Design and Synthesis of Fungal-Selective Resorcylate Aminopyrazole Hsp90 Inhibitors

Article abstract of DOI:10.1021/acs.jmedchem.9b00826

The molecular chaperone Hsp90, essential in all eukaryotes, plays a multifaceted role in promoting survival, virulence, and drug resistance across diverse pathogenic fungal species. The chaperone is also critically important, however, to the pathogen's hu

Full text of DOI:10.1021/acs.jmedchem.9b00826

Subscriber access provided by University of Texas Libraries
Article
Design and synthesis of fungal-selective
resorcylate aminopyrazole Hsp90 inhibitors
David Huang, Emmanuelle LeBlanc, Tanvi Shekhar-Guturja, Nicole Robbins, Damian
Krysan, Juan Carlos Pizarro, Luke Whitesell, Leah Cowen, and Lauren E. Brown
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b00826 • Publication Date (Web): 12 Sep 2019
Downloaded from pubs.acs.org on September 12, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 163
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Design and synthesis of fungal-selective resorcylate
aminopyrazole Hsp90 inhibitors
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
A†‡
B‡
B
B
David S. Huang , Emmanuelle V. LeBlanc , Tanvi Shekhar-Guturja , Nicole Robbins ,
C
D
B
B
A
Damian J. Krysan , Juan Pizarro , Luke Whitesell *, Leah E. Cowen * and Lauren E. Brown *
A
Department of Chemistry and Center for Molecular Discovery (BU-CMD), Boston University,
Boston, MA, 02215, USA
B
Department of Molecular Genetics, University of Toronto, Toronto, Ontario, M5G 1M1,
Canada.
C
Departments of Pediatrics and Microbiology/Immunology, Carver College of Medicine,
University of Iowa, Iowa City, Iowa, 52242, USA
D
Department of Tropical Medicine, School of Public Health and Tropical Medicine and Vector-
Borne Infectious Disease Research Center, Tulane University, New Orleans, LA, 70112, USA.
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 163
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ABSTRACT
The molecular chaperone Hsp90, essential in all eukaryotes, plays a multifaceted role in promoting
survival, virulence and drug resistance across diverse pathogenic fungal species. The chaperone is
also critically important, however, to the pathogen’s human host, preventing the use of known
clinical Hsp90 inhibitors in antifungal applications due to concomitant host toxicity issues. With
the goal of developing Hsp90 inhibitors with acceptable therapeutic indices for the treatment of
invasive fungal infections, we initiated a program to design and synthesize potent inhibitors with
selective activity against fungal Hsp90 isoforms over their human counterparts. Building on our
previously-reported derivatization of resorcylate natural products to produce fungal-selective
compounds, we have developed a series of synthetic aminopyrazole-substituted resorcylate amides
with broad, potent, and fungal-selective Hsp90 inhibitory activity. Herein we describe the
synthesis of this series, as well as biochemical structure-activity relationships driving selectivity
for the Hsp90 isoforms expressed by Cryptococcus neoformans and Candida albicans, two
pathogenic fungi of major clinical importance.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
INTRODUCTION
The morbidity and mortality caused by fungal infections cripple human health across the
globe. Over a billion people are affected by superficial infections, such as ringworm and athlete’s
foot. Adding to these numbers are the burden of oral and other mucosal infections. Of most concern
is the increasing number of invasive systemic infections, which leads to over one million deaths
1
each year. People with compromised immune function, such as patients receiving cancer
chemotherapies, organ transplant recipients and those infected with HIV, are most vulnerable to
invasive fungal infections. The pathogens responsible for > 90% of invasive mycoses are Candida
albicans, Aspergillus fumigatus and Cryptococcus neoformans. Once diagnosed, treatment options
2
ACS Paragon Plus Environment

