O-alkylhydroxylamines as rationally-designed mechanism-based inhibitors of indoleamine 2,3-dioxygenase-1

Article abstract of DOI:10.1016/j.ejmech.2015.12.028

Indoleamine 2,3-dioxygenase-1 (IDO1) is a promising therapeutic target for the treatment of cancer, chronic viral infections, and other diseases characterized by pathological immune suppression. Recently important advances have been made in understanding IDO1's catalytic mechanism. Although much remains to be discovered, there is strong evidence that the mechanism proceeds through a heme-iron bound alkylperoxy transition or intermediate state. Accordingly, we explored stable structural mimics of the alkylperoxy species and provide evidence that such structures do mimic the alkylperoxy transition or intermediate state. We discovered that O-benzylhydroxylamine, a commercially available compound, is a potent sub-micromolar inhibitor of IDO1. Structure-activity studies of over forty derivatives of O-benzylhydroxylamine led to further improvement in inhibitor potency, particularly with the addition of halogen atoms to the meta position of the aromatic ring. The most potent derivatives and the lead, O-benzylhydroxylamine, have high ligand efficiency values, which are considered an important criterion for successful drug development. Notably, two of the most potent compounds demonstrated nanomolar-level cell-based potency and limited toxicity. The combination of the simplicity of the structures of these compounds and their excellent cellular activity makes them quite attractive for biological exploration of IDO1 function and antitumor therapeutic applications.

Full text of DOI:10.1016/j.ejmech.2015.12.028

European Journal of Medicinal Chemistry 108 (2016) 564e576
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Research paper
O-alkylhydroxylamines as rationally-designed mechanism-based
inhibitors of indoleamine 2,3-dioxygenase-1
,
William P. Malachowski a *, Maria Winters a, James B. DuHadaway b,
Ariel Lewis-Ballester e, Shorouk Badir a, Jenny Wai a, Maisha Rahman a, Eesha Sheikh a,
,
, c, d
Judith M. LaLonde a ***, Syun-Ru Yeh e, George C. Prendergast b
,
Alexander J. Muller b
, d, **
a Department of Chemistry, Bryn Mawr College, 101 N. Merion Ave., Bryn Mawr, PA 19010-2899, USA
b Lankenau Institute for Medical Research, Wynnewood, PA 19096, USA
c Department of Pathology, Anatomy & Cell Biology, Thomas Jefferson University, Philadelphia, PA 19104, USA
d Kimmel Cancer Center, Thomas Jefferson University, Philadelphia, PA 19104, USA
e Department of Physiology and Biophysics, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY 10461, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Indoleamine 2,3-dioxygenase-1 (IDO1) is a promising therapeutic target for the treatment of cancer,
chronic viral infections, and other diseases characterized by pathological immune suppression. Recently
important advances have been made in understanding IDO1's catalytic mechanism. Although much
remains to be discovered, there is strong evidence that the mechanism proceeds through a heme-iron
bound alkylperoxy transition or intermediate state. Accordingly, we explored stable structural mimics
of the alkylperoxy species and provide evidence that such structures do mimic the alkylperoxy transition
or intermediate state. We discovered that O-benzylhydroxylamine, a commercially available compound,
is a potent sub-micromolar inhibitor of IDO1. Structureeactivity studies of over forty derivatives of O-
benzylhydroxylamine led to further improvement in inhibitor potency, particularly with the addition of
halogen atoms to the meta position of the aromatic ring. The most potent derivatives and the lead, O-
benzylhydroxylamine, have high ligand efficiency values, which are considered an important criterion for
successful drug development. Notably, two of the most potent compounds demonstrated nanomolar-
level cell-based potency and limited toxicity. The combination of the simplicity of the structures of
these compounds and their excellent cellular activity makes them quite attractive for biological explo-
ration of IDO1 function and antitumor therapeutic applications.
Received 31 July 2015
Received in revised form
12 November 2015
Accepted 14 December 2015
Available online 17 December 2015
Keywords:
IDO1 inhibition
O-alkylhydroxylamines
Antitumor therapy
Rational drug design
© 2015 Elsevier Masson SAS. All rights reserved.
1. Introduction
engender a tolerogenic microenvironment that actively suppresses
the ability of the immune system to mount an effective response
Immune escape by tumors is a fundamental aspect of disease
progression resulting from immunoediting of tumors as they
interact with the host immune system [1,2]. Key to this process
from a therapeutic standpoint is that tumors are selected to
[3]. Ongoing progress in understanding the cellular and molecular
mechanisms that govern the pathological state of tumor immune
tolerance has revealed several protein targets which provide the
potential for therapeutic intervention [4]. One central player is the
immunomodulatory enzyme indoleamine 2,3-dioxygenase, IDO1,
formerly known as IDO before the discovery of a second isoform
[5,6]. IDO1 can contribute to immune escape when expressed
directly in tumor cells or when expressed in immunosuppressive
antigen presenting cells such as tolerogenic dendritic cells or tumor
associated macrophages [7,8]. For both cases, experimental results
suggest that IDO1 inhibition may restore an effective antitumor
immune response and thus provide a method to treat malignant
Abbreviations: IDO1, indoleamine 2,3-dioxygenase isoform-1; Trp, tryptophan;
UVevis, ultravioletevisible; CYP3A4, cytochrome P-450 isoform 3A4.
* Corresponding author.
** Corresponding author. Lankenau Institute for Medical Research, Wynnewood,
PA 19010 USA.
*** Corresponding author.
E-mail addresses: wmalacho@brynmawr.edu (W.P. Malachowski), mullera@
mlhs.org (A.J. Muller).
http://dx.doi.org/10.1016/j.ejmech.2015.12.028
0223-5234/© 2015 Elsevier Masson SAS. All rights reserved.

W.P. Malachowski et al. / European Journal of Medicinal Chemistry 108 (2016) 564e576
565
Table 1
diseases in combination with chemotherapeutic agents and/or
IC50 of commercially available Ar-X-Y or Ar-C-X-Y Compounds.a
immunotherapy-based strategies [9]. In fact, there are currently
three drugs in clinical trials testing IDO1 inhibition as a strategy for
the treatment of cancer [10]. Clearly there is an interest and a need
for further development of potent IDO1 inhibitors to adequately
address this therapeutic opportunity [11,12].
Compound structure
IC50 (mM)
PheNHeNH2
0.23 0.25
0.81 0.081
6.0
9.2
16
PheCH2eO-NH2
PheCH2eNH-OH
PheNHeOH
IDO1 is an extrahepatic, tryptophan (Trp) metabolizing enzyme
[13e15], which catalyzes the initial and rate-limiting step along the
kynurenine pathway. The oxidative metabolism of Trp by IDO1
involves the coordination of molecular oxygen to a ferrous heme
iron and its subsequent addition across the C-2/C-3 bond of the
indole ring. Two alkylperoxy transition or intermediate states
resulting from dioxygen insertion to C-2 or C-3 carbon of the indole
ring (Fig. 1, compound 1 and 2) have been proposed [13,16e18].
Here we sought to rationally design a new family of IDO1 inhibitors
by mimicking the alkylperoxy species as illustrated in Fig. 1, in
which one oxygen atom of the peroxo moiety is substituted with an
electronegative atom, N (compound 3). To simplify compound 3,
the alkylperoxy moiety is further reduced to Ar-C-X-Y, where Ar is
an aryl ring and C is a linker containing a carbon moiety (such as
methylene or carbonyl). In related work, two previously reported
IDO1 inhibitors in the scientific literature demonstrate a similar
theme in their inhibitor design: The hydroxyamidines reported in
2009 by Incyte [19] and the recent report of alkylhydrazine deriv-
ative inhibitory activity by Ching et al. [20]. Neither of these reports
discussed the potential mimicry of the proposed alkylperoxy in-
termediate or related rational design idea. Contemporaneous with
our studies, O-alkylhydroxylamine IDO1 inhibitors were reported
in the patent literature by NewLink Genetics [21]. With the rise of
chemical library screening for drug lead development, mechanism-
based rational drug design ideas are less frequently used, but the
current work illustrates there is still merit in their use as they
proved a simple and efficient way to identify a new IDO1 inhibitor
lead.
Ph-C(¼O)eNHeOH
PheCH2eNHeNH2
71
a
IC50 values are based on single point inhibition curves run
once for each compound, but repeated 2ꢀ for the most potent
compounds (initial IC50 values below 1.0
reported as the averages SD.
mM). These results are
benzylhydoxylamine, the second best inhibitor with an IC50 of
0.90 M, as we were more interested in developing a reversible
m
inhibitor of IDO1. Nonetheless, all the compounds in Table 1
represent a new class of IDO1 inhibitors: alkylperoxy mimics. The
therapeutic potential of IDO1 inhibition, as well as the commercial
availability and sub-micromolar level potency of these compounds,
made this discovery an important opportunity. Our efforts at
developing peroxy structural mimics, in particular, optimizing the
O-alkylhydroxylamine structural class, are described below.
2.2. Chemistry
The O-alkylhydroxylamine derivatives were synthesized from
the analogous alcohol through a one pot process involving the
Mitsunobu reaction with N-hydroxyphthalimide [27] and subse-
quent deprotection of the phthalimide group with hydrazine
(Scheme 1) [28]. The O-alkyl hydroxylamines were purified and
isolated as their hydrochloride salts. Alcohols that were not
commercially available were synthesized via reduction of the
appropriate aldehyde with NaBH4 (Scheme 2).
3. Results and discussion
2. Results and discussion
3.1. IDO1 inhibition by O-alkylhydroxylamines
2.1. Discovery of new structural class of IDO1 inhibitors
In an effort to further optimize the O-benzylhydroxylamine
lead, we explored two general modifications of the O-benzylhy-
droxylamine structure: (1) alteration of the carbon linker between
the aryl ring and the hydroxylamine group; (2) substitution of the
aryl ring (Fig. 2).
To explore this rational drug design idea, we screened
commercially available compounds with a general structure of Ar-
X-Y or Ar-C-X-Y (Table 1). Gratifyingly, a screen of a few dozen
commercial compounds yielded several micromolar or sub-
micromolar potency molecules. Similar to the recent report from
the Ching group [20], phenylhydrazine was found to be a highly
potent inhibitor of IDO1. This is not surprising since phenyl-
hydrazine has been found to interact with other heme-containing
proteins, such as hemoglobin [22], cytochrome P450 [23], cata-
lase [24] and myoglobin [25]. In fact, phenylhydrazine can be
oxidized by heme iron to generate phenyl radicals that can subse-
quently react to covalently modify the heme prosthetic group [26].
Although we did not see any signs of irreversible inhibition by
phenylhydrazine, we chose to focus our efforts on O-
3.1.1. Alternation of the carbon linker between the aryl ring and the
peroxy mimic moiety
Linker length and flexibility was explored with the compounds
shown in Table 2. It quickly became clear that any modifications
made to the carbon linker in the lead compound resulted in a
dramatic decrease in inhibitor potency. Attempts at rigidifying the
hydroxylamine moiety in the benzylic position (compounds 3, 4
and 6) were unsuccessful as well.
Fig. 1. Proposed alkylperoxy transition or intermediate state of IDO1 (1 and 2) and their mimic (3).

