Mathematical and Structural Characterization of Strong Nonadditive Structure-Activity Relationship Caused by Protein Conformational Changes

Article abstract of DOI:10.1021/acs.jmedchem.8b00713

In medicinal chemistry, accurate prediction of additivity-based structure-activity relationship (SAR) analysis rests on three assumptions: (1) a consistent binding pose of the central scaffold, (2) no interaction between the R group substituents, and (3) a relatively rigid binding pocket in which the R group substituents act independently. Previously, examples of nonadditive SAR have been documented in systems that deviate from the first two assumptions. Local protein structural change upon ligand binding, through induced fit or conformational selection, although a well-known phenomenon that invalidates the third assumption, has not been linked to nonadditive SAR conclusively. Here, for the first time, we present clear structural evidence that the formation of a hydrophobic pocket upon ligand binding in PDE2 catalytic site reduces the size of another distinct subpocket and contributes to strong nonadditive SAR between two otherwise distant R groups.

Full text of DOI:10.1021/acs.jmedchem.8b00713

Subscriber access provided by Gothenburg University Library
Article
Mathematical and Structural Characterization of Strong Non-
additive SAR Caused by Protein Conformational Changes
Laurent Gomez, Rui Xu, William Sinko, Brandon Selfridge, William Francois Vernier,
Kiev Ly, Richard Truong, MArkus Metz, Tami Marrone, Kristen Sebring, Yingzhuo Yan,
brent appleton, Kathleen Aertgeerts, eben massari, and James Guy Breitenbucher
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b00713 • Publication Date (Web): 02 Aug 2018
Downloaded from http://pubs.acs.org on August 2, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Mathematical and Structural Characterization of Strong Non-
additive SAR Caused by Protein Conformational Changes
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Laurent Gomez,* Rui Xu, William Sinko, Brandon Selfridge, William Vernier, Kiev Ly, Richard
Truong, Markus Metz, Tami Marrone, Kristen Sebring, Yingzhou Yan, Brent Appleton, Kathꢀ
leen Aertgeerts, Mark Eben Massari, J. Guy Breitenbucher.
Dart Neuroscience LLC, 12278 Scripps Summit Drive, San Diego, CA 92131, USA
ABSTRACT: In medicinal chemistry, accurate prediction of additivityꢀbased SAR analysis rests
on three assumptions: (1) a consistent binding pose of the central scaffold, (2) no interaction beꢀ
tween the R group substituents, and (3) a relatively rigid binding pocket in which the R group
substituents act independently. Previously, example of nonꢀadditive SAR have been documented
in systems that deviate from the first two assumptions. Local protein structural change upon ligꢀ
and binding, through induced fit or conformational selection, although a wellꢀknown phenomeꢀ
non that invalidates the third assumption, has not been linked to nonꢀadditive SAR conclusively.
Here, for the first time, we present clear structural evidence that the formation of a hydrophobic
pocket upon ligand binding in PDE2 catalytic site reduces the size of another distinct subꢀpocket,
and contributes to strong nonꢀadditive SAR between two otherwise distant R groups.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
INTRODUCTION
Rational small molecule drug discovery is based on the underlying assumption that the biological
properties of a molecule can be predictably altered by incorporating specific structural modificaꢀ
tions to that molecule (Structure Activity Relationships, SAR). The process of lead optimization
relies on a level of intuitive pattern recognition by the chemist in the form of SAR. Given the
significance of identifying predictable SAR patterns, it was inevitable that chemists would atꢀ
tempt to improve their intuition by mathematically quantifying the relationship between structure
and biological effect (QSAR). One of the first proposed quantitative models to exploit the addiꢀ
tivity of SAR relationships was published in 1964 by Free and Wilson, a model now referred to
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
as FreeꢀWilson Additivity. Although the FreeꢀWilson model of additivity was quite a simple
mathematical concept, the idea has had far reaching implications in drug discovery. Today the
medicinal chemistry literature is dominated by papers which optimize one R group, followed by
a second R group sequentially (Linear Analoging) rather than creating a matrix of compounds
altering multiple R groups simultaneously (Combinatorial Analoging). When chemists engage in
linear analoging they make the implicit assumption of additivity articulated in Free and Wilson’s
1
2
original work (i.e. the best R is the best regardless of the nature of R ). Additionally, this asꢀ
sumption underlies many chemical and physical descriptors that are calculated to correlate a
compound’s calculated physical properties and biological effect (MW is innately additive; most
2
methods of calculating tPSA and cLogP also rely on the assumption of additivity). In fact, the
value of nearly all QSAR models which seek to mathematically correlate a molecule’s calculated
physical properties with a biological effect would be greatly compromised if the system being
3
studied was found to be inconsistent with classical FreeꢀWilson additivity.
ACS Paragon Plus Environment

