A Series of 2-((1-Phenyl-1H-imidazol-5-yl)methyl)-1H-indoles as Indoleamine 2,3-Dioxygenase 1 (IDO1) Inhibitors

Article abstract of DOI:10.1002/cmdc.202100107

Indoleamine 2,3-dioxygenase 1 (IDO1) is a promising therapeutic target in cancer immunotherapy and neurological disease. Thus, searching for highly active inhibitors for use in human cancers is now a focus of widespread research and development efforts. I

Full text of DOI:10.1002/cmdc.202100107

Accepted Article
Title: Discovery of a series of 2-((1-phenyl-1H-imidazol-5-
yl)methyl)-1H-indoles as indoleamine 2,3-dioxygenase 1 (IDO1)
inhibitors
Authors: Donald F Weaver, Yong Zheng, Paul M. Stafford, Kurt R.
Stover, Darapaneni Chandra Mohan, Mayuri Gupta, Eric
C. Keske, Paolo Schiavini, Laura Villar, Fan Wu, Alexander
Kreft, Kiersten Thomas, Elana Raaphorst, Jagadeesh P.
Pasangulapati, Siva R. Alla, Simmi Sharma, Ramana R.
Mittapalli, Irina Sagamanova, Mark A. Reed, and Shea L.
Johnson
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemMedChem 10.1002/cmdc.202100107
Link to VoR: https://doi.org/10.1002/cmdc.202100107
01/2020

10.1002/cmdc.202100107
ChemMedChem
COMMUNICATION
Discovery of a series of 2-((1-phenyl-1H-imidazol-5-yl)methyl)-1H-
indoles as indoleamine 2,3-dioxygenase 1 (IDO1) inhibitors
Yong Zheng,[a] Paul M. Stafford,[a] Kurt R. Stover,[a] Darapaneni Chandra Mohan,[a] Mayuri Gupta,[a] Eric
C. Keske,[a] Paolo Schiavini,[a] Laura Villar,[a] Fan Wu,[a] Alexander Kreft,[a] Kiersten Thomas,[a] Elana
Raaphorst,[a] Jagadeesh P. Pasangulapati,[a] Siva R. Alla,[a] Simmi Sharma,[a] Ramana R. Mittapalli,[a]
Irina Sagamanova,[a] Shea L. Johnson,[a] Mark A. Reed,[a],[b] Donald F. Weaver*[a],[b],[c]
Dedicated to the memory of Prof. Walter A. Szarek
[a]
Dr. Y. Zheng, P. M. Stafford, Dr. K. R. Stover, Dr. D. C. Mohan, Dr. M. Gupta, Dr. E. C. Keske, Dr. P. Schiavini, Dr. L. Villar, Dr. F. Wu, Dr. A. Kreft, K.
Thomas, E. Raaphorst, J. P. Pasangulapati, Dr. S. R. Alla, Dr. S. Sharma, Dr. R. R. Mittapalli, Dr. I. Sagamanova, Dr. S. L. Johnson, Prof. M. A. Reed,
Prof. D. F. Weaver
Krembil Research Institute
University Health Network
60 Leonard Avenue, Toronto, ON M5T 2S8, Canada
Corresponding author’s E-mail: donald.weaver@uhnresearch.ca
Prof. M. A. Reed, Prof. D. F. Weaver
Department of Chemistry
[b]
[c]
University of Toronto, Toronto, ON M55 3H6, Canada
Prof. D. F. Weaver
Department of Medicine, Division of Neurology
University of Toronto, Toronto, ON M5S 1A8, Canada
Supporting information for this article is given via a link at the end of the document.
Abstract: Indoleamine 2,3-dioxygenase 1 (IDO1) is a promising Introduction
therapeutic target in cancer immunotherapy and neurological disease.
Thus, searching for highly active inhibitors for use in human cancers
Over the last two decades, immunotherapy has arisen as a
is now a focus of widespread research and development efforts. In
this study, we report the structure-based design of 2-(5-
imidazolyl)indole derivatives, a series of novel IDO1 inhibitors which
have been designed and synthesized based on our previous study
using N1-substituted 5-indoleimidazoles. Among these, we have
identified one with a strong IDO1 inhibitory activity (IC50 = 0.16 µM,
EC50= 0.3 µM). Structural-activity relationship (SAR) and
computational docking simulations suggest that a hydroxyl group
favorably interacts with a proximal Ser167 residue in Pocket A,
improving IDO1 inhibitory potency. The brain penetrance of potent
compounds was estimated by calculation of the Blood Brain Barrier
(BBB) Score and Brain Exposure Efficiency (BEE) Score. Many
compounds had favorable scores and the two most promising
powerful approach to the treatment of cancers.[1,2] IDO1 is a
heme-containing monomeric enzyme which catalyzes the
catabolism of tryptophan in the first rate-limiting step to N-
formylkynurenine via the oxidative cleavage of the C2−C3 indole
double bond.[3] N-formylkynurenine is further converted to
kynurenine and a series of biologically active metabolites,
including the final product nicotinamide adenine dinucleotide
(NAD).[4] IDO1 regulates the immune response by producing
biologically active kynurenine pathway metabolites.[5] In the tumor
microenvironment, IDO1 is found as a key mediator of innate and
adaptive immunity.[5]
IDO1 is highly expressed by tumor cells to escape a
potentially effective immune response via depletion of L-
tryptophan in the tumor microenvironment and by production of
the catabolic product kynurenine, which selectively impairs the
growth and survival of T-cells.[6] Dysregulation of IDO1 expression
is also involved in several inflammatory diseases, arthritis, and
neurological disorders including Alzheimer’s disease, Parkinson’s
disease, and cerebral ischemia.[7]
compounds were advanced to
a pharmacokinetic study which
demonstrated that both compounds were brain penetrant. We have
thus discovered a flexible scaffold for brain penetrant IDO1 inhibitors,
exemplified by several potent, brain penetrant, agents. With this
promising scaffold, we provide herein a basis for further development
of brain penetrant IDO1 inhibitors.