566
W.P. Malachowski et al. / European Journal of Medicinal Chemistry 108 (2016) 564e576
Scheme 1. Mitsunobu reaction to convert alcohols to O-alkyl hydroxylamines.
imidazole study [29]. Addition of iodine, bromine, chlorine or tri-
fluoromethyl substituents to either ortho, meta or para positions of
O-benzylhydroxylamine (8e15) proved beneficial with the greatest
gain in IDO1 inhibitor potency seen with halogens in the meta
position. In contrast to other halogen substitutions, the more
electronegative fluorine containing inhibitors (15, 16 and 18) were
essentially equipotent or slightly less potent than the lead.
Scheme 2. Synthesis of monoaryl alcohols.
The benefits of halogen aromatic substitution have been
recognized before in several previous reports of structureeactivity
relationship studies [19,30e35]. There are three potential expla-
nations for the benefits of halogen substitution: Favorable pi
stacking interactions between the halo-aromatics and aromatic
amino acid side chains in the IDO1 active site; halogen bonding
Fig. 2. Two modifications of O-benzylhydroxylamine lead compound.
Table 2
Inhibition data for monoaryl Hydroxylamines with various linkers between the aryl and the Hydroxylamine Group.a
Compd
Lead
Structure
IC50
(
m
M)
Compd
Structure
IC50
10
(mM)
0.81 0.081
4
1
5.2
5
20
2
3
7.1
10
6
7
26
82
a
IC50 values are based on single point inhibition curves run once for each compound except for the lead compound which was repeated 2ꢀ. These results are reported as the
average SD.
3.1.2. Substitution of the aryl ring
between the halogen atoms of the inhibitor and Lewis basic sites in
the IDO1 active site [36]; or beneficial hydrophobic interactions
between the halogen and a complementary pocket in the active
site. Favorable pi stacking interactions between the three relatively
more electron rich Phe residues (F163, 164 and 226) in the active
site and the electron poor halo-aromatic pi systems in the in-
hibitors would benefit binding. Contradicting that assessment
would be the fluorobenzene derivatives (15, 16, and 18) which
afforded inhibition essentially equivalent to the lead, O-benzylhy-
droxylamine, or slightly less potent. The weaker activity of the
fluorine substitutions versus the chlorine, bromine and iodine
would be consistent with halogen bonding events since fluorine is
too electronegative and has too small a sigma hole for effective
Derivatization of the O-benzylhydroxylamines via substitution
on the aromatic ring initially explored potential interactions with
the amino acid residues located in the active site (C129, S167, Y126)
by adding hydrogen donating or accepting substitutions to the
phenyl ring. Previous work from our laboratory had successfully
exploited these interactions in the optimization of the phenyl-
imidazole structural series [29]. However, addition of a hydroxyl
group at the ortho (29) or para (31) position substantially reduced
inhibition (Table 3). Likewise, ortho, meta, or para-substitution with
a methoxy also proved deleterious (24, 26 and 30). Halogen sub-
stitution was the only successful modification that increased in-
hibitor potency, which contrasts with the results of the phenyl-

W.P. Malachowski et al. / European Journal of Medicinal Chemistry 108 (2016) 564e576
567
Table 3
Inhibition data for monoaryl Hydroxylamines with ring Substitution.a
Compd
X
IC50
(m
M)
Compd
X
IC50
(m
M)
Compd
X
IC50(mM)
8
9
4-I
0.22 0.066
0.30 0.015
0.32 0.042
0.34 0.12
0.41
0.57
0.63
0.81 0.081
0.98
16
17
18
19
20
21
22
23
3-F
4-Cl
2-F
4-Br
2-Br
2-CF3
3-CH3
4-NO2
1.0
1.3
1.4
1.6
1.6
2.5
2.7
2.7
24
25
26
27
28
29
30
31
2-OCH3
4-CF3
3-OCH3
4-Ph
4-CH(CH3)2
2-OH
4-OCH3
4-OH
5.4
8.9
9.5
3-Cl
3-Br
3-I
3-CF3
2-I
2-Cl
H
4-F
10
11
12
13
14
Lead
15
12
15
19
24
61
a
IC50 values are based on single point inhibition curves run once for each compound, but repeated at least 2ꢀ for the most potent compounds (initial IC50 values below
mM). These results are reported as the averages SD.
0.4
halogen bonding [37,38]. A more traditional explanation of simple
favorable hydrophobic interactions by the halogens seems less
likely due to the weaker activity seen with 3-methyl (22) and 4-
isopropyl (28), which are considered roughly isosteric with chlo-
rine and iodine respectively [39].
further studies, the result strongly supports the proposed mimicry
of the alkylperoxy transition or intermediate state by the monoaryl
hydroxylamines.
In addition to the deoxy enzyme, we also studied inhibitor
binding to the CO-bound ferrous enzyme, where CO was used as a
surrogate for O2. The CO-bound IDO1 exhibits Soret and Q band at
421 and 538/568 nm, typical for a CO bound six-coordinate low-
spin heme species (Fig. 3C). The addition of the lead compound
resulted in a spectrum similar to that of the inhibitor-bound deoxy
species shown in (B), indicating that inhibitor binding prevents CO
binding to the heme iron. In contrast, the addition of 8 or 40 led to a
spectrum, consistent with a mixture of inhibitor-bound deoxy
spectrum (B) and CO-bound spectrum (C), indicating that inhibitor
binding partially prevents CO binding to the heme iron. Compara-
tive studies suggested that the affinity of these inhibitors towards
the ferric enzyme (Fig. 3A) is much weaker than the ferrous
enzyme.
3.1.3. Disubstitution of the aryl ring
Synergistic effects of multiple substituents were also explored
with a particular focus on maximizing the halogen benefits. How-
ever, as depicted in Table 4, minimal additive effects were observed
when combining beneficial substitutions from the monoaryl sys-
tem. None of the disubstituted derivatives was more potent than
the most potent monosubstituted derivatives, i.e. the C-3
substituted compounds.
3.1.4. Spectroscopic analysis of monoaryl hydroxylamine binding
Given the perplexing results, especially the contradictions with
the phenyl-imidazole work, we further explored the binding
mechanism of the monoaryl hydroxylamines, including the lead
compound, a monosubstituted derivative (8) and a disubstituted
derivative (40), by UVeVis absorption spectroscopic studies. As
shown in Fig. 3B, in the absence of inhibitors, the deoxy ferrous
enzyme exhibits Soret and Q band at 427 and 556 nm, characteristic
for a typical five-coordinate high-spin heme species. Binding of the
lead compound, 8 and 40, led to shift of the Soret band to
419e421 nm, and the appearance of two new Q bands at 531/
556 nm, characteristic of a six-coordinate low-spin species, indi-
cating the coordination of the monoaryl hydroxylamines to the
heme iron, possibly via the amine group. Although definitive proof
of the coordination of the amine group to the heme iron requires
3.1.5. Modeling of monoaryl hydroxylamine binding to guide drug
development
With confirmation that the monoaryl hydroxylamine com-
pounds inhibit IDO1 by coordinating to the heme iron, we then
turned to docking studies using Gold (v5.1) to determine their
potential binding modes (Fig. 4). Unsubstituted monoaryl inhibitor,
O-benzylhydroxylamine, was predicted to have two equally plau-
sible binding modes: in the interior of the cavity near residues F163
and S167, or in the entrance of the cavity between residues F226
and the heme propionate group. Ortho-substituted inhibitors (C-2
substitution, Fig. 5) were predicted to bind in the back of the cavity
(blue structure in Fig. 4), while the para-substituted inhibitors (C-4
Table 4
Inhibition data for monoaryl hydroxylamines with ring disubstitution.a
Compd
R
IC50
(
mM)
Compd
R
IC50
(
m
M)
Compd
R
IC50(mM)
32
33
34
35
36
Lead
2,3-Cl2
3-CF3,4-Cl
2-F, 5-CF3
2,4-Cl2
2-Cl, 4-I
H
0.45
0.49
0.57
0.63
0.73
37
38
39
40
41
3,5-Cl2
2-F,4-CF3
2,4-F2
2,5-(OCH3)2
3,4-Cl2
0.90
1.1
1.4
1.7
2.1
42
43
44
45
46
2-CF3, 4-F
3,5-F2
3,5-Br2
3,5-(CF3)2
2-OH, 3-OCH3
2.4
3.4
4.2
4.6
55
0.81 0.081
a
IC50 values are based on single point inhibition curves run once for each compound.

568
W.P. Malachowski et al. / European Journal of Medicinal Chemistry 108 (2016) 564e576
Fig. 3. Absorption spectra of ferric (A), deoxy ferrous (B) and CO-bound (C) IDO1 in the presence of 500 mM inhibitors. Ref spectra were obtained in the absence of inhibitors. The
616 nm band labeled with an asterisk is of unknown origin.
4. Detailed enzyme inhibition studies
As further confirmation of the inhibitory characteristics of the
hydroxylamine structural class, inhibitory constants (Ki) were
determined for two of the most potent hydroxylamine inhibitors.
Analysis showed Ki values of 164 and 154 nM for two O-alkylhy-
droxylamines, 8 and 9, respectively These potency values yield
ligand efficiencies of 0.93 for both compounds [40e42]. Given the
strong correlation between successful drugs and high ligand effi-
ciencies, the O-alkylhydroxylamines represent a promising class of
IDO1 inhibitors [43]. Based on LineweavereBurk graphical analysis,
both molecules demonstrated an uncompetitive mode of inhibi-
tion. Additionally, it was determined that inhibition of IDO1 activity
by these two compounds was reversible and appeared to be the
result of one-to-one interaction between the O-alkylhydroxyl-
amine inhibitor and IDO1 based on doseeresponse studies. Details
of the procedures and graphs for these studies can be found in the
Supplementary Data section (see Figs. S1eS3). An uncompetitive
mode of inhibition would seem inconsistent with the demon-
strated binding at the heme iron from the spectroscopic studies.
However, 4-phenylimidazole has been crystallized in IDO1 bound
to the heme iron [44] and it demonstrates noncompetitive inhibi-
tion [45]. Furthermore, 4-phenylimidazole derivatives have also
demonstrated an uncompetitive inhibition mode [29] and pre-
sumably they are also binding at the IDO1 active site, in direct
competition with the Trp binding location. Therefore, uncompeti-
tive or non-competitive inhibition mode does not preclude binding
in the active site or to the heme iron. One explanation for such
behavior is that the inhibitors are actually in direct competition for
binding at the heme iron with the other substrate in the reaction,
oxygen. In an assay that modifies the concentration of tryptophan,
an inhibitor that competes with oxygen would likely afford an
uncompetitive inhibition mode.
Fig. 4. Compounds 8 and 14 docked in the IDO1 active site. Compound 14 (blue;
chlorine atom green) binds in the interior of the active site, while 8 (magenta; iodine
atom red) prefers to bind in the active site entrance when docked. The heme is shown
in white and the green represents active site structure of IDO1.
substitution, Fig. 5) appeared to preferentially bind in the cavity
entrance (magenta structure in Fig. 4). The two outliers to this trend
were compound 29 (2-OH) and compound 31 (4-OH). For com-
pound 31, docking predicted the formation of a hydrogen bond
with the S167 in the back of the cavity, while for compound 29
docking suggested hydrogen bonding with the heme propionate in
the front of the cavity. Nonetheless, assessment of the inhibition
data does not support that either of these predicted hydrogen
bonds enhance the binding affinity in comparison with the halo-
genated derivatives; this stands in contrast to earlier reported
hydroxyl-substituted 4-phenyl-imidazole inhibitors [29]. Docking
of the meta substituted analogs to IDO1, however, did not define a
consistent binding mode as the aryl ring could be equally posi-
tioned either in the back or the entrance of the cavity (C-3 substi-
tution, Fig. 5).
5. Selectivity studies
The simplicity of the O-alkylhydroxylamines makes concerns
about their selectivity warranted, especially given their primary
mechanism of attraction to IDO1, heme iron binding. To analyze the
selectivity of two of the top compounds, 8 and 9, two additional
heme iron containing enzymes, catalase and CYP3A4, were
screened for inhibition. As noted in Table 5, neither 8 nor 9 showed
inhibition of catalase; perhaps this is not surprising given catalase's
small natural substrate, hydrogen peroxide, however, this result
does demonstrate that the inhibitory activity of these compounds is