Page 3 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4
Furthermore, most if not all computational scoring functions including Free Energy Perturbation
+
5
Methods such as FEP , work best under the conditions of additivity. However, challenges such
as binding mode changes of the ligand and binding site changes are currently being actively inꢀ
6
ꢀ7
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
vestigated.
In 2006 we described a method in which the FreeꢀWilson additivity of a combinatorial matrix
8ꢀ9
could be used to evaluate structure activity relationships. The method we developed allowed
one to determine the overall FreeꢀWilson additivity within a combinatorial matrix of compounds.
In cases where nonꢀadditivity was verified, we reasonably hypothesized that this indicated a
change in binding mode between the ligands and the target protein, or suggested an interaction
between the varied substituents within the set of ligands. Although this hypothesis was intuitive,
we had no structural confirmation that nonꢀadditive behavior actually resulted in observable
binding mode changes. In fact, although a fundamental change in binding mode is often cited as
the cause of nonꢀadditive behavior in medicinal chemistry papers, there are actually very few
1
0
examples where this is rigorously confirmed with structural data.
In 2008 researchers from Janssen published an article whereby they used a similar method to
quantify the additivity intrinsic to small molecule interactions with a variety of biological targets.
The authors came to the conclusion that some targets were prone to display nonꢀadditive SAR
4
while others were more generally additive in their interactions. However, this work did not conꢀ
tain any structural data that could have provided useful rational and understanding for specific
observed nonꢀadditive behavior.
With increased knowledge such as available ligand SAR and structural data about the target in
question nonꢀadditive behavior can be studied, and in several cases it was explained by alternaꢀ
11ꢀ12
tive intraꢀmolecular substituent interactions or switched binding modes.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Recently, Kramer and coworkers published a very clever procedure for determining nonꢀadditive
13
SAR. They applied this method to publicly available SAR data. In this work Kramer was able
to explain nonꢀadditive data in terms of changes in ligand binding modes within the ligandꢀ
protein crystal structures in a few cases. In none of the 15 cases analyzed did the nonꢀadditivity
described involve a significant induced conformational change in the protein’s structure. Given
the large number of examples of nonꢀadditive behavior described in the medicinal chemistry litꢀ
erature, we suggest that a medicinal chemistry project should never be attempted without first
quantifying the degree of nonꢀadditivity intrinsic to the particular system being studied. Here we
employ Kramer’s method as means of quantitating the degree of nonꢀadditivity for any set of
compounds using PDE2 as a specific example system. We also demonstrate how obtaining crysꢀ
tal structure of ligands bound the PDE2 enzyme lead to the identification of putative protein conꢀ
formational changes responsible for the nonꢀadditive effects seen.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
RESULTS AND DISCUSSION
In this current work we demonstrate convincingly through the use of quantitative additivity analꢀ
ysis, that nonꢀadditivity was significant within the matrix studied. Computational tools and Xꢀ
ray crystallography were then used to demonstrate protein conformational differences among ligꢀ
and coꢀcomplexes which unambiguously explained the source of the nonꢀadditive behavior. Takꢀ
ing together, these findings constitute a case of nonꢀadditive SAR caused by an imposed strucꢀ
tural requirement on the protein.
ACS Paragon Plus Environment