Since IDO1 plays an important role in cancer immunotherapy and
neurological diseases, searching for highly active inhibitors is now
a focus of research and development efforts by pharmaceutical
1
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202100107
ChemMedChem
COMMUNICATION
companies and academic research institutions.[8-30] Various
molecular scaffolds with inhibitory activity against IDO1 have
been documented in the academic and patent literature.[8-30] To
date, several IDO1 inhibitors such as Epacadostat, PF-06840003,
NLG-919 and BMS-986205 have advanced to clinical trials
(Fig.1).[16,22,27,31,32] Indoximod (NewLink Genetics) was the first
IDO1 inhibitor reported. NewLink Genetics also announced an
indoximod prodrug with improved pharmacokinetic properties, but
the structure has not yet been divulged.[33] Epacadostat (Incyte,
INCB024360) and Linrodostat (Bristol-Myers Squibb, BMS-
986205) showed promising outcomes in phase I/II clinical trials;
however, a phase III clinical trial combining Epacadostat with the
PD-1 antibody pembrolizumab in melanoma failed to demonstrate
efficacy.[34] PF-06840003 (Pfizer) and NLG-919 (NewLink
Genetics) also failed in phase I and II clinical trials, respectively.
No IDO1 inhibitor has been approved for clinical use to date.
Despite these recent clinical failures, there is a continuing need
for further research into the mechanism of IDO1 in cancer biology
and as a druggable target for cancer. Interest in the industrial
sector in IDO1 also continues with multiple companies
maintaining IDO1 inhibitors in their pipelines, chiefly for non-CNS
malignancies, keeping the door open for the development of brain
penetrant IDO1 inhibitors for CNS indications.
In our previously published study we discovered a series of
potent IDO1 inhibitors which structurally featured an imidazole
ring connected to a phenyl ring through a C-C bond, and to an
indole moiety through an N-CH2 bridge (Fig. 2).[17] Computational
modeling demonstrated that the hydroxyl group interacts with
Ser167 in Pocket A via a hydrogen bond, the 5-halogen
substituted indole group binds in pocket B. Imidazole combines
with heme iron atom and indole-NH forms interactions with the
propionate of the heme by hydrogen bonding (Fig. 3). We
envisaged swapping the orientation of the imidazole ring, which
would allow the phenyl and indole rings to occupy the same IDO1
binding pockets as in the original orientation (Fig. 2). This
modification may provide an opportunity for improving the IDO
inhibitory activity and ADME/PK properties and producing brain
penetrant compounds. In our continuous efforts to identify more
potent IDO1 inhibitors, herein we report a series of 2-((1-phenyl-
1H-imidazol-5-yl)methyl)-1H-indole
inhibitors.
compounds
as
IDO1
Figure 2. Design strategy for IDO1 inhibitors
Figure 1. Representative IDO1 inhibitors
In accord with available IDO1 co-crystal structures, the
IDO1 active site is structurally divided into three regions: pocket
A, pocket B and a heme-cofactor.[16,22,28,35-38] Pocket A is a narrow
lipophilic pocket which is located in the distal heme site. Most
inhibitors with published co-crystal structures occupy pocket A
with an aromatic ring. Pocket B is near the entrance of the active
site, which contains hydrophilic (Arg231) and hydrophobic
(Phe226) residues. The heme cofactor structurally consists of two
binding sites: the porphyrin-bound iron atom and a propionate
group. Most potent inhibitors feature imidazole,[22,25]
imidazothiozole[28] or triazole moieties[39] tightly bound with the
heme iron via an N-atom, except Epacadostat, which coordinates
to the heme with an O-atom.[27] In addition, the propionate of the
heme is also a binding site which contributes to the biological
activity; for example, the hydroxyl group on the side chain of NLG-
919 forms a hydrogen bond with the propionate of the heme.[22]
Key interactions with the propionate of the heme were also found
in Epacadostat and the imidazothiazole series.[27,28] Through
crystallographic studies of substrate-bound IDO1, A third pocket
(pocket C) has been characterized, which is located below the
heme plane, accounting for the inhibition by substrate
phenomenon and the binding of a positive allosteric modulator of
the enzyme (Phe270, Asp274 and Arg343). Interestingly, a heme-
free form (holo-) of IDO1 has been also identified, wherein the
three pockets form a unique and larger ligand binding pocket in
which potent suicide inhibitors bind to the enzyme.[36b,c] In contrast
to other inhibitors, BMS-986205 acts as a suicide inhibitor and
irreversibly inhibits hIDO1 by binding to the apo-form followed by
heme release.[34b,c]
Figure 3. Computational docking for our previously reported IDO1 inhibitor
(IDO1 crystal structure: PDB entry 4PK5).[17]
Results and Discussion
Chemistry: The synthetic procedure for compounds 9a-i,
11a-p and 12a-d is shown in Scheme 1. The anilines 1e and 1q
were synthesized from 1-fluoro-2-nitrobenzene and 2-fluoro-5-
chloro-1-nitrobenzene, respectively. Anilines 1a-r and ethyl
glyoxalate underwent a Van Leusen reaction to yield esters 2a-
q.[40a, b] Reduction of esters 2a-q with LiAlH4 followed by oxidation
with Dess-Martin periodinane provided aldehydes 3a-q. The
regioselective lithiation of N-Ts-indoles 4a-i with nBuLi followed by
nucleophilic addition to the aldehydes gave rise to the
2
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202100107
ChemMedChem
COMMUNICATION
corresponding alcohols 5a-i and 6a-p. The final dehydroxylation
and detosylation afforded the target molecules 9a-i and 11a-p.
Compounds 12e-g were prepared by deprotection of isopropyl or
methyl ether groups with BCl3.[40c] Compound 9j was synthesized
from 13 in three steps. Compound 13 was prepared according to
the published procedure.[41] Compound 9k was synthesized by
oxidation of 5d followed by N-tosyl deprotection. Esters 16a and
16b were prepared by a Mitsunobu reaction of ethyl imidazole-4-
carboxylate with alcohols 15a and 15b.[42] The remaining steps for
the preparation of 12a and 12b were similar those of 9a-i and 11a-
p. Aldehydes 17a and 17b were synthesized from 2- and 3-iodo-
thiophene according to published procedures.[41] The remaining
steps for the preparation of 12c and 12d were similar as that for
preparation of 9a-i and 11a-p.