W.P. Malachowski et al. / European Journal of Medicinal Chemistry 108 (2016) 564e576
569
Fig. 5. Structureeactivity relationships (SAR) of substituted monoaryl hydroxylamines with predicted binding mode. Plot of pIC50 values for the substituted hydroxylamine
compounds. The x-axis lists the substitution number on the phenyl ring with disubstituted represented by the sum of the two numbers. The data points are colored based on docked
binding mode in the back (blue), front (pink) or either location (gray) of the IDO1 active site.
Table 5
has been seen repeatedly with IDO1 studies [19,46] and has been
attributed to deficiencies in the isolated enzyme assay, and spe-
cifically problems with controlling IDO1 redox activity.
Cytotoxicity studies with HeLa cells also demonstrated good cell
viability up to concentrations of 100 mM of 8 and 9 (see Supple-
Inhibition of catalase and CYP3A4 by 8 and 9.
Compound
IDO1
Catalase
IC50 M)
CYP3A4
IC50
(
m
M)
(
m
Fold
D
IC50
(
m
M)
Fold D
8
9
0.33
0.31
>100
>100
>300
>320
7.2
15
22
48
mentary Data, Fig. S4). Assessment of impact of human serum
protein binding indicated a substantial reduction in inhibitory ac-
tivity for one of the compounds, 8, while the inhibitory activity of 9
was relatively unaffected (see Supplementary Data, Fig. S5). This
analysis indicates that it should be possible to develop O-alkylhy-
droxylamine inhibitors that are not unduly compromised by serum
protein binding.
Cumulatively, the cell-based potency and the cytotoxicity
studies dramatically illustrate the therapeutic potential for the
hydroxylamines. To the best of our knowledge, there are only two
other structural classes reported to date with similar nanomolar-
level cell-based potency [19,32,33,46] and both of these have
been developed for clinical evaluation.
not attributable to an indirect effect on the catalase in the reaction
mixture that is required to protect IDO1 from oxidative damage.
CYP3A4 was inhibited at the low micromolar level by both 8 and 9.
However, this is still 22- and 48-fold less potent than IDO1 based on
the IC50 values for 8 and 9, respectively. This suggests that the O-
alkylhydroxylamines are selective, but they do demonstrate some
promiscuity and as studies advance, they will need to be closely
monitored for off-target activity.
6. Cell-based assays
7. Conclusion
To assess the therapeutic potential of the hydroxylamine
structural class, two of the most potent compounds, 8and 9, were
tested for cell-based activity in two cell systems: HeLa, expressing
A new structural class of IDO1 inhibitors, O-alkylhydroxyl-
amines, has been discovered. The parent compound, O-benzylhy-
droxylamine, exhibits sub-micromolar inhibition and, unlike many
similarly potent IDO1 inhibitors, is an inexpensive, commercially
available compound. Structureeactivity studies of the O-benzyl-
hydroxylamine lead have explored the chemical space around this
structure and led to some modest improvements in potency.
Halogenation of the aromatic ring was particularly successful in
improving potency. Spectroscopic studies have shown that the O-
alkylhydroxylamine compounds coordinate to the heme iron,
which provide preliminary support for these compounds as alkyl-
peroxy transition or intermediate state mimics. More importantly,
among the strongest enzymatic inhibitors in the O-alkylhydroxyl-
amine class, two exhibited submicromolar cell-based potency with
minimal toxicity. With ligand efficiencies of 0.9, these compounds
hold significant promise as therapeutic tools. Translating enzymatic
potency to cell-based potency has been a major roadblock to in-
hibitor development for other structural classes such as the
native human IDO1 induced with IFNg, and Trex, expressing re-
combinant human IDO1 induced with doxycycline (Table 6; graphs
found in Supplementary Data, Fig. S4). Both compounds demon-
strated nanomolar level activity in both the HeLa and Trex cell lines.
Compounds 8 and 9 were more potent in the cell-based assay than
in the isolated enzyme assay. Although such a result is generally
considered inconsistent with normal drug activity, this irregularity
Table 6
IDO1 IC50 values in HeLa and Trex cell-based assays.
Compound
HeLa (
m
M)a
Trex (m
M)b
8
9
0.10 0.024
0.14 0.023
0.18
0.077
a
HeLa assay results are the averages of three runs with standard deviations.
Trex assay was a single run.
b

570
W.P. Malachowski et al. / European Journal of Medicinal Chemistry 108 (2016) 564e576
naphthoquinones [47]. Furthermore, two O-alkylhydroxylamine
derivatives, 8 and 9, demonstrated exceptional potency on par with
the two compounds INCB024360 [48] and NLG919 [49] that are
currently being evaluated in clinical trials. Given the exciting po-
tential of these rather simple molecules to meet the burgeoning
interest in IDO1 inhibition for both biological investigations of the
kynurenine pathway and therapeutic applications, we believe that
the O-alkyl hydroxylamines will be a useful structural class of
inhibitors.
and treated with HCl (2.0 M solution in ether) to afford the HCl salt
of the O-alkylhydroxylamine. Contaminating diisopropyl hydrazi-
nodicarboxylate could be washed away from the HCl salt with
dichloromethane.
8.1.3. O-(
Synthesized from
general procedure to afford 1 as white crystals in 96% yield. 1H NMR
(400 MHz, CD3OD) 7.46e7.41 (br s, 5H, ArH), 4.86 (s, 1H, ArCH,
a-cyclopropyl)benzylhydroxylamine hydrochloride (1)
a
-cyclopropyl benzyl alcohol according to the
d
J ¼ 8 Hz), 1.36e1.29 (m, 1H, ArCH(ONH2)CH), 0.87e0.81 (m, 1H,
8. Experimental section
CH2), 0.65e0.57 (m, 2H, CH2), 0.36e0.31 (m, 1H, CH2). 13C NMR
(100 MHz, CD3OD)
d 138.36, 130.66, 130.19, 128.63, 92.98, 16.05,
8.1. Chemical synthesis
6.14, 2.64. Anal. calcd. for C10H14ClNO: C ¼ 60.15; H ¼ 7.07; N ¼ 7.01.
Found: C ¼ 60.05%, H ¼ 7.35%, N ¼ 6.91%.
8.1.1. General procedures
All reactants and reagents were commercially available and
were used without further purification unless otherwise indicated.
Anhydrous THF was freshly distilled from Na and benzophenone.
All reactions were carried out under an inert atmosphere of argon
or nitrogen unless otherwise indicated. Concentrated refers to the
removal of solvent with a rotary evaporator at normal water aspi-
rator pressure followed by further evacuation with a direct-drive
rotary vane vacuum pump. Thin layer chromatography was per-
formed using silica gel 60 Å precoated glass or aluminum backed
plates (0.25 mm thickness) with fluorescent indicator, which were
cut. Developed TLC plates were visualized with UV light (254 nm),
iodine, p-anisaldehyde or ninhydrin. Flash column chromatography
was conducted with the indicated solvent system using normal
phase silica gel 60 Å, 230e400 mesh. Yields refer to chromato-
graphically and spectroscopically pure (>95%) compounds except
as otherwise indicated. Melting points were determined using an
open capillary and are uncorrected. 1H NMR spectra were recorded
at 400 MHz and 13C NMR spectra were recorded at 100 MHz.
8.1.4. O-(3-phenyl-2-propenyl)-1-hydroxylamine hydrochloride (2)
Synthesized from 3-phenyl-2-propen-1-ol according to the
general procedure to afford 2 as white crystals in 90% yield. 1H NMR
(400 MHz, CD3OD)
d 7.53e7.47 (m, 2H, ArH), 7.34e7.29 (m, 3H,
ArH), 6.85 (d, 1H, ArCH, J ¼ 16 Hz) 6.40e6.33 (m, 1H, ArCHCH), 4.69
(dd, 2H, ArCHCHCH2, J ¼ 6.9, 1.1 Hz). 13C NMR (100 MHz, CD3OD)
d
139.57, 137.10, 129.93, 129.88, 128.13, 121.47, 77.00. Anal. calcd. for
C9H12ClNO: C, 58.23; H, 6.52; N, 7.54Found: C ¼ 58.71, H ¼ 6.95,
N ¼ 7.49.
8.1.5. O-(1,2,3,4-tetrahydro-1-naphthalene)-hydroxylamine
hydrochloride (3)
Synthesized from 1,2,3,4-tetrahydro-1-naphthalenol according
to the general procedure to afford 3 in 62% yield as white crystals.
1H NMR (400 MHz, CD3OD)
ArH, J ¼ 4 Hz), 7.25e7.17 (m, 2H, ArH), 5.14 (t, 1H, ArCH, J ¼ 2 Hz),
2.90e2.75 (m, 2H, ArCH2), 2.32e2.27 (m, 1H, ArCH2CH2), 2.01e1.94
(m, 2H, ArCH2CH2CH2), 1.87e1.81 (m, 1H, ArCH2CH2). 13C NMR
d
7.40 (d, 1H, ArH, J ¼ 8 Hz), 7.31 (t, 1H,
Chemical shifts are reported in
d
values (ppm) relative to an in-
(100 MHz, CD3OD) d 139.93, 132.27. 131.63, 130.44, 130.33, 127.16,
ternal reference (0.05% v/v) of tetramethylsilane (TMS, 0.0) for 1H
NMR and the CD3OD solvent peak (49.0) for 13C NMR. Peak splitting
patterns in the 1H NMR are reported as follows: s, singlet; d,
doublet; t, triplet; q, quartet; m, multiplet; br,broad. The attached
proton test (APT) experiment was conducted for 13C NMR analysis;
methylene and quaternary carbons were identified as negative
peaks, while methyl and methine had positive peaks. HPLC was
conducted on an Agilent 1100 with an Ascentis Express C-18 col-
82.10, 29.55, 27.51, 18.67. Anal. calcd. for C10H14ClNO: C ¼ 60.15;
H ¼ 7.07; N ¼ 7.01Found: C ¼ 60.13, H ¼ 7.34, N ¼ 6.91.
8.1.6. O-(2,3-dihydro-1H-inden-2-yl)-hydroxylamine hydrochloride
(4)
Synthesized from 2,3-dihydro-1H-inden-2-ol according to the
general procedure to afford 4 as white crystals in 81% yield.
Mp ¼ 132e133.5 ꢁC. 1H NMR (400 MHz, CD3OD)
d 7.29e7.25 (m, 2H,
umn (100 ꢀ 4.6 mm, 2.7
m
m) and a mobile phase of 80:20
ArH), 7.22e7.18 (m, 2H, ArH), 5.07e5.02 (m, 1H, ArCH2CH), 3.37 (d,
1H, ArCHCH2, J ¼ 4 Hz), 3.33 (d, 1H, ArCHCH2, J ¼ 4 Hz), 3.19 (d, 1H,
ArCHCH2, J ¼ 4 Hz), 3.15 (d, 1H, ArCHCH2, J ¼ 4 Hz. 13C NMR
MeCN:H2O. GC analyses were performed on the free hydroxyl-
amine with an EI-MS detector fitted with a 30 m ꢀ 0.25 mm column
filled with cross-linked 5% PH ME siloxane (0.25
mm film thickness);
(100 MHz, CD3OD) d 139.18, 126.73, 124.25, 86.05, 37.65. GC
gas pressure 7.63 psi He. Analysis of samples involved heating from
70 to 250 ꢁC (10 ꢁC/min) and finally holding at 250 ꢁC for 7 min. The
free hydroxylamine compounds were generated by treating the
hydrochloride salts with saturated NaHCO3 and extracting with
EtOAc. All compounds were found to be >95% purity by elemental
analysis, GC, or HPLC as indicated.
tR ¼ 3.182 min.
8.1.7. O-phenethylhydroxylamine hydrochloride (5)
Synthesized from phenethyl alcohol according to the general
procedure to afford
5
as white crystals in 56% yield.
Mp ¼ 110e111 ꢁC. 1H NMR (400 MHz, CD3OD)
d 7.28e7.21 (m, 5H,
ArH), 4.25 (t, 2H, ArCH2, J ¼ 8 Hz), 2.98 (t, 2H, ArCH2CH2, J ¼ 8 Hz).
8.1.2. General synthesis of O-alkyl hydroxylamines
13C NMR (100 MHz, CD3OD)
d 139.28, 129.97, 129.61, 127.83, 76.86,
To a solution of alcohol (1 mmol) in freshly distilled THF (5 ml)
was added triphenylphosphine (1.1 mmol) and N-hydroxylph-
thalimide (1.1 mmol). After the solution was cooled to 0 ꢁC diiso-
propylazodicarboxylate (1.1 mmol) was added dropwise. The
solution was allowed to warm to room temperature over 3 h. Re-
action progress was monitored by TLC (1:1 heptanes:ethyl acetate).
Hydrazine monohydrate (1.1 mmol) was then added and the so-
lution was allowed to stir for 30 min. The resulting reaction mixture
was filtered to remove the white precipitate. The filtrate was
concentrated and subjected to flash chromatography (1:1 hep-
tanes/ethyl acetate). The resulting product was dissolved in ether
35.08. Anal. calcd. for C8H11NO-0.9375 HCl: C ¼ 56.07, H ¼ 7.03,
N ¼ 8.06. Found: C ¼ 55.88%, H ¼ 6.56%, N ¼ 8.08%.
8.1.8. O-(2-3-dihydro-1H-inden-1-yl)hydroxylamine (6)
Synthesized from 2,3-dihydro-1H-inden-ol according to the
general procedure to afford 6 as white crystals in 78% yield.
Mp ¼ 150.5e152 ꢁC. 1H NMR (CD3OD)
d
7.49 (d, 1H, ArH, J ¼ 8 Hz),
7.40e7.33 (m, 2H, ArH), 7.30e7.26 (m, 1H, ArH), 5.59 (d, 2H,
CHONH2, J ¼ 4 Hz), 3.18e3.10 (m, 1H, ArCH2), 2.96e2.89 (m, 1H,
ArCH2), 2.48e2.39 (m, 1H, ArCH2CH2), 2.36e2.31 (m, 1H,
ArCH2CH2). 13C NMR (100 MHz, CD3OD)
d 146.69, 139.10, 131.42,

