), 4.37 (t, J = 7.3

Hz, 2H), 2.27−2.30 (m, 2H), 2.05 (q, J = 6.9 Hz, 2H), 2.01−2.03 (t,

1H). ¹³C NMR (150 MHz, CDCl₃/MeOH-d₄): δ 154.7, 153.8, 151.7,

142.2, 136.9, 136.1, 134.3, 120.5, 115.8, 115.2, 82.2, 69.8, 43.1, 28.3,

15.9. HRMS (ESI): m/z [M + H]⁺calcd for C₁₈H₁₄N₇S 360.1031,

found 360.1028.

General Procedure for the Synthesis of 13a−k,m,o,q,r. A mixture

of 8-((3,5-dichlorophenyl)thio)-9H-purin-6-amine (1.0 mmol),

Cs₂CO₃(1.2 mmol), and the respective alkyl bromides or alkyl

benzenesulfonate (2.0 mmol) in DMF (1.5 mL) under nitrogen

atmosphere was heated at 80 °C for 30 min. Following solvent

removal, the crude material was puriﬁed by preparatory TLC with

CHCl₃:MeOH:NH₄OH at 10:1:0.5 or CHCl₃:MeOH:AcOH at

10:1:0.5 to provide the corresponding 9-alkyl-8-arylsulfanyladenine

derivatives.

8-((3,5-Dichlorophenyl)thio)-9-(pent-4-yn-2-yl)-9H-purin-6-

1

amine (13m). Yield: 12%. H NMR (600 MHz, CDCl₃): δ 8.27 (s,

9-(But-3-yn-1-yl)-8-((3,5-dichlorophenyl)thio)-9H-purin-6-amine

(13a). Yield: 47%. ¹H NMR (400 MHz, CDCl₃): δ 8.37 (1H, s), 7.28

(3H, s), 6.04 (2H, br s), 4.45 (2H, t, J = 7.0 Hz), 2.76 (2H, td, J = 6.8,

2.3 Hz), 1.95 (1H, t, J = 2.4 Hz). ¹³C NMR (CDCl₃): δ 154.9, 153.6,

151.3, 143.7, 135.8, 134.9, 128.3, 127.9, 120.4, 79.4, 71.5, 42.3, 19.5.

1H), 7.26−7.28 (m, 3H), 6.28 (br s, 2H), 4.96 (m, 1H), 3.31 (ddd, J =

2.6, 9.6, 16.9 Hz, 1H), 2.82 (ddd, J = 2.6, 5.9, 16.9 Hz, 1H), 1.83 (t, J =

2.6 Hz, 1H), 1.66 (d, J = 6.9 Hz, 3H). ¹³C NMR (150 MHz, CDCl₃):

δ 155.0, 152.4, 151.1, 143.6, 135.8, 135.3, 128.2, 127.9, 120.6, 80.0,

3937

DOI: 10.1021/acs.jmedchem.5b00197

J. Med. Chem. 2015, 58, 3922−3943

Journal of Medicinal Chemistry

Article

70.8, 53.4, 24.9, 19.3. HRMS (ESI): m/z [M + H]⁺calcd for

C₁₆H₁₄N₅SCl₂378.0347, found 378.0336.

Synthesis of 14a and 14b. A mixture of 9g (20 mg, 0.053 mmol)

and m-CPBA (18.2 mg, 0.106 mmol) in THF:CH₂Cl₂(2 mL) under

nitrogen protection was stirred at room temperature for 30 min.

Following solvent removal, the crude material was puriﬁed by

preparative TLC [CH₂Cl₂:CH₃OH−NH₃(7 N), 20:1] to aﬀord 14a

and 14b.

9-Butyl-8-((2,4-dichlorophenyl)thio)-9H-purin-6-amine (13o).

1

Yield: 23%. H NMR (600 MHz, CDCl₃): δ 8.36 (s, 1H), 7.47 (d, J

= 2.1 Hz, 1H), 7.17−7.19 (m, 1H), 7.13−7.15 (m, 1H), 5.71 (br s,

2H), 4.22 (t, J = 7.5 Hz, 2H), 1.72−1.77 (m, 2H), 1.32−1.36 (m, 2H),

0.92 (t, J = 7.3 Hz, 3H). ¹³C NMR (150 MHz, CDCl₃): δ 154.6, 153.4,

151.6, 143.8, 134.9, 134.6, 132.3, 130.1, 129.9, 128.1, 120.4, 43.8, 31.4,

19.9, 13.5. HRMS (ESI): m/z [M + H]⁺calcd for C₁₅H₁₆N₅SCl₂

368.0503, found 368.0493.

9-(But-3-yn-1-yl)-8-((2,4-dichlorophenyl)thio)-9H-purin-6-amine

(13q). Yield: 20%. ¹H NMR (600 MHz, CDCl₃): δ 8.33 (s, 1H), 7.47

(s, 1H), 7.21 (m, 2H), 5.96 (br s, 2H), 4.45 (t, J = 7.0 Hz, 2H), 2.75−

2.78 (m, 2H), 1.95 (t, J = 2.5 Hz, 1H). ¹³C NMR (150 MHz, CDCl₃):

δ 154.6, 153.1, 151.3, 144.4, 135.2, 134.8, 132.6, 130.2, 129.8, 128.1,

120.3, 79.3, 71.5, 42.3, 19.6. HRMS (ESI): m/z [M + H]⁺calcd for

C₁₅H₁₂N₅SCl₂364.0190, found 364.0177.

8-((2,4-Dichlorophenyl)sulﬁnyl)-9-(pent-4-yn-1-yl)-9H-purin-6-

1

amine (14a). Yield: 23%. H NMR (600 MHz, CDCl₃/MeOH-d₄): δ

8.29 (s, 1H), 8.04 (d, J = 8.5 Hz, 1H), 7.58 (dd, J = 8.5, 2.0 Hz, 1H),

7.41 (d, J = 2.0 Hz, 1H), 4.57 (t, J = 7.1 Hz, 2H), 2.30 (td, J = 7.0, 2.6

Hz, 2H), 2.06−2.20 (m, 2H), 1.98 (t, J = 2.6 Hz, 1H). ¹³C NMR (150

MHz, CDCl₃/MeOH-d₄): δ 155.1, 152.6, 150.6, 139.2, 136.3, 131.9,

130.1, 128.7, 128.6, 119.1, 81.9, 70.0, 43.5, 28.7, 16.0. HRMS (ESI):

m/z [M + H]⁺calcd for C₁₆H₁₄Cl₂N₅OS 394.0296, found 394.0279.

8-((2,4-Dichlorophenyl)sulfonyl)-9-(pent-4-yn-1-yl)-9H-purin-6-

1

amine (14b). Yield: 34%. H NMR (600 MHz, CDCl₃): δ 8.39 (s,

1H), 8.24 (d, J = 8.4 Hz, 1H), 7.48−7.51 (m, 2H), 6.24 (br s, 2H),

4.88 (t, J = 7.6 Hz, 2H), 2.29 (td, J = 7.0, 2.6 Hz, 2H), 2.10−2.16 (m,

2H), 1.92 (t, J = 2.6 Hz, 1H). ¹³C NMR (150 MHz, CDCl₃): δ 155.4,

153.3, 150.7, 146.2, 142.2, 134.9, 134.8, 132.6, 132.0, 127.9, 119.3,

82.1, 69.6, 44.4, 28.9, 16.1. HRMS (ESI): m/z [M + H]⁺calcd for

C₁₆H₁₄Cl₂N₅O₂S 410.0245, found 410.0228.

8-((2,4-Dichlorophenyl)thio)-9-(pent-4-yn-2-yl)-9H-purin-6-

1

amine (13r). Yield: 12%. H NMR (600 MHz, CDCl₃): δ 8.27 (s,

1H), 7.47 (d, J = 2.0 Hz, 1H), 7.14−7.18 (m, 2H), 6.13 (br s, 2H),

4.98 (m, 1H), 3.31 (ddd, J = 2.6, 9.2, 16.9 Hz, 1H), 2.82 (ddd, J = 2.6,

6.2, 16.9 Hz, 1H), 1.83 (t, J = 2.6 Hz, 1H), 1.66 (d, J = 6.9 Hz, 3H).