The syntheses of 10a-h are shown in Scheme 2. A Horner-
Wadsworth-Emmons reaction between 3a and 19 afforded 10a.
Reductive amination of aldehyde 3a and secondary amines 20a-
d provided 10b-e. Coupling of secondary amines 21a-c with Boc-
glycine and further deprotection of tert-butyloxycarbonyl group
provided 22a-c. Compounds 22a-c reacted with isothiocyanate
23 to give the thiourea 24a-c. Selective methylation of 24a-c and
subsequent cyclization of 25a-c with Lawesson’s reagent
afforded 26a-c. Final removal of SMe group of 26a-c with nickel
Raney gave 10f-h.[43]
Fig. 4. Computational docking of 9d (IDO1 crystal structure: PDB entry 4PK5).
Structure-activity relationship studies: We first examined a
series of substituted indole moieties to investigate the effect of
indole ring substitution on potency (Table 1). The determination
of IC50 and EC50 values is described in Supporting Information 4.2
and 4.3. Compound 9a with a non-substituted indole moiety has
an IC50 of 16 μM. We explored whether substituents on the indole
ring would increase the potency by either forming interactions with
Arg231 or hydrophobic interactions with additional residues in
pocket B. Compounds containing 4- or 6-chloro indole moieties
(9b and 9c) did not produce a significant change in IDO1 inhibition
(IC50 = 11 and 22.15 μM, respectively) relative to 9a (IC50 =16.12).
Compared with 9a-c (IC50 = 16.12, 11.77 and 22.15 μM
respectively), the IC50 of 9d was improved to 0.44 μM (EC50 = 0.85
μM) by introducing a chloro group on C5-position of the indole.
This can be rationalized by the the chlorine acting as both a
halogen bond donor and hydrogen bond acceptor with Arg231 in
pocket B (Fig. 4).[43a] This result was consistent with published
results which demonstrated that π-cation interaction between
chloro group and Arg231 is important for IDO inhibitory
activity.[17,19,28] Other functional groups on the C5-position of
indole were also investigated: compound 9e containing 5-bromo
caused a slight loss of activity (IC50 = 0.46 μM, EC50 = 1.9 μM).
However, fluoro, cyano, methoxy and methyl groups (9f-i) caused
a slight decrease in activity relative to 9d with IC50 values ranging
from 2.9 to 8.7 μM. A further loss of activity (IC50 = 28 μM) was
observed for 9j when the CH2 bridge was switched from C2-
position to C3-position of the indole ring. Replacement of CH2
linker with a ketone group led to a complete loss of activity for 9k,
probably because of the electron pair repulsion created between
the carbonyl oxygen’s lone pair of 9k and the propionate of the
heme which may cause a decrease in the binding efficiency.[23] In
addition, several replacements of indole ring with heterocycles
were attempted. However, replacement of indole with pyrrolidine-
Next, we attempted to optimize the phenyl ring in pocket A (Table
2). The unsubstituted phenyl compound 11a had an IC50 of 90 μM.
To improve IDO1 inhibitory activity substitutions on the phenyl ring
were investigated. Ortho-substituted phenyl groups (11b-e) had
poor activity with IC50 values between 10 and 200 μM. Switching
the substitution from the ortho- to para-position did not produce
any improvement in activity (11f-h). Surprisingly, meta-
substituted phenyl yielded more potent compounds (9d, 11i).
Compounds 9d and 11i have IC50 values of 0.44 and 0.98 μM,
respectively. Meta-substitution may be the optimal position to
occupy the small lipophilic cavity at the top of pocket A and, at the
same time, generate an interaction between the sulfur atom of
Cys129 and the chlorine of the 3-chlorophenyl moiety.[11,17,25] The
3-F, 3-OMe substitutions produced compounds with IC50 values
greater than 10 μM (11j and 11k). We next studied the influence
of di-substituted groups on phenyl rings and found di-substituted
compounds 11l-n displaying decreased IDO inhibitory activities
(IC50 > 4.9 μM), compared with most potent mono-substituted
compound 9d. We then explored replacement of the phenyl
functionality with aliphatic and heterocyclic substituents and
observed that compound 12a, containing the cyclopentyl moiety,
had an IC50 of 45 μM (EC50 = 0.7 μM). It is hypothesized this
phenomenon arises due to complications in regulating the IDO1
redox activity and/or an environmental effect, which likely also
accounts for the reported lower EC50 values as compared to the
IC50 values in the cell-based activity assay.[44] As we discussed
above, a 3-halogen phenyl ring is necessary to maintain the
activity by establishing interactions with the amino acid residues
in pocket A. The cyclopentyl ring in 12a cannot interact with
residues in pocket A, which may be the cause of reduced
potency. Changing from a cyclopentyl to cyclohexyl ring in pocket
2,5-dione (10a, IC50 > 200 μM), anilines (10b and 10c, IC50
=
A
led to an inactive compound (12b, IC50 > 200 μM).
53.61 and 82.3 μM respectively) and tetrahydroisoquinoline (10d
and 10e, IC50 > 200 and 200 μM respectively) led to a significant
drop in IDO1 inhibition activities compared with 9d (IC50 = 0.44
μM, EC50 = 0.85 μM). Considering that the basicity of the ligand is
crucial to form a metal complex, increasing the basicity of the
imidazole may result in a stronger binding to the heme iron of
IDO1, which may improve IDO1 inhibitory activity from bottom. To
increase the basicity of the imidazole, a N-atom was introduced
at the C5-position of imidazole ring (10f-h). Compared with 10d
(IC50 > 200 μM) and 10e (IC50 > 200 μM), compounds 10f-h
displayed improved IDO1 inhibitory activities with IC50 of 1.24 to
8.4 μM, respectively, but were less potent than 9d (IC50 = 0.44
μM, EC50 = 0.85 μM).
Replacements of phenyl with a thiophene led to a loss activity
(12c, IC50 = 60 μM and 12d, IC50 = 18 μM).