Page 5 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The system investigated here uses a triazoloquinozaline scaffold developed by researchers at
14
Janssen R&D as PDE2 inhibitors. Each compound was synthesized according to the synthetic
route shown in Scheme 1. Condensation of alanine with 4ꢀfluoroꢀ3ꢀnitrobenzoic acid 1 under
basic conditions afforded the desired intermediate 2 in 60% yield. A subsequent reduction of the
nitro group followed by a spontaneous intramolecular cyclization yielded tetrahydroquinoxaline
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
. The intermediate was then oxidized to the desired dihydroquinoxaline 4 using hydrogen perꢀ
oxide under basic conditions. Esterification of the acid moiety was accomplished by treating
compound 4 with ethanol and sulfuric acid. The corresponding dihydroquinoxaline 5 was then
chlorinated under standard conditions, phosphorus oxychloride and N,Nꢀdimethylaniline, to afꢀ
ford the key intermediate 6 in moderate yields (17%) after column chromatography. The first set
of substitutions was introduced after condensation of five differently substituted 3ꢀ
chlorobenzhydrazides with intermediate 6 to provide all six triazoloquinoxaline intermediates, 7ꢀ
12, in moderate to good yields. Saponification of the ester functionality under basic conditions
followed by amide bond formation afforded the desired matrix of compounds, 19ꢀ48, which is
described in Figure 1. The design was intended to test the steric tolerance of the PDE2 active site
1
2
at the variable R and R positions in the matrix. It should be noted that a 3D energy minimized
1
2
model of the compounds suggests no possible direct interaction between R and R due to the
1
rigid nature of the triazoloquinozaline scaffold (Figure 1). A crystal structure of compound 1 (R
2
1
2
=
Me, R = H) confirmed this observation and suggests that R and R fit into two distinct, nonꢀ
overlapping protein pockets present in the catalytic site of PDE2 (Figure 2). The triazoloquinozaꢀ
line core is sandwiched in the πꢀclamp formed by Phe862, Phe830 and Ile826. The amide funcꢀ
1
tionality is situated above the πꢀclamp, positioning the R group in a solventꢀexposed region surꢀ
rounded by lipophilic residues such as Met847. The chlorophenyl group is buried deep inside the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
binding pocket, near the bound metals and orients R substitutions toward a waterꢀfilled channel
between bound metals and Leu770.
Scheme 1: Synthetic route
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
o
Reagents and conditions: a. alanine, H O, NaHCO , 95 C; b. Raney Ni, NaHCO , H O, 60 psi; c. i) 2N NaOH,
2
3
3
2
o
o
o
3
0% H O , 100 C; ii) HCl; d. EtOH, H SO , 80 C; e. POCl , N,Nꢀdimethyl aniline, 120 C; f. ArCONHNH , nꢀ
2 2 2 4 3 2
o
1
BuOH, 160 C; g. LiOH, THF/H O; h) DIEA, HATU, R NH , DMF.
2
2
Figure 1: Matrix of compounds and lowest energy conformation for compound 19
ACS Paragon Plus Environment

Page 7 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2: Crystal structure of compound 19 (PDB ID: 6C7E)
All compounds were tested for their inhibition of the PDE2 enzyme and reported as pIC and
5
0
run in quadruplicate with a mean standard error of all experiments of 0.10 (Range of errors =
1
0.16ꢀ0.01; Table 1). We initially plotted our data in a 3D bar graph format by varying the R subꢀ
2
stituent on the yꢀaxis and the R substituent on the xꢀaxis (Figure 3). Upon analysis, this matrix
represented a good data set for determining additivity due to measurable enzyme inhibition for
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
all compounds in the matrix, and an acceptable range of activities (pIC = 7.6ꢀ9.8; range = 2.2
5
0
log units) well above the mean standard error of the assay.
Figure 3. PDE2 pIC values of the compounds in the matrix.
50
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Simple visual inspection of the data depicted in Figure 3 suggests that there may be some nonꢀ
additive behavior in the matrix. To quantify the degree of nonꢀadditivity we utilized the doubleꢀ
1
3
transformation procedure described by Kramer. Using this method nonꢀadditivity can be quanꢀ
tified for any set of four compounds which undergo two substituent changes in two dimensions.
This is illustrated in Figure 4.
Figure 4. PDE2 pIC values of selected compounds in the matrix demonstrating a lack of
5
0
additivity.
ACS Paragon Plus Environment