¹³C NMR (150 MHz, CDCl₃): δ 154.8, 152.3, 151.2, 144.1, 134.8,

8-(2,4-Dichlorophenoxy)-9-(pent-4-yn-1-yl)-9H-purin-6-amine

(14c). A mixture of 2,4-dichlorophenol (0.069 mmol) and KOt-Bu

(0.069 mmol) in DMF (1.5 mL) was stirred for 15 min at room

temperature. Compound 12 (0.057 mmol) was added and the reaction

mixture was allowed to stir at 80 °C for 2 h. Following solvent

removal, the crude material was puriﬁed by preparatory TLC

(CH₂Cl₂:MeOH, 20:1) to aﬀord 14c. Yield: 51%. ¹H NMR (600

MHz, CDCl₃): δ 8.22 (s, 1H), 7.45 (d, J = 2.5 Hz, 1H), 7.43 (d, J = 8.7

Hz, 1H), 7.27 (dd, J = 8.7, 2.5 Hz, 1H), 5.42 (br s, 2H), 4.26 (t, J = 7.1

Hz, 2H), 2.27 (td, J = 7.0, 2.6 Hz, 2H), 2.09−2.14 (m, 2H), 1.90 (t, J =

2.6 Hz, 1H). ¹³C NMR (150 MHz, CDCl₃): δ 153.2, 152.7, 151.4,

149.9, 147.4, 132.1, 130.6, 128.9, 128.3, 127.1, 123.7, 115.4, 82.5, 69.4,

41.2, 27.9, 16.1. HRMS (ESI): m/z [M + H]⁺calcd for C₁₆H₁₄Cl₂N₅O

362.0575, found 362.0570.

8-(2,4-Dichlorobenzyl)-9H-purin-6-amine (17). In a conical-

bottomed Smith process vial, the mixture of 4,5,6-triaminopyrimidine

(15, 0.21 g, 1.7 mmol), 2,4-dichlorophenylacetic acid (16; 0.25 g, 1.4

mmol), and triphenyl phosphite (0.52 g, 443 μL, 1.7 mmol) in 1.5 mL

of anhydrous pyridine was charged. The sealed vial was irradiated in

the microwave for 30 min at 220 °C. After cooling, the reaction

mixture was concentrated under vacuum and the residue puriﬁed by

column chromatography (CH₂Cl₂:MeOH, 10:0 to 10:1) to give the

desired product 17. Yield: 79%. MS (ESI): m/z 294.0 [M + H]⁺.

8-(2,4-Dichlorobenzyl)-9-(pent-4-yn-1-yl)-9H-purin-6-amine

(14d). A mixture of 17 (1.0 mmol), Cs₂CO₃(1.2 mmol), and 1-

chloro-pent-4-yne (2.0 mmol) in DMF (1.5 mL) under nitrogen

protection was heated at 80 °C for 2 h. Following solvent removal, the

crude material was puriﬁed by preparatory TLC

(CH₂Cl₂:MeOH:AcOH, 15:1:0.1) to provide 14d. Yield: 13%. ¹H

NMR (600 MHz, CDCl₃): δ 8.25 (s, 1H), 7.38 (d, J = 2.2 Hz, 1H),

7.13 (dd, J = 8.3 and 2.2 Hz, 1H), 7.01 (d, J = 8.3 Hz, 1H), 5.86 (br s,

2H), 4.29 (s, 2H), 4.15 (t, J = 7.4 Hz, 2H), 2.17 (td, J = 6.8 and 2.6

Hz, 2H), 1.90−1.96 (m, 3H). ¹³C NMR (150 MHz, CDCl₃): δ 154.3,

151.7, 151.4, 150.2, 134.7, 134.1, 132.4, 131.4, 129.8, 127.8, 118.9,

82.5, 70.1, 42.3, 31.3, 28.4, 15.9. HRMS (ESI): m/z [M + H]⁺calcd

for C₁₇H₁₆Cl₂N₅360.0783, found 360.0772.

134.5, 132.3, 130.2, 130.0, 127.9, 120.7, 80.0, 70.8, 53.4, 24.5, 19.1.

HRMS (ESI): m/z [M + H]⁺calcd for C₁₆H₁₄N₅SCl₂378.0347, found

378.0354.

General Procedure for the Synthesis of 13l,n,p,s. To a suspension

of 8-((2,4-dichlorophenyl)thio)-9H-purin-6-amine (1.0 mmol) in

CH₂Cl₂:toluene (0.5:2.5 mL) were added PPh₃(4.0 mmol) and

alcohol (2.0 mmol) under nitrogen protection. After stirring for 10

min, DBAD (6 mmol) was added and the reaction mixture was stirred

at rt for 2−5 h. Following solvent removal, the crude material was

puriﬁed by preparative TLC [CH₂Cl₂:CH₃OH:AcOH, 20:1:0.1 or

CH₂Cl₂:NH₃−CH₃OH (7 N), 20:1] to aﬀord the desired compounds.

8-((3,5-Dichlorophenyl)thio)-9-(pentan-2-yl)-9H-purin-6-amine

(13l). Yield: 15%. ¹H NMR (600 MHz, CDCl₃/MeOH-d₄, two

rotamers): δ 8.17−8.21 (m, 1H), 7.40−7.56 (m, 3H), 4.76−4.79 (m,

0.4H), 4.65−4.69 (m, 0.6H), 2.21−2.27 (m, 0.6H), 1.97−2.03 (m,

0.4H), 1.81−1.92 (m, 1H), 1.60−1.63 (m, 3H), 1.03−1.28 (m, 2H),

0.87−0.89 (m, 3H). ¹³C NMR (150 MHz, CDCl₃/MeOH-d₄): δ

153.3, 151.2, 150.0, 145.6, 135.9, 135.2, 133.5, 130.2, 129.1, 128.1,

120.4, 54.4, 36.7, 19.8, 19.7, 13.6. HRMS (ESI): m/z [M + H]⁺calcd

for C₁₆H₁₈Cl₂N₅S 382.0660, found 382.0663.

8-((3,5-Dichlorophenyl)thio)-9-(hex-5-yn-3-yl)-9H-purin-6-amine

1

(13n). Yield: 37%. H NMR (600 MHz, CDCl₃): δ 8.25 (s, 1H),

7.35−7.37 (m, 2H), 7.33 (t, J = 1.7 Hz, 1H), 4.71−4.74 (m, 1H),

3.27−3.33 (m, 1H), 2.80−2.84 (m, 1H), 2.32−2.35 (m, 1H), 2.02−

2.05 (m, 1H), 1.84 (t, J = 2.5 Hz, 1H), 0.79 (t, J = 7.4 Hz, 3H). ¹³C

NMR (150 MHz, CDCl₃): δ 153.6, 150.5, 147.7, 135.8, 133.6, 132.1,

131.9, 129.3, 128.9, 128.4, 79.6, 71.1, 59.9, 26.2, 23.3, 10.9. HRMS

(ESI): m/z [M + H]⁺calcd for C₁₇H₁₅Cl₂N₅S 392.0503, found

392.0503.

8-((2,4-Dichlorophenyl)thio)-9-(pentan-2-yl)-9H-purin-6-amine

(13p). Yield: 15%. ¹H NMR (600 MHz, CDCl₃): δ 8.22 (s, 1H), 7.43

(d, J = 1.8 Hz, 1H), 7.13−7.20 (m, 2H), 6.14 (br s, 2H), 4.65−4.70

(m, 1H), 2.17−2.23 (m, 1H), 1.80−1.86 (m, 1H), 1.54 (d, J = 6.9 Hz,

3H), 1.12−1.19 (m, 1H), 0.97−1.02 (m, 1H), 0.79 (t, J = 7.4 Hz, 3H).

¹³C NMR (150 MHz, CDCl₃): δ 153.3, 151.2, 150.0, 145.6, 135.9,

135.2, 133.5, 130.2, 129.1, 128.1, 120.4, 54.4, 36.7, 19.8, 19.7, 13.6.

HRMS (ESI): m/z [M + H]⁺calcd for C₁₆H₁₈Cl₂N₅S 382.0660, found

382.0663.