Based on our previous results and literature precedent,
IDO1 inhibitory activity can be increased by the introduction of a
hydroxyl group at the ortho-position of the phenyl ring.[17,23,25] The
hydroxyl group is thought to interact with Ser167 in the pocket A
through a hydrogen bond, which contributes to the inhibitory
activity. Inspired by this observation, we installed a 2-hydroxyl
group on the phenyl ring of 11a to produce compound 12e and as
3
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202100107
ChemMedChem
COMMUNICATION
Scheme 1. Synthetic routes of 9a-i, 11a-p and 12a-g.
4
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202100107
ChemMedChem
COMMUNICATION
O
P
O
EtO
EtO
O
Cl
NH
NH
HN
N
LiOH monohydrate
EtOH, rt
H
N
Cl
10a
N
CHO
N
N
H
N
R
1 = H, R2 = 4-ClC6H4
10b
Cl
R4
R4
10c R1 = H, R2 = 4-CNC6H4
R3
3a
N
NaBH3CN, AcOH
Cl
Cl
R3
Ph
N
N
N
MeOH, rt
10d
10e
10b-e
N
1 = H, R2 = 4-ClC6H4
1 = H, R2 = 4-ClC6H4
Ph
15a
R
R
N
N
N
N
15b
15c
15d
HN
HN
Cl
NCS
O
O
H
N
H
1) Boc-glycine, HOBT, EDCI, DIPEA
CH2Cl2, 0ºC to rt
N
Cl
Et3N
N
N
NH
NH2·TFA
S
EtOH
R1
R1
2) TFA
CH2Cl2, 0ºC to rt
R1
22a-c
24a-c
21a-c
MeI, K2CO3 AcCN, rt
R1
R1
Cl
O
H
N
N
R1 = Ph
N
N
Cl
10f
Raney-Ni
N
Lawesson’s reagent
DME, rt
N
10g R1 = 4-FC6H4
N
Cl
SMe
R1
N
EtOH, rt
10h
R1 = 3-ClC6H4
MeS
N
25a-c
10f-h
26a-c
Scheme 2. Synthetic routes of 10a-h.
predicted, it displayed a significantly improved IC50 of 2.29 μM,
compared with the non-hydroxyl compound 11a (IC50 = 90 μM).
Encouraged by this result, we further explored whether inclusion
of an additional substituent on the phenyl ring of 12e would
substantially improve binding affinity. A 3-chloro group was
foundto be the best substituent in pocket A for IDO1 inhibitory
potency (9d). 12f was synthesized by installing a 3-chloro group
on the phenyl ring of 12e. Significantly, 12f displayed an inhibitory
activity of 0.16 μM, approximately 2.8-fold more potent than 9d,
with an EC50 of 0.3 μM. Further switching from a 2-hydroxyl to 3-
hydroxyl group (12g) led to decreased activity (IC50 = 1.45 μM),
which suggested that the 2-hydroxyl group played an important
role in maintaining IDO1 inhibitory activity.[17,25,39] The SAR
studies of di-substituted compounds indicated that both a 2-
hydroxyl and 5-chloro in Pocket A (12f) are necessary for potent
IDO inhibition, and other di-substituted compounds (11l-n)
showed decreased IDO inhibitory activities.
hydrogen bond. The interaction with propionate of the heme was
found to be important for the IDO inhibitory activity.[17,22,27,28]
The binding modes of potent compounds 9d and 12f with
IDO1 were further evaluated by computational modelling. As
shown in Fig. 4, computational modelling of 9d reveals that the
N1-phenyl ring likely occupies pocket A, and the indole ring
extends to pocket B. The imidazole ring coordinates with the
heme iron via an N-atom. The binding model of 12f was similar
with that of 9d. Computational modeling of 12f (Fig. 5) further
suggests that the 2-hydroxyphenyl group fits in pocket A and the
hydroxyl group acts as a hydrogen bond donor to Ser167, which
may account for the high potency against IDO1. The indole-NH
was found to interact with the propionate of the heme through a
Fig. 5. Computational docking of 12f (IDO1 crystal structure: PDB entry
4PK5).[17]
5
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202100107
ChemMedChem
COMMUNICATION
Table 1. SAR studies of pocket B.
GBVI/WSA dG
Compound
9a
R
IC50 (µM)
EC50 (µm)
ClogP
LE[a]
BBB Score[b]
docking Score
(kcal/mol)
16.12
4.63
0.31
4.24
4.18
-8.4039
-8.5461
9b
11.77
5.51
0.30
9c
9d
9e
9f
22.15
0.44
0.46
2.90
2.32
8.67
2.57
10.0
0.85
1.90
3.34
2.12
1.66
>10
5.51
5.51
5.51
5.66
4.45
4.88
5.36
0.30
0.39
0.39
0.34
0.33
0.30
0.34
4.18
4.18
4.13
4.2
-8.7656
-8.9499
-8.9346
-8.5877
-9.2024
-9.1466
-8.9140
-8.5745
Cl
N
H
9g
9h
9i
4.11
4.31
4.2
Me
N
H
9j
26.86
>200
>2.50
5.51
5.24
0.28
4.19
3.94
9k
-8.4426
6
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202100107
ChemMedChem
COMMUNICATION
-7.2551
10a
200
1.92
4.07
10b
10c
10d
10e
10f
53.61
82.3
>10
4.77
3.89
5.24
4.69
5.60
0.28
0.26
4.98
4.67
4.95
4.97
4.94
-8.2282
-8.6985
>200
>200
1.24
-7.3541
-8.1877
-8.8333
-9.1090
>2.50
>2.50
0.35
0.33
10g
10h
1.30
8.40
5.74
6.31
4.89
4.88
-7.5222
0.28
The cellular assay that was employed can be found in the 4.3 section of the Supporting Information. [a] Ligand efficiency (LE) calculated using IDO1 IC50. [b] BBB
Score calculated according to Ref [49].
Table 2. SAR studies of pocket A.