Page 3 of 163
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
are limited to only three major classes of antifungal drugs, notoriously hampered by problems with
2
host toxicity, the emergence of resistance, or limited spectrum of activity. In fact, the only new
class of antifungals to reach the clinic in decades has no efficacy against C. neoformans and related
species.3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Selective targeting of fungal stress responses provides a promising therapeutic strategy to
mitigate resistance and more effectively combat invasive mycoses. The essential molecular
chaperone Hsp90 has been extensively validated as a regulator of virulence and antifungal drug
resistance in Candida and Aspergillus species.4, 5 For instance, in C. albicans, genetic depletion or
pharmacological inhibition of Hsp90 increases the efficacy of current antifungal drugs, reduces
acquired antifungal resistance in clinical isolates, and improves clearance in a mouse model of
6
disseminated candidiasis. Recent studies have demonstrated the critical importance of Hsp90 for
C. neoformans thermotolerance and shown that Hsp90 inhibition alters capsule assembly and
7
, 8
sensitivity to antifungals, influencing virulence of the pathogen. While targeting Hsp90 offers a
promising but relatively unexplored strategy for antifungal drug development, the chaperone has
been intensively explored as a target in oncology. A structurally diverse array of drugs targeting
the ATP-binding pocket of human Hsp90 continue to be evaluated for anticancer activity in
patients. In contrast, allosteric approaches to targeting the function of Hsp90 at sites other than its
9
N-terminal ATPase have only been explored in preclinical studies, the exception being a putative
C-terminal inhibitor (RTA901) which has recently completed Phase I testing in humans
(NCT0266693).
Unfortunately, dose-limiting toxicities coupled with relatively limited therapeutic
efficacy have so far precluded FDA approval of any N-terminal Hsp90 inhibitor either alone or in
combination with other therapeutic agents. In the course of these anticancer drug development and
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 163
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
testing campaigns, no effort has been devoted to the pursuit of fungal selectivity and an Hsp90
inhibitor with the properties required for use as an antifungal has yet to be reported.
Fungal selectivity is a crucial feature for an Hsp90 inhibitor to be developed as an
antifungal given that Hsp90 is essential in all eukaryotes. Its function supports protein quality
control mechanisms, productive folding and the stability of conformationally labile proteins, many
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
0
involved in key signaling cascades. The chaperoning by Hsp90 of its so-called client proteins is
ATP-dependent and coordinated by a suite of co-chaperones and accessory factors that impart
client selectivity and help regulate progression through the chaperoning cycle. Although Hsp90
is highly conserved across phylogenetic kingdoms, species-specific variations are observed at the
level of conformational flexibility, intrinsic ATPase activity, chaperoning dynamics, and the
involvement of specific co-chaperone/accessory proteins.11 Therefore, despite a very high degree
of conservation at the primary sequence level, these important functional differences provide hope
that species-selectivity can be achieved, either at the classical N-terminal ATP-binding pocket or
alternatively via allosteric inhibitors acting at other sites.12
While efforts to achieve species-selectivity are just beginning, the pursuit of human
paralog-specific Hsp90 inhibitors has already achieved considerable success. These efforts have
been focused on achieving selectivity at the N-terminal nucleotide-binding domain (NBD) across
the four family members expressed in humans: Hsp90α, Hsp90β, Trap1 and Grp94.13, 14 For
example, Blagg and coworkers have described successful efforts to modify the resorcylate scaffold
to confer selectivity towards specific human paralogs, including selective Grp94 inhibitors with
applications in oncology and glaucoma,15-19and more recently, the first Hsp90β selective inhibitor
20
with applications in cancer. In addition, isoform-selective purine mimetics, such as Hsp90α/β-
2
1
14
specific inhibitor TAS-116 and modified analogs of BIIB021 selectively targeting Trap1 have
4
ACS Paragon Plus Environment

Page 5 of 163
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
been described. Modified benzamides resembling SNX-2112 have also been diverted to both
2
2
23
Hsp90α/β-specific and Trap1-specific activities for neurological applications.
11
We recently disclosed the discovery of the first fungal-selective Hsp90 inhibitors, with
activity against the C. albicans Hsp90 isoform, based on semi-synthetic oxime-derivatization of
the resorcylate macrocycle natural products radicicol (1) and monocillin I (2). For therapeutic
applications, fungal-selectivity is critical as current inhibitors targeting host Hsp90 have
deleterious effects that preclude their use in the context of systemic infection. Our most promising
lead from this series, monocillin-derived oxime CMLD013075 (3) (Figure 1A), has >25-fold
binding selectivity for the C. albicans Hsp90 NBD compared to the human ortholog, limits fungal
proliferation in whole cell assays, and is less toxic to human cells compared to the non-selective
compound radicicol. Importantly, the co-crystal structure of C. albicans Hsp90 NBD with
CMLD013075 displayed unique structural rearrangements, including remodeling of the ATP-
binding site, N-terminus, and lid region of the fungal chaperone. Aided by structural insights, key
residues were identified as critical for the fungal selectivity of this derivative. Encouraged by these
findings and using 3 as a point of departure, we now report the structure activity relationship
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(SAR)-guided efforts to develop fully synthetic, non-macrocyclic resorcylate inhibitor
chemotypes, focusing on selectivity toward both C. neoformans and C. albicans Hsp90.
Replacement of the macrolactone of radicicol with acyclic isosteres including amides
(
Onalespib (4)24-27), oxazoles (Luminespib (5)28-33), triazolones (Ganetespib (6)34-42), and ketones
(KW-2478 (7)43-46) has been a widely successful strategy for the development of multiple classes
of synthetic Hsp90 inhibitors currently in clinical evaluation (Figure 1B). Using our macrocyclic
oxime CMLD013075 as a lead template, our initial efforts focused on the replacement of the
selectivity-imparting oxime with a suitable heterocyclic isostere, with the parallel goals of
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 163
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
removing the isomerizable oxime (which we postulated could obfuscate selectivity analysis), and
reducing rotational degrees of freedom to enhance binding affinity. After evaluating various
heterocyclic options for similarity and synthetic tractability, we selected the aminopyrazole of
general type 8 (Figure 1C) for initial development. We hypothesized that a pendant aminopyrazole
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
could project substituents (R /R /R ) in orientations similar to that of the CMLD013075 oxime, to
impart fungal selectivity in the binding of Hsp90. In addition to the attractiveness of the pyrazole
4
7
from the standpoint of developability, we also postulated that structure-activity relationships at
1
2
3
three points of diversity (R /R /R ) could be easily elaborated through the coupling of aryl bromide
9
with a combination of commercial and synthetic aminopyrazoles (10).
6
ACS Paragon Plus Environment