(6-Amino-9-(pent-4-yn-1-yl)-9H-purin-8-yl)(2,4-dichlorophenyl)-

methanone (14e). A mixture of 14d (1.0 mmol) and Cs₂CO₃(2.0

mmol) in DMF (1.3 mL) was heated at 80 °C for 3 h. Following

solvent removal, the crude material was puriﬁed by preparatory TLC

[CH₂Cl₂:CH₃OH−NH₃(7 N), 20:1] to provide the corresponding

8-((2,4-Dichlorophenyl)thio)-9-(hex-5-yn-3-yl)-9H-purin-6-amine

1

(13s). Yield: 24%. H NMR (600 MHz, MeOH-d₄): δ 8.27 (s, 1H),

7.69 (d, J = 2.2 Hz, 1H), 7.63 (d, J = 8.5 Hz, 1H), 7.42 (d, J = 8.5 and

2.3 Hz, 1H), 4.75−4.79 (m, 1H), 3.24−3.28 (m, 1H), 2.89−2.94 (m,

1H), 2.35−2.41 (m, 1H), 2.25 (t, J = 2.6 Hz, 1H), 2.05−2.10 (m, 1H),

0.84 (t, J = 7.4 Hz, 3H). ¹³C NMR (150 MHz, MeOH-d₄): δ 153.3,

151.2, 150.0, 145.6, 135.9, 135.2, 133.5, 130.2, 129.1, 128.1, 120.4,

54.4, 36.7, 19.8, 19.7, 13.6. HRMS (ESI): m/z [M + H]⁺calcd for

C₁₇H₁₆Cl₂N₅S 392.0660, found 392.0663.

1

aryl ketone 14e. Yield: 50%. H NMR (600 MHz, CDCl₃): δ 8.25 (s,

1H), 7.49 (d, J = 8.2 Hz, 1H), 7.43 (d, J = 1.9 Hz, 1H), 7.33 (dd, J =

8.3, 1.9 Hz, 1H), 6.26 (br s, 2H), 4.73 (t, J = 7.2 Hz, 2H), 2.28 (td, J =

7.0, 2.6 Hz, 2H), 2.07−2.13 (m, 2H), 1.88 (t, J = 2.6 Hz, 1H). ¹³C

NMR (150 MHz, CDCl₃): δ 185.5, 155.9, 153.5, 151.3, 144.4, 137.9,

135.5, 133.5, 131.4, 130.3, 127.1, 119.6, 82.4, 69.3, 44.1, 28.9, 16.1.

3938

DOI: 10.1021/acs.jmedchem.5b00197

J. Med. Chem. 2015, 58, 3922−3943

Journal of Medicinal Chemistry

Article

HRMS (ESI): m/z [M + H]⁺calcd for C₁₇H₁₄Cl₂N₅O 374.0575,

found 374.0571.

puriﬁed by preparatory TLC [CH₂Cl₂:MeOH−NH₃(7 N), 15:1] to

aﬀord corresponding deprotected intermediates 21d−f.

2-((2,4-Dichlorophenyl)thio)-1H-imidazo[4,5-c]pyridin-4-amine

General Procedure for the Synthesis of 18a and 18b′. A mixture

of 8-arylsulfanyladenine (100 mmol), Cs₂CO₃(100 mmol), and 3-

(tert-butoxycarbonyl-isopropyl-amino)propyl tosylate (200 mmol) in

DMF (1.3 mL) under nitrogen protection was heated at 80 °C for 30

min. Following solvent removal, the crude material was puriﬁed by

preparatory TLC with CH₂Cl₂:MeOH:AcOH at 20:1:0.1 to aﬀord the

Boc-protected N9- and N3-alkylated compounds. They were

separately treated with TFA (1 mL) at 0 °C for 1.5 h to provide

with corresponding 9-alkyl-8-arylsulfanyladenine and 3-alkyl-8-arylsul-

fanyladenine derivatives.

8-((2,4-Dichlorophenyl)thio)-9-(3-(isopropylamino)propyl)-9H-

purin-6-amine (18a). Yield: 35%. ¹H NMR (600 MHz, CDCl₃/

MeOH-d₄): δ 8.23 (s, 1H), 7.55 (s, 1H), 7.38 (d, J = 8.3 Hz, 1H), 7.30

(d, J = 8.3 Hz, 1H), 4.34 (t, J = 6.7 Hz, 2H), 2.92 (septet, J = 5.8 Hz,

1H), 2.69 (t, J = 6.4 Hz, 2H), 2.13 (quintet, J = 6.5 Hz, 2H), 1.16 (d, J

= 5.7 Hz, 6H). HRMS (ESI): m/z [M + H]⁺calcd for C₁₇H₂₁Cl₂N₆S

411.0925, found 411.0907.

1

(21d). Yield: 92%. H NMR (500 MHz, CDCl₃/MeOH-d₄): δ 7.46−

7.48 (m, 2H), 7.30 (d, J = 8.5 Hz, 1H), 7.02 (dd, J = 8.5, 2.2 Hz, 1H),

6.85 (d, J = 6.4 Hz, 1H). MS (ESI): m/z 310.8 [M + H]⁺.

2-((3,5-Dichlorophenyl)thio)-1H-imidazo[4,5-c]pyridin-4-amine

(21e). Yield: 71%. ¹H NMR (500 MHz, MeOH-d₄): δ 7.47 (d, J = 6.7

Hz, 1H), 7.41−7.43 (m, 3H), 6.96 (d, J = 6.7 Hz, 1H). MS (ESI): m/z

310.8 [M + H]⁺.

2-((6-Iodobenzo[d][1,3]dioxol-5-yl)thio)-1H-imidazo[4,5-c]-

pyridin-4-amine (21f). Yield: 78%. MS (ESI): m/z 412.8 [M + H]⁺.

General Procedure for the Synthesis of 18d−f. To the appropriate

intermediate (21d−f; 0.0477 mmol) in dry DMF (1.5 mL) were

added Cs₂CO₃(0.0572 mmol) and 1,3-dibromopropane (0.238

mmol), and the reaction mixture was stirred at rt for 2 h. The solvent

was removed under reduced pressure and the crude product was

puriﬁed by preparatory TLC [CH₂Cl₂:MeOH−NH₃(7 N), 20:1] to

aﬀord corresponding alkylated derivatives (conﬁrmed by LC−MS). To

these 3-bromopropyl alkylated intermediates (0.0175 mmol) in dry

DMF was added isopropylamine (0.35 mmol), and the reaction

mixture was stirred at rt for 24 h. Then, the solvent was removed

under reduced pressure and the crude product was puriﬁed by

preparatory TLC [CH₂Cl₂:MeOH−NH₃(7 N), 15:1] to aﬀord the

desired ﬁnal products 18d−f.

2-((2,4-Dichlorophenyl)thio)-1-(3-(isopropylamino)propyl)-1H-

imidazo[4,5-c]pyridin-4-amine (18d). Yield: 35%. ¹H NMR (500

MHz, CDCl₃): δ 7.86 (d, J = 5.9 Hz, 1H), 7.43 (d, J = 2.2 Hz, 1H),

7.13 (dd, J = 8.5, 2.2 Hz, 1H), 7.02 (d, J = 8.6 Hz, 1H), 6.75 (d, J = 5.9

Hz, 1H), 5.29 (br s, 2H), 4.25 (t, J = 7.1 Hz, 2H), 2.71 (m, 1H), 2.56

(t, J = 6.8 Hz, 2H), 1.88 (m, 2H), 1.02 (d, J = 6.3 Hz, 6H).

2-((3,5-Dichlorophenyl)thio)-1-(3-(isopropylamino)propyl)-1H-

imidazo[4,5-c]pyridin-4-amine (18e). Yield: 39%. ¹H NMR (500

MHz, CDCl₃): δ 7.88 (d, J = 5.9 Hz, 1H), 7.21−7.25 (m, 3H), 6.75 (d,

J = 5.9 Hz, 1H), 5.26 (br s, 2H), 4.26 (t, J = 7.1 Hz, 2H), 2.71 (m,

1H), 2.56 (t, J = 6.8 Hz, 2H), 1.88 (m, 2H), 1.02 (d, J = 6.2 Hz, 6H).