GBVI/WSA dG
Compound
11a
R
IC50 (µM)
EC50 (µM)
ClogP
LE[a]
BBB Score[b]
docking Score
(kcal/mol)
90.31
5.01
0.26
4.24
-8.4455
-9.0412
-8.4820
11b
11c
144.6
156.1
5.74
4.94
0.23
0.23
4.17
4.19
7
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202100107
ChemMedChem
COMMUNICATION
11d
10.07
>200
58.66
>200
45.74
4.83
5.66
5.51
4.94
4.82
0.29
4.36
-9.3821
-7.7936
-8.3626
-8.4238
-8.9892
-9.2475
11e
11f
4.23
4.18
4.31
4.2
0.26
11g
11h
0.25
0.37
11i
11j
0.98
12.85
25.90
7.20
5.89
4.94
4.82
4.13
4.2
-8.6657
-9.3143
-9.1159
0.30
0.27
OMe
11k
>2.5
4.31
11l
114.0
5.44
0.23
0.31
4.16
-8.6350
-8.367
11m
11n
4.85
6.34
4.99
4.12
4.27
34.40
0.25
0.30
12a
12b
44.81
>200
0.71
4.39
4.95
4.8
-8.998
-7.898
4.77
8
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202100107
ChemMedChem
COMMUNICATION
12c
60.0
18.22
2.29
4.69
4.69
4.44
0.28
0.32
0.36
4.22
-7.989
-8.321
-9.001
-9.206
12d
12e
>2.5
0.64
4.24
4.22
12f
0.16
1.45
0.30
5.21
5.21
0.40
0.36
4.21
4.29
-8.739
12g
> 2.5
The cellular assay that was employed can be found in the 4.3 section of the Supporting Information. [a] Ligand efficiency (LE) calculated using IDO1 IC50. [b] BBB
Score calculated according to Ref [49].
Table 3. Pharmacokinetic parameters of compounds 9d and 12f. PO dose was administered at 5 mg/kg and IV dose was administered at 2.5 mg/kg.
9d
12f
IV
PO
IV
PO
Parameter
Plasm
a
Brain
Plasma
Brain
Plasma
Brain
Plasma
Brain
R2
0.94
1,233
-
0.93
4,825
-
0.98
235
0.5
1.9
-
0.96
1,114
0.5
0.96
1,251
-
0.98
269
-
0.87
135
0.5
2.1
-
0.75
50
adj
Cmax (ng/mL or g)
Tmax (h)
2
T1/2 (h)
1.0
3.4
3.7
6
0.9
1.4
1.8
5.2
4.1
8
2.2
14.0
-
VdSS (L)
0.6
-
8.3
Cl (mL/min/kg)
TLast
0.5
2.4
6
2.2
2.2
3.6
8
20.7
8
6
8
8
AUC0-Tlast (ng·h/mL or g)
666
675
5,023
5,118
357
395
2,115
2,171
593
608
1,032
1,159
318
342
324
1,042
AUC0-∞
MRT0-Tlast (h)
0.84
29
1.05
-
-
-
-
-
1.03
28
2.93
-
-
-
-
-
Oral Bioavailability (%)
AUC0-Tlast Brain/Plasma
7.5
5.9
1.7
1.0
Selected potent compounds (9c-i, 11n, 12a, 12c-f) were
advanced to solubility and in vitro intrinsic clearance ClInt using
mouse liver microsomes (MLM) studies (9f-h, 11n, 12a, 12e-f).
Non-hydroxyl compounds, 9d-f bearing a 5-halogen in Pocket B
had low solubility (Ksol < 10 µM). The solubility for 9c (6-Cl) and
9g-i (5-CN, 5-OMe, 5-Me) were improved with Ksol values
ranging from 10 to 31.2 μM. Compounds containing an aliphatic
ring (12a) and heterocycle (12c and d) in Pocket A exhibited poor
solubility (Ksol < 1.2 µM), likely due to high lipophilicity. The
phenolic compound 12e had good solubility with Ksol = 28 µM
possibly because the hydroxyl group increases hydrophilicity. The
most potent compound 12e displayed improved solubility with
Ksol = 9.38 μM.
9f-h resulted in high MLM clearance (CLint > 348 mL/min/kg)
possibly because of high lipophilicity. Compound 12a had poor
MLM stability (CLint 1337 mL/min/kg), possibly due to poor
=
metabolic stability of the cyclohexyl group as first pass
metabolism of cycloalkyl groups is a well-documented issue.[23]
The phenolic compound 12e had MLM data with CLint 319
=
mL/min/kg, whilst 12f showed acceptable MLM data (31.7
mL/min/kg). Compounds 9d and 12f were also assayed for brain
protein binding via equilibrium dialysis (percentage bound = 98.29
and 99.92% respectively). The high protein binding percentages
9
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202100107
ChemMedChem
COMMUNICATION
for 9d and 12f also arose from the high lipophilicity of these
compounds.[17]
We calculated the LE for our compounds (Tables 1 and 2)
to assess whether they are in the appropriate range for druglike
compounds (LE > 0.3). The most potent compound, 12f,
displayed the best LE value (0.4), which is comparable to the
mean LE value (0.45) for most approved drugs administered
orally.[45] The ‘GBVI/WSA dG’ docking scores[46a] for our
compounds are given in Table 1 and 2 to estimate the binding
affinity of each ligand in the active site of IDO1. We can see that
most of the active compounds, including 9d and 12f, have low
GBVI/WSA binding free energy. Detailed information of docking
simulations protocols, ligand structure optimization, dataset
curation and descriptor calculation has been given in supporting
information and in prior publications.[46b-d] There are several
potential applications for a brain penetrant IDO1 inhibitor.[16,45,47]
To evaluate our compounds for their ability to penetrate the brain,
we used two assessment tools, one called the Brain Exposure
Efficiency (BEE) score, and the other called the Blood Brain
Barrier (BBB) score; both developed by our group.[48,49] The BBB
score estimates passive diffusion across the BBB, whereas the
BEE score predicts active transport across the BBB as a function
of efflux and influx transporters. The BBB score values of our
active compounds ranged between 4 to 5, which is optimal for
BBB penetration. The BBB scores for 9d and 12f are 4.18 and
4.21 respectively, which predict a good brain penetration for these
analogs. The BBB scores for all other analogs are presented in
Tables 1 and 2, respectively. The predicted Kp,uu values (the
BEE scores) for both 9d and 12f are above 0.10, which is optimal
for brain penetration.[49] The BEE Score efflux and influx
transporter activity calculations predicts that the analog 9d is
predicted to be a substrate of OCT1 influx transporter and 12f is
predicted to be a substrate of both OCT1 and OCT2 influx
transporters. Compound 9d is predicted not to be a substrate of
either the P-gp or BCRP efflux transporters. More details about
these BEE calculations are provided in supporting information
Tables S1-S6. Brain penetration and transporter affinity
predictions could be employed for further studies in CNS drug
development. Based on the above results we examined the
pharmacokinetic profiles of compounds 9d and 12f.
