Discovery of Hydroxyamidine Based Inhibitors of IDO1 for Cancer Immunotherapy with Reduced Potential for Glucuronidation

Article abstract of DOI:10.1021/acsmedchemlett.9b00572

Following the impressive success of checkpoint inhibitors in the treatment of cancer, combinations of IDO1 inhibitors with PD-1/PD-L1 antibodies are in clinical development aiming to increase response rates. Using the hydroxyamidine pharmacophore of the IDO1 inhibitor INCB14943 as a starting point for the design of new inhibitors, the potential shortcomings of extensive hydroxyamidine glucuronidation in humans was addressed. Compounds were optimized using a stability assay with recombinant UGT1A9 enzyme together with the measurement of glucuronide formation in human hepatocytes. Optimized analog 24 showed cellular and biochemical IDO1 IC₅₀ values in the low nanomolar range, a suitable in vitro ADME/PK profile, and efficacy in an animal model of cancer. In a humanized liver mouse model the lead compound exhibited significantly reduced glucuronidation compared to epacadostat (2).

Full text of DOI:10.1021/acsmedchemlett.9b00572

pubs.acs.org/acsmedchemlett
Letter
Discovery of Hydroxyamidine Based Inhibitors of IDO1 for Cancer
Immunotherapy with Reduced Potential for Glucuronidation
Christoph Steeneck,* Olaf Kinzel, Simon Anderhub, Martin Hornberger, Sheena Pinto,
Barbara Morschhaeuser, Floriane Braun, Gerald Kleymann, and Thomas Hoffmann
Cite This: https://dx.doi.org/10.1021/acsmedchemlett.9b00572
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: Following the impressive success of checkpoint
inhibitors in the treatment of cancer, combinations of IDO1
inhibitors with PD-1/PD-L1 antibodies are in clinical development
aiming to increase response rates. Using the hydroxyamidine
pharmacophore of the IDO1 inhibitor INCB14943 as a starting
point for the design of new inhibitors, the potential shortcomings
of extensive hydroxyamidine glucuronidation in humans was
addressed. Compounds were optimized using a stability assay
with recombinant UGT1A9 enzyme together with the measure-
ment of glucuronide formation in human hepatocytes. Optimized analog 24 showed cellular and biochemical IDO1 IC50 values in
the low nanomolar range, a suitable in vitro ADME/PK profile, and efficacy in an animal model of cancer. In a humanized liver
mouse model the lead compound exhibited significantly reduced glucuronidation compared to epacadostat (2).
KEYWORDS: Indoleamine-2,3-dioxygenase-1, IDO1, glucuronidation, UGT1A9, tryptophan-kynurenine-AhR-pathway,
cancer immunotherapy
inflammatory signals such as interferon gamma (IFNγ), and
mpressive progress in cancer immunotherapy has been
Imade since the emergence of checkpoint inhibitors such as immune suppression results when high IDO1 activity causes a
local tryptophan depletion leading to decreased proliferation of
effector T-cells.9 In addition, kynurenine and further down-
stream metabolites are agonists of the ligand-controlled nuclear
transcription factor arylhydrocarbon receptor (AhR). In the
tumor microenvironment activated AhR causes an immune
tolerant state by reducing effector T-cells and increasing
immune suppressor cells.10,11
Following the success with cancer immunotherapies based
on checkpoint inhibitors, the development of IDO1 inhibitors
as immune modulators has gained great attention in recent
years. Results from several clinical phase 1/2 trials in
combination with checkpoint inhibitors were encouraging.12
Selected examples of IDO1 inhibitors that went into clinical
development are shown in Figure 1. Reported inhibitors fall
into two classes depending on their mode of action: heme-
binding inhibitors exemplified by navoximod (1)13 and
epacadostat (2)14 next to heme-displacing inhibitors exempli-
fied by linrodostat (3).15
antibodies against programmed cell death protein 1 (PD-1)
and its ligand (PD-L1) as well as cytotoxic T-lymphocyte
associated antigen 4 (CTLA-4). However, these new
therapeutics are still associated with shortcomings which
include moderate response rates,1 adaptive and intrinsic
resistance especially of some solid tumors,2 as well as strong
side effects that can result in treatment termination.3 Besides
the PD-1/PD-L1 axis the so-called tryptophan-kynurenine-
arylhydrocarbon receptor (Trp-Kyn-AhR) pathway has been
intensively investigated as a potential central mediator of
immune evasion.4
Indoleamine-2,3-dioxygenase-1 (IDO1) was first described
by Munn and Mellor in 1998 as an immune modulating
enzyme involved in the prevention of fetal rejection.5 The
same group, along with other researchers, showed that IDO1 is
strongly involved in tumor immune resistance.6,7
The immune suppressive effect of IDO1 results from its
capacity to degrade tryptophan to N-formylkynurenine, the
first and rate limiting step of the kynurenine pathway. The
same degradation reaction is catalyzed by tryptophan 2,3-
dioxygenase (TDO), a related enzyme with slightly different
substrate specificity. In contrast to IDO1, which is expressed in
various tissues throughout the body, TDO is mainly expressed
in the liver.8 In subsequent steps of the kynurenine pathway N-
formylkynurenine is converted to kynurenine and other
downstream metabolites. IDO1 expression can be induced by
The initial hopes regarding IDO1 inhibition as a therapeutic
concept were disappointed by the outcome of ECHO-301, a
Received: December 4, 2019
Accepted: January 22, 2020
https://dx.doi.org/10.1021/acsmedchemlett.9b00572
© XXXX American Chemical Society
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX
A

ACS Medicinal Chemistry Letters
pubs.acs.org/acsmedchemlett
Letter
Figure 1. Examples of IDO1 Inhibitors in clinical development.