¹³C NMR (150 MHz, CDCl₃): δ 151.3, 142.7, 141.7, 140.8, 135.9,

8-((2,4-Dimethylphenyl)thio)-3-(3-(isopropylamino)propyl)-3H-

1

purin-6-amine (18b′). Yield: 29%. H NMR (500 MHz, CDCl₃): δ

7.96 (s, 1H), 7.48 (d, J = 7.9 Hz, 1H), 7.11 (s, 1H), 7.00 (d, J = 7.8

Hz, 1H), 4.48 (t, J = 6.3 Hz, 2H), 2.91 (septet, J = 6.3 Hz, 1H), 2.65

(t, J = 6.1 Hz, 2H), 2.25−2.29 (m, 2H), 1.18 (d, J = 6.3 Hz, 6H).

HRMS (ESI): m/z [M + H]⁺calcd for C₁₉H₂₇N₆S 371.2018, found

371.2035.

8-((3,5-Dichlorophenyl)thio)-9-(3-(isopropylamino)propyl)-9H-

purin-6-amine (18c). Yield: 32%. ¹H NMR (600 MHz, CDCl₃/

MeOH-d₄): δ 8.21 (s, 1H), 7.26 (t, J = 1.7 Hz, 1H), 7.24 (d, J = 1.9

Hz, 2H), 4.23 (t, J = 6.9 Hz, 2H), 2.63 (septet, J = 6.2 Hz, 1H), 2.46

(t, J = 6.4 Hz, 2H), 1.89 (quintet, J = 6.9 Hz, 2H), 0.97 (d, J = 6.3 Hz,

6H). ¹³C NMR (150 MHz, CDCl₃): δ 154.7, 153.1, 151.3, 144.1,

135.9, 133.7, 128.8, 128.7, 119.7, 48.6, 43.4, 41.6, 29.8, 22.4. HRMS

(ESI): m/z [M + H]⁺calcd for C₁₇H₂₁Cl₂N₆S 411.0925, found

411.0917. HPLC: (a) t_R= 6.28 min, 99.5%.

8-((3,5-Dichlorophenyl)thio)-3-(3-(isopropylamino)propyl)-3H-

purin-6-amine (18c′). Yield: 18%. ¹H NMR (600 MHz, CDCl₃/

MeOH-d₄): δ 7.98 (s, 1H), 7.33 (d, J = 1.8 Hz, 2H), 7.06 (t, J = 1.8

Hz, 1H), 4.40 (t, J = 6.7 Hz, 2H), 2.66 (septet, J = 6.2 Hz, 1H), 2.51

(t, J = 6.4 Hz, 2H), 2.06 (quintet, J = 6.5 Hz, 2H), 0.96 (d, J = 6.2 Hz,

6H). ¹³C NMR (150 MHz, CDCl₃): δ 157.9, 153.4, 151.0, 142.8,

137.8, 134.9, 127.9, 126.8, 122.6, 48.9, 48.1, 43.0, 29.5, 22.9. HRMS

(ESI): m/z [M + H]⁺calcd for C₁₇H₂₁Cl₂N₆S 411.0925, found

411.0928. HPLC: (a) t_R= 4.95 min, 99.7%.

General Procedure for the Synthesis of 20d−f. To 4-amino-1-(4-

methoxybenzyl)-1H-imidazo[4,5-c]pyridine-2-thiol (19;²⁰50 mg,

0.174 mmol) were added the respective aryl iodide (0.348 mmol),

neocuproine hydrate (3.6 mg, 0.0174 mmol), CuI (3.3 mg, 0.0174

mmol), sodium tert-butoxide (25 mg, 0.261 mmol), and DMF (5 mL),

and the reaction mixture was stirred at 110 °C for 24 h. Then, the

solvent was removed under reduced pressure and the crude product

was puriﬁed by preparatory TLC [CH₂Cl₂:MeOH−NH₃(7 N), 10:1]

to aﬀord desired coupling derivatives 20d−f.

2-((2,4-Dichlorophenyl)thio)-1-(4-methoxybenzyl)-1H-imidazo-

[4,5-c]pyridin-4-amine (20d). Yield: 40%. ¹H NMR (500 MHz,

CDCl₃/MeOH-d₄): δ 7.78 (d, J = 5.9 Hz, 1H), 7.36 (d, J = 2.1 Hz,

1H), 7.02 (dd, J = 8.5, 2.1 Hz, 1H), 6.98 (d, J = 8.6 Hz, 2H), 6.91 (d, J

= 8.6 Hz, 1H), 6.73 (d, J = 8.6 Hz, 2H), 6.65 (d, J = 6.0 Hz, 1H), 5.30

(s, 2H), 3.75 (s, 3H). MS (ESI): m/z 430.9 [M + H]⁺.

135.7, 127.8, 127.5, 127.0, 97.6, 48.8, 44.0, 43.3, 30.5, 22.8. HRMS

(ESI): m/z [M + H]⁺calcd for C₁₈H₂₂N₅SCl₂410.0973, found

410.0978.

2-((6-Iodobenzo[d][1,3]dioxol-5-yl)thio)-1-(3-(isopropylamino)-

propyl)-1H-imidazo[4,5-c]pyridin-4-amine (18f). Yield: 40%. ¹H

NMR (600 MHz, CDCl₃): δ 7.83 (d, J = 5.9 Hz, 1H), 7.26 (s, 1H),

6.77 (s, 1H), 6.73 (d, J = 5.9 Hz, 1H), 5.97 (s, 2H), 5.31 (br s, 2H),

4.23 (t, J = 7.2 Hz, 2H), 2.74−2.77 (m, 1H), 2.60 (t, J = 6.9 Hz, 2H),

1.90−1.93 (m, 2H), 1.06 (d, J = 6.2 Hz, 6H). ¹³C NMR (150 MHz,

CDCl₃): δ 151.2, 149.3, 148.6, 141.3, 140.9, 129.2, 127.4, 118.9, 111.2,

102.2, 97.5, 89.3, 48.9, 44.0, 43.3, 30.4, 22.8. HRMS (ESI): m/z [M +

H]⁺calcd for C₁₉H₂₃IN₅O₂S 512.0617, found 512.0628.

Reagents. Recombinant Hsp90α (ADI-SPP-776), Hsp90β (ADI-

SPP-777), and Trap-1 (ADI-SPP-848) were purchased from Enzo Life

Sciences. Canine Grp94 was generated as previously reported.⁵⁰

FP Measurements. The Hsp90 FP competition assays were

performed on an Analyst GT instrument (Molecular Devices,

Sunnyvale, CA) and carried out in black 96-well microplates (Corning

# 3650) in a total volume of 100 μL in each well. A stock of 10 μM

Cy3B-GM²¹and PU-FITC3²²was prepared in DMSO and diluted

with Felts buﬀer [20 mM Hepes (K), pH 7.3, 50 mM KCl, 2 mM

DTT, 5 mM MgCl₂, 20 mM Na₂MoO₄, and 0.01% NP40 with 0.1 mg/

mL BGG]. To each well was added the ﬂuorescent dye labeled Hsp90

ligand (6 nM Cy3B-GM for Hsp90α, Hsp90β, and Grp94 and 3 nM

PU-FITC3 for Trap-1), protein (10 nM Hsp90α, 10 nM Hsp90β, 10

nM Grp94, 30 nM Trap-1) and the tested inhibitor (initial stock in

DMSO) in a ﬁnal volume of 100 μL of Felts buﬀer. Compounds were

added in duplicate or triplicate wells. For each assay, background wells

(buﬀer only), tracer controls (free, ﬂuorescent dye labeled Hsp90

ligand only), and bound controls (ﬂuorescent dye labeled Hsp90

ligand in the presence of protein) were included on each assay plate.

The assay plate was incubated on a shaker at 4 °C for 24 h and the FP

values (in mP) were measured. The fraction of ﬂuorescent dye labeled

2-((3,5-Dichlorophenyl)thio)-1-(4-methoxybenzyl)-1H-imidazo-

[4,5-c]pyridin-4-amine (20e). Yield: 39%. MS (ESI): m/z 431.0 [M +

H]⁺.