Fig. 6. Pharmacokinetic profiles of Compounds 9d and 12f. A) Mice were
administered 9d 5 mg/kg PO and 2.5 mg/kg IV. Brain and plasma were collected
at each indicated time point and concentration of the compound was
determined. B) Mice were administered 12f 5 mg/kg PO and 2.5 mg/kg. Brain
and plasma were collected at each indicated time point and concentration of the
compound was determined. Full PK parameters for each compound are
presented in Table 3.
Conclusion
In summary, on the basis of computational docking simulations, a
series of 2-((1-phenyl-1H-imidazol-5-yl)methyl)-1H-indoles was
designed, synthesized, and evaluated as IDO1 inhibitors.
Optimization of pocket A and B substituents led to identification of
compound 12f as our most potent IDO1 inhibitor. A structure
activity relationship analysis indicated that a 5-chloro group on the
indole ring is essential for the IDO1 inhibitory activity in pocket B.
Both the 3-chloro and 2-hydroxylgroups on the phenyl ring are
important for the IDO1 inhibitory activity targeting pocket A.
Computer docking studies demonstrated that the hydroxyl group
may interact with the Ser167 residue in pocket A. Potent
compounds 9d and 12f were predicted to have good brain
penetrance as calculated by the BBB score and BEE score. We
then performed a pharmacokinetic study, which demonstrated
that 9d and 12f were both brain penetrant, though they had
relatively low exposure and half-lives. The scaffold we have
developed could be further developed to produce brain penetrant
IDO1 inhibitor compounds for CNS indications, such as neuro-
oncology or Alzheimer’s disease. Towards this goal, studies on
the replacement of the 2-hydroxyl group with other H-donors,
which may enhance both IDO1 inhibitory activity and ADME/PK
properties (such as solubility, MLM and PK), are ongoing in our
laboratory.
Our pharmacokinetic studies on 9d and 12f (Fig.6; Table 3)
demonstrated that both compounds were reasonably brain
penetrant (AUCBrain/Plasma > 1) but that both compounds had
somewhat low oral bioavailability (< 30%). The brain exposure of
9d was several fold that of 12f (AUC
Brain = 2115, 324 for
0-Tlast
9d and 12f, respectively) The 2-hydroxyl group in 12f decreases
lipophilicity and increase the number of hydrogen bond donors
relative to 9d, which may explain its lower brain exposure relative
to 12f. Additionally, we attribute the short half-lives of the
compounds to possible oxidation of the methylene between the
imidazole and indole on the pocket B side, or, oxidation of the
chlorobenzene or phenol on the pocket A side of 9d and 12f
respectively.[50]
Acknowledgements
This work was supported by a CIHR operating grant and by
funding from the Krembil Foundation. DFW acknowledges salary
support from a Canada Research Chair, Tier 1. We thank Dr.
Christopher J. Barden and Dr. Jake Goodwin-Tindall for helpful
discussions.
Keywords: IDO1 inhibitors • imidazole • structure-activity
relationship • computational modelling •
[1]
D. N. Khalil, E. L. Smith, R. J. Brentjens, J. D. Wolchok, Nat. Rev. Clin.
Oncol. 2016, 13, 273-290.
[2]
[3]
I. Mellman, G. Coukos, G. Dranoff, Nature, 2011, 480, 480-489.
H. Sugimoto, S. Oda, T. Otsuki, T. Hino, T. Yoshida, Y. Shiro, Proc. Natl.
Acad. Sci. U. S. A. 2006, 103, 2611−2616.
[4]
[5]
A. B. Dounay, J. B. Tuttle, P. R. Verhoest, J. Med. Chem. 2015, 58,
8762−8782.
G. C. Prendergast, C. Smith, S. Thomas, L. Mandik-Nayak, L. Laury-
Kleintop, R. Metz, A. J. Muller, Cancer Immunol. Immunother, 2014, 63
721-735.
10
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202100107
ChemMedChem
COMMUNICATION
[6]
[7]
[8]
[9]
J. Godin-Ethier, L. A. Hanafi, C. A. Piccirillo, R. Lapointe, Clin. Cancer
[39] U. F. Röhrig, S. R. Majjigapu, A. Grosdidier, S. Bron, V. Stroobant, L.
Pilotte, D. Colau, P. Vogel, B. J. Van den Eynde, V. Zoete, O. Michielin,
J. Med. Chem. 2012, 55, 5270−5290.
[40] a) A. M. Van Leusen, J. Wildeman, O. H. Oldenziel, J. Org. Chem. 2012,
42, 1153-1159. b) B. Chen, M. S. Bednarz, R. Zhao, J. E. Sundeen, P.
Chen, Z. Shen, A. P. Skoumbourdis, J. C. Barrish, Tetrahedron Lett.
2000, 41, 5453-5456. c) A. V. Kalinin, M. A. Reed, B. H. Norman, V.
Snieckus, J. Org. Chem. 2003, 68, 5992-5999.
[41] Y. Wu, S. Izquierdo, P. Vidossich, A. Lledýs, A. Shafir, Angew. Chem.
Int. Ed. 2016, 55, 7152 – 7156.
[42] J. K. Laha, G. D. Cuny, J. Org. Chem. 2011, 76, 8477–8482.
[43] a) C. Lamberth, R. Dumeunier, S. Trah, S. Wendeborn,J. Godwin, P.
Schneiter, A. Corran, Bioorg. Med. Chem. 2013, 21,127-134. b) S.
Panda, A. Roy, S. J. Deka, V. Trivedi, D. Manna, ACS Med. Chem. Lett.
2016, 7, 12, 1167–1172.
[44] A. L. Hopkins, G. M. Keseru, P. D. Leeson, D. C. Rees, C. H. Reynolds,
Nat. Rev. Drug Discovery 2014, 13, 105−121.