Page 9 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
Starting with the root compound 19 in the back left (R = Me, R = H; pIC = 8.6) we can transꢀ
50
1
1
2
form R to get compound 21 (R = iBu, R = H; pIC = 9.5) which results in a pIC increase of
5
0
50
2
1
2
0
.9. Additionally, when we transform R providing compound 34 (R = Me, R = OiPr; pIC =
50
9.2) this results in a pIC increase of 0.6. However, when we perform both transformations to
5
0
1
2
get compound 36 (R = iBu, R = OiPr; pIC = 8.9) we get a compound which is significantly
5
0
less potent than we would expect if there was perfect additivity (Additive pIC = 8.6 + 0.9 + 0.6
5
0
= 10.1). Thus we quantitatively characterize this transformation as nonꢀadditive by 1.2 log units
10.1 – 8.9 = 1.2). Using the equation below allows for the quantification of nonꢀadditivity in
(
any doubleꢀtransformation of this type, as discussed by Kramer.
nonꢀadditivity = NAd = (pAct – pAct ) – (pAct – pAct )
3
4
2
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
In order to quantify the nonꢀadditive behavior of the entire combinatorial matrix, one can use the
equation above on all doubleꢀtransformations. Each compound is used as starting point for 20
1
2
doubleꢀtransformations considering R of 5 and R of 6 varied substituents. After removal of duꢀ
plicate cycles 300 NAd values are produced, and their frequencies as a function of nonꢀadditivity
are shown as histogram in Figure 5. As well, the experimentally observed and theoretical nonꢀ
additivity distribution with a mean standard error of 0.1 are shown. All nonꢀadditivity values
larger than +/ꢀ 0.52 cannot be explained with experimentally uncertainty. Thus we can determine
with 99% confidence that 54% of the double transformation data contained in the current matrix
is truly nonꢀadditive, and not the result of experimental uncertainty. However, we cannot judge
the NAd value < 0.52 as being innately additive, there is simply too much experimental variance
to make a judgement on these transformations.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 1. pIC data for all compounds described in the combinatorial matrix.
50
1
R
Compound number
and
pIC50
Me
iPr
iBu
tBu
Adamantyl
1
9
20
9.57
25
21
9.52
26
22
9.47
27
23
9.77
28
H
F
8.55
2
4
8
.69
29
.74
9.17
30
9.37
31
9.30
32
9.58
33
2
R
OMe
8
8.96
35
9.42
36
9.00
37
9.49
38
3
4
OiPr
9
.16
8.99
40
8.93
41
8.30
42
7.94
43
OiBu
39
ACS Paragon Plus Environment

Page 11 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
9
.61
9.46
45
9.29
46
8.75
47
7.74
48
OtBu
44
9.16
9.10
8.88
8.35
7.58
Figure 5. Theoretical non-additivity distribution expected from mean standard error of the
assay of 0.1 log units (red line) and real non-additivity distribution ( blue line). A histogram
for experimental data with y axis on the right side is also shown.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Once the nonꢀadditivity was quantitatively confirmed, we sought to understand the physical naꢀ
ture of the observed nonꢀadditivity within the set of compounds. Upon inspection of the double
transformation data the largest nonꢀadditive value obtained in the matrix arose from the transꢀ
1
2
formation; R = H to OiBu, and R = Me to adamantyl, resulting in an NAd = 3.1. Interestingly
this represented the largest change in molecular volume within the matrix resulting in a highly
significant change in expected activity. In addition, the second largest nonꢀadditive transforꢀ
mations also suggested that the size of the Rꢀgroups was playing a significant role in the nonꢀ
1
2
additive behavior; R = H to OtBu, and R = Me to Ad. Because of this, we rearranged the matrix
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
in order of the molecular volume of the substituents and plotted the NAd calculated for each
1
2
compound relative to the smallest compound in the matrix (19; R = Me, R = H) using Kramer’s
double transformation method (Figure 6).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6. Graphical representation of the non-additivity values for each compound relative
to compound 19, the dotted line at NAd = 0.5 represents the value that provides 95% confi-
dence that the non-additivity observed cannot be explained by the experimental uncertain-
ty.
Figure 6 clearly illustrates that there is a consistent nonꢀadditive relationship within the matrix,
and that the nonꢀadditivity is related to the size of the variable groups in the matrix. Seemingly,
one can increase the size of each of the Rꢀgroups independently and increase the activity of the
compounds, but if the size of both groups are increased together a drop in activity occurs.
ACS Paragon Plus Environment