2-((6-Iodobenzo[d][1,3]dioxol-5-yl)thio)-1-(4-methoxybenzyl)-

1H-imidazo[4,5-c]pyridin-4-amine (20f). Yield: 20%. MS (ESI): m/z

533.0 [M + H]⁺.

General Procedure for the Synthesis of 21d−f. To coupling

products 20d−f (0.067 mmol) was added triﬂuoroacetic acid (3 mL),

and the reaction mixture was stirred at 80 °C for 3 h. Then, the solvent

was removed under reduced pressure and the crude product was

3939

DOI: 10.1021/acs.jmedchem.5b00197

J. Med. Chem. 2015, 58, 3922−3943

Journal of Medicinal Chemistry

Article

Hsp90 ligand bound to Hsp90 was correlated to the mP value and

plotted against values of competitor concentrations. The inhibitor

concentration at which 50% of bound ﬂuorescent dye labeled Hsp90

ligand was displaced was obtained by ﬁtting the data. For Cy3B-GM,

an excitation ﬁlter at 530 nm and an emission ﬁlter at 580 nm were

used with a dichroic mirror of 561 nm. For PU-FITC3, an excitation

ﬁlter at 485 nm and an emission ﬁlter at 530 nm were used with a

dichroic mirror of 505 nm. All experimental data were analyzed using

SOFTmax Pro 4.3.1 and plotted using Prism 4.0 (Graphpad Software

Inc., San Diego, CA).

Molecular Modeling. All computations were carried out on a HP

workstation xw8200 with the Ubuntu 8.10 operating system using

Maestro v8.5. Protein sequence and crystal structures were down-

loaded from NCBI and RCSB databases, respectively.

Ligand Preparation. All the compounds were constructed using

the fragment dictionary of Maestro 8.5. The geometry of compounds

was optimized using the Macromodel program v9.6 and the OLPS-AA

force ﬁeld.⁵¹Ensuing ligands were further prepared using the Ligprep

HEK293 cells were purchased from ATCC and cultured in DMEM in

the presence of 10% FBS and 1% penicillin/streptomycin. RAW 264.7

cells was purchased from ATCC and cultured in DMEM with 2 mM ^L-

glutamine, 10 mM HEPES, 1 mM sodium pyruvate, 10% low

endotoxin FBS, and 1% Pen/Strep. When cultured, cells were

incubated in the humidiﬁed cell incubators with CO₂at 37 °C.

Immunoblotting. Cells were either treated with DMSO (vehicle)

or indicated compounds for 24 h and lysed in RIPA buﬀer (50 mM

Tris-HCl, pH 7.4, 150 mM NaCl, 0.5% sodium deoxycholate, and

0.5% NP40) supplemented with cocktail protease inhibitors (Roche)

to produce whole-cell lysates. Protein concentrations were determined

using the BCA kit (Pierce) according to the manufacturer’s

instructions. The protein lysates (5−50 μg) were electrophoretically

resolved by SDS−PAGE, transferred onto nitrocellulose membranes,

and probed with primary antibodies against HER2 (Zymed, 28004;

1:500), Hsp70 (Stressgen, SPA-810; 1:2500), cleaved PARP

(Promega, G7341; 1:2500), and β-actin (Sigma, A1978; 1:3000).

After washing oﬀ the excess antibodies, the membranes were incubated

with the corresponding horseradish peroxidase (HRP)-conjugated

secondary antibody. Blots were visualized by autoradiography using

the Enhanced Chemiluminescence Detection System (GE Healthcare)

according to the manufacturer’s instructions. Films were scanned in

Adobe Photoshop CS5, and quantitative densitometric analysis was

performed using ImageJ (US National Institutes of Health).

(v2.2) utility provided by Schrodinger LLC.

̈

Docking. The X-ray crystal structure of PU-H54-bound Grp94

(PDB ID: 3O2F) was ﬁrst aligned using a protein structure alignment

tool and then optimized for subsequent grid generation and docking

using the default parameters in Protein Preparation Wizard provided

by Schrodinger LLC. Grids were then prepared using the Receptor

̈

Grid Generation tool in Glide 5.0.^52,53Next, the extraprecision (XP)

Glide docking method was used to dock compounds ﬂexibly into the

ATP binding site of Grp94 according to the previously reported

procedure.⁵⁴Upon completion of each docking calculation, at most

100 poses per docking were run and at most 10 poses per ligand were

allowed to be generated. Top-scored docking poses (orientation plus

conformation) based on the Glide scoring (Gscore) function were

analyzed. Docking accuracy was measured by resemblance of the

lowest energy pose (binding conformation) predicted by the object

scoring function Gscore with the experimental binding mode as

determined by X-ray crystallography. Bound inhibitors PU-H54 (4)

was removed from the Grp94 (PDB ID: 3O2F) binding site and

redocked into the respective binding sites. Very good agreement was

found between the localization of the inhibitor from docking and from

the crystal structure with a RMSD of 0.04 Å (PDB ID: 3O2F) between

the predicted conformation and the observed X-ray crystallographic

conformation, validating the docking strategy.

Kinase Assays. For most assays, kinase-tagged T7 phage strains

were grown in parallel in 24-well blocks in an Escherichia coli host

derived from the BL21 strain. E. coli cells were grown to log-phase and

infected with T7 phage from a frozen stock (multiplicity of infection =

0.4) and incubated with shaking at 32 °C until lysis (90−150 min).

The lysates were centrifuged (6000g) and ﬁltered (0.2 μm) to remove

cell debris. The remaining kinases were produced in HEK-293 cells

and subsequently tagged with DNA for qPCR detection. Streptavidin-

coated magnetic beads were treated with biotinylated small molecule

ligands for 30 min at room temperature to generate aﬃnity resins for

kinase assays. The liganded beads were blocked with excess biotin and

washed with blocking buﬀer [SeaBlock (Pierce), 1% BSA, 0.05%

Tween 20, 1 mM DTT] to remove unbound ligand and to reduce

nonspeciﬁc phage binding. Binding reactions were assembled by

combining kinases, liganded aﬃnity beads, and test compounds in 1×

binding buﬀer (20% SeaBlock, 0.17× PBS, 0.05% Tween 20, 6 mM

DTT). Test compounds were prepared as 40× stocks in 100% DMSO

and directly diluted into the assay. All reactions were performed in

polypropylene 384-well plates in a ﬁnal volume of 0.04 mL. The assay

plates were incubated at room temperature with shaking for 1 h, and

the aﬃnity beads were washed with wash buﬀer (1× PBS, 0.05%

Tween 20). The beads were then resuspended in elution buﬀer (1×

PBS, 0.05% Tween 20, 0.5 μM nonbiotinylated aﬃnity ligand) and

incubated at room temperature with shaking for 30 min. The kinase

concentration in the eluates was measured by qPCR.

Cell Viability Assessment. Cells were treated for 72 h with

vehicle or inhibitors. Cell viability was assessed using CellTiter-Glo

luminescent cell viability assay (Promega). The method determines

the number of viable cells in culture based on quantiﬁcation of the

ATP present, which signals the presence of metabolically active cells.

Drugs were added to quadruplicate wells.

C2C12 Diﬀerentiation and IGF-II Secretion Assay. C2C12 is

an immortal line of mouse skeletal myoblasts originally derived from

satellite cells from the thigh muscle of a 2-month-old female mouse

donor. These cells diﬀerentiate well into myocytes under appropriate

culture conditions. The cells were induced to diﬀerentiate by replacing

the culture medium with DMEM supplemented with 2% horse serum

and 1% penicillin/streptomycin (diﬀerentiation medium) for 36−48 h.

Secreted IGF-II was quantiﬁed by using an IGF-II mouse ELISA kit

(Abcam, AB100696) according to the manufacturer’s instructions.

Brieﬂy, after shifting the culture medium to diﬀerentiation medium,

C2C12 cells were treated for 24 h with the indicated compounds.

Medium from each experimental condition was then transferred into

ELISA plates coated with anti-IGF-II and incubated overnight at 4 °C.

The bound IGF-II was detected with a biotinylated anti-IGF-II

antibody provided in the kit. After the sequential incubation with

HRP-conjugated streptavidin, TMB one-step substrate reagent, and

the stop solution, the absorbance was measured at 450 nm. The

secreted IGF-II was quantiﬁed against a standard curve generated with

recombinant IGF-II provided by the kit.