Figure 2. Structure based design from INCB14943 (4). (A) Overlay of 5 with INCB14943 (4). (B) The methyl group of 5 points toward the
pocket entry and solvent, suggesting SAR tolerance of a variety of polar groups. ICM 3.8-6a modeling software (Molsoft, San Diego, CA) was used
with standard parameters for docking. The cocrystal structure of an NLG-919 analog (navoximod scaffold) with IDO1 (PDB: 5ek4) was used as
reference structure. (C) IDO1 potencies of INCB14943 analogues devoid of an amino substituent.
large phase 3 clinical trial with epacadostat in combination
with the PD-1 antibody pembrolizumab. No significant
differences of the epacadostat/pembrolizumab group were
observed versus placebo/pembrolizumab group in patients
with advanced melanoma.16 Since then, the strong interest in
IDO1 inhibitors has diminished, but clinical investigation of
both epacadostat17 and linrodostat17,18 continues, and at the
same time other approaches to inhibit the Trp-Kyn-AhR
pathway such as Kyn degrading enzymes and AhR antagonists
are being pursued.19−22
A recent publication on clinical trial data of the PD-1
antibody nivolumab describes a strong correlation of serum
kynurenine/tryptophan levels with overall survival of patients
receiving nivolumab. This suggests a patient stratification and
metabolic monitoring in future trials with combinations of
PD1 blockage and IDO or IDO/TDO inhibitors as a
prerequisite.23 Moreover, these results underline that kynur-
enine signaling could still be regarded as a relevant therapeutic
pathway.
caught our interest. Among several attempts to identify new
chemotypes by ligand-based design, one approach was based
on hydroxyamidines. From modeling of hydroxyamidine
NCB14943 (4),8,9 a prototypic heme binding IDO1 inhibitor
and scaffold of epadadostat, we envisioned analogues devoid of
an amino group attached to the 1,2,5-oxadiazolyl ring as
exemplified by analogs 5 and 6 (Figure 2C). In a published
binding pose of 4 binding to the active site of IDO1, the
contribution of a hydrogen bridge between the amino group
and the heme propionic acid is suggested as a possibility based
on the vicinity of these two groups.8 Docking of 4 and a
corresponding ether analog (5) into the binding pocket
showed that the SAR might allow other groups than amino,
with absent hydrogen bonding to the heme propionic acid
(Figure 2A). Moreover, ether and thioether analogues, in our
model, showed binding conformations with the alkyl group
pointing toward the entry of the B-pocket and solvent (Figure
2B). This anticipated a certain level of SAR tolerance for
diverse moieties in this region that could be exploited for the
identification of new heme binding IDO1 inhibitors with
improved properties. To test the hypothesis that analogs of 4
In the course of our ongoing AhR antagonist program the
mechanistically related and growing field of IDO1 inhibitors
B
https://dx.doi.org/10.1021/acsmedchemlett.9b00572
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
pubs.acs.org/acsmedchemlett
Letter
Figure 3. IDO1 potencies and in vitro stability data of epacadostat (2) compared with 7 and 8.
Table 1. Optimization of B-Pocket Substituents
a
b
c
d
IC50 values are the mean of at least two runs. Human microsomal stability. Mouse microsomal stability. Stability in UGT1A9 recombinant
e
f
enzyme assay. % Glucuronide formation after 120 min incubation with human hepatocytes. Pure enantiomer, absolute configuration unknown.
without an amino group attached to the oxadiazole might still
be active IDO1 inhibitors, compounds 5 and 6 were
synthesized and tested. In the biochemical assay ether 5
showed a similar potency to compound 4 (5: IC50 = 230 nM,
C
https://dx.doi.org/10.1021/acsmedchemlett.9b00572
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
pubs.acs.org/acsmedchemlett
Letter
Table 2. Optimization of A-Pocked Substituents
a
b
c
d
IC50 values are the mean of at least two runs. Human microsomal stability. Mouse microsomal stability. Stability in UGT1A9 recombinant
e
enzyme assay. % Glucuronide formation after 120 min incubation with human hepatocytes.
4: IC50 = 210 nM), while thioether 6 was less active (IC50
=
with recombinant UGT1A9 as well as in human hepatocytes
early in the screening path.