W.P. Malachowski et al. / European Journal of Medicinal Chemistry 108 (2016) 564e576
571
127.95, 127.02, 126.28, 90.29, 31.90, 30.83. GC tR ¼ 6.781 min.
general procedure to afford 14 as white crystals in 75% yield.
Mp ¼ 148e150 ꢁC. 1H NMR (400 MHz, CD3OD)
d 7.57 (d, 1H, ArH,
8.1.9. O-(3-phenylpropyl)-hydroxylamine hydrochloride (7)
J ¼ 8 Hz), 7.48e7.35 (m, 3H, ArH), 5.22 (s, 2H, ArH). 13C NMR
Synthesized from 3-phenylpropanol according to the general
(100 MHz, CD3OD) d 135.63, 132.68, 132.21, 131.89, 130.74, 128.37,
procedure to afford
7
as white crystals in 95%yield.
74.92. Anal. calcd. for C7H9Cl2NO: C, 43.33; H, 4.67; N, 7.22. Found:
C ¼ 43.01, H ¼ 4.72, N ¼ 7.03.
Mp ¼ 168e169 ꢁC.1H NMR (400 MHz, CD3OD)
d
7.33e7.25 (m, 2H,
ArH), 7.21e7.13 (m, 3H, ArH), 4.02 (t, 2H, CH2ONH2, J ¼ 8 Hz), 2.72 (t,
2H, ArCH2, J ¼ 8 Hz), 2.03e1.96 (m, 2H, ArCH2CH2). 13C NMR
8.1.17. O-(4-fluorobenzyl)hydroxylamine hydrochloride (15)
Synthesized from 4-fluorobenzyl alcohol according to the gen-
eral procedure to afford 15 as white crystals in 85% yield.
(100 MHz, CD3OD)
d 142.15, 129.66, 129.51, 127.33, 75.63, 32.63,
30.57. GC tR ¼ 6.425 min.
Mp ¼ 219e220 ꢁC. 1H NMR (400 MHz, CD3OD)
d 7.49 (t, 2H, ArH,
8.1.10. 4-(iodobenzyl)hydroxylamine hydrochloride (8)
J ¼ 4 Hz) 7.16 (t, 2H, ArH, J ¼ 4 Hz), 5.02 (s, 2H, ArCH2). 13C NMR
Synthesized from 4-iodobenzyl alcohol according to the general
(100 MHz, CD3OD)
d
164.90 (d, JC-F ¼ 240 Hz), 132,87 (d, JC-F ¼ 9 Hz),
procedure to afford
8
as white crystals in 70% yield.
130.53 (d, JC-F ¼ 3 Hz), 116.78 (d, JC-F ¼ 22 Hz), 77.31. Anal. calcd. for
C7H9ClFNO: C, 47.34; H, 5.11; N, 7.89. Found: C ¼ 47.43, H ¼ 4.94,
N ¼ 7.81.
Mp ¼ 210e212 ꢁC 1H NMR (400 MHz, CD3OD)
d 7.80 (d, 2H, ArH,
J ¼ 4 Hz), 7.22 (d, 2H, ArH J ¼ 4 Hz), 4.99 (s, 1H, ArCH2). 13C NMR
(100 MHz, CD3OD)
d 139.03, 133.86, 132.02, 96.09, 77.16. GC
tR ¼ 8.828 min.
8.1.18. O-(3-fluorobenzyl)hydroxylamine hydrochloride (16)
Synthesized from 3-fluorobenzyl alcohol according to the gen-
eral procedure to afford 16 as white crystals in 81% yield.
8.1.11. O-(3-chlorobenzyl)hydroxylamine hydrochloride (9)
Synthesized from the respective alcohol according to the gen-
Mp ¼ 245 ꢁC (decomp).1H NMR (400 MHz, CD3OD)
d 7.45 (t, 1H,
eral procedure to afford
9
as white crystals in 50% yield.
ArH, J ¼ 8 Hz) 7.28e7.15 (m, 3H, ArH), 5.07 (s, 2H, ArCH2). 13C NMR
Mp ¼ 144e148 ꢁC. 1H NMR (400 MHz, CD3OD)
d
7.49 (s, 1H, ArH),
(100 MHz, CD3OD)
131.80 (d, JC-F ¼ 8 Hz), 126.04 (d, JC-F ¼ 3 Hz), 117.31 (d, JC-F ¼ 21 Hz),
116.85 (d, JC-F ¼ 22 Hz), 77.12 (d, JC-F ¼ 2 Hz). GC tR ¼ 3.981 min.
d
164.22 (d, JC-F ¼ 240 Hz), 136.91 (d, JC-F ¼ 8 Hz),
7.45e7.35 (m, 1H, ArH), 5.04 (s, 2H, ArCH2). 13C NMR (100 MHz,
CD3OD)
d 139.49, 135.59, 131.42, 130.55, 130.12, 128.56, 77.02. GC
tR ¼ 6.251 min.
8.1.19. O-(4-chlorobenzyl)hydroxylamine hydrochloride (17)
Synthesized from 4-chlorobenzyl alcohol according to the gen-
eral procedure to afford 17 as white crystals in 70% yield.
8.1.12. O-(3-bromobenzyl)hydroxylamine hydrochloride (10)
Synthesized from 3-bromobenzenyl alcohol according to the
general procedure to afford 10 as white crystals in 98%yield.
Mp ¼ 226e228.5 ꢁC. 1H NMR (400 MHz, CD3OD)
d
7.44 (br s, 4H,
Mp ¼ 210e211 ꢁC. 1H NMR (400 MHz, CD3OD)
d
7.66e7.56 (m, 2H,
ArH), 5.05 (s, 2H, ArCH2). 13C NMR (100 MHz, CD3OD)
d 136.54,
ArH), 7.43 (d, 1H, ArH, J ¼ 8 Hz),7.37 (t, 1H, ArCH, J ¼ 8 Hz), 5.03 (s,
133.11, 132.05, 130.01, 77.13. Anal. calcd. for C7H9Cl2NO: C ¼ 43.44;
H ¼ 4.67; N ¼ 7.22. Found: C ¼ 43.32%, H ¼ 4.59%, N ¼ 7.15%.
2H). 13C NMR (100 MHz, CD3OD)
d 136.78, 133.62, 133.14, 131.71,
129.02, 123.59, 77.03. Anal. calcd. for C7H9ClBrNO:C ¼ 35.25;
H ¼ 3.80; N ¼ 5.87. Found: C ¼ 35.37%, H ¼ 3.52%, N ¼ 5.76%.
8.1.20. O-(2-fluorobenzyl)hydroxylamine hydrochloride (18)
Synthesized from the respective alcohol according to the gen-
eral procedure to afford 18 as white crystals in 90% yield.
8.1.13. O-(3-iodobenzyl)hydroxylamine hydrochloride (11)
Synthesized from the respective alcohol according to the gen-
eral procedure to afford 11 as white crystals in 41% yield.
Mp ¼ 182e184 ꢁC. 1H NMR (400 MHz, CD3OD)
d 7.58e7.45 (m, 2H,
ArH), 7.27e7.19 (m, 2H, ArH), 5.17 (s, 2H, ArCH2). 13C NMR (100 MHz,
Mp ¼ 209e214 ꢁC. 1H NMR (400 MHz, CD3OD)
d
7.82e7.77 (m, 2H,
CD3OD)
d
162.93 (d, JC-F ¼ 247 Hz), 133.26 (d, JC-F ¼ 8 Hz), 133.19 (d,
ArH),
d
7.44 (d, 1H, ArH, J ¼ 8 Hz), 7.21 (t, 1H, ArH, J ¼ 8 Hz), 4.98 (s,
JC-F ¼ 4 Hz), 125.82 (d, JC-F ¼ 3 Hz), 121.48 (d, JC-F ¼ 14 Hz),
2H, ArCH2). 13C NMR (100 MHz, CD3OD)
d 139.79, 139.27, 136.74,
116.71 Hz (d, JC-F ¼ 21 Hz), 71.73 (d, JC-F ¼ 4 Hz). GC tR ¼ 4.067 min.
131.77, 129.62, 95.02, 77.04. GC tR ¼ 8.732 min.
8.1.21. O-(4-bromobenzyl)hydroxylamine hydrochloride (19)
Synthesized from 4-bromobenzyl alcohol according to the
general procedure to afford 19 as white crystals in 92% yield.
8.1.14. O-(3-trifluoromethylbenzyl)hydroxylamine hydrochloride
(12)
Synthesized from 3-trifluoromethylbenzyl alcohol according to
the general procedure to afford 12 as white crystals in 88% yield.
Mp ¼ 232e236 ꢁC. 1H NMR (400 MHz, CD3OD)
d 7.58 (d, 2H, ArH,
J ¼ 4 Hz), 7.37 (d, 2H, ArH, J ¼ 4 Hz), 5.02 (s, 2H, ArCH2). 13C NMR
Mp ¼ 164e165 ꢁC 1H NMR (400 MHz, CD3OD)
d
7.77e7.72 (m, 3H,
(100 MHz, CD3OD) d 133.60, 133.11, 132.29, 124.70, 77.23. GC
ArH), 7.65 (t, 1H, ArH, J ¼ 8 Hz), 5.14 (s, 2H, ArCH2). 13C NMR
tR ¼ 7.383 min.
(100 MHz, CD3OD)
d
134.40, 132.62, 130.7 (q, JC-F ¼ 40 Hz), 129.50,
125.92 (q, JC-F ¼ 4 Hz), 125.51 (q, JC-F ¼ 4 Hz), 123.99 (d, JC-
8.1.22. O-(2-bromobenzyl)hydroxylamine hydrochloride (20)
Synthesized from 2-bromobenzyl alcohol according to the
general procedure to afford 20 as white crystals in 87% yield.
¼ 270 Hz), 75.71. GC tR ¼ 4.170 min.
F
8.1.15. O-(2-iodobenzyl)hydroxylamine hydrochloride (13)
Synthesized from the respective alcohol according to the gen-
eral procedure to afford 13 as white crystals in 43% yield.
Mp ¼ 95e98 ꢁC. 1H NMR (400 MHz, CD3OD)
d 7.70 (dd, 1H, ArH,
J ¼ 4, 1 Hz), 7.54 (dd, 1H, ArH, J ¼ 4, 1 Hz), 7.46 (dt, 1H, ArH, J ¼ 7,
1 Hz), 7.37 (dt, 1H, ArH, J ¼ 7, 1 Hz), 5.20 (s, 2H, ArCH2). 13C NMR
Mp ¼ 119e120 ꢁC. 1H NMR (400 MHz, CD3OD)
d
7.95 (d, 1, ArH,
(100 MHz, CD3OD) d 134.38, 133.83, 132.71, 132.56, 129.19, 125.47,
J ¼ 8 Hz), 7.51e7.44 (m, 2H, ArH), 7.16 (t, 1H, ArH, J ¼ 8 Hz), 5.14 (s,
77.32. Anal. calcd. for C7H9ClBrNO C ¼ 35.25; H ¼ 3.80; N ¼ 5.87.
1H, ArCH2). 13C NMR (100 MHz, CD3OD)
d 141.12, 137.07, 132.35,
Found: C ¼ 35.14%, H ¼ 3.57%, N ¼ 5.86%.
132.01, 129.83, 100.05, 81.35. Anal. calcd. for C7H9ClINOeH2O: C,
27.70; H, 3.65; N, 4.61. Found: C ¼ 27.84, H ¼ 3.66, N ¼ 4.48.
8.1.23. O-(2-trifluoromethylbenzyl)hydroxylamine hydrochloride
(21)
8.1.16. O-(2-chlorobenzyl)hydroxylamine hydrochloride (14)
Synthesized from 2-trifluoromethylbenzyl alcohol according to
the general procedure to afford 21 as white crystals in 78% yield.
Synthesized from the respective alcohol according to the