[45] E. Fertan, K. R.J. Stover, M. G. Brant, P. M. Stafford, B. Kelly, E. Diez-
Cecilia, A. A. Wong, D. F. Weaver, R. E. Brown, Front. Pharmacol. 2019,
24, 1044.
[46] a) C. R. Corbeil, C. I. Williams, P. Labute, J. Comput.-Aided Mol. Des.
2012, 26(6), 775-786. b) Z. Wang, Y. Wang, P. Vilekar, S. P. Yang, M.
Gupta, M. I. Oh, A. Meek, L. Doyle, L. Villar, A. Brennecke, I. Liyanage,
M. Reed, C. J. Barden, D. F. Weaver, PeerJ 2020, 7, e9533. c)M. Gupta,
E. F. da Silva, H. F. Svendsen, J. Phys. Chem. B. 2013,117(25), 7695-
7709. d)I. S. Liyanage, M. Gupta, F. Wu, M. Taylor, M. D. Carter, D. F.
Weaver, Chemotherapy. 2019, 64(1), 22-27.
Res. 2011, 17, 6985−6991.
L. Vécsei, L. Szalárdy, F. Fülöp, J. Toldi, Nat. Rev. Drug Discovery, 2013,
12, 64−82.
T. Weng, X. Qiu, J. Wang, Z. Li, J. Bian, Eur. J. Med. Chem. 2018, 143,
656-669.
X. Wang, S. Sun, Q. Dong, X. Wu, W. Tang, Y. Xing, MedChemComm
2019, 10, 1740–1754.
[10] M. Sono, S. G. Cady, Biochemistry, 1989, 28, 5392−5399.
[11] S. Fallarini, A. Massarotti, A. Gesù, S. Giovarruscio, G. Coda Zabetta, R.
Bergo, B. Giannelli, A. Brunco, G. Lombardi, G. Sorba and T. Pirali.
MedChemComm 2016, 7, 409–419.
[12] J. A. C. Alexandre, M. K. Swan, M. J. Latchem, D. Boyall, J. R. Pollard,
S. W. Hughes, J. Westcott, ChemBioChem 2018, 19, 552–561.
[13] S. Y. Lin, T. K. Yeh, C. C. Kuo, J. S. Song, M. F. Cheng, F. Y. Liao, M.
W. Chao, H. L. Huang, Y. L. Chen, C.Y. Yang, M. H. Wu, C. L. Hsieh, W.
Hsiao, Y. H. Peng, J. S. Wu, L. M. Lin, M. Sun, Y. S. Chao, C. Shih, S.
Y. Wu, S. L. Pan, M. S. Hung, S. H. Ueng. J. Med. Chem. 2016, 59, 419-
430.
[14] W. P. Malachowski, M. Winters, J. B. DuHadaway, A. Lewis-Ballester, S.
Badir, J. Wai, M. Rahman, E. Sheikh, J. M. LaLonde, S. R. Yeh, G. C.
Prendergast, A. J. Muller, Eur. J. Med. Chem. 2016, 108, 564-576.
[15] S. Paul, A. Roy, S.J. Deka, S. Panda, V. Trivedi, D. Manna, Eur. J. Med.
Chem. 2016, 121, 364-375.
[16] S. Crosignani, P. Bingham, P. Bottemanne, H. Cannelle, S.
Cauwenberghs,M. Cordonnier, D. Dalvie, F. Deroose, J.L. Feng, B.
Gomes, S. Greasley, S. E. Kaiser, M. Kraus, M. Negrerie, K. Maegley,
N. Miller, B.W. Murray, M. Schneider, J. Soloweij, A.E. Stewart, J.
Tumang, V. R. Torti, B. Van Den Eynde, M. Wythes, J. Med. Chem. 2017,
60, 9617-9629.
[17] M. G. Brant, J. Goodwin-Tindall, K. R. Stover, P. M. Stafford, F. Wu, A.
R. Meek, P. Schiavini, S. Wohnig, D. F. Weaver, ACS Med. Chem. Lett.
2018, 9, 131-136.
[18] K. Fang, G. Dong, Y. Li, S. He, Y. Wu, S. Wu, W. Wang, C. Sheng, ACS
Med. Chem. Lett. 2018, 9, 312-317.
[47] L. Zhai, K. L. Lauing, A. L. Chang, M. Dey, J. Qian, Y. Cheng, M. S.
Lesniak, D. A. Wainwright, J. Neurooncol. 2015, 123, 395–403.
[48] M. Gupta, H. J. Lee, C. J. Barden, D. F. Weaver, J. Med. Chem. 2019,
62, 9824−9836.
[49] M. Gupta, T. Bogdanowicz, M. A. Reed, C. J. Barden, D. F. Weaver, ACS
Chem. Neurosci. 2020, 11, 205-224.
[50] A. Mandal, M. Patel, Y. Sheng, A.K. Mitra, Curr Drug Targets, 2016, 17,
1773 - 1798.
[19] Y. Zou, F. Wang, Y. Wang, Q. Sun, Y. Hu, Y. Li, W. Liu, W. Guo, Z.
Huang, Y. Zhang, Q. Xu, Y. Lai, Eur. J. Med. Chem. 2017, 140, 293-304.
[20] A. Coluccia, S. Passacantilli, V. Famiglini, M. Sabatino, A. Patsilinakos,
R. Ragno, C. Mazzoccoli, L. Sisinni, A. Okuno, O. Takikawa, R. Silvestri,
G. La Regina, J. Med. Chem. 2016, 59, 9760-9773.
[21] K. Fang, S. Wu, G. Dong, Y. Wu, S. Chen, J. Liu, W. Wang, C. Sheng,
Org. Biomol. Chem. 2017, 15, 9992-9995.
[22] Y. H. Peng, S.H. Ueng, C.T. Tseng, M. S. Hung, J. S. Song, J. S. Wu,
F.Y. Liao, Y.S. Fan, M. H. Wu, W. C. Hsiao, C. C. Hsueh, S. Y. Lin, C.
Y. Cheng, C. H. Tu, L. C. Lee, M. F. Cheng, K. S. Shia, C. Shih, S. Y.
Wu, J. Med. Chem. 2016, 59, 282-293.