Page 7 of 163
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Figure 1. Design of aminopyrazole resorcylate-type inhibitor chemotype 8 based on precedented
fungal-selective natural-product-derived inhibitors (A) and truncated resorcylates under clinical
evaluation in oncology (B).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
RESULTS AND DISCUSSION
Synthesis of resorcylate aminopyrazole analogs
Our initial synthesis of aminopyrazole resorcylates began with 1-bromo-3,5-
dimethoxybenzene 11 (Scheme 1). Formylation, de-methylation, MOM protection, and Pinnick
oxidation afforded carboxylic acid 12, which was then subjected to HATU-mediated amidation
with isoindoline to produce amide 13. We initially selected the isoindoline amide as it is conserved
across multiple classes of acyclic resorcylate heat shock protein inhibitors, 48-53 providing a simple,
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 163
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
precedented model scaffold on which our selectivity-inducing strategy could be evaluated. We
next installed the aminopyrazole using Pd-mediated coupling; after a brief exploration of coupling
54
conditions we ultimately settled on Pd
2
(dba) /Xantphos/NaOPh in dioxane under microwave
3
5
5
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
irradiation as the optimal conditions across a wide scope of substrates. Following amination,
acid-mediated MOM deprotection produced the desired aminopyrazole-substituted resorcylates.
Scheme 1. First-generation synthetic route to aminopyrazole/isoindoline resorcylate amides
Conditions: a) POCl
3
, DMF, 100 °C; b) BBr
3
, CH
2
Cl , -78 °C to RT; c) MOMCl, DIPEA, DMF;
2
t
d) NaOCl
Et N, THF/CH
Pd (dba)
2
, NaH
2
PO •H O, 2-methyl-2-butene, THF/ BuOH/H O; e) isoindoline•HCl, HATU,
4
2
2
t
t
t
t
3
2
Cl
2
; f) BuXphos Pd G1 (10 mol%), BuXPhos (10 mol%), NaO Bu, BuOH, or
(dba) (10
2
3
(4 mol%), Xantphos (8 mol%), NaOPh, dioxane, 60 °C to 120 °C, or Pd
2
3
mol%), Xantphos (10 mol%), NaOPh, dioxane, 170 °C, microwave; g) HCl, methanol, 50 °C.
8
ACS Paragon Plus Environment

Page 9 of 163
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Scheme 2. Early exploration of amide SAR using first-generation synthetic route
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
5
Conditions: a) HNR R , HATU, Et
3
N, THF/CH
2
Cl
2
; b) 10a, Pd
2
(dba) (4 mol%), Xantphos (10
3
mol%), NaOPh, dioxane, 170 °C, microwave; c) HCl, methanol, 50 °C.
We also applied this first-generation synthetic sequence to explore replacement of the
isoindoline amide for several early compounds (Scheme 2). During the course of analog synthesis,
however, we found that reversing the order of coupling/amidation resulted in a more efficient
procedure with improved yields and product purities; the resultant second-generation route is
depicted in Scheme 3. Following esterification of carboxylic acid 12, the resulting ester 17 was
subjected to Pd-mediated coupling with 10 to afford intermediate 18. Following ester hydrolysis,
carboxylic acid 19 was subsequently amidated, which was initially performed using the HATU-
mediated conditions, and later optimized to employ polymer-supported carbonyldiimidazole (PS-
CDI) as a coupling reagent for improved parallel processing. Finally global MOM-deprotection
provided the desired products for testing. All tested compounds were purified by mass-targeted
HPLC.
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 163
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Scheme 3. Second-generation synthetic route to aminopyrazole resorcylate amides
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Conditions: a) CH
NaOPh; c) KOH, EtOH, 95 °C; d) HNR R , HATU, Et
HOBt•xH O, Et N, THF/CH Cl ; e) HCl, methanol, 50 °C.
3
I, K
2
CO
3
, DMF, 80 °C; b) 10, Pd
2
(dba)
3
(4 mol%), Xantphos (10 mol%),
4
5
4 5
3
N, CH
2
Cl /THF, RT or HNR R , PS-CDI,
2
2
3
2
2
Measurement of fungal Hsp90 binding affinity and selectivity
All analogs were assessed for Hsp90 binding affinity using a fluorescence polarization
(FP)-based equilibrium competition assay in fungal and human whole cell lysates. Notably, this
approach allows for the assessment of compound binding while the target protein is in native
complexes with co-chaperones; and, in the case of human cell lysate, in a biologically relevant mix
of Hsp90 paralogs. Using lysates, we were able to measure the relative potency and selectivity for
fungal Hsp90 versus the entire ensemble of human Hsp90 isoforms in microplate format using
small amounts of test materials. To confirm target engagement with an alternative biochemical
approach, the most selective analogs were also assessed by protein thermal shift assays using
purified recombinant Hsp90 nucleotide binding domains (NBD) of the relevant fungal species.
Thermal shift assays were performed under saturating ligand conditions, i.e. equimolar
concentrations (10 μM) of protein and ligand. As a result, they provided qualitative evidence of
1
0
ACS Paragon Plus Environment