560 nM). Both analogues showed significantly higher potency
in an IDO1 assay based on SKOV3 cells. This unusual
phenomenon has been observed by others in the IDO1 field
and is attributed to differences between the isolated enzyme
and native IDO1 in cells including the presence of a methylene
blue ascorbate system for regeneration of ferrous IDO, that is
present in the biochemical assay.8,24 With an IC50 of 39 nM in
the cell-based assay ether 5 was 2.5-fold less potent than
INCB14943 (4) whereas thioether 6 was nearly equally potent
with an IC50 of 15 nM. The higher cell-based IDO1 potency of
thioether analogues compared to corresponding ethers was
generally observed.
At the time we planned to further invest in the optimization
of the new hydroxyamidine based IDO1 inhibitors, a
publication on clinicial data of epacadostat appeared, showing
that, next to atypical reductive metabolism by gut bacteria, its
hydroxyamidine moiety is extensively glucuronidated in
humans. At steady state after 10 days, 50 mg epacadostat
BID p.o., the ratio of AUC1−12h glucuronide/parent was
8.1:1.25 In vitro tests with a panel of 12 recombinant UGT
isoforms (UDP-glucuronosyltransferase) identified UGT1A9
as the major enzyme involved. Other isoforms showed little or
no turnover. Given the strong implications of this regarding the
effective clinical doses as well as the potential for drug−drug
interaction, we decided to explore compounds with signifi-
cantly reduced glucuronidation in humans. For this we
incorporated in vitro measurement of glucuronide formation
In Figure 3 the IDO1 activity and in vitro glucuronidation
data of epacadostat (2) compared to the corresponding
thioether and ether analogs are shown. An almost 3-fold
increased potency of thioether 8 compared to 2 in the cellular
assay made this new chemotype interesting per se to us.
However, UGT1A9 glucuronidation of both the ether analog 7
(t1/2 = 2.9 min) and thioether analog 8 (t1/2 = 11 min) was
increased compared to 2 (t1/2 = 23 min). From this benchmark
result we started a broad search to optimize the compounds
aiming at a reduced glucuronidation. Due to the higher
potency in the IDO1 cell-based assay, we focused on
thioethers.
Expecting a higher level of SAR tolerance, we started with
variations for B-pocket substituents. Based on docking results
of some of our early analogs into a published homology model
of the protein UGT1A9,26 we envisioned that affinity for this
enzyme might be reduced by modifying the sulfamide moiety.
This functional group seemed to undergo a series of polar
interactions with polar amino acid side chains in the binding
pocket of the active site of the enzyme (see Supporting
Information). In addition to that we assumed that further
reducing the hydrophobicity of compounds could reduce
binding and turnover by UGT1A9 since the published
3DQSAR model proposes that hydrophobic groups in
proximity to the site of glucuronidation enhance glucuronida-
tion.26
D
https://dx.doi.org/10.1021/acsmedchemlett.9b00572
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
pubs.acs.org/acsmedchemlett
Letter
Table 3. Compounds Selected for Further Profiling
17
24
IDO1(enzyme) IC50 [nM]
IDO1(SKOV3) IC50 [nM]
93
21
4
9
21
10
9
6
TDO (HEK) IC50 [nM]
>10000
>10000
IDO1 human whole blood IC50 [nM]
HLM/MLM/RLM t1/2 [min]
UGT1A9 assay t1/2 [min]
320 40
>180/165/−
>180
16
220
>180/75.7/33.7
54
22
7
8
% glucuronide (human hepatocytes at 120 min)
a
PPB h/m/r [% bound]
-
95.4/97.4/95.7
CYP inh. (3A4, 2D6, 2C9, 2C19) IC50 [μM]
Safety panel (44 targets, >50% inhib. at 10 μM)
mouse PK (10 mg/kg po) :
all >10
-
all >10
2
b
Cmax [μM]
AUC0−24h [μM·h]
t1/2 [h]
3.3
30
3.6
5.0
11.9
2.6
a
b
= human/mouse/rat. Blood levels obtained from capillary microsampling.
Our screening path started with a biochemical IDO1 assay
high potency in both IDO1 assays together with further
improved in vitro glucuronidation compared to 15 (18.0%
glucuronide in hepatocytes at 120 min). The mouse liver
microsomal stability of 16 was only moderate though.