[23] S. Kumar, J. P. Waldo, F. A. Jaipuri, A. Marcinowicz, C. Van Allen, J.
Adams, T. Kesharwani, X. Zhang, R. Metz, A. J. Oh, S. F. Harris, M. R.
Mautino, J. Med. Chem. 2019, 62, 6705−6733.
[24] S. Chen, W. Guo, X. Liu, P. Sun, Y. Wang, C. Ding, L. Meng, A. Zhang,
Eur. J. Med. Chem. 2019, 179, 38-55.
[25] a) S. Kumar, D. Jaller, B. Patel, J. M. LaLonde, J. B. DuHadaway, W. P.
Malachowski, G. C. Prendergast, A. J. Muller, J. Med. Chem. 2008, 51,
4968-4977; b) Y-H. Peng, F-Y. Liao, C-T. Tseng, R. Kuppusamy, A-S.
Li, C-H. Chen, Y-S. Fan, S-Y. Wang, M-H. Wu, C-C. Hsueh, J-Y. Chang,
L-C. Lee, C. Shih, K-S Shia, T-K. Yeh, M-S. Hung, C-C. Kuo, J-S. Song,
S-Y. Wu, S-H . Ueng J. Med. Chem. 2020, 63, 4, 1642–1659.
[26] W. Tu, F. Yang, G. Xu, J. Chi, Z. Liu, W. Peng, B. Hu, L. Zhang, H. Wan,
N. Yu, F. Jin, Q. Hu, L. Zhang, F. He, W. Tao, ACS Med. Chem. Lett.
2019, 10, 949-953.
[27] E. W. Yue, R. Sparks, P. Polam, D. Modi, B. Douty, B. Wayland, B.
Glass, A. Takvorian, J. Glenn, W. Zhu, M. Bower, X. Liu, L. Leffet, Q.
Wang, K. J. Bowman, M.J. Hansbury, M. Wei, Y. Li, R. Wynn, T. C. Burn,
H. K. Koblish, J. S. Fridman, T. Emm, P.A. Scherle, B. Metcalf, A.P.
Combs, ACS Med. Chem. Lett. 2017, 8, 486-491.
[28] S. Tojo, T. Kohno, T. Tanaka, S. Kamioka, Y. Ota, T. Ishii, K. Kamimoto,
S. Asano, Y. Isobe. ACS Med. Chem. Lett. 2014, 5, 1119−1123.
[29] Q. Du, X. Feng, Y. Wang, X. Xu, Y. Zhang, X. Qu, Z. Li, J. Bian, Eur. J.
Med. Chem.2019, 182, 111629.
[30] S. Cai, X. Yang, P. Chen, X. Liu, J. Zhou, H. Zhang, Bioorganic
Chemistry, 2020, 94, 103356.
[31] G. L. Beatty, P. J. O’Dwyer, J. Clark, J. G. Shi, K. J. Bowman, P. A.
Scherle, R. C. Newton, R. Schaub, J. Maleski, L. Leopold, T. F. Gajewski,
Clin. Cancer Res. 2017, 23, 3269.
[32] J. E. Cheong, A. Ekkati, L. Sun, Expert Opin. Ther. Pat. 2018, 28, 317-
330.
[33] M.R. Mautino, S. Kumar, H. Zhuang, J. Waldo, F. Jaipuri, H. Potturi, E.
Brincks, J. Adams, A. Marcinowicz, C. Van Allen, N. Vahanian, C. J. Link,
Cancer Res. 2017, 77, 4076.
[34] a) V. G. Long, R. Dummer, O. Hamid, T. F. Gajewski, C. Caglevic, S.
Dalle, A. Arance, M. S.Carlino, J.-J. Grob, T. M. Kim, L. Demidov, C.
Robert, J. Larkin, J. R. Anderson, J. Maleski, M. Jones, S. J. Diede, T.
C. Mitchell, Lancet Oncol. 2019, 20, 1083–1097. b) M. T. Nelp, P. A.
Kates, J. T. Hunt, J. A. Newitt, A. Balog, D. Maley, X. Zhu, L. Abell, A.
Allentoff, R. Borzilleri, H. A. Lewis ,Z. Lin, S. P. C. Seitz, J. T. Yan,
Groves, Proc. Natl. Acad. Sci. U.S.A. 2018, 115 3249−3254. c) K. N.
Pham, S-R Yeh, J. Am. Chem. Soc. 2018, 140, 14538–14541.
[35] U. F. Röhrig, L. Awad, A. Grosdidier, P. Larrieu, V. Stroobant, D. Colau,
V. Cerundolo, A. J. Simpson, P. Vogel, B. J. Van den Eynde, V. Zoete,
O. Michielin, J. Med. Chem. 2010, 53, 1172−118.
[36] U. F. Röhrig, A. Reynaud, Somi Reddy Majjigapu, P. Vogel, Florence
Pojer, V. Zoete, J. Med. Chem. 2019, 62, 8784−8795.b) A. Coletti, M.
Ballarotto, A. Riccio, A. Carotti, U. Grohmann, E. Camaioni, A.
Macchiarulo, ChemMedChem. 2020, 15, 891-899. c) A. Lewis-Ballester,
K. N. Pham, D. Batabyal, S. Karkashon, J. B. Bonanno, T. L. Poulos, S.-
R. Yeh, Nature Commun, 2017, 8, 1693.
[37] H. Zhang, K. Liu, Q. Pu, A. Achab, M. J. Ardolino, M. Cheng, Y. Deng,
A. C. Doty, H. Ferguson, X. Fradera, I. Knemeyer, R. Kurukulasuriya, Y.
Lam, C. A. Lesburg, T. A. Martinot, M. A. McGowan, J. R. Miller, K. Otte,
P. J. Biju, N. Sciammetta, N. Solban, W. Yu, H. Zhou, X. Wang, D. J.
Bennett, Y. Han, ACS Med. Chem. Lett. 2019, 10, 1530−1536.
[38] K. N. Pham, A. Lewis-Ballester, S. Yeh, J. Am. Chem. Soc. 2019, 141,
18771−18779.
11
This article is protected by copyright. All rights reserved.