572
W.P. Malachowski et al. / European Journal of Medicinal Chemistry 108 (2016) 564e576
Mp ¼ 115e119 ꢁC.1H NMR (400 MHz, CD3OD)
d
7.81 (d, 1H, ArH,
8.1.30. O-(4-isopropylbenzyl)hydroxylamine hydrochloride (28)
Synthesized from the respective alcohol according to the gen-
eral procedure to afford 28 as white crystals in 89%yield.
J ¼ 8 Hz), 7.76e7.66 (m, 2H, ArH), 7.63 (t, 1H, ArH, J ¼ 8 Hz), 5.26 (s,
2H, ArCH2). 13C NMR (100 MHz, CD3OD)
d 133.89, 132.94, 132.41,
131.13, 130.10 (q, JC-F 40 Hz), 127.50 (q, JC-F ¼ 6 Hz), 125.53 (d, JC-
Mp ¼ 159e160 ꢁC. 1H NMR (400 MHz, CD3OD)
d 7.39 (d, 2H, ArH,
¼ 271 Hz), 74.25. Anal. calcd. for C8H9ClF3NOC ¼ 42.22; H ¼ 3.99;
J ¼ 8 Hz), 7.33 (d, 2H, ArH, J ¼ 8 Hz), 5.03 (s, 2H, ArCH2),2.99e2.92
F
N ¼ 6.15. Found: C ¼ 42.10%, H 4.15%, N ¼ 6.05%.
(m, 1H, (CH3)2CHAr), 1.27 (d, 6H, J ¼ 8 Hz). 13C NMR (100 MHz,
CD3OD)
d 150.55, 130.30, 129.36, 126.60, 76.69, 33.86, 22.89. Anal.
calcd. for C10H16ClNO C ¼ 59.55; H ¼ 8.00; N ¼ 6.94. Found:
8.1.24. O-[(3-methylphenyl)-methyl]-hydroxylamine hydrochloride
(22)
C ¼ 59.46%, H ¼ 7.77%, N ¼ 7.02%.
Synthesized from 3-methyl benzenemethanol according to the
general procedure to afford 22 as white crystals in 94% yield.
8.1.31. O-(2-hydroxybenzyl)hydroxylamine hydrochloride (29)
Synthesized from the respective alcohol according to the gen-
eral procedure to afford 29 as white crystals in 79% yield.
Mp ¼ 134e135 ꢁC. 1H NMR (400 MHz, CD3OD) 7.35e7.25, (m, 2H,
ArH) 6.93e6.85 (m, 2H, ArH), 5.10 (s, 2H, ArCH2). 13C NMR
Mp ¼ 187e188.5 ꢁC. 1H NMR (400 MHz, CD3OD)
d 7.33e7.22 (m, 4H,
ArH), 5.01 (s, 2H, ArCH2), 2.37 (s, 3H, ArCH3). 13C NMR (100 MHz,
CD3OD)
d 139.84, 134.12, 131.30, 131.02, 129.79, 127.49, 78.17, 21.23.
Anal. calcd. for C8H12ClFNO C ¼ 55.34; H ¼ 6.97; N ¼ 8.07. Found:
(100 MHz, CD3OD)
d 157.79, 132.82, 132.37, 120.86, 120.78, 116.46,
C ¼ 55.47%, H ¼ 7.22%, N ¼ 8.06%.
73.63. GC tR ¼ 7.582 min.
8.1.25. O-(4-nitrobenzyl)hydroxylamine hydrochloride (23)
Synthesized from the respective alcohol according to the gen-
eral procedure to afford 23 as off-white crystals in 54% yield.
8.1.32. O-(4-methoxybenzyl)hydroxylamine hydrochloride (30)
Synthesized from 4-methoxy benzyl alcohol according to the
general procedure to afford 30 as white crystals in 87%yield.
Mp ¼ 199e199.5 ꢁC. 1H NMR (400 MHz, CD3OD)
d 8.29 (d, 2H, ArH,
Mp ¼ 125e127 ꢁC. 1H NMR (400 MHz, CD3OD)
d 7.39 (d, 2H, ArH,
J ¼ 12 Hz), 7.70 (d, 2H, ArH, J ¼ 12 Hz), 5.19 (s, ArCH2, 2H). 13C NMR
J ¼ 8 Hz), 6.99 (d, 2H, ArH, J ¼ 8 Hz), 4.98 (s, 2H, ArCH2), 3.83 (s, 3H,
ArOCH3). 13C NMR (100 MHz, CD3OD)
d 162.17, 135.55, 132.21,
(100 MHz, CD3OD)
d 149.92, 141.55, 130.98, 124.90, 76.56. Anal.
calcd. for C7H9ClN2O3:C ¼ 41.09; H ¼ 4.43; N ¼ 13.69. Found:
126.00, 124.20, 115.13, 77.75, 55.64. Anal. calcd. for C8H12ClNO2:
C ¼ 50.67; H ¼ 6.38; N ¼ 7.39. Found: C ¼ 50.84%, H ¼ 6.53%,
N ¼ 7.39%.
C ¼ 41.20%, H ¼ 4.42%, N ¼ 13.42%.
8.1.26. O-(2-methoxybenzyl)hydroxylamine hydrochloride (24)
Synthesized from respective alcohol according to the general
procedure to afford 24 as white crystals in 64% yield.
8.1.33. O-(4-hydroxylbenzyl)hydroxylamine hydrochloride (31)
Synthesized from 4-hydroxybenzyl alcohol according to the
general procedure to afford 31 as white crystals in 44% yield.
Mp ¼ 102e104 ꢁC. 1H NMR (400 MHz, CD3OD)
d 7.42 (t, 1H, ArH,
J ¼ 8 Hz), 7.37 (d, 1H, ArH,, J ¼ 4 Hz), 7.06 (d, 1H, ArH, J ¼ 8 Hz), 6.99
Mp ¼ 189e191 ꢁC. 1H NMR (400 MHz, CD3OD)
d 7.29 (d, 2H, ArH,
(t, 1H, ArH, J ¼ 8 Hz), 5.09 (s, 2H, ArCH2), 3.88 (s, 3H, CH3). 13C NMR
J ¼ 8 Hz), 6.85 (d, 2H, ArH, J ¼ 8 Hz), 4.93 (s, 2H, ArCH2). 13C NMR
(100 MHz, CD3OD)
d
159.72, 132.67, 122.43, 121.75, 112.14, 73.34,
(100 MHz, CD3OD) d 160.14, 132.50, 124.86, 116.63, 78.11. HPLC
56.09. GC tR ¼ 6.953 min.
tR ¼ 0.475 min.
8.1.34. O-(2,3-dichlorobenzyl)hydroxylamine hydrochloride (32)
Synthesized from 2,3-dichlorobenzyl alcohol according to the
general procedure to afford 32 as white crystals in 48% yield.
8.1.27. O-(4-Trifluoromethylbenzyl)hydroxylamine hydrochloride
(25)
Synthesized from the respective alcohol according to the gen-
eral procedure to afford 25 as white crystals in 99% yield.
Mp ¼ 184.5e185.5 ꢁC. 1H NMR (400 MHz, CD3OD)
d 7.64 (dd, 1H,
Mp ¼ 179e181 ꢁC. 1H NMR (400 MHz, CD3OD)
d
7.77 (d, 2H, ArH,
ArH, J ¼ 8, 1 Hz), 7.51 (dd, 2H, ArH, J ¼ 8, 1 Hz), 7.39 (t, 1H, ArH,
J ¼ 8 Hz), 5.24 (s, 2H, ArCH2). 13C NMR (100 MHz, CD3OD)
d 134.63,
J ¼ 8 Hz), 7.67 (d, 2H, ArH, J ¼ 8 Hz),5.16 (s, 2H, ArCH2). 13C NMR
134.61, 133.84, 132.88, 130.96, 129.32, 75.38. GC tR ¼ 8.752 min.
(100 MHz, CD3OD)
(q, JC-F ¼ 4 Hz),124.09, 77.07. Anal. calcd. for C8H9ClF3NO: C ¼ 42.22;
d
138.81, 132.52 (d, JC-F ¼ 32 Hz), 130.73, 126.84
8.1.35. O-(4-chloro-3-trifluoromethyl)benzyl)hydroxylamine
hydrochloride (33)
H ¼ 3.99; N ¼ 6.15 Found: C ¼ 41.98%, H ¼ 3.89%, N ¼ 6.34%.
Synthesized from the respective alcohol according to the gen-
eral procedure to afford 33 (12% yield) as white crystals.).
8.1.28. O-(3-methoxybenzyl)hydroxylamine hydrochloride (26)
Synthesized from 3-methoxybenzenyl alcohol according to the
general procedure to afford 26 as white crystals in 98% yield.
Mp ¼ 120e122 ꢁC. 1H NMR (400 MHz, CD3OD)
d 7.86 (s, 1H, ArH),
7.69 (2, 2H, ArH), 5.10 (s, 2H, ArCH2). 13C NMR (100 MHz, CD3OD)
Mp ¼ 126e127.5 ꢁC. 1H NMR (400 MHz, CD3OD)
d 7.34 (t, 1H, ArH,
d
135.33, 134.30 (d, JC-F ¼ 3 Hz), 133.35, 129.54 (q, JC-F ¼ 30 Hz),
J ¼ 8 Hz), 7.03e6.97 (m, 3H, ArH), 5.04 (s, 2H, ArCH2), 3.81 (s, 3H,
129.49, 124.13 (d, JC-F 271 Hz), 76.39. Anal. calcd. for
C8H8Cl2F3NO-0.25 HCl: C ¼ 35.43; H ¼ 3.08; N ¼ 5.17. Found:
C ¼ 35.11%, H ¼ 2.81%, N ¼ 5.11%.
¼
CH3OAr). 13C NMR (100 MHz, CD3OD)
d 161.44, 135.64, 130.99,
122.46, 116.15, 116.74, 78.00, 55/76. Anal. calcd. for C8H12ClNO2.
C ¼ 50.67; H ¼ 6.38; N ¼ 7.39. Found: C ¼ 50.52%, H ¼ 6.07%,
N ¼ 7.36%.
8.1.36. O-(2-fluoro-5-Trifluoromethylbenzyl)hydroxylamine
hydrochloride (34)
8.1.29. O-(4-phenylbenzyl)hydroxylamine hydrochloride (27)
Synthesized from the respective alcohol according to the gen-
eral procedure to afford 27 as white crystals in 75% yield.
Synthesized from the respective alcohol according to the gen-
eral procedure to afford 34 as white crystals in 40% yield.
174.5e175 ꢁC. 1H NMR (400 MHz, CD3OD)
d 7.89e7.83 (m, 2H, ArH),
4
Mp ¼ 187e188 ꢁC. 1H NMR (400 MHz, CD3OD)
d
7.71 (d, 2H, ArH,
7.42 (t, 1H, ArH, J ¼ 8 Hz), 5.22 (d, 2H, ArCH2, JC-F ¼ 12 Hz). 13C NMR
J ¼ 8 Hz), 7.62 (d, 2H, ArH, J ¼ 8 Hz), 7.52 (d, 2H, ArH, J ¼ 8 Hz), 7.44
(100 MHz, CD3OD)
d
164.70 (d, JC-F ¼ 254 Hz), 130.56, 130.50 (d, JC-
(t, 2H, ArH, J ¼ 4 Hz), 7.37 (t, 1H, ArH,, J ¼ 4 Hz), 5.07 (s, 2H, ArCH2).
¼ 15 Hz), 128.42, 125.02 (d, JC-F ¼ 269 Hz), 123.03 (d, JC-F ¼ 16 Hz),
F
13C NMR (100 MHz, CD3OD)
d
143.93, 141.55, 133.21, 131.05, 129.98,
117.95 (d, JC-F ¼ 22 Hz), 70.95 (d, JC-F ¼ 4 Hz). Anal. calcd. for
C8H8ClF4NO-0.25 HCl: C ¼ 37.72, H ¼ 3.27, N ¼ 5.50. Found:
128.83, 128.03, 127.94, 77.98. GC tR ¼ 12.779 min.