Page 11 of 163
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
target binding, but not a quantitative measurement of ligand affinity. For quantitation, a ligand
dissociation constant (K ) for key compounds was also determined using purified NBDs in FP
i
assays and KD measurements were made by surface plasmon resonance (SPR) using a Biacore
instrument. Finally, all analogs were assessed for whole cell antifungal activity against the
pathogens C. albicans and C. neoformans. Quantitative dose-response assays were performed for
all compounds found to inhibit growth at a concentration < 50 μM.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Structure activity relationships for resorcylate aminopyrazoles
We first examined N-(para)-methoxybenzyl substituted aminopyrazoles, designed to
mimic our parent Candida-selective inhibitor CMLD013075. We began by making systematic
2
alterations to the resorcylate amide, with R substitution limited to methyl and phenyl. Our initial
amide diversification utilized several pyrrolidine/isoindoline-based heterocycles, which are
prevalent among resorcylate amide Hsp90 inhibitors reported by Astex and Pfizer (20-21, 26-29,
3
2-36, 38),49, 50 as well as new isoindoline isosteres (pyrido- and pyrazolopyrrolidines 22-25). We
also pursued a small series of acyclic mono- and disubstituted amides, both new (30-31) and
5
6
precedented (37, 39). From this initial set, we were pleased to find a number of compounds had
200 nM EC50 values against one or both fungal species (Table 1). Consistent with published
<
inhibitors in this space, larger, substituted isoindoline-type moieties (32-36) generally exhibited
excellent potency, but with no apparent selectivity for the fungal Hsp90 isoforms. In contrast, we
2
found that the pairing of smaller heterobicylic amides with a phenyl group at the R position
(compounds 21, 23, 25, 27, and 29) afforded modestly fungal-selective compounds; as a general
2
trend, their R = CH
3
analogs (20, 22, 24, 26 and 28) were more potent but nonselective. Activity
was mainly relegated to the heterobicyclic amides; our limited set of acyclic and monocyclic
1
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 163
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
amides (30-31, 37-39) were for the most part less active and also nonselective, with the interesting
exception of low potency cryptococcal-selective compound 30. Based on these results, and given
our early hypothesis that the installation of functionality at the aminopyrazole would be the key
driver in imparting selectivity, we opted to progress forward with the lower-molecular weight
isoindoline, pyridopyrrolidine, and pyrazolopyrrolidine amides, selected to represent both
precedented and novel resorcylate amide substitutions with varying basicities.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 1. Structure-activity relationships for N-(4-methoxybenzyl)-substituted aminopyrazoles,
2
exploring variation of the resorcylate amide with methyl- and phenyl-substitution at R . Fold-
selectivity > 5 for any compound is highlighted in red.
1
2
ACS Paragon Plus Environment

Page 13 of 163
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
C. neoformans
fold-
C. albicans
fold-
C. neoformans
EC50 (μM)
C. albicans
EC50 (μM)
Entry Compound
X
R2
a
c
selectivityb
selectivityb
1
2
3
4
5
6
7
8
9
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
CH
3
0.040
0.877
0.087
0.142
0.063
0.121
0.109
0.787
0.592
0.705
1.330
> 9
0.8
2.5
1.2
4.0
1.7
2.7
0.6
2.2
0.1
2.7
5.8
-
0.011
0.511
0.184
0.068
0.157
0.063
0.117
1.089
0.054
1.043
> 9
0.9
2.2
0.4
6.2
0.5
3.9
0.4
1.2
0.5
1.3
-
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Ph
CH
3
Ph
CH
3
Ph
CH
3
Ph
CH
3
1
0
1
Ph
1
CH
3
12
13
Ph
> 9
-
CH
3
0.096
0.146
0.086
0.091
1.0
0.8
1.2
0.8
0.171
0.023
0.115
0.014
0.4
1.8
0.7
1.7
1
1
1
4
5
6
Ph
CH
3
Ph
1
1
7
8
36
CH
CH
3
0.115
0.9
0.143
0.6
37
38
39
3
4.814
0.464
>10
1.6
2.1
-
> 6
0.282
>10
0.0
1.2
-
19
Ph
Ph
2
0
EC50 values were determined by FP-based equilibrium competition assay performed in 384-well
format using whole cell lysates prepared from C. neoformans (a) and C. albicans (c) and serial
compound dilutions. All determinations were performed in duplicate. To calculate fold-selectivity
(
b), the EC50 value determined in human HepG2 cell lysate was divided by the EC50 value
1
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 163
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
determined in fungal cell lysate. The resulting ratio was then normalized to values determined in
the same assay for the non-selective inhibitor geldanamycin using lysate of each cell type. Results
for key selective compounds were confirmed by repeat assay (see Supplemental Table 1).
2
3
Our next series of analogs explored additional R /R substitutions on the aminopyrazole,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
again keeping the R para-methoxybenzyl group intact (Table 2). For the R unsubstituted
3
pyrazoles (40-44), we found that substitution at R was tolerated, but with decreasing potency as
steric bulk increased. Several of these compounds also exhibited modest undesired selectivity
toward the human isoform. Based on these results, we opted not to pursue this substitution pattern
further. In contrast, and similar to our initial cohort, we identified wider tolerance for substitution
2
at the R position with several acyclic (45-48) and cyclic (50-53) aliphatic groups, as well as furan
2
t
(
54-55) substitution. A drop in potency was limited to the bulkier R = Bu analog 49.
Disappointingly, however, none of the inhibitors exhibited the modest fungal selectivity that had
2
been observed in their R = Ph substituted counterparts 21, 23 and 25 (Table 1).
1
4
ACS Paragon Plus Environment