TLR9 Assay. HEK 293T cells were transfected with pUNO-

hTLR9-HA (Invivogen) using X-tremgene HP (Roche) according to

the manufacturer’s instructions. At 24 h after transfection, cells were

split onto cell culture chamber slides (Lab-Tek). Cells were then

treated for 24 h with compounds at varying concentrations. After

treatment, cells were ﬁxed for 20 min in 4% paraformaldehyde in PBS,

permeabilized with 0.1% Triton X-100 in PBS for 10 min, and blocked

with 3% BSA in PBS for 30 min, followed by staining for 1 h with anti-

HA (Abcam, ab9110; 1:250) or a normal rabbit IgG (Abcam,

ab37415; 1:250). Cells were washed with PBS, stained with an anti-

rabbit-Cy3 antibody (Invitrogen, A10520; 1:400), and ﬁnally mounted

in the dark at 4 °C with with one drop of Prolong Gold Antifade

reagent (with DAPI, Life Technologies, P36935). Cells were visualized

under a confocal microscope (Leica Upright Confocal SP5).

Fluorescence intensity was quantiﬁed using MetaMorph Microscopy

Automation and Image Analysis Software (Molecular Devices Inc.)

and normalized to the cell number.

Cell Lines. The HER2-overexpressing breast cancer cells SKBr3

were obtained from the American Type Culture Collection (ATCC).

Cells were cultured routinely in McCoy’s 5A supplemented with 10%

FBS and 1% penicillin and streptomycin (Pen/Strep). C2C12 and

TNFα ELISA. Raw264.7 cells were stimulated with 10 ng/mL LPS

or 2.5 μM ODN1585 (G*G*G* GTC AAC GTT GAG G*G*G*

G*G), followed by the 18 h treatments with the indicated inhibitors.

3940

DOI: 10.1021/acs.jmedchem.5b00197

J. Med. Chem. 2015, 58, 3922−3943

Journal of Medicinal Chemistry

Article

The mouse TNFα ELISA MAX Set (BioLegend) was used to quantify

the TNFα according to the manufacturer’s instructions.

(4) Altieri, D. C.; Stein, G. S.; Lian, J. B.; Languino, L. R. TRAP-1,

the mitochondrial Hsp90. Biochim. Biophys. Acta 2012, 1823, 767−

773.

(5) Kang, B. H.; Plescia, J.; Dohi, T.; Rosa, J.; Doxsey, S. J.; Altieri, D.

C. Regulation of tumor cell mitochondrial homeostasis by an

organelle-specific Hsp90 chaperone network. Cell 2007, 131, 257−

270.

Animal Studies. Four- to 6-week-old nu/nu athymic female mice

were obtained from Harlan Laboratories. Experiments were carried out

under an Institutional Animal Care and Use Committee approved

protocol, and institutional guidelines for the proper and humane use of

animals in research were followed. MDA-MB-468 cells (1 × 10⁷) were

sc implanted in the right ﬂank of mice using a 20-gauge needle and

allowed to grow. All mice received Augmentin (amoxicillin/clavulanate

potassium; SmithKline Beecham) in their drinking water while on

therapy. Mice were sacriﬁced by CO₂euthanasia.

Pharmacodynamic (PD) and Pharmacokinetic (PK) Studies.

18c (75 or 150 mg/kg, ip) was administered to mice (n = 4) bearing

breast tumors, and mice were sacriﬁced at 2 and 24 h

postadministration. Drug levels in tissues and plasma were determined

by LC−MS/MS. Compound 18c was injected as a single dose

intravenously or orally at 10 mg/kg to B6D2F1 mice (n = 2 male mice

per experimental point, 20 mice total). The agent was formulated in

30% captisol, 5 mM citrate buﬀer at pH 4.2. Drug levels in plasma

were determined by LC−MS/MS.

(6) Costantino, E.; Maddalena, F.; Calise, S.; Piscazzi, A.; Tirino, V.;

Fersini, A.; Ambrosi, A.; Neri, V.; Esposito, F.; Landriscina, M. TRAP1,

a novel mitochondrial chaperone responsible for multi-drug resistance

and protection from apoptotis in human colorectal carcinoma cells.

Cancer Lett. 2009, 279, 39−46.

(7) Lee, A. S. Glucose-regulated proteins in cancer: Molecular

mechanisms and therapeutic potential. Nat. Rev. Cancer 2014, 14,

263−276.

(8) Yang, Y.; Liu, B.; Dai, J.; Srivastava, P. K.; Zammit, D. J.;

Lefranco̧ is, L.; Li, Z. Heat shock protein gp96 is a master chaperone

for toll-like receptors and is important in the innate function of

macrophages. Immunity 2007, 26, 215−226.

(9) Hua, Y.; White-Gilbertson, S.; Kellner, J.; Rachidi, S.; Usmani, S.

Z.; Chiosis, G.; Depinho, R.; Li, Z.; Liu, B. Molecular chaperone gp96

is a novel therapeutic target of multiple myeloma. Clin. Cancer Res.

2013, 19, 6242−6251.

(10) Patel, P. D.; Yan, P.; Seidler, P. M.; Patel, H. J.; Sun, W.; Yang,

C.; Que, N. S.; Taldone, T.; Finotti, P.; Stephani, R. A.; Gewirth, D.

T.; Chiosis, G. Paralog-selective Hsp90 inhibitors define tumor-specific

regulation of HER2. Nat. Chem. Biol. 2013, 9, 677−684.

(11) Soldano, K. L.; Jivan, A.; Nicchitta, C. V.; Gewirth, D. T.

Structure of the N-terminal domain of GRP94. J. Biol. Chem. 2003,

279, 48330−48338.

(12) Rosser, M. F.; Nicchitta, C. V. Ligand interactions in the

adenosine nucleotide-binding domain of the Hsp90 chaperone,

GRP94. I. Evidence for allosteric regulation of ligand binding. J. Biol.

Chem. 2000, 275, 22798−22805.

ASSOCIATED CONTENT

* Supporting Information

Experimental details for the synthesis of 18g and Figures S1−

S4. This material is available free of charge via the Internet at

http://pubs.acs.org.

■

S

AUTHOR INFORMATION

Corresponding Authors

*T.T.: phone, 646-888-3180. e-mail, taldonet@mskcc.org.

*G.C.: phone: 646-888-2235. e-mail: chiosisg@mskcc.org.

■

Author Contributions

H.J.P., P.D.P., S.O.O., P.Y., W.S., and M.R.P. contributed

equally.

(13) Immormino, R. M.; Dollins, D. E.; Shaffer, P. L.; Soldano, K. L.;

Walker, M. A.; Gewirth, D. T. Ligand-induced conformational shift in

the N-terminal domain of GRP94, an Hsp90 chaperone. J. Biol. Chem.

2004, 279, 46162−46171.

Notes

The authors declare no competing ﬁnancial interest.

(14) Immormino, R. M.; Metzger, L. E.; Reardon, P. N.; Dollins, D.

E.; Blagg, B. S.; Gewirth, D. T. Different poses for ligand and

chaperone in inhibitor-bound Hsp90 and GRP94: Implications for

paralog-specific drug design. J. Mol. Biol. 2009, 388, 1033−1042.

(15) Cristalli, G.; Lambertucci, C.; Taffi, S.; Vittori, S.; Volpini, R.

Medicinal chemistry of adenosine A2A receptor agonists. Curr. Top.

Med. Chem. 2003, 3, 387−401.

(16) Clevenger, R. C.; Blagg, B. S. Design, synthesis, and evaluation

of a radicicol and geldanamycin chimera, radamide. Org. Lett. 2004, 6,

4459−4462.

(17) Duerfeldt, A. S.; Peterson, L. B.; Maynard, J. C.; Ng, C. L.;

Eletto, D.; Ostrovsky, O.; Shinogle, H. E.; Moore, D. S.; Argon, Y.;

Nicchitta, C. V.; Blagg, B. S. Development of a Grp94 inhibitor. J. Am.

Chem. Soc. 2012, 134, 9796−9804.

(18) Stothert, A. R.; Suntharalingam, A.; Huard, D. J.; Fontaine, S.