W.P. Malachowski et al. / European Journal of Medicinal Chemistry 108 (2016) 564e576
573
C ¼ 37.61%, H ¼ 3.05%, N ¼ 5.36%.
ArH), 7.60 (d, 1H, ArH, J ¼ 8 Hz), 7.39 (dd, 1H, ArH, J ¼ 8, 1 Hz), 5.06
(s, 2H, ArCH2). 13C NMR (100 MHz, CD3OD)
135.13, 134.57, 133.78,
d
8.1.37. O-(2,4-dichlorobenzyl)hydroxylamine hydrochloride (35)
Synthesized from 2,4-dichloro benzyl alcohol according to the
general procedure to afford 35 as white crystals in 69% yield.
132.28, 132.11, 130.06, 76.42. Anal. calcd. for C7H8Cl3NO C ¼ 36.80;
H ¼ 3.53; N ¼ 6.13. Found: C ¼ 36.40%, H ¼ 3.08%, N ¼ 6.10%.
Mp ¼ 184.5e185.5 ꢁC. 1H NMR (400 MHz, CD3OD)
d
7.57e7.54 (m,
8.1.44. O-(2-Trifluoromethyl,4-fluorobenzyl)hydroxylamine
hydrochloride (42)
2H, ArH), 7.44e7.41 (m, 1H, ArH), 7.40 (dd, 1H, ArH, J ¼ 2.0 Hz,
8.2 Hz), 5.19 (s, 2H, ArCH2). 13C NMR (100 MHz, CD3OD)
d
137.50,
Synthesized from respective alcohol according to the general
procedure to afford 42 as white crystals in 61% yield.
136.74, 133.94, 131.07, 130.71, 128.88, 74.40. Anal. calcd. for
C7H8Cl3NO C ¼ 36.80; H ¼ 3.53; N ¼ 6.13. Found: C ¼ 36.89%,
H ¼ 3.50%, N ¼ 5.88%.
Mp ¼ 118e119 ꢁC. 1H NMR (400 MHz, CD3OD)
d 7.59 (dd, 1H, ArH,
J ¼ 8, 1 Hz),
d
7.51 (dd, 1H, ArH, J ¼ 8, 1 Hz), 7.47 (dt, 1H, ArH, J ¼ 4,
1 Hz), 5.22 (s, 2H, ArCH2). 13C NMR (100 MHz, CD3OD)
d 164.21 (d,
8.1.38. O-(2-Chloro,4-iodobenzyl)hydroxylamine hydrochloride
(36)
JC-F ¼ 249 Hz),135.99 (d, JC-F ¼ 8 Hz),132.41 (q, JC-F ¼ 32 Hz),128.62,
124.57 (d, JC-F ¼ 272 Hz), 120.65 (d, JC-F ¼ 21 Hz), 115.35 (m), 73.55
(d, JC-F ¼ 2 Hz). GC tR ¼ 3.903 min.
Synthesized from the respective alcohol according to the gen-
eral procedure to afford 36 as white crystals in 80% yield.
Mp ¼ 117e119 ꢁC. 1H NMR (400 MHz, CD3OD)
d
7.91 (s, 1H, ArH),
8.1.45. O-(3,5-diflourobenzyl)hydroxylamine hydrochloride (43)
Synthesized from 3,5-diflourobenzyl alcohol according to the
general procedure to afford 43 as white crystals in 92% yield.
7.78 (d, 1H, ArH, J ¼ 8 Hz), 7.31 (d, 1H, ArH, J ¼ 8 Hz), 5.17 (s, 2H,
ArCH2). 13C NMR (100 MHz, CD3OD)
d 139.26, 137.91, 136.43, 133.89,
131.89, 96.49, 74.51. Anal. cald. for C7H8Cl2INO: C ¼ 26.28; H ¼ 2.52;
Mp ¼ 200.5e201 ꢁC.1H NMR (400 MHz, CD3OD)
d 7.10 (d, 2H,
N ¼ 4.38. Found: C ¼ 26.62%, H ¼ 2.15%, N ¼ 4.34%.
ArH, J ¼ 4 Hz), 7.03 (t, ArH, J ¼ 8 Hz), 5.09 (s, 2H, ArCH2). 13C NMR
(100 MHz, CD3OD)
d
164.66 (dd, JC-F ¼ 247, 13 Hz), 138.65 (t, JC-
8.1.39. O-(3,5-dichlorobenzyl)hydroxylamine hydrochloride (37)
Synthesized from respective alcohol according to the general
procedure to afford 37 as white crystals in 43% yield. 1H NMR
¼ 9 Hz), 112.89 (dd, JC-F ¼ 12, 7 Hz), 105.57 (t, JC-F ¼ 25 Hz), 76.45.
F
GC tR ¼ 3.717 min.
(400 MHz, CD3OD)
d
7.52 (s, 1H, ArH), 7.45 (d, 2H, ArH), 5.05 (s, 2H,
8.1.46. O-(3,5-dibromobenzyl)hydroxylamine hydrochloride (44)
Synthesized from 3,5-dibromo benzyl alcohol according to the
general procedure to afford 44 as white crystals in 79%yield.
ArCH2). 13C NMR (100 MHz, CD3OD)
76.35. GC tR ¼ 8.248 min.
d 138.30, 136.58, 130.41, 128.66,
Mp ¼ 197e200 ꢁC. 1H NMR (400 MHz, CD3OD)
d 7.79 (t, 1H, ArH,
8.1.40. O-(2-fluoro, 4-trifluoromethylbenzyl)hydroxylamine
hydrochloride (38)
J ¼ 1 Hz), 7.62 (d, 2H, ArH, J ¼ 1 Hz), 5.03 (s, 2H, ArCH2). 13C NMR
(100 MHz, CD3OD)
d 138.71, 135.95, 132.02, 124.22, 76.16. GC
Synthesized from 2-fluoro-4-trifluoromethyl benzyl alcohol
according to the general procedure to afford 38 as white crystals in
tR ¼ 10.513 min.
67% yield. Mp ¼ 228e231 ꢁC.1H NMR (CD3OD)
d
7.75 (t, 1H, ArH,
8.1.47. O-(3,5-Bis(triflouromethyl-benzyl)hydroxylamine
hydrochloride (45)
J ¼ 8 Hz), 7.61 (t, 2H, ArH, J ¼ 8 Hz), 5.23 (s, 2H, ArCH2). 13C NMR
(100 MHz, CD3OD)
d
162.41 (d, JC-F ¼ 250 Hz), 134.84 (dd, JC-F ¼ 42,
Synthesized from 3,5-(bis)triflouromethyl benzyl alcohol ac-
cording to the general procedure to afford 45 as white crystals in
8 Hz), 133.79 (d, JC-F ¼ 4 Hz), 126.00 (d, JC-F ¼ 14 Hz), 124.46 (d, JC-
¼ 270 Hz), 122.67 (t, JC-F ¼ 4 Hz), 114.20 (dq, JC-F ¼ 25, 4 Hz), 70.94
93% yield. Mp-195-197 ꢁC.1H NMR (400 MHz, CD3OD)
d
8.08 (s, 2H,
ArH), 8.04 (s, 1H, ArH), 5.26 (s, 2H, ArCH2). 13C NMR (100 MHz,
CD3OD)
F
(d, JC-F ¼ 3 Hz). Anal. calcd. for C8H8ClF4NO: C ¼ 39.12; H ¼ 3.28;
N ¼ 5.70. Found: C ¼ 39.06%, H ¼ 3.02%, N ¼ 5.66%.
d
137.90, 133.22 (q, JC-F ¼ 33 Hz), 130.60 (d, JC-F ¼ 3 Hz),
124.59 (d, JC-F ¼ 271 Hz), 124.13 (t, JC-F ¼ 3 Hz), 76.14. Anal. calcd. for
C9H8ClF6NO C ¼ 36.57; H ¼ 2.73; N ¼ 4.74. Found: C ¼ 36.55%,
H ¼ 2.49%, N ¼ 4.64%.
8.1.41. O-2,4-(difluorobenzyl)hydroxylamine hydrochloride (39)
Synthesized from the respective alcohol according to the gen-
eral procedure to afford 39 as white crystals in 30% yield.
Mp ¼ 155e157 ꢁC. 1H NMR (400 MHz, CD3OD)
d
7.54 (q, 1H, ArH,
8.1.48. O-(2-Hydroxy-3-methoxybenzyl)hydroxylamine
hydrochloride (46)
J ¼ 4 Hz), 7.10e7.03 (m, 2H, ArH), 5.12 (s, 2H, ArCH2). 13C NMR
(100 MHz, CD3OD)
d
165.63 (d, JC-F ¼ 249 Hz), 163.23 (d, JC-
Synthesized from 2-hydroxy-3-methoxybenzyl alcohol accord-
ing to the general procedure to afford 46 as white crystals in 45%
¼ 231 Hz), 134.70 (d, JC-F ¼ 5 Hz), 117.83, 112.92 (d, JC-F ¼ 4 Hz),
F
105.16 (t, JC-F ¼ 26 Hz), 71.12 (d, JC-F ¼ 4 Hz). Anal. calcd. for
C7H8ClF2NO: C ¼ 42.99; H ¼ 4.12; N ¼ 7.16. Found: C ¼ 43.15%,
H ¼ 3.74%, N ¼ 6.89%.
yield. Mp ¼ 176e178 ꢁC.1H NMR (400 MHz, CD3OD)
d 7.03 (d, 1H,
ArH, J ¼ 8 Hz), 6.91e6.82 (m, 2H, ArH), 5.08 (s, 2H, ArCH2), 3.87 (s,
3H, ArOCH3). 13C NMR (100 MHz, CD3OD)
d 149.13, 147.21, 124.15,
120.61, 120.56, 113.91, 73.28, 56.56. Anal. Calcd. for C8H12ClNO3
C ¼ 46.73; H ¼ 5.88; N ¼ 6.81. Found: C ¼ 46.38%, H ¼ 5.71%,
N ¼ 6.72%.
8.1.42. O-(2,5-dimethoxybenzyl)hydroxylamine hydrochloride (40)
Synthesized from 2,5-dimethoxybenzyl alcohol according to the
general procedure to afford 40 as white crystals in 90% yield.
Biochemical assays. Recombinant human IDO1 was expressed
and purified as described [50]. Inhibition assays to determine IC50
and Ki values were performed in a 96-well microtiter plate as
described by Littlejohn et al. [50]. with some modification. Under
these assay conditions, we have previously determined that the
recombinant hIDO1 follows MichaeliseMenten kinetics [51].
Briefly, the reaction mixture contained 50 mM potassium phos-
Mp ¼ 115e120 ꢁC. 1H NMR (400 MHz, CD3OD)
d 6.99e6.95 (m, 3H,
ArH), 5.05 (s, 2H), 3.83 (s, 3H), 3.76 (s, 3H). 13C NMR (100 MHz,
CD3OD)
d 155.11, 153.71, 123.23, 118.26, 116.95, 113.27, 73.22, 56.56,
56.20. Anal. calcd. for C9H14Cl3NO3: C ¼ 49.21; H ¼ 6.42; N ¼ 6.38.
Found: C ¼ 49.14%, H ¼ 6.76%, N ¼ 5.94%.
8.1.43. O-(3,4-dichlorobenzy)lhydroxylamine hydrochloride (41)
Synthesized from 3,4-dichlorobenzyl alcohol according to the
general procedure to afford 41 as white crystals in 48%yield.
phate buffer (pH 6.5), 40 mM ascorbic acid, 400
20 M methylene blue and ~27 nM purified recombinant IDO1 per
reaction. The reaction mixture was added to the substrate, -tryp-
tophan (L-Trp), and the inhibitor. For IC50 determinations, inhibitor
mg/ml catalase,
m
L
Mp ¼ 193.5e195.5 ꢁC. 1H NMR (400 MHz, CD3OD)
d 7.64 (s, 1H,