Page 15 of 163
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
3
1
Table 2. Exploration of SAR at R /R for R = p-methoxybenzyl substituted aminopyrazoles
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
C. neoformans
C. albicans
C. neoformans
EC50 (μM)
C. albicans
EC50 (μM)
Entry Compound
X
R2
R3
fold-
fold-
a
c
selectivityb
selectivityb
1
2
3
4
5
6
7
8
9
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
A
A
A
A
A
B
C
B
C
A
B
C
B
C
B
C
H
H
H
H
H
H
0.072
0.094
0.396
1.623
1.756
0.022
0.025
0.025
0.023
1.026
0.026
0.023
0.021
0.041
0.039
0.040
0.8
0.3
0.4
1.6
0.5
0.9
0.7
1.0
0.8
1.4
0.9
0.9
0.7
0.6
1.3
0.9
0.041
0.022
0.147
0.615
0.465
0.014
0.012
0.013
0.009
0.816
0.014
0.014
0.006
0.009
0.022
0.018
1.0
0.7
0.5
2.1
1.1
0.5
0.6
0.7
0.8
1.0
0.6
0.5
1.0
1.0
0.9
0.7
CH
3
iPr
Ph
Bn
Et
H
iPr
H
H
H
1
0
1
2
tBu
1
1
13
H
H
1
1
1
4
5
6
EC50 and selectivity values were determined as described for Table 1.
1
We next assessed replacement of the p-methoxybenzyl group at R . Initially, this group
had been chosen based on analogy to our Candida-selective inhibitor CMLD013075. Our X-ray
1
1
crystallographic analysis (PDB ID: 6CJP) indicates that the aryl ring participates in a key binding
1
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 163
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
interaction following a major structural rearrangement of the Candida Hsp90 lid region, serving
as a donor in an N-H…π interaction with C. albicans Asn40. However, given the limited scope of
radicicol- and monocillin-derived analogs that we previously explored, coupled with a current lack
of structural information about the cryptococcal isoform, it is not clear that this group represents
an “ideal” binding moiety for either fungal species. As an initial probe, we focused on varying
solely the R1 group across the isoindoline, tetrahydropyrrolopyridine and
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
3
tetrahydropyrrolopyrazole amides, leaving the R and R sites unsubstituted. The results for this
series are summarized in Table 3. We once again identified a wide array (aliphatic, aromatic,
heteroaromatic) of aminopyrazole substitutions that afforded in most cases sub-125 nM potencies
for both fungal species (56, 59, 62-77 and 80-86), but all broadly nonselective with the exception
of isoindoline 83. This compound was exemplary as the first compound in our aminopyrazole
series to exhibit sub-100 nM EC50 with greater than 10-fold selectivity. Interestingly, however, in
isolated cases the tetrahydropyrrolopyridine and tetrahydropyrrolopyrazole amides diverged from
their isoindoline counterparts with a slight decrease in potency (compounds 78-79), which was in
some cases coupled with a slight increase in cryptococcal selectivity (57-58 and 60-61). These
compounds, bearing aliphatic N-substitutions of varying size, showed 2- to 5-fold selectivity
toward C. neoformans, with no apparent selectivity toward C. albicans.
1
2
3
Table 3. Exploring alternative R substituents on R /R -unsubstituted aminopyrazoles. Fold
selectivity >5 for any compound is highlighted in red.
1
6
ACS Paragon Plus Environment

Page 17 of 163
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
C. neoformans
C. albicans
fold-
C. neoformans
EC50 (μM)
C. albicans
EC50 (μM)
Entry Compound
R1
X
fold-
a
c
selectivityb
selectivityb
1
2
3
4
5
6
7
8
9
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
0.088
0.286
0.142
0.125
0.250
0.097
0.062
0.103
0.041
0.051
0.026
0.018
0.061
0.044
0.037
0.087
0.045
0.035
0.052
0.043
0.036
0.058
0.207
0.199
0.074
0.106
0.058
1.5
4.4
5.1
1.3
2.6
3.6
1.1
1.7
2.3
0.7
1.7
1.3
0.5
0.9
0.8
0.4
1.0
1.0
0.8
1.1
1.1
1.7
1.6
1.8
0.9
1.2
1.4
0.111
0.624
0.377
0.076
0.309
0.161
0.035
0.117
0.057
0.009
0.015
0.008
0.014
0.021
0.015
0.013
0.016
0.014
0.015
0.018
0.015
0.041
0.143
0.164
0.033
0.080
0.041
0.4
0.7
0.7
0.7
0.7
0.7
0.7
0.6
0.6
1.2
0.9
0.8
0.8
0.7
0.8
0.9
1.0
1.0
0.8
0.8
0.8
0.7
0.7
0.7
0.7
0.6
0.8
CH
3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
iPr
iBu
1
0
11
1
1
1
1
1
2
3
4
5
6
Ph
Cy
Bn
17
1
1
2
2
8
9
0
1
22
23
2
2
2
2
4
5
6
7
1
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 163
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
8
83
84
85
86
87
A
B
C
B
C
0.044
0.040
0.036
0.063
0.048
12.8
0.8
0.8
0.9
1.1
0.366
0.012
0.012
0.025
0.021
1.1
0.9
0.8
0.8
0.9
29
3
3
3
0
1
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
EC50 and selectivity values were determined as described for Table 1.
1
We next progressed to examining the combined modifications of the R N-substitution with
2
additional groups at R (Table 4). Again mindful of keeping physicochemical properties such as
molecular weight and lipophilicity within an acceptable “druglike” range, we imposed a limitation
1
2
for this series that each pyrazole should contain a maximum of one aryl ring at either R or R , but
57
not at both. This series produced a number of analogs with more modest sub-micromolar potency
and cryptococcal selectivity greater than 4-fold (91-95). Of these, compounds 94 and 95 also
exhibited modest selectivity for C. albicans Hsp90 over human Hsp90 paralogs, which was
consistent with their early near neighbor analogs 21, 23 and 25 (Table 1).
1
8
ACS Paragon Plus Environment