Replacing the sulfonamide of 15 by a sulfamide group led to
analog 17 with good IDO1 potency, high microsomal stability,
and improved glucuronidation data compared to 15 (UGT1A9
followed by a cell-based assay using SKOV3 cells (see
Supporting Information). Compounds showing cell-based
potency below an IC50 of 50 nM were tested for stability in
the UGT1A9 enzyme assay and in microsomes followed by
glucuronide formation measurement in human hepatocytes. In
Table 1 a selection of the B-pocket SAR plus in vitro metabolic
stability data is shown. The sulfonamide 9 showed a high
potency in the SKOV3 IDO1 assay (9.3 nM) but the
biochemical potency was lower compared to 7. Next to a
low mouse liver microsomal stability (t1/2 = 17.4 min), the
stability in the UGT1A9 assay (t1/2 = 21.4 min) was not
promising. Analog 10 demonstrated increased stability in the
UGT1A9 assay (t1/2 = 51.5 min) as well as in microsomes but
did not sufficiently improve biochemical IDO1 potency. It was
productive to find that B-pocket structures composed of a
carboxamide linked to polar substituted alkylene groups, like in
10, were generally very potent. Sulfonamide 11 showed high
potency in both IDO1 assays together with increased stability
in both assays for glucuronidation (UGT1A9 t1/2 = 45.5 min,
26.3% glucuronide in hepatocytes after 120 min). A methyl
branch at the ethylene linker led to compounds nearly equally
potent to 11 (both enantiomers 12 and 13) and markedly
increased the UGT1A9 stability. Enantiomer 13 showed
decreased glucuronidation in human hepatocytes (10.9%
glucuronide after 120 min), but mouse microsomal stability
(t1/2 = 23.9 min) was low. The sulfoximine analog 14 was
highly potent and stable in microsomes together with
significantly reduced glucuronidation (UGT1A9 t1/2 = 91.8
min, 8.8% glucuronide in hepatocytes after 120 min). Since we
obtained comparably optimized compounds with other, achiral
moieties (see below) we did not further evaluate this specific
racemic example or its enantiomers. In compound 15 the
bromo substituent of 11 was replaced by a difluoromethyl
group. The analog was equally potent to 11 but more stable in
the UGT1A9 assay (t1/2 > 180 min) and hepatocytes (24.2%
glucuronide after 120 min). Therefore, the 3-difluoromethyl-4-
fluorophenyl substituent was used in the following side chain
variations as A-pocket substituent. Sulfone-analog 16 showed
t
1/2 > 180 min, 15.5% glucuronide in hepatocytes at 120 min).
Based on these data we decided to further profile 17 (see
below). Overall, we found that mostly improved structures
combining high potency and the desired metabolic stability
had sulfonyl derived groups at the end of the B-pocket side
chain. Attempts to broaden this scope are exemplified by
compounds 18, 19, 20, and 21. All four analogues showed high
IDO1 potency in combination with excellent UGT1A9
stability (18: t1/2 > 180 min, 19: t1/2 = 104 min, 20 and 21:
t1/2 > 180 min). However, based on the low mouse liver
microsomal stability these analogues were not further
evaluated.
Encouraged by the improvement of glucuronidation stability
from bromo analog 11 to difluoro analog 15, we systematically
explored the effect of substituents of the phenyl group binding
into the A-pocket of IDO1, keeping the 2-sulfamoylacetamide
moiety constant. Variation of A-pocket groups indeed had a
pronounced effect as shown in Table 2. Replacing the bromo
substituent of 11 with a trifluoromethyl group led to 22 which
was equally potent compared to 11 and showed increased in
vitro stability toward glucuronidation (UGT1A9 t1/2 = 86.7
min, 17.5% glucuronide in hepatocytes at 120 min). To test
the impact of further increased polarity, nitrile analog 23 was
made. While this modification caused a potency drop (SKOV3
IDO1 IC50 = 39 nM), probably due to limited acceptance for
polar residues in the lipophilic A-pocket, the glucuronidation
was further decreased (UGT1A9 t1/2 > 180 min, 14.3%
glucuronidation in hepatocytes at 120 min). The effect of
removing the para-fluoro substituent was studied with analogs
24 and 25. Regarding glucuronidation both compounds were
comparable to 11; however, there was a further strong potency
increase seen in both IDO1 activity assays with these
E
https://dx.doi.org/10.1021/acsmedchemlett.9b00572
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
pubs.acs.org/acsmedchemlett
Letter
monosubstituted compounds especially in the biochemical
assay (IC50 = 21 nM for 24 and IC50 = 30 nM for 25).
Following up on this result, the monosubstitution was tested
with a trifluoromethyl, difluoromethyl, and nitrile group,
respectively (analogues 26, 27, and 28). While all these
analogues showed an increased stability toward UGT1A9 (t1/2
> 180 min) they were less potent than 24 and 25, especially in
the biochemical assay. With regard to potency and stability
toward glucuronidation in hepatocytes, 24 was an improved
analog of 11 and 15. Therefore it was selected for further
characterization in addition to compound 17 (see Table 3
below).
Table 3 shows key data for compounds 17 and 24. In the
biochemical assay compound 24 was more than 4-fold more
potent (17: IC50 = 93 nM, 24: IC50 = 21 nM) while the
difference in cell-based activity was less pronounced (17: IC50
= 21 nM, 24: IC50 = 10 nM). Both analogues were selective
over TDO (IC50 > 10 μM). The human whole blood potencies
were similar (17: IC50 = 320 nM, 24: IC50 = 220 nM) and
comparable with epacadostat (2) (IC50 = 300 nM) (for details
see Supporting Information). Microsomal stability was very
high for both analogues across species. Based on both
UGT1A9 assay and in human hepatocytes, compound 17
The pharmacokinetics of compound 24 were investigated in
more detail with a special emphasis on human glucuronidation
(Figure 4). For this the PK in a humanized mouse liver model
was studied using PXB mice (PhoenixBio). These are chimeric
mice with a humanized liver that is highly repopulated by
human hepatocytes so that the major human drug-metaboliz-
ing enzymes and transporters are present in the liver.27,28 The
use of these mice as animal models for predicting human drug
metabolism and pharmacokinetics has been reported.29 We
used this model for compound 24 benchmarked against
epacadostat (2). In a single dose PK (10 mg/kg, po),
compound 24 displayed a more than 3-fold higher Cmax and
more than 2-fold higher AUC compared to epacadostat (2).