574
W.P. Malachowski et al. / European Journal of Medicinal Chemistry 108 (2016) 564e576
solutions were made fresh and serially diluted and the L-Trp sub-
strate was tested at 100 M). For Ki determinations,
Assessment of cell viability. The single point dilution series used
to determine IC50 values were also evaluated for cell viability. The
TCA-fixed cells remaining in the 96-well plate following transfer of
the media were processed essentially as described [52]. Fixed cells
were washed four times in tap water, blotted, air-dried, and treated
m
M (Km ¼ 80
m
the assay was set up similarly except that serial dilutions of in-
hibitors were matrixed with serial dilutions of L-Trp substrate. The
reaction was carried out at 37 ꢁC for 60 min and stopped by the
addition of 30% (w/v) trichloroacetic acid. The plate was then
incubated at 65 ꢁC for 15 min to convert N-formylkynurenine to
for 15 min at room temperature with 100 mL of 0.4% (w/v) sulfar-
hodamine B (SRB) (SigmaeAldrich, St. Louis, MO) prepared in 1%
kynurenine and centrifuged at 1250 g for 10 min. Lastly, 100
ml
acetic acid. Wells were then rinsed four times in 1% acetic acid, air-
supernatant from each well was transferred to a new 96 well plate
and mixed at equal volume with 2% (w/v) p-dimethylamino-
benzaldehyde (Ehrlich's reagent) in acetic acid. The yellow color
generated from the reaction with kynurenine was measured at
490 nm using a Synergy HT microtiter plate reader (Bio-Tek,
Winooski, VT). As an initial compound screen, assessments of IC50
values were performed as single point dilution series. The most
potent compounds were subsequently retested two or more times
with the results reported as averages. The data were analyzed using
Graph Pad Prism 4 software (Graph Pad Software Inc., San Diego,
CA) and the Enzyme Kinetics module in SigmaPlot version 10
(Systat Software Inc, San Jose, CA). (Note that a separate experiment
analyzing the possible competition between kynurenine and four
different O-alkylhydroxylamine inhibitors for p-dimethylamino-
benzaldehyde showed no impact on the visible absorbance reading
from the presence of any one of the four inhibitors.)
dried, and developed by adding 200 mL of 10 mM unbuffered
TriseHCl and incubating for 15 min at room temperature with
gentle shaking. Staining intensity, proportional to cell number, was
determined by reading the absorbance at 570 nm on a plate reader.
Graphs of cell viability curves were generated using Prism v.5.0
(GraphPad Software, Inc.).
Docking calculations. Small molecules were constructed in
MOE V2011.1 (Chemical Computing Group, Inc.) and ionized and
hydrogens added using MOE's WashMDB function. The small
molecule conformation was minimized to a gradient of 0.01 with
the MMFF94x [53,54]using a distance-dependent dielectric con-
stant of 1. The crystal structure of IDO1 bound with 4-
phenylimidizole was used for docking calculations [44]. The 2-[N-
cyclohexylamino]ethanesulfonic acid and 4-phenyl-1-imidazole
ligands were removed from the active site docking, hydrogen
atoms were added and tautomeric states and orientations of Asn,
Gln, His residues were determined with Molprobity [55,56]. Hy-
drogens were then added to crystallographic waters using MOE.
The Amber99 [57] force field in MOE was used, and iron was
parameterized in the Feþ3 state. Dioxygen was not added to the
iron. All hydrogens were minimized to an rms gradient of 0.01,
holding the remaining heavy atoms fixed. A stepwise minimization
followed for all atoms using a quadratic force constant (100 kcal-
mole/Å) to tether the atoms to their starting geometries; for each
subsequent minimization, the force constant was reduced by a half
until reached 0.01 and was then followed by a cycle of minimization
without applying any constraints. GOLD version 5.1 (Cambridge
Crystallographic Data Centre) was used with coordination binding
of the hydroxylamine nitrogen to the Feþ3 atom of the heme.
Chemscore parameters, adapted for metaleligand interactions,
were used for scoring [58]. Fifty genetic algorithm (GA) docking
runs were performed with the initial_virtual_pt_match_max ¼ 2.5,
all other parameters were set as defaults.
Spectroscopic measurements. The UVeVis spectra were ob-
tained at room temperature using a UV2401 spectrophotometer
(Shimadzu Scientific Instruments, Inc., Columbia, MD) with a slit
width of 1 nm in a 1 cm path-length quartz cuvette. The proteins
(5 mM) were buffered with a 100 mM pH of 7.4 Tris buffer. The
deoxy samples were prepared by injecting ~10-fold excess sodium
dithionite into the protein samples pre-purged with nitrogen gas
with a gas-tight syringe. The CO adducts were prepared by adding
200 ml CO gas (1 atm) into the deoxy samples.
Cell-based IDO1 inhibition and cytotoxicity assays. Com-
pounds were evaluated for inhibitory activity against human IDO1
expressed endogenously in HeLa cells and exogenously in T-Rex
cells. HeLa cells were seeded in a 96 well plate at a density of
10,000 cells per well in 100
Penicillin-Streptomycin and IDO1 expression was induced by the
addition of IFN to a final concentration of 100 ng/mL. T-Rex cells
containing a tet-regulated human IDO1 cDNA were seeded in a 96-
well plate at a density of 10,000 cells per well in 100 L of
DMEM þ10% FBS þ1% Penicillin-Streptomycin and IDO1 expression
was induced by the addition of 100 L of media containing 20 ng/
ml DMEM þ10% fetal bovine serum þ1%
g
m
Conflict of interest
m
mL doxycycline. With both assays, IDO1 induction was allowed to
proceed for 24 h after which the media was discarded, the wells
W.P.M., J.B.D., A.J.M. and G.C.P. declare a conflict of interest due
to their various relationships with New Link Genetics Corporation,
which has licensed IDO1-related intellectual property from the
Lankenau Institute for Medical Research and Georgia Regents
University. W.P.M., J.B.D. and A.J.M. are inventors and shareholders
in the company. G.C.P. is an inventor and shareholder who has
received grant support and compensation from the company in his
role as a scientific advisor.
rinsed once, and serial dilutions of compound in 200
DMEM þ10% FBS with the final concentration of tryptophan
adjusted to 100
M. Following incubation at 37 ꢁC for an additional
24 h, the assay was stopped by the addition of 50 L of 50% (w/v)
mL of
m
m
TCA to each well, and the cells were fixed by incubating for 1 h at
4 ꢁC.
Assessment of IDO1 activity. Compound IC50 values were assessed
from single point dilution series with the most potent compounds
subsequently retested two or more times and the results reported
as averages. Following the TCA fixation step, the supernatants were
transferred to a round-bottomed 96-well plate and incubated at
65 ꢁC for 15 min. The plates were then centrifuged at 1250ꢀ g for
Acknowledgment
Financial support for this work was provided by the National
Institutes of Health, NCI R01 CA109542-04A2 to J. M. L., G. C. P., A. J.
M. and W. P. M., and GM086482 to S. R. Y. A.J.M. is also the recipient
of grants from the Susan G. Komen for the Cure and receives
additional funding support through NIH grant CA159337. G.C.P. is
also the recipient of NIH grants CA159337 and CA159315. Addi-
tional support for this project was provided by grants to G.C.P. from
the Charlotte Geyer Foundation and the Lankenau Hospital Foun-
dation. W.P.M. receives additional financial support through NIH
grant GM087291. Shreekari Tadepalli is recognized for
10 min, and 100 mL of clarified supernatant was transferred to a new
flat-bottomed 96-well plate and mixed at equal volume with 2% (w/
v) p-dimethylaminobenzaldehyde in acetic acid. The yellow reac-
tion was measured at 490 nm using a Synergy HT microtiter plate
reader (Bio-Tek, Winooski,VT). Graphs of inhibition curves with
IC50 values were generated using Prism v.5.0 (GraphPad Software,
Inc.).