Page 19 of 163
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
2
Table 4. Examining varied parings of R /R substitutions on the aminopyrazole ring. Fold
selectivity >5 for any compound is highlighted in red.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
C. neoformans
fold-
C. albicans
fold-
C. neoformans
C. albicans
EC50 (μM)
Entry Compound
X
R1
R2
a
c
EC50 (μM)
selectivityb
selectivityb
1
2
3
88
89
90
A
A
A
CH
CH
CH
3
3
3
0.252
0.560
0.700
0.2
0.2
0.2
0.121
0.116
0.134
0.3
0.4
0.4
4
5
6
7
8
9
91
92
93
94
95
96
97
98
99
A
B
C
B
C
B
C
B
C
0.078
0.127
0.066
0.517
0.379
0.615
0.419
0.315
0.147
9.2
8.2
6.7
9.7
6.7
0.4
0.3
1.7
1.8
0.328
0.395
0.213
0.379
0.182
0.059
0.034
0.100
0.070
0.4
0.8
0.6
3.9
4.1
1.7
1.4
1.6
1.4
CH
3
Ph
tBu
Ph
Ph
Ph
1
0
1
1
iBu
12
EC50 and selectivity values were determined as described for Table 1.
Among our initial fungal-selective leads from this effort, compounds 91-93 stood out as
having high cryptococcal selectivity without a concomitant loss in cryptococcal potency as seen
1
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 163
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
in earlier analogs. To follow up, we designed a final array of analogs N-methylated at R , probing
2
more diverse aliphatic and aryl substituents at R (Table 5).
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 5. Variation of R substituent for N-methylated aminopyrazoles yields C. neoformans- and
C. albicans-selective Hsp90 inhibitors with diverging isoform selectivities. Fold selectivities > 5
are highlighted in red.
C. neoformans
fold-
C. albicans
fold-
C. neoformans
EC50 (μM)
C. albicans
EC50 (μM)
Entry Compound
R2
iPr
X
a
c
selectivityb
selectivityb
1
2
100
101
102
103
104
105
106
107
108
109
110
111
112
113
A
B
C
A
B
C
A
B
C
A
B
C
A
B
0.087
0.062
0.040
0.156
0.110
0.037
0.065
0.139
0.084
0.267
0.421
0.176
0.281
0.852
0.9
1.6
0.048
0.156
0.094
0.033
0.670
0.096
0.599
0.795
0.398
0.573
1.030
0.396
1.642
5.000
0.6
0.2
0.5
1.4
0.4
0.6
0.7
0.8
0.9
3.0
2.4
2.5
2.0
1.7
3
3.0
4
0.7
5
Cy
6.5
6
3.7
7
14.1
12.9
12.4
16.3
15.1
14.6
33.3
27.6
8
9
10
1
1
1
1
1
2
3
4
2
0
ACS Paragon Plus Environment

Page 21 of 163
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
5
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
0.244
4.601
1.523
0.434
1.318
1.781
0.424
0.630
1.289
0.489
9.530
5.815
1.385
>10
26.5
3.2
5.5
9.8
6.3
9.5
14.9
4.4
6.9
6.8
1.0
0.8
1.1
-
1.881
2.236
0.594
0.294
1.067
1.779
0.636
0.186
1.139
0.376
1.084
0.319
0.103
1.197
0.237
0.071
0.134
0.050
0.016
1.3
2.1
4.4
4.6
2.6
4.0
3.4
5.8
3.1
3.4
3.1
4.8
4.9
2.1
4.5
5.0
15.3
18.2
15.9
16
17
1
1
2
2
8
9
0
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
22
23
2
2
2
2
4
5
6
7
28
29
8.337
2.262
5.402
1.262
0.514
0.4
0.5
1.0
1.0
1.3
3
3
3
3
0
1
2
3
EC50 and selectivity values were determined as described for Table 1.
Gratifyingly, this series produced highly selective inhibitors for both the C. neoformans
and C. albicans isoforms of Hsp90. While our exploration of aliphatic substitution was limited,
high cryptococcal potency (EC50 < 160 nM), and in some cases modestly Cryptococcus-selective
compounds (3- to 6.5-fold) were observed with isopropyl (100-102) and cyclohexyl (103-105)
2
2
substitution at R . The most highly selective compounds, however, were observed among the R
arylated analogs, with diverging species-selectivity based on the nature and position of the aryl
ring substituent. The ortho-methylated analogs 106-108 displayed slightly enhanced cryptococcal
2
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 163
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
selectivity and similar cryptococcal potency (<150 nM) as compared to their unsubstituted
congeners 91-93 (Table 4), with no apparent selectivity and significantly lower potencies (≥ 400
nM) in lysate of C. albicans. Movement of the methyl substituent from ortho- to meta- (compounds
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
09-111) afforded similarly Cryptococcus-selective compounds, albeit with lower potencies.
Interestingly, the meta-methoxy substituted 112-114 exhibited a significant improvement in
cryptococcal selectivity (27- to 33-fold) despite only modest cryptococcal potency (EC50s all >250
nM). Trifluoromethylation at the same meta- position (compounds 115-117), resulted in a dramatic
reduction in both cryptococcal selectivity and activity.
Moving from testing in C. albicans lysate to C. neoformans lysate, the aforementioned
meta-substituted compounds 109-117 also exhibited modest selectivity, with the best Candida-
selectivity observed m-trifluoromethylated analogs 116 and 117 (4.4- and 4.6-fold, respectively).
The meta-substituted series also exhibited consistently poor C. albicans potencies, with EC50
values ranging from ~300 nM to 5 μM. In contrast, improved C. albicans selectivities and
2
potencies were observed among the analogs that were para-substituted on the R phenyl ring. para-
Methylated (119-120) and para-methoxy substituted (121-123) aminopyrazoles exhibited
moderate selectivities and, in most cases, equivalently low potencies against both fungal species,
with EC50 values generally ranging from 0.5-2 μM. Incorporation of larger lipophilic substituents
at the para-position such as trifluoromethyl (124-126) and tert-butyl (127-129) further depressed
cryptococcal potency, with EC50s ranging from 2 to >10 μM and no apparent selectivity. In contrast
these compounds (124-129) maintained improved potencies and similar selectivities against
Candida Hsp90 to their para-methyl- and para-methoxy- counterparts 124-127. This series also
highlights what we have observed to be an occasional sensitivity to the nature of the
amide/aminopyrazole pairing; for example in direct contrast to the cryptococcal potency trends
2
2
ACS Paragon Plus Environment