N.; Crowley, V. M.; Mishra, S.; Blagg, B. S.; Lieberman, R. L.; Dickey,

C. A. Exploiting the interaction between Grp94 and aggregated

myocilin to treat glaucoma. Hum. Mol. Genet. 2014, 23, 6470−6480.

(19) Bao, R.; Lai, C. J.; Qu, H.; Wang, D.; Yin, L.; Zifcak, B.; Atoyan,

R.; Wang, J.; Samson, M.; Forrester, J.; DellaRocca, S.; Xu, G. X.; Tao,

X.; Zhai, H. X.; Cai, X.; Qian, C. CUDC-305, a novel synthetic HSP90

inhibitor with unique pharmacologic properties for cancer therapy.

Clin. Cancer Res. 2009, 15, 4046−4057.

ACKNOWLEDGMENTS

■

H.J.P. is funded by Susan G. Komen, Award # PDF14301695,

S.O.O. is supported by PF-14-154-01-CDD from the American

Cancer Society, P.Y. is supported by AACI Translational

Cancer Research Fellowship. G.C. is funded by the Leukemia

and Lymphoma Society, Breast Cancer Research Fund, U01

AG032969-01A1, R21 AI090501, R01 CA172546, and R01

CA155226. P.D.P and R.A.S. are supported by funds and

resources from St. John’s University. Z.L. is funded by R01

AI070603 and R01 AI077283, C.L. and J.B. by R03 AI097671

and R01 AI076588, and D.G. by R01 CA095130.

ABBREVIATIONS USED

■

Hsp90, heat shock protein 90; Grp94, glucose regulated protein

94; ER, endoplasmic reticulum; Trap-1, tumor necrosis factor

receptor-associated protein-1; Abs, antibodies; TLRs, toll-like

receptors; IgG, immunoglobulin; RA, rheumatoid arthritis;

HUVECs, human umbilical vein endothelial cells

REFERENCES

■

(20) Cai, X.; Qian, C.; Zhai, H. Preparation of imidazo[4,5-c]pyridine

derivatives as HSP90 inhibitors. Patent WO2008115719A1, 2008.

(21) Moulick, K.; Clement, C. C.; Aguirre, J.; Kim, J.; Kang, Y.; Felts,

S.; Chiosis, G. Synthesis of a red-shifted fluorescence polarization

probe for Hsp90. Bioorg. Med. Chem. Lett. 2006, 16, 4515−4518.

(22) Taldone, T.; Patel, P. D.; Patel, M.; Patel, H. J.; Evans, C. E.;

Rodina, A.; Ochiana, S.; Shah, S. K.; Uddin, M.; Gewirth, D.; Chiosis,

G. Experimental and structural testing module to analyze paralogue-

(1) Sorger, P. K.; Pelham, H. R. The glucose-regulated protein Grp94

is related to heat shock protein Hsp90. J. Mol. Biol. 1987, 194, 341−

344.

(2) Marzec, M.; Eletto, D.; Argon, Y. GRP94: An HSP90-like protein

specialized for protein folding and quality control in the endoplasmic

reticulum. Biochim. Biophys. Acta 2012, 1823, 774−787.

(3) Whitesell, L.; Lindquist, S. L. Hsp90 and the chaperoning of

cancer. Nat. Rev. Cancer 2005, 5, 761−772.

3941

DOI: 10.1021/acs.jmedchem.5b00197

J. Med. Chem. 2015, 58, 3922−3943

Journal of Medicinal Chemistry

Article

specificity and affinity in the Hsp90 inhibitors series. J. Med. Chem.

2013, 56, 6803−6818.

(41) Matts, R. L.; Brandt, G. E.; Lu, Y.; Dixit, A.; Mollapour, M.;

Wang, S.; Donnelly, A. C.; Neckers, L.; Verkhivker, G.; Blagg, B. S. A

systematic protocol for the characterization of Hsp90 modulators.

Bioorg. Med. Chem. 2011, 19, 684−692.

(42) Moulick, K.; Ahn, J. H.; Zong, H.; Rodina, A.; Cerchietti, L.;

Gomes DaGama, E. M.; Caldas-Lopes, E.; Beebe, K.; Perna, F.; Hatzi,

K.; Vu, L. P.; Zhao, X.; Zatorska, D.; Taldone, T.; Smith-Jones, P.;

Alpaugh, M.; Gross, S. S.; Pillarsetty, N.; Ku, T.; Lewis, J. S.; Larson, S.

M.; Levine, R.; Erdjument-Bromage, H.; Guzman, M. L.; Nimer, S. D.;

Melnick, A.; Neckers, L.; Chiosis, G. Affinity-based proteomics reveal

cancer-specific networks coordinated by Hsp90. Nat. Chem. Biol. 2011,

7, 818−826.

(43) Beebe, K.; Mollapour, M.; Scroggins, B.; Prodromou, C.; Xu,

W.; Tokita, M.; Taldone, T.; Pullen, L.; Zierer, B. K.; Lee, M. J.;

Trepel, J.; Buchner, J.; Bolon, D.; Chiosis, G.; Neckers, L.

Posttranslational modification and conformational state of heat

shock protein 90 differentially affect binding of chemically diverse

small molecule inhibitors. Oncotarget 2013, 4, 1065−1074.

(44) Mollapour, M.; Bourboulia, D.; Beebe, K.; Woodford, M. R.;

Polier, S.; Hoang, A.; Chelluri, R.; Li, Y.; Guo, A.; Lee, M. J.; Fotooh-

Abadi, E.; Khan, S.; Prince, T.; Miyajima, N.; Yoshida, S.; Tsutsumi, S.;

Xu, W.; Panaretou, B.; Stetler-Stevenson, W. G.; Bratslavsky, G.;

Trepel, J. B.; Prodromou, C.; Neckers, L. Asymmetric Hsp90 N

domain SUMOylation recruits Aha1 and ATP-competitive inhibitors.

Mol. Cell 2014, 53, 317−329.

(45) Kamal, A.; Thao, L.; Sensintaffar, J.; Zhang, L.; Boehm, M. F.;

Fritz, L. C.; Burrows, F. J. A high-affinity conformation of Hsp90

confers tumour selectivity on Hsp90 inhibitors. Nature 2003, 425,

407−410.

(46) Cerchietti, L. C.; Lopes, E. C.; Yang, S. N.; Hatzi, K.; Bunting,

K. L.; Tsikitas, L. A.; Mallik, A.; Robles, A. I.; Walling, J.; Varticovski,

L.; Shaknovich, R.; Bhalla, K. N.; Chiosis, G.; Melnick, A. A purine

scaffold Hsp90 inhibitor destabilizes BCL-6 and has specific antitumor

activity in BCL-6-dependent B cell lymphomas. Nat. Med. 2009, 15,

1369−1376.

(47) Breinig, M.; Caldas-Lopes, E. M.; Goeppert, B.; Malz, M.;

Bergmann, F.; Rieker, R.; Schirmacher, P.; Chiosis, G.; Kern, M. A.

Targeting the heat shock protein Hsp90 with non-quinone inhibitors:

A novel chemotherapeutic approach in human hepatocellular

carcinoma. Hepatology 2009, 50, 102−112.

(48) Caldas-Lopes, E.; Cerchietti, L.; Ahn, J. H.; Clement, C. C.;

Robles, A. I.; Rodina, A.; Moulick, K.; Taldone, T.; Gozman, A.; Guo,

Y.; Wu, N.; de Stanchina, E.; White, J.; Gross, S. S.; Ma, Y.;

Varticovski, L.; Melnick, A.; Chiosis, G. Hsp90 inhibitor PU-H71, a

multimodal inhibitor of malignancy, induces complete responses in

triple-negative breast cancer models. Proc. Natl. Acad. Sci. U. S. A.

2009, 106, 8368−8373.

(23) He, H.; Llauger, L.; Rosen, N.; Chiosis, G. General method for

the synthesis of 8-arylsulfanyl adenine derivatives. J. Org. Chem. 2004,

69, 3230−3232.