W.P. Malachowski et al. / European Journal of Medicinal Chemistry 108 (2016) 564e576
575
phenylhydrazine with microsomal cytochrome P-450. Catalysis of heme
modification, J. Biol. Chem. 257 (1982) 4404.
[24] P.R. Ortiz De Montellano, D.E. Kerr, Inactivation of catalase by phenyl-
hydrazine. Formation of a stable aryl-iron heme complex, J. Biol. Chem. 258
(1983) 10558e10563.
[25] D.P. Ringe, D.E. Kerr, P.R. Ortiz De Montellano, D.E. Kerr, Reaction of
myoglobin with phenylhydrazine: a molecular doorstop, Biochemistry 23
(1984) 2e4.
[26] P.R. Ortiz de Montellano, K.L. Kunze, Formation of N-phenylheme in the he-
molytic reaction of phenylhydrazine with hemoglobin, J. Am. Chem. Soc. 103
(1981) 6534e6536.
contributions to the characterization of several compounds. A
generous award (CHE-0958996) from the National Science Foun-
dation enabling acquisition of the 400 MHz NMR spectrometer
used in these studies is gratefully acknowledged. The authors
would also like to thank Bryn Mawr College for financial support of
this work. The authors are grateful to Richard Metz for providing
purified recombinant IDO1 enzyme.
Appendix A. Supplementary data
[27] B.R. Castro, Replacement of Alcoholic Hydroxyl Groups by Halogens and Other
Nucleophiles via Oxyphosphonium Intermediates, In Organic Reactions, John
Wiley & Sons, Inc., 2004.
[28] M. Macchia, N. Jannitti, G. Gervasi, R. Danesi, Geranylgeranyl diphosphate-
based inhibitors of post-translational geranylgeranylation of cellular pro-
teins, J. Med. Chem. 39 (1996) 1352e1356.
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.ejmech.2015.12.028.
[29] S. Kumar, D. Jaller, B. Patel, J.M. LaLonde, J.B. DuHadaway, W.P. Malachowski,
G.C. Prendergast, A.J. Muller, Structure based development of
phenylimidazole-derived inhibitors of indoleamine 2,3-dioxygenase, J. Med.
Chem. 51 (2008) 4968e4977.
References
[1] R.D. Schreiber, L.J. Old, M.J. Smyth, Cancer immunoediting: integrating
immunity's roles in cancer suppression and promotion, Science 331 (2011)
1565e1570.
[2] G.P. Dunn, L.J. Old, R.D. Schreiber, The immunobiology of cancer immuno-
surveillance and immunoediting, Immunity 21 (2004) 137e148.
[3] W. Zou, Immunosuppressive networks in the tumour environment and their
therapeutic relevance, Nat. Rev. Cancer 5 (2005) 263e274.
[4] A.J. Muller, P.A. Scherle, Targeting the mechanisms of tumoral immune
tolerance with small-molecule inhibitors, Nat. Rev. Cancer 6 (2006) 613e625.
[5] R. Metz, J.B. Duhadaway, U. Kamasani, L. Laury-Kleintop, A.J. Muller,
G.C. Prendergast, Novel tryptophan catabolic enzyme IDO2 is the preferred
biochemical target of the antitumor indoleamine 2,3-dioxygenase inhibitory
compound D-1-methyl-tryptophan, Cancer Res. 67 (2007) 7082e7087.
[6] H.J. Ball, H.J. Yuasa, C.J.D. Austin, S. Weiser, N.H. Hunt, Indoleamine 2,3-
dioxygenase-2; a new enzyme in the kynurenine pathway, Int. J. Biochem.
Cell Biol. 41 (2009) 467e471.
[7] A.J. Muller, G.C. Prendergast, Marrying immunotherapy with chemotherapy:
why say IDO? Cancer Res. 65 (2005) 8065e8068.
[8] D.H. Munn, A.L. Mellor, Indoleamine 2,3-dioxygenase and tumor-induced
tolerance, J. Clin. Investig. 117 (2007) 1147e1154.
[9] A.J. Muller, J.B. DuHadaway, P.S. Donover, E. Sutanto-Ward, G.C. Prendergast,
Inhibition of indoleamine 2,3-dioxygenase, an immunoregulatory target of the
cancer suppression gene Bin1, potentiates cancer chemotherapy, Nat. Med. 11
(2005) 312e319.
[10] ClinicalTrials.gov. National Institutes of Health, 2013; Vol. 2013; pp a list of
clinical trials.
[30] K. Matsuno, K. Takai, Y. Isaka, Y. Unno, M. Sato, O. Takikawa, A. Asai, S-ben-
zylisothiourea derivatives as small-molecule inhibitors of indoleamine-2,3-
dioxygenase, Bioorg. Med. Chem. Lett. 20 (2010) 5126e5129.
[31] Q. Huang, M. Zheng, S. Yang, C. Kuang, C. Yu, Q. Yang, Structureeactivity
relationship and enzyme kinetic studies on 4-aryl-1H-1,2,3-triazoles as
indoleamine 2,3-dioxygenase (IDO) inhibitors, Eur. J. Med. Chem. 46 (2011)
5680e5687.
€
[32] U.F. Rohrig, S.R. Majjigapu, A. Grosdidier, S. Bron, V. Stroobant, L. Pilotte,
D. Colau, P. Vogel, B.J. Van den Eynde, V. Zoete, O. Michielin, Rational design of
4-Aryl-1,2,3-triazoles for indoleamine 2,3-Dioxygenase 1 inhibition, J. Med.
Chem. 55 (2012) 5270e5290.
[33] M.-F. Cheng, M.-S. Hung, J.-S. Song, S.-Y. Lin, F.-Y. Liao, M.-H. Wu, W. Hsiao, C.-
L. Hsieh, J.-S. Wu, Y.-S. Chao, C. Shih, S.-Y. Wu, S.-H. Ueng, Discovery and
structureeactivity relationships of phenyl benzenesulfonylhydrazides as
novel indoleamine 2,3-dioxygenase inhibitors, Bioorg. Med. Chem. Lett. 24
(2014) 3403e3406.
[34] S. Tojo, T. Kohno, T. Tanaka, S. Kamioka, Y. Ota, T. Ishii, K. Kamimoto, S. Asano,
Y. Isobe, Crystal structures and structureeactivity relationships of imidazo-
thiazole derivatives as IDO1 inhibitors, ACS Med. Chem. Lett. (2014).
[35] S. Serra, L. Moineaux, C. Vancraeynest, B. Masereel, J. Wouters, L. Pochet,
ꢀ
ꢀ
R. Frederick, Thiosemicarbazide, a fragment with promising indolamine-2,3-
dioxygenase (IDO) inhibition properties, Eur. J. Med. Chem. 82 (2014) 96e105.
[36] C. Bissantz, B. Kuhn, M. Stahl, A medicinal chemist's guide to molecular in-
teractions, J. Med. Chem. 53 (2010) 5061e5084.
[37] T. Clark, M. Hennemann, J. Murray, P. Politzer, Halogen bonding: the s-hole,
J. Mol. Model 13 (2007) 291e296.
[11] Dolusic, E. F., Raphael Indoleamine 2,3-dioxygenase inhibitors: a patent re-
view (2008-2012). Expert Opinion on Therapeutic Patents 2013, 23,
1367e1381.
[38] J.P.M. Lommerse, A.J. Stone, R. Taylor, F.H. Allen, The nature and geometry of
intermolecular interactions between halogens and oxygen or nitrogen, J. Am.
Chem. Soc. 118 (1996) 3108e3116.
[39] R.B. Silverman, The Organic Chemistry of Drug Design and Drug Action,
Elsevier Academic Press, 2004.
€
[12] U.F. Rohrig, S.R. Majjigapu, P. Vogel, V. Zoete, O. Michielin, Challenges in the
discovery of indoleamine 2,3-dioxygenase 1 (IDO1) inhibitors, J. Med. Chem.
(2015).
[13] M. Sono, M.P. Roach, E.D. Coulter, J.H. Dawson, Heme-containing oxygenases,
Chem. Rev. 96 (1996) 2841e2888.
[14] N.P. Botting, Chemistry and neurochemistry of the kynurenine pathway of
tryptophan metabolism, Chem. Soc. Rev. 24 (1995) 401e412.
[15] M. Sono, O. Hayaishi, The reaction mechanism of indoleamine 2,3-
dioxygenase, Biochem. Rev. 50 (1980) 173e181.
[40] Ligand efficiency was calculated as follows: LE¼1.37*pKi/number of heavy
atoms.
[41] I.D. Kuntz, K. Chen, K.A. Sharp, P.A. Kollman, The maximal affinity of ligands,
Proc. Natl. Acad. Sci. 96 (1999) 9997e10002.
[42] A.L. Hopkins, C.R. Groom, A. Alex, Ligand efficiency: a useful metric for lead
selection, Drug Discov. Today 9 (2004) 430e431.
[43] A.L. Hopkins, G.M. Keseru, P.D. Leeson, D.C. Rees, C.H. Reynolds, The role of
ligand efficiency metrics in drug discovery, Nat. Rev. Drug Discov. 13 (2014)
105e121.
[44] H. Sugimoto, S. Oda, T. Otsuki, T. Hino, T. Yoshida, Y. Shiro, Crystal structure of
human indoleamine 2,3-dioxygenase: catalytic mechanism of O2 incorpora-
tion by a heme-containing dioxygenase, Proc. Natl. Acad. Sci. U. S. A. 103
(2006) 2611e2616.
[16] L.W. Chung, X. Li, H. Sugimoto, Y. Shiro, K. Morokuma, Density functional
theory study on a missing piece in understanding of heme chemistry: the
reaction mechanism for indoleamine 2,3-dioxygenase and tryptophan 2,3-
dioxygenase, J. Am. Chem. Soc. 130 (2008) 12299e12309.
[17] A. Lewis-Ballester, D. Batabyal, T. Egawa, C. Lu, Y. Lin, M.A. Marti, L. Capece,
D.A. Estrin, S.-R. Yeh, Evidence for a ferryl intermediate in a heme-based
dioxygenase, Proc. Natl. Acad. Sci. 106 (2009) 17371e17376.
[45] M. Sono, S.G. Cady, Enzyme kinetic and spectroscopic studies of inhibitor and
effector interactions with indoleamine 2,3-dioxygenase. 1. Norharman and 4-
phenylimidazole binding to the enzyme as inhibitors and heme ligands,
Biochemistry 28 (1989) 5392e5399.
[46] S. Yang, X. Li, F. Hu, Y. Li, Y. Yang, J. Yan, C. Kuang, Q. Yang, Discovery of
tryptanthrin derivatives as potent inhibitors of indoleamine 2,3-dioxygenase
with therapeutic activity in Lewis Lung Cancer (LLC) tumor-bearing mice,
J. Med. Chem. 56 (2013) 8321e8331.
[47] S. Kumar, W.P. Malachowski, J.B. DuHadaway, J.M. LaLonde, P.J. Carroll,
D. Jaller, R. Metz, G.C. Prendergast, A.J. Muller, Indoleamine 2,3-dioxygenase is
the anticancer target for a novel series of potent naphthoquinone-based in-
hibitors, J. Med. Chem. 51 (2008) 1706e1718.
[18] L. Capece, A. Lewis-Ballester, S.-R. Yeh, D.A. Estrin, M.A. Marti, Complete re-
action mechanism of indoleamine 2,3-dioxygenase as revealed by QM/MM
simulations, J. Phys. Chem. B 116 (2012) 1401e1413.
[19] E.W. Yue, B. Douty, B. Wayland, M. Bower, X. Liu, L. Leffet, Q. Wang,
K.J. Bowman, M.J. Hansbury, C. Liu, M. Wei, Y. Li, R. Wynn, T.C. Burn,
H.K. Koblish, J.S. Fridman, B. Metcalf, P.A. Scherle, A.P. Combs, Discovery of
potent competitive inhibitors of indoleamine 2,3-dioxygenase with in vivo
pharmacodynamic activity and efficacy in a mouse melanoma model, J. Med.
Chem. 52 (2009) 7364e7367.
[20] S.-P.S. Fung, H. Wang, P. Tomek, C.J. Squire, J.U. Flanagan, B.D. Palmer,
D.J.A. Bridewell, S.M. Tijono, J.F. Jamie, L.-M. Ching, Discovery and character-
isation of hydrazines as inhibitors of the immune suppressive enzyme,
[48] Incyte Corporation, A phase 1/2 randomized, blinded, placebo controlled
study of ipilimumab in combination with INCB024360 or placebo in subjects
with unresectable or metastatic melanoma, in: In: ClinicalTirals.Gov
[Internet], National Library of Medicine (US), Bethesda (MD), 2012. Available
indoleamine 2,3-dioxygenase
7595e7603.
1 (IDO1), Bioorg. Med. Chem. 21 (2013)
[21] Mautino, M. J. F.; Marcinowicz-Flick, A.; Kesharwani, T.; Waldo, J.: IDO in-
hibitors.; Organization, W. I. P., Ed., 2009; Vol. WO2009/073620.
[22] O. Augusto, KentL. Kunze, P.R. Ortiz de Montellano, N-phenylprotoporphyrin
IX formation in the hemoglobin-phenylhydrazin reaction, J. Biol. Chem. 257
(1982) 6231.
from:
http://clinicaltrials.gov/show/NCT01604889.
NLM
Identifier:
NCT01604889.
[49] NewLink Genetics Corporation, IDO inhibitor study for advanced solid tumors,
in: In: ClinicalTrials.Gov [Internet], National Library of Medicine (US),
[23] H.G. Jonen, J. Werringloer, R.A. Prough, R.W. Estabrook, The reaction of

576
W.P. Malachowski et al. / European Journal of Medicinal Chemistry 108 (2016) 564e576
Bethesda (MD), 2014. Vol. [cited 2014 May 30]; pp Available from: http://
clinicaltrials.gov/show/NCT02048709. NLM Identifier: NCT02048709.
intermolecular-interaction energies and geometries, J. Comput. Chem. 20
(1999) 740e774.
[50] T.K. Littlejohn, O. Takikawa, D. Skylas, J.F. Jamie, M.J. Walker, R.J. Truscott,
Expression and purification of recombinant human indoleamine 2, 3-dioxy-
genase, Protein Expr. Purif. 19 (2000) 22e29.
[55] J.M. Word, S.C. Lovell, J.S. Richardson, D.C. Richardson, Asparagine and
glutamine: using hydrogen atom contacts in the choice of side-chain amide
orientation, J. Mol. Biol. 285 (1999) 1735e1747.
[51] H.E. Flick, J.M. LaLonde, W.P. Malachowski, A.J. Muller, The tumor-selective
[56] S.C. Lovell, I.W. Davis, W.B. Arendall III, P.I.W. de Bakker, J.M. Word,
cytotoxic agent
b-lapachone is a potent inhibitor of IDO1, Int. J. Tryptophan
M.G. Prisant, J.S. Richardson, D.C. Richardson, Structure validation by C
a
ge-
Res. 6 (2013) 35e45.
ometry: 4,
437e450.
j and Cb deviation, Proteins Struct. Funct. Genet. 50 (2003)
[52] P. Skehan, R. Storeng, D. Scudiero, A. Monks, J. McMahon, D. Vistica,
J.T. Warren, H. Bokesch, S. Kenney, M.R. Boyd, New colorimetric cytotoxicity
assay for anticancer-drug screening, J. Natl. Cancer Inst. 82 (1990) 1107e1112.
[53] T.A. Halgren, MMFF VI. MMFF94s option for energy minimization studies,
J. Comput. Chem. 20 (1999) 720e729.
[57] J.W. Ponder, D.A. Case, Force fields for protein simulations, Adv. Protein Chem.
66 (2003) 27e85.
[58] S.B. Kirton, C.W. Murray, M.L. Verdonk, R.D. Taylor, Prediction of binding
modes for ligands in the cytochromes P450 and other heme-containing pro-
teins, Proteins 58 (2005) 836e844.
[54] T.A. Halgren, MMFF MMFF VII. Characterization of MMFF94, MMFF94s, and
other widely available force fields for conformational energies and for