Page 13 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
2
Since a direct interaction between R and R substituents is unlikely to be the source of the nonꢀ
additive behavior as we discussed previously, we decided to rely on crystallographic methods to
investigate the cause of this observed phenomenon. Therefore, we selected six candidates with
1
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
varying sizes of R and R , including all four compounds presented in Figure 4 for structural deꢀ
termination in the coꢀcomplex of PDE2 enzyme (see supporting information). The results indiꢀ
cate that all six compounds adopt a similar binding mode (RMSD values ≤ 1Å; see supporting
information) and rule out the possibility of ligand rearrangement in the binding site which has
traditionally been a common explanation for nonꢀadditivity. Analyses of our crystal structures
1
2
indicate that both R and R pockets can independently accommodate large hydrophobic funcꢀ
1
tional groups. For example, the adamantyl group which is the largest R in our tested matrix fits
1
in the R pocket and is surrounded by Leu770, Met847 and other hydrophobic residues at the enꢀ
trance of the catalytic site (see supporting information). These favorable interactions contribute
1
to about 16ꢀfold improvement for PDE2 when compared to analog 19 and 23 (R = Me and Ad
2
1
respectively with R = H, NAd = 1.2). More importantly and unlike the R pocket, significant
2
conformational changes are necessary to accommodate large R groups.
2
In ligand coꢀcomplexes where a large R group is presented to this part of the PDE2 enzyme,
protein rearrangement is observed which primarily involves the shift of Leu770 residue and open
2
a hydrophobic pocket near the bound metals. For example the compound with the largest R
1
2
group of the matrix (39; R = Me and R = OiBu) inserts into the newly formed pocket (see supꢀ
porting information) and improves PDE2 inhibition by 7.9ꢀfold when comparing analogs 19 and
2
1
3
9 (R = H and OiBu respectively with R = Me, NAd = 0.9). Similar structural changes have
1
5a
previously been reported for PDE2 in complex with inhibitor BAY60ꢀ7550 as well as with
15b
similar PDE2 analogs described herein.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
This “leucine flip” is mostly localized in the side chain of Leu770. Similar side chain flexibility
as well as larger scale conformational changes concomitant with ligand binding have been well
1
6
documented in the literature. Protein structural changes upon ligand binding can originate from
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
either conformational selection of one of several preꢀexisting conformations or an induced fit of
17
the protein by the ligand.
Comparing the structures of compounds 23 and 39 highlights the structural basis for the observed
1
2
nonꢀadditivity. Favorable binding of large R and R substituents requires close interaction with
residue Leu770, which can adapt different side chain conformations (Figure 7). In this case,
1
2
Leu770 works as a molecular divider that distributes ligand binding volume between R and R
1
2
pockets. A large R group inevitably results in a small R pocket, and vice versa. For this series
1
2
of compounds, size increase of both R and R groups leads to steric clashes around Leu770 and
deviation from prediction by perfect additivity. A threeꢀdimensional representation of the nonꢀ
additive SAR as a function of Rꢀgroups total surface area is shown in Figure 8. The x and y axes
1
2
represent the total surface area (tSA) for R and R respectively plotted against the deviation
1
from additivity (Z axis; NAd). As can be seen, the largest analog 43 (R = Ad, tSA = 312.9 Å and
2
R = OiBu, tSA = 244.9) lead to the largest NAd of 3.1. The structural determination of comꢀ
2
pound 43 reveals an isobutyl group buried in the R hydrophobic pocket and a highly disordered
1
adamantyl group in the R solventꢀexposed pocket that displays ambiguous electron density and
therefore cannot be modelled confidently. We reasoned that this is due to the steric clash between
2
the adamantyl group and Leu770 when the R hydrophobic pocket is formed. Compound 43
shows the largest difference between the actual PDE2 pIC and that predicted by perfect addiꢀ
5
0
tivity. Thus, the nonꢀadditivity we observed can be explained by an indirect, targetꢀmediated inꢀ
1
2
teraction between the size R and R groups and the two pockets present in the enzyme.
ACS Paragon Plus Environment