The ratio of AUC glucuronide/AUC parent is reduced from
3.6:1 for epacadostat (2) to 0.4:1 for 24.
In order to evaluate the antitumor effects of the IDO1
inhibitor 24, we studied its efficacy in a syngeneic CT26 model
either alone or in combination with a CTLA-4 mAb. Tumor
size and animal survival were regularly monitored. In the single
treatment study mice were dosed orally with the IDO1
inhibitor alone at 100 mg/kg BID. The treament group
showed reduced tumor growth compared to the vehicle group.
In a separate study mice were dosed orally with the IDO1
inhibitor at 100 mg/kg BID in combination with the antibody
(100 μg ip, days 3, 6, and 9) over a 15 days period and survival
was monitored for an additional 45 days. CTLA-4 mAb
therapy alone led to a marked delay in tumor growth but
improved survival only slightly (10%). Cotreatment with
CTLA-4 mAb and IDO1 inhibitor 24 resulted in further
reduced tumor growth and led to a significant improvement of
survival indicating a synergy of the combination. 40% of mice
were still alive at day 70 after tumor inoculation (Figure 5).
To confirm on-target effects, we tested compound 24 in a
PK/PD model at 100 mg/kg po which demonstrated
kynurenine reduction (for details see Supporting Information).
In conclusion, we have discovered new IDO1 inhibitors
which were designed from the hydroxyamidine pharmacophore
of INCB14943 (4) using a thioether linker group attached to
the oxadiazole together with selected A- and B-pocket residues.
A similar approach incorporating thioether next to amino
linkers has recently been published.30 The publication reports
the SAR of an intereresting amino-cyclobutarene motif as A-
pocket moiety leading to IDO1 inhibitors of high potency and
selectivity. Our lead compound demonstrated high IDO1
was more stable toward glucuronidation (17: UGT1A9 t1/2
>
180 min, 16% glucuronide in hepatocytes after 120 min, 24:
UGT1A9 t1/2 = 54 min, 22% glucuronide in hepatocytes after
120 min). Compound 24 had moderately high plasma protein
binding across species, and both compounds were clean in a
small CYP inhibition panel. 24 displayed a favorable in vitro
safety panel (see Supporting Information). Oral mouse PK
data for 17 and 24 were suitable for further evaluation in mice.
A single dose PK of 24 in rats showed high exposure (AUC =
26.0 μM·h) and a bioavailability of 28% at 5 mg/kg po (Table
4). Based on the higher biochemical and cell-based IDO1
potency, compound 24 was selected for further in vivo
characterization.
Table 4. Single Dose Rat Pharmacokinetic Data of 24 (1
a
mg/kg iv; 5 mg/kg po)
C
AUC0−24h
b
Cl
Vd
t1/2
F
[%]
maxb
Cpd [μM]
c
c
c
[μM·h]
[mL/min/kg] [L/kg]
[h]
24 5.6
26.0
3.65 0.33
1.5
28
a
b
Blood levels obtained from capillary microsampling. = calculated
c
from po arm. = calculated from iv arm.
Figure 4. Compound 24 has a significantly improved PK profile over epacadostat (2) in mice with a humanized liver (PXB mice) with regard to
hydroxyamidine glucuronidation.
F
https://dx.doi.org/10.1021/acsmedchemlett.9b00572
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
pubs.acs.org/acsmedchemlett
Letter
Figure 5. Antitumor efficacy of compound 24 in the CT26 syngeneic model. CT26 cells were implanted subcutaneously into Balb/c mice which
were then randomized and treated with either hydroxyamidine IDO1 inhibitorcompound 24 alone or CTLA-4 antibody alone or in combination
with hydroxyamidine IDO1 inhibitorcompound 24. (A) Tumor growth curves of the vehicle versus single agentcompound 24. Single agent
compound 24 alone showed a reduction in tumor volume p = 0.033 (Unpaired t test). (B) Tumor growth curves of compound 24 in combination
with anti-CTLA-4. (C) Kaplan−Meier survival plot representing mice survival during the 15-day treatment period and post-treatment until day 70.
potency in biochemical and cell-based assays as well as in a
human whole blood assay. The in vitro potency translated into
efficacy in an animal model of cancer in combination with a
checkpoint inhibitor. Compound 24 shows in vitro/in vivo data
that are predictive for a markedly reduced glucuronidation of
the hydroxyamidine in humans when compared to epacadostat.