(24) Llauger, L.; He, H.; Kim, J.; Aguirre, J.; Rosen, N.; Peters, U.;

Davies, P.; Chiosis, G. Evaluation of 8-arylsulfanyl, 8-arylsulfoxyl, and

8-arylsulfonyl adenine derivatives as inhibitors of the heat shock

protein 90. J. Med. Chem. 2005, 48, 2892−2905.

(25) Biamonte, M. A.; Shi, J.; Hong, K.; Hurst, D. C.; Zhang, L.; Fan,

J.; Busch, D. J.; Karjian, P. L.; Maldonado, A. A.; Sensintaffar, J. L.;

Yang, Y. C.; Kamal, A.; Lough, R. E.; Lundgren, K.; Burrows, F. J.;

Timony, G. A.; Boehm, M. F.; Kasibhatla, S. R. Orally active purine-

based inhibitors of the heat shock protein 90. J. Med. Chem. 2006, 49,

817−828.

(26) Taldone, T.; Patel, P. D.; Patel, H. J.; Chiosis, G. About the

reaction of aryl fluorides with sodium sulfide: Investigation into the

selectivity of substitution of fluorobenzonitriles to yield mercapto-

benzonitriles via SNAr displacement of fluorine. Tetrahedron Lett.

2012, 53, 2548−2551.

(27) Tao, H.; Kang, Y.; Taldone, T.; Chiosis, G. Microwave-assisted

one step synthesis of 8-arylmethyl-9H-purin-6-amines. Bioorg. Med.

Chem. Lett. 2009, 19, 415−417.

(28) He, H.; Zatorska, D.; Kim, J.; Aguirre, J.; Llauger, L.; She, Y.;

Wu, N.; Immormino, R. M.; Gewirth, D. T.; Chiosis, G. Identification

of potent water soluble purine-scaffold inhibitors of the heat shock

protein 90. J. Med. Chem. 2006, 49, 381−390.

(29) Chiosis, G.; Lucas, B.; Shtil, A.; Huezo, H.; Rosen, N.

Development of a purine-scaffold novel class of Hsp90 binders that

inhibit the proliferation of cancer cells and induce the degradation of

Her2 tyrosine kinase. Bioorg. Med. Chem. 2002, 10, 3555−3564.

(30) Immormino, R. M.; Kang, Y.; Chiosis, G.; Gewirth, D. T.

Structural and quantum chemical studies of 8-aryl-sulfanyl adenine

class Hsp90 inhibitors. J. Med. Chem. 2006, 49, 4953−4960.

(31) Fabian, M. A.; Biggs, W. H., 3rd; Treiber, D. K.; Atteridge, C. E.;

Azimioara, M. D.; Benedetti, M. G.; Carter, T. A.; Ciceri, P.; Edeen, P.

T.; Floyd, M.; Ford, J. M.; Galvin, M.; Gerlach, J. L.; Grotzfeld, R. M.;

́

Herrgard, S.; Insko, D. E.; Insko, M. A.; Lai, A. G.; Lelias, J. M.; Mehta,

S. A.; Milanov, Z. V.; Velasco, A. M.; Wodicka, L. M.; Patel, H. K.;

Zarrinkar, P. P.; Lockhart, D. J. A small molecule-kinase interaction

map for clinical kinase inhibitors. Nat. Biotechnol. 2005, 23, 329−336.

(32) Taldone, T.; Ochiana, S.; Patel, P. D.; Chiosis, G. Selective

targeting of the stress chaperome as a therapeutic strategy. Trends

Pharmacol. Sci. 2014, 35, 592−603.

(33) Medzhitov, R. Toll-like receptors and innate immunity. Nat. Rev.

Immunol. 2001, 1, 135−145.

(34) Brooks, J. C.; Sun, W.; Chiosis, G.; Leifer, C. A. Heat shock

protein gp96 regulates toll-like receptor 9 proteolytic processing and

conformational stability. Biochem. Biophys. Res. Commun. 2012, 421,

780−784.

(35) Ostrovsky, O.; Ahmed, N. T.; Argon, Y. The chaperone activity

of GRP94 towards of insulin-like growth factor II is necessary for the

stress response to serum deprivation. Mol. Biol. Cell 2009, 20, 1855−

1864.

(49) Rodina, A.; Vilenchik, M.; Moulick, K.; Aguirre, J.; Kim, J.;

Chiang, A.; Litz, J.; Clement, C. C.; Kang, Y.; She, Y.; Wu, N.; Felts, S.;

Wipf, P.; Massague, J.; Jiang, X.; Brodsky, J. L.; Krystal, G. W.; Chiosis,

G. Selective compounds define Hsp90 as a major inhibitor of apoptosis

in small cell lung cancer. Nat. Chem. Biol. 2007, 3, 498−507.

(50) Dollins, D. E.; Immormino, R. M.; Gewirth, D. T. Structure of

unliganded GRP94, the endoplasmic reticulum Hsp90. Basis for

nucleotide-induced conformational change. J. Biol. Chem. 2005, 280,

30438−30447.

(36) Neckers, L. Heat shock protein 90: The cancer chaperone. J.

Biosci. 2007, 32, 517−530.

(37) Supko, J. G.; Hickman, R. L.; Grever, M. R.; Malspeis, L.

Preclinical pharmacologic evaluation of geldanamycin as an antitumor

agent. Cancer Chemother. Pharmacol. 1995, 36, 305−315.

(38) Neckers, L. Chaperoning oncogenes: Hsp90 as a target of

geldanamycin. Handb. Exp. Pharmacol 2006, 259−277.

(39) Jhaveri, K.; Ochiana, S. O.; Dunphy, M. P.; Gerecitano, J. F.;

Corben, A. D.; Peter, R. I.; Janjigian, Y. Y.; Gomes-DaGama, E. M.;

Koren, J., 3rd; Modi, S.; Chiosis, G. Heat shock protein 90 inhibitors

in the treatment of cancer: Current status and future directions. Expert

Opin. Invest. Drugs 2014, 23, 611−628.

(51) Jorgensen, W. L.; Maxwell, D.; Tirado-Rives, J. Development

and testing of the OPLS-All Atom force field on conformational

energetics and properties of organic liquids. J. Am. Chem. Soc. 1996,

118, 11225−11236.

(52) Friesner, R. A.; Banks, J. L.; Murphy, R. B.; Halgren, T. A.;

Klicic, J. J.; Mainz, D. T.; Repasky, M. P.; Knoll, E. H.; Shelley, M.;

Perry, J. K.; Shaw, D. E.; Francis, P.; Shenkin, P. S. Glide: A new

approach for rapid, accurate docking and scoring. 1. Method and

assessment of docking accuracy. J. Med. Chem. 2004, 47, 1739−1749.

(53) Halgren, T. A.; Murphy, R. B.; Friesner, R. A.; Beard, H. S.;

Frye, L. L.; Pollard, W. T.; Banks, J. L. Glide: A new approach for

rapid, accurate docking and scoring. 2. Enrichment factors in database

screening. J. Med. Chem. 2004, 47, 1750−1759.

(40) Patel, H. J.; Modi, S.; Chiosis, G.; Taldone, T. Advances in the

discovery and development of heat-shock protein 90 inhibitors for

cancer treatment. Expert Opin. Drug Discovery 2011, 6, 559−587.

3942

DOI: 10.1021/acs.jmedchem.5b00197

J. Med. Chem. 2015, 58, 3922−3943

Journal of Medicinal Chemistry

Article

(54) Patel, P. D.; Patel, M. R.; Kaushik-Basu, N.; Talele, T. T. 3D

QSAR and molecular docking studies of benzimidazole derivatives as

hepatitis C virus NS5B polymerase inhibitors. J. Chem. Inf. Model.

2008, 48, 42−55.

3943

DOI: 10.1021/acs.jmedchem.5b00197

J. Med. Chem. 2015, 58, 3922−3943

Article Doi

Structure-activity relationship in a purine-scaffold compound series with selectivity for the endoplasmic reticulum Hsp90 paralog Grp94

DOI: 10.1021/acs.jmedchem.5b00197

Source and publish data:

Authors:

Article abstract of DOI:10.1021/acs.jmedchem.5b00197

Full text of DOI:10.1021/acs.jmedchem.5b00197

Products guided by the article

R&D Labs maybe for 1360066-73-4

Relevant to this article

Hot Product