Page 15 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Figure 7: Overlay of compounds 23 (in cyan, PDB code: 6C7D) and 39 (in orange, PDB
code: 6C7F) bound to PDE2 enzyme.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
Figure 8: Correlation between R and R surface areas and deviation from additive SAR
(NAd)
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
CONCLUSIONS
Given the importance and complexity of ligandꢀprotein interactions in drug discovery, medicinal
chemists continuously rely on structure activity relationships to rationally design molecules and
improve their binding properties. It is therefore crucial that SAR analyses are carefully conductꢀ
ed in order to properly and accurately determine the optimization strategy. In this study, we have
demonstrated that a process that includes a mathematical analysis of nonꢀadditivity coupled with
proteinꢀligand coꢀcomplex structural information reveals unprecedented protein conformational
change as the cause of the nonꢀadditive SAR. Our study shows that optimal inhibition of PDE2
1
2
can be achieved by balancing the size of either R or R substituents. Our results indicate howevꢀ
er that the optimization of these two vectors will diverge due to nonꢀadditive SAR and therefore
ACS Paragon Plus Environment

Page 17 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
no assumptions should be made. This also represents a challenge for computational scientists to
predict the binding energies as conformational changes in the binding site introduce an additional
source of error. In general free energy calculations seem to work better if only the ligand is modiꢀ
fied in the binding site.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
We expect that the approach and the findings discussed herein should be generally applicable to
other biological targets. In effect, these results give medicinal chemists the ability to analyze and
rationalize nonꢀadditive SAR which unlocks the potential of more efficiently improving ligandꢀ
protein interactions.
EXPERIMENTAL SECTION
General Methods. Solvents and reagents were purchased and used without further purification
unless otherwise noted. Flash column chromatography (FCC) was performed on Teledyne Isco
CombiFlash® instruments equipped with silica gel columns. Preparative HPLC purification was
performed on a Waters Preparative LCMS system equipped with a Waters XBridge Prep C18 5
µm OBD 19x50 mm column. Elution was achieved with increasing concentration of acetonitrile
in water and 0.05% TFA added as modifier. Analytical HPLC was performed on an Acquity
UPLC system equipped with a BEH C18 1.7 ꢀM 2.1 x 50 mm column. Elution was achieved
with increasing concentration of acetonitrile in water and 0.1% formic acid added as modifier.
NMR was performed on a Bruker Ultrashield 400 Plus spectrometer. Chemical shifts (δ) are reꢀ
1
13
ported in ppm relative to the residual solvent signal. Data for H/ C NMR spectra are reported
as follows: chemical shift (multiplicity, coupling constants, number of hydrogens). Abbreviations
are as follows: s (singlet), d (doublet), t (triplet), q (quartet), m (multiplet), br (broad).The purity
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
of all compounds tested was determined by LCMS and H NMR to be >95%. Highꢀresolution
mass spectral data were obtained from the University of California, San Diego.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
-((1-Carboxyethyl)amino)-3-nitrobenzoic acid (2). To a stirring solution of 4ꢀfluoroꢀ3ꢀ
nitrobenzoic acid (1; 50.0 g, 270 mmol, 1.0 eq.) in H O (500 mL) was added Lꢀalanine (48 g,
2
5
40 mmol, 2.0 eq.) followed by NaHCO (68 g, 810 mmol, 3.0 eq.). The reaction mixture was
3
o
o
heated to 100 C. After heating the reaction mixture for 36 h at 100 C, the solution was cooled
to room temperature and acidified with 6N HCl. The resultant precipitate was filtered and dried
1
under vacuum to afford the title compound (2; 41 g, 60% yield) as a light red solid. H NMR
(300 MHz, DMSOꢀd ) δ 13.16 (brs, 2H), 8.70 (d, J = 6.9 Hz, 1H), 8.64 (d, J = 1.8 Hz, 1H), 7.99
6
(
dd, J = 9.0, 2.1 Hz, 1H), 7.09 (d, J = 9.3 Hz, 1H), 4.62–4.53 (m, 1H), 1.50 (d, J = 6.9 Hz, 3H).
+
MS(ESI): m/z 255.16 [M+H] .
2
-Methyl-3-oxo-3,4-dihydroquinoxaline-6-carboxylic acid (4). To a stirring solution of 4ꢀ((1ꢀ
carboxyethyl)amino)ꢀ3ꢀnitrobenzoic acid (2; 41.0 g, 161 mmol, 1 eq.) in H O (400 mL) in a
2
bomb apparatus was added Raney Ni (30 g wet) followed by NaHCO (40.7 g, 484 mmol, 3.0
3
eq.). The apparatus was evacuated and refilled with N (x 2) before charging with H at 60 psi.
2
2
After stirring under a H atmosphere at 60 psi for 60h, the reaction mixture was filtered through a
2
pad of celite and and concentrated in vacuo to afford 2ꢀmethylꢀ3ꢀoxoꢀ3,4ꢀdihydroquinoxalineꢀ6ꢀ
carboxylic acid (3; 40.0 g) crude which was carried on without further purification. MS(ESI):
+
m/z 207.04 [M+H] .
ACS Paragon Plus Environment