Page 23 of 163
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
observed with unsubstituted analogs 56-58, pairing of the pyrido- and pyrazolopyrrolidine with
t
2
the bulkier 3-CF
3
-Ph (116-117), 4-CF -Ph (125-126) and 4- Bu-Ph (128-129) substituents at R
3
improved C. albicans potency and selectivity relative to their isoindoline counterparts 115, 124
and 127. This trend did not hold, however, for all analogs. Perhaps most intriguingly, the para-
trifluoromethoxy substituted compounds 130-132, which were completely nonselective and only
modestly potent toward cryptococcal Hsp90, exhibited dramatic improvements in potency toward
C. albicans, with EC50 values ranging from 16-134 nM and 15- to 18-fold Candida selectivity.
These divergent structure-selectivity trends, wherein ortho/meta-methyl and meta-methoxy
compounds exhibited high Cryptococcus selectivity and poor Candida selectivity, whereas para-
trifluoromethoxy substitution rendered high Candida selectivity and poor Cryptococcus activity,
are summarized in Figure 2.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 163
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 2. Summary of iterative progression to fungal selective inhibitors 106-132 with divergent
patterns of species selectivity dependent on position of substitution on the aminopyrazole phenyl
ring (red)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Relationship of fungal to human selectivity
To better understand the phylogenetic origins of the divergent selectivity between fungi,
we performed protein::protein BLAST sequence alignments across the different species studied.
This analysis indicated that C. neoformans and C. albicans share 69% sequence identity across the
entire Hsp90 protein, and 71% identity across their NBD (residues 1-240). As a comparison,
human Hsp90α and Hsp90β share 69% and 67% identity with C. albicans Hsp90 across their NBD,
respectively. Thus, the two fungal species diverge in primary sequence as greatly from one another
as they do from human Hsp90. In light of such sequence divergence, perhaps it is not surprising
that while we set out to discriminate against human Hsp90, the potency and selectivity of our
2
4
ACS Paragon Plus Environment

Page 25 of 163
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
synthetic inhibitors also diverged between the two fungal species studied. A graphic summary of
inhibitor potency/selectivity relationships found by screening compounds in C. neoformans lysate
(Fig 3A) and C. albicans lysate (Fig. 3B) highlights the progress made in achieving our goal of
achieving fungal selectivity, while the divergence between compound selectivity in regards to C.
neoformans vs. C. albicans is best demonstrated by plotting the selectivity of compounds for one
fungus vs. human against selectivity for the other (Fig. 3C). To more accurately define their
potency and selectivity, the activity of 27 compounds with a screening EC50 < 1μM in lysate of
either fungal species was confirmed by repeat testing in two additional experiments, with results
provided in Supplemental Table 1.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3. Scatter plots depict fungal potency (x-axis) vs. fungal selectivity (y-axis) relationships
for all aminopyrazoles when screened using human cell lysate and lysate of either C. neoformans
(Panel A) or C. albicans (Panel B). All potencies are reported as the inverse log10 of compound
EC50 (pEC50 as measured by FP assay). The scatter plot in Panel C compares compound selectivity
patterns between the two fungi. Key fungal-selective compounds for each species (112-114 and
1
30-132) are highlighted in color to underscore their divergence in potency and selectivity. Each
point represents the mean of duplicate determinations in a single experiment.
2
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 163
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Validation of whole cell lysate FP results
To confirm the FP results obtained in lysate for our most potent and selective compounds, we
prepared recombinant Candida, Cryptococcus, and human Hsp90 NBDs by expression and
purification in E. coli. Using recombinant proteins, we were able to define assay-independent
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
nanomolar inhibitory constants (K ) for these compounds (Supplemental Fig. 1). We also
i
confirmed binding of the compounds to their relevant NBD by qualitative thermal shift
denaturation assays performed under saturating ligand conditions (Supplemental Fig. 1). Thermal
shift assays were performed under saturating ligand conditions, i.e. equimolar concentrations (10
µM) of protein and ligand. As a result, they can provide only qualitative evidence of target binding,
but not a quantitative measurement of ligand affinity. This feature of the thermal shift method is
well demonstrated in Supplemental Table 2, which presents K
and low potency compounds. Here, compounds with K values of less than 50 nM for a particular
NBD increase its ΔTm to a similar extent irrespective of absolute potency. In contrast, lower
affinity compounds (K > 100 nM) fail to increase the Tm of the respective NBD.
i
and thermal shift data for both high
i
i
As an orthogonal, highly quantitative approach to FP, we measured the binding affinities
of our six lead compounds for C. albicans, C. neoformans and human Hsp90 NBDs by surface
plasmon resonance (SPR, Supplemental Table 3). The affinity values determined for compounds
varied by less than an order of magnitude between the two different experimental techniques. The
same pattern of fungal selectivity for compounds demonstrated by FP assay in whole cell lysates
was also seen by SPR. The magnitude of selectivity determined by SPR assays compared to FP
assays in lysate, however, was reduced. Such a difference might be expected given the absence in
SPR assays of native co-chaperone containing complexes and, in the case of human cell lysate, a
biologically relevant mix of Hsp90 paralogs.
2
6
ACS Paragon Plus Environment