This is an important improvement taking into account the
possible advantages of a reduced risk of drug−drug
interactions and interindividual variations due to polymor-
phism.26,31 In addition to that, the absence of excessive
metabolism would significantly lower the effective doses.
compounds into UGT1A9, safety panel data, exper-
imental details of CT26 model, PK/PD model of
compound 24, experimental details for PK studies in a
humanized mouse liver model (PDF)
AUTHOR INFORMATION
■
Corresponding Author
Christoph Steeneck − Phenex Pharmaceuticals AG 69123
Heidelberg, Germany; orcid.org/0000-0002-7746-735X;
Email: christoph.steeneck@phenex-pharma.com
ASSOCIATED CONTENT
Authors
■
Olaf Kinzel − Phenex Pharmaceuticals AG 69123 Heidelberg,
Germany
Simon Anderhub − Phenex Pharmaceuticals AG 69123
Heidelberg, Germany
Martin Hornberger − Phenex Pharmaceuticals AG 69123
Heidelberg, Germany
Sheena Pinto − Phenex Pharmaceuticals AG 69123 Heidelberg,
Germany
sı
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acsmedchemlett.9b00572.
Chemistry and synthetic schemes, synthetic procedures
and analytical data for selected compounds, experimen-
tal details for IDO1 activity assays and in vitro stability
assays for glucuronidation, proposed binding mode of
G
https://dx.doi.org/10.1021/acsmedchemlett.9b00572
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
pubs.acs.org/acsmedchemlett
Letter
suppress systemic tryptophan catabolism and the growth of IDO-
expressing tumors. Mol. Cancer Ther. 2010, 9, 489−498.
(10) Mezrich, J. D.; Fechner, J. H.; Zhang, X.; Johnson, B. P.;
Burlingham, W. J.; Bradfield, C. A. An interaction between kynurenine
and the aryl hydrocarbon receptor can generate regulatory T cells. J.
Immunol. 2010, 185, 3190−3198.
(11) Opitz, C. A.; Litzenburger, U. M.; Sahm, F.; Ott, M.; Tritschler,
I.; Trump, S.; Schumacher, T.; Jestaedt, L.; Schrenk, D.; Weller, M.;
Jugold, M.; Guillemin, G. J.; Miller, C. L.; Lutz, C.; Radlwimmer, B.;
Lehmann, I.; von Deimling, A.; Wick, W.; Platten, M. An endogenous
tumourpromoting ligand of the human aryl hydrocarbon receptor.
Nature 2011, 478, 197−203.
(12) Prendergast, G. C.; Malachowski, W. P.; DuHadaway, J. B.;
Muller, A. J. Discovery of IDO1 inhibitors: from bench to bedside.
Cancer Res. 2017, 77, 6795−6811.
(13) Kumar, S.; Waldo, J. P.; Jaipuri, F. A.; Marcinowicz, A.; Van
Allen, C.; Adams, J.; Kesharwani, T.; Zhang, X.; Metz, R.; Oh, A. J.;
Harris, S. F.; Mautino, M. R. Discovery of clinical candidate (1R,4r)-
4-((R)-2-((S)-6-fluoro-5H-imidazo[5,1-a]isoindol-5-yl)-1-
hydroxyethyl)cyclohexan-1-ol (navoximod), a potent and selective
inhibitor of indoleamine 2,3-dioxygenase 1. J. Med. Chem. 2019, 62,
6705−6733.
(14) Yue, E. W.; Sparks, R.; Polam, P.; Modi, D.; Douty, B.;
Wayland, B.; Glass, B.; Takvorian, A.; Glenn, J.; Zhu, W.; Bower, M.;
Liu, X.; Leffet, L.; Wang, Q.; Bowman, K. J.; Hansbury, M. J.; Wei,
M.; Li, Y.; Wynn, R.; Burn, T. C.; Koblish, H. K.; Fridman, J. S.;
Emm, T.; Scherle, P. A.; Metcalf, B.; Combs, A. P. INCB24360
(epacadostat), a highly potent and selective indoleamine-2,3-
dioxygenase 1 (IDO1) inhibitor for immuno-oncology. ACS Med.
Chem. Lett. 2017, 8, 486−491.
(15) Nelp, M. T.; Kates, P. A.; Hunt, J. T.; Newitt, J. A.; Balog, A.;
Maley, D.; Zhu, X.; Abell, L.; Allentoff, A.; Borzilleri, R.; Lewis, H. A.;
Lin, Z.; Seitz, S. P.; Yan, C.; Groves, J. T. Immune-modulating
enzyme indoleamine 2,3-dioxygenase is effectively inhibited by
targeting its apo-form. Proc. Natl. Acad. Sci. U. S. A. 2018, 115,
3249−3254.
(16) Muller, A. J.; Manfredi, M. G.; Zakharia, Y.; Prendergast, G. C.
Inhibiting IDO pathways to treat cancer: lessons from the ECHO-301
trial and beyond. Semin. Immunopathol. 2019, 41, 41−48.
(17) Liu, M.; Wang, X.; Wang, L.; Ma, X.; Gong, Z.; Zhang, S.; Li, Y.
Targeting the IDO1 pathway in cancer: from bench to bedside. J.
Hematol. Oncol. 2018, 11, 100.