Page 19 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The above crude (3; 40.0 g, 194 mmol theoretical, 1.0 eq.) was added to a stirring solution of
sodium hydroxide (38.8 g, 5.0 eq.) in H O (400 mL) and 30% H O solution (40 mL) at 0 °C.
2
2
2
The reaction mixture was heated to 100 °C. After stirring at 100°C for 16h, the reaction mixture
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
was cooled to 0 °C and acidified with 6N HCl. The precipitated solid was filtered, dried and tritꢀ
1
urated with Et O (40 mL) to afford the title compound (4; 20.0 g, 61% yield overall yield). H
2
NMR (300 MHz, DMSOꢀd ) δ 12.46 (brs, 1H), 7.89 (s, 1H), 7.77 (brs, 2H), 3.86 (brs, 1H), 2.43
6
+
(
s, 3H). MS(ESI): m/z 205.09 [M+H] .
Ethyl 2-methyl-3-oxo-3,4-dihydroquinoxaline-6-carboxylate (5). To a stirring solution of 2ꢀ
methylꢀ3ꢀoxoꢀ3,4ꢀdihydroquinoxalineꢀ6ꢀcarboxylic acid (4; 20.0 g, 86.1 mmol, 1.0 eq.) in EtOH
o
(
400 mL) at 0 C was added H SO (10 mL). The reaction mixture was warmed to 80 °C. After
2
4
o
stirring at 80 C for 16h, the reaction mixture was cooled to room temperature, and concentrated
in vacuo. The crude material was basified with NaHCO (aq.) and extracted with EtOAc (3 X
3
2
50 mL). The combined organics were dried over Na SO , filtered, and concentrated in vacuo to
2 4
afford the title compound (5; 13.5 g, 68% yield) as a pink solid which was carried on without any
1
further purification. H NMR (400 MHz, DMSOꢀd6) δ 12.42 (brs, 1H), 7.89 (s, 1H), 7.78 (brs,
2
H), 4.35 (q, J = 7.6 Hz, 2H), 2.43 (s, 3H), 1.34 (t, J = 7.2 Hz, 3H). MS(ESI): m/z 233.15
+
[M+H] .
Ethyl 3-chloro-2-methylquinoxaline-6-carboxylate (6). To a solution of ethyl 2ꢀmethylꢀ3ꢀoxoꢀ
3
,4ꢀdihydroquinoxalineꢀ6ꢀcarboxylate (5; 13.5 g, 58.1 mmol, 1.0 eq.) in N,Nꢀdimethyl aniline
(
74 mL) was added POCl (54.0 mL, 581 mmol, 10 eq.) dropwise. The resulting mixture was
3
stirred at room temperature for 1 hours then heated to 120 °C. After heating the reaction mixture
for 1h at 120 °C, the solution was cooled to 0 °C, diluted with NaHSO (aq.), and extracted with
4
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
EtOAc (3 X 250 mL). The combined organics were dried over Na SO , filtered, and concentrated
2
4
in vacuo to obtain a crude solid. The crude solid was purified by neutral alumina gel chromatogꢀ
raphy (2% DCM in petroleum ether) to afford the title compound (6; 2.5 g, 17%) as an offꢀwhite
1
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
solid. H NMR (400 MHz, CDCl ) δ 8.70 (d, J = 2.0 Hz, 1H), 8.35 (dd, J = 8.8, 2.0 Hz, 1H),
3
8
.06 (d, J = 8.8 Hz, 1H), 4.46 (q, J = 6.8 Hz, 2H), 2.87 (s, 3H), 1.45 (t, J = 6.8 Hz, 3H).
+
MS(ESI): m/z 251.04 [M+H] .
Ethyl 1-(2-chlorophenyl)-4-methyl-[