Page 27 of 163
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Whole cell antifungal activity
Having achieved promising potency and species-selectivity for several compounds at the
level of fungal target engagement, we next examined the ability of these compounds to inhibit
fungal growth. We found that minimal inhibitory concentrations (MICs) for most of the potent and
selective analogs highlighted in Table 5 were much higher than their EC50 values in lysate,
generally > 50 μM. The disparity between whole cell antifungal activity and the EC50 values we
determined in FP assays is undoubtedly due to poor permeability/accumulation of the compounds
in fungal cells. This common problem in the development of antifungals occurs because the fungal
cell wall and membrane as well as the diverse drug efflux pumps expressed by fungi render it a
challenge to achieve intracellular concentrations of experimental compounds sufficient to inhibit
the function of their targets.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Of the fungal Hsp90-selective compounds tested, only the 14-fold C. neoformans-selective
analog 106 inhibited growth of the organism below 10 μM (Fig. 4A). While triazole antifungals in
current clinical use against Cryptococcus do have MICs in excess of this range, they also possess
far greater selectivity than we have achieved so far and are much less toxic to human cells. As
single agents, the MICs of all our Candida-selective compounds were >50 μM. To provide a more
sensitive read-out, however, we took advantage of the well-established ability of Hsp90 inhibitors
to potentiate the activity of conventional antifungals against drug-resistant isolates of C. albicans.5
Testing compounds 130 and 131 in combination with the widely used antifungal fluconazole, we
found an MIC of 12.5 μM for the 15-fold Candida-selective analog 130 against a moderately
fluconazole-resistant clinical isolate of C. albicans. This compound also converted the fungistatic
activity of fluconazole to fungicidal against the same isolate, an effect consistent with Hsp90
inhibitory activity (Figure 4B).
2
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 163
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 4. Biological activity of fungal-selective inhibitors.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Panel A: Growth inhibition by fungal-selective aminopyrazoles of C. neoformans reference strain
H99 cultured in RPMI 1640 medium at 37 °C. Panel B: Growth inhibition by fungal-selective
aminopyrazoles of a C. albicans clinical isolate (CaCi2) with or without a background
concentration of 8 µg/mL fluconazole. The effect of 48-hour exposure to inhibitors over a twofold
dilution series of concentrations is displayed in heat-map format. Color scale bar: no growth
inhibition (green) to complete inhibition (black). Each colored box represents the mean of
technical duplicates. The experiment was repeated as an independent biological replicate to
confirm results. Following exposure to compounds, aliquots of the cultures in each well were
spotted onto compound-free YPD agar and plates incubated at 30 °C for an additional 24 hours
before imaging to assess fungicidal activity (Panel B, right).
2
8
ACS Paragon Plus Environment

Page 29 of 163
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Thus, although their potency and selectivity require further improvement, the whole cell
activity of these resorcylate aminopyrazoles remains consistent with an Hsp90-targeted mode of
action. Encouraged by this finding and to aid future efforts in developing the scaffold, we
performed an initial evaluation of its stability to P450-mediated metabolism in liver microsomes,
a major pharmacological liability of our previous fungal selective macrocyclic oxime
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
1
CMLD013075 . Although all the compounds tested suffered from relatively rapid metabolism
(Fig. 5 and Supplemental Table 4), important insights were gained into the basis of their instability.
Comparing the half-lives of cryptococcal-selective compounds 112-114 reveals an apparent
stabilizing effect of the pyrazolopyrrolidine amide, which is consistent with the previously
24
reported metabolic instability of isoindolines due to oxidation at the 5/6 position. The isoindoline
was chosen for this study despite its known downstream pharmacological liabilities, as it
represented a low molecular-weight starting point allowing for the methodical assessment of the
relative potency and selectivity of different aminopyrazole substitutions. Assessment of additional
analogs 105, 111, 129, and 131-132 indicate that additional metabolic liabilities are also likely
present at the aminopyrazole, with the para-trifluoromethoxy substitution clearly inhibiting
metabolism. Still, the relatively short half-life of compound 132 (31 minutes) underscores the need
for further optimization of metabolic stability, in addition to fungal penetration, as we advance in
future work to compounds with suitable properties for testing in vivo. Metabolic stability
optimization for resorcylate Hsp90 inhibitors via modification of the amide is well-precedented;24,
5
0
similar strategies, paired with targeted alterations of the aminopyrazole aryl substituent, are
currently under study in our laboratory and will be reported in due course.
2
9
ACS Paragon Plus Environment