Barbara Morschhaeuser − Phenex Pharmaceuticals AG 69123
Heidelberg, Germany
Floriane Braun − Phenex Pharmaceuticals AG 69123
Heidelberg, Germany
Gerald Kleymann − Phenex Pharmaceuticals AG 69123
Heidelberg, Germany
Thomas Hoffmann − Phenex Pharmaceuticals AG 69123
Heidelberg, Germany
Complete contact information is available at:
https://pubs.acs.org/10.1021/acsmedchemlett.9b00572
Author Contributions
All authors have given approval to the final version of the
manuscript.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We would like to thank Christina Sonnek and Dr. Michael
Albers for support and contributions to this work.
■
ABBREVIATIONS
■
ADME, absorption, distribution, metabolism, and excretion;
CYP, cytochrome P450; HLM, human liver microsomes; Kyn,
kynurenine; MLM, mouse liver microsomes; PK, pharmaco-
kinetics; PPB, plasma protein binding; RLM, rat liver
microsomes; SAR, structure−activity relationship; Trp,
tryptophan
REFERENCES
■
(1) Alsaab, H. O.; Sau, S.; Alzhrani, R.; Tatiparti, K.; Bhise, K.;
Kashaw, S. K.; Iyer, A. K. PD-1 and PD-L1 checkpoint signaling
inhibition for cancer immunotherapy: mechanism, combinations, and
clinical outcome. Front. Pharmacol. 2017, 8, 561.
(2) Zhao, X.; Subramanian, S. Intrinisc resistance of solid tumors to
immune checkpoint blockade therapy. Cancer Res. 2017, 77, 817−
822.
(3) Chen, T. W.; Razak, A. R.; Bedard, P. L.; Siu, L. L.; Hansen, A.
R. A systematic review of immune-related adverse event reporting in
clinical trials of immune checkpoint inhibitors. Ann. Oncol. 2015, 26,
1824−1829.
(18) NCT03661320: A phase 3, randomized study of neoadjuvant
chemotherapy alone versus neoadjuvant chemotherapy plus nivolu-
mab or nivolumab and BMS-986205, followed by continued post-
surgery therapy with nivolumab or nivolumab and BMS-986205 in
participants with muscle-invasive bladder cancer, ClinicalTrials.gov,
September 7, 2018.
(19) NCT04069026: A first-in-humans dose finding study for an aryl
hydrocarbon receptor inhibitor (AhRi) in patients with advanced
cancer, ClinicalTrials.gov, August 26, 2019.
̈
(4) Platten, M.; Nollen, E. A. A.; Rohrig, U. F.; Fallarino, F.; Opitz,
C. A. Tryptophan metabolism as a common therapeutic target in
cancer, neurodegeneration and beyond. Nat. Rev. Drug Discovery
2019, 18, 379−401.
(5) Munn, D. H.; Zhou, M.; Attwood, J. T.; Bodarev, L.; Conway, S.
J.; Marshall, B.; Brown, C.; Mellor, A. L. Prevention of allogeneic fetal
rejection by tryptophan catabolism. Science 1998, 281, 1191−1193.
(6) Munn, D. H.; Mellor, A. L. IDO and tolerance in tumors. Trends
Mol. Med. 2004, 10, 15−18.
(20) Triplett, T. A.; Garrison, K. C.; Marshall, N.; Donkor, M.;
Blazeck, J.; Lamb, C.; Qerqez, A.; Dekker, J. D.; Tano, Y.; Lu, W.;
Karamitros, C. S.; Ford, K.; Tan, B.; Zhang, M.; McGovern, K.;
Coma, S.; Kumada, Y.; Yamany, M. S.; Sentandreu, E.; Fromm, G.;
Tiziani, S.; Schreiber, T. H.; Manfredi, M.; Ehrlich, L. I. R.; Stone, E.;
Georgiou, G. Reversal of IDO-mediated cancer immune suppression
by systemic kynurenine depletion with a therapeutic enzyme. Nat.
Biotechnol. 2018, 36, 758−764.
́
(7) Uyttenhove, C.; Pilotte, L.; Theate, I.; Stroobant, V.; Colau, D.;
Parmentier, N.; Boon, T.; Van den Eynde, B. Evidence for a tumoral
immune resistance mechanism based on tryptophan degradation by
indoleamine 2,3- dioxygenase. Nat. Med. 2003, 9, 1269−1274.
(8) Yue, E. W.; Douty, B.; Wayland, B.; Bower, M.; Liu, X.; Leffet,
L.; Wang, Q.; Bowman, K. J.; Hansbury, M. J.; Liu, C.; Wei, M.; Li, Y.;
Wynn, R.; Burn, T. C.; Koblish, H. K.; Fridman, J. S.; Metcalf, B.;
Scherle, P. A.; Combs, A. P. Discovery of potent competitive
inhibitors of indoleamine 2,3-dioxygenase with in vivo pharmacody-
namic activity and efficacy in a mouse melanoma model. J. Med. Chem.
2009, 52, 7364−7367.
(21) Hoffmann, T. Discovery of small molecule arylhydrocarbon
receptor(AhR) antagonists for cancer immunotherapy. Discovery on
Target Boston, Sept. 17, 2019.
(22) Pinto, S.; Steeneck, C.; Albers, M.; Anderhub, S.; Birkel, M.;
̈
Buselic-Wolfel, L.; Eisenhardt, G.; Kremoser, K.; Hoffmann, T.;
Deuschle, U. Targeting the IDO1-kynurenine-AhR pathway for
cancer immunotherapy [abstract]. In Proceedings of the American
Association for Cancer Research Annual Meeting 2019; 2019 Mar 29−
Apr 3; Atlanta, GA, Philadelphia (PA): AACR; Cancer Res 2019; 79
(13 Suppl.): Abstract nr 1210.
(9) Koblish, H. K.; Hansbury, M. J.; Bowman, K. J.; Yang, G.;
Neilan, C. L.; Haley, P. J.; Burn, T. C.; Waeltz, P.; Sparks, R. B.; Yue,
E. W.; Combs, A. P.; Scherle, P. A.; Vaddi, K.; Fridman, J. S.
Hydroxyamidine inhibitors of indoleamine-2,3-dioxygenase potently
H
https://dx.doi.org/10.1021/acsmedchemlett.9b00572
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
pubs.acs.org/acsmedchemlett
Letter
́
(23) Li, H.; Bullock, K.; Gurjao, C.; Braun, D.; Shukla, S. A.; Bosse,
D.; Lalani, A. A.; Gopal, S.; Jin, C.; Horak, C.; Wind-Rotolo, M.;
Signoretti, S.; McDermott, D. F.; Freeman, G. J.; Van Allen, E. M.;
Schreiber, S. L.; Hodi, F. S.; Sellers, W. R.; Garraway, L. A.; Clish, C.
B.; Choueiri, T. K.; Gioannakis, M. Metabolomic adaptations and
correlates of survival to immune checkpoint blockade. Nat. Commun.
2019, 10, 4346.
̈
(24) Rohrig, U. F.; Majjigapu, S. R.; Vogel, P.; Zoete, V.; Michielin,
O. Challenges in the Discovery of Indoleamine 2,3-Dioxygenase 1
(IDO1). J. Med. Chem. 2015, 58, 9421−9437.
(25) Boer, J.; Young-Sciame, R.; Lee, F.; Bowman, K. J.; Yang, X.;
Shi, J. G.; Nedza, F. M.; Frietze, W.; Galya, L.; Combs, A. P.;
Yeleswaram, S.; Boer, S. Roles of UGT, P450, and gut microbiota in
the metabolism of epacadostat in humans. Drug Metab. Dispos. 2016,
44, 1668−1674.
(26) Wu, B.; Wang, X.; Zhang, S.; Hu, M. Accurate prediction of
glucuronidation of structurally diverse phenolics by human UGT1A9
using combined experimental and in silico approaches. Pharm. Res.
2012, 29, 1544−1561.
(27) Katoh, M.; Matsui, T.; Nakajima, M.; Tateno, C.; Kataoka, M.;
Soeno, Y.; Horie, T.; Iwasaki, K.; Yoshizato, K.; Yokoi, T. Expression
of human cytochromes P450 in chimeric mice with humanized liver.
Drug Metab. Dispos. 2004, 32, 1402−1410.
(28) Katoh, M.; Matsui, T.; Okumura, H.; Nakajima, M.; Nishimura,
M.; Naito, S.; Tateno, C.; Yoshizato, K.; Yokoi, T. Expression of
human phase II enzymes in chimeric mice with humanized liver. Drug
Metab. Dispos. 2005, 33, 1333−1340.
(29) Sanoh, S.; Ohta, S. Chimeric mice transplanted with human
hepatocytes as a model for prediction of human drug metabolism and
pharmacokinetics. Biopharm. Drug Dispos. 2014, 35, 71−86.
(30) Zhang, H.; Liu, K.; Pu, Q.; Achab, A.; Ardolino, M. J.; Cheng,
M.; Deng, Y.; Doty, A. C.; Ferguson, H.; Fradera, X.; Knemeyer, I.;
Kurukulasuriya, R.; Lam, Y. H.; Lesburg, C. A.; Martinot, T. A.;
McGowan, M. A.; Miller, J. R.; Otte, K.; Biju, P. J.; Sciammetta, N.;
Solban, N.; Yu, W.; Zhou, H.; Wang, X.; Bennett, D. J.; Han, Y.
Discovery of amino-cyclobutarene-derived indoleamine-2,3-dioxyge-
nase 1 (IDO1) inhibitors for cancer immunotherapy. ACS Med. Chem.
Lett. 2019, 10, 1530−1536.
(31) Mano, E. C. C.; Scott, A. L.; Honorio, K. M. UDP-
glucuronosyltransferases: structure, function and drug design studies.
Curr. Med. Chem. 2018, 25, 3247−3255.
NOTE ADDED AFTER ASAP PUBLICATION
■
This letter published ASAP on January 27, 2020. Changes have
been made to Figure 4 and corresponding text. The corrected
version was reposted the same day.
I
https://dx.doi.org/10.1021/acsmedchemlett.9b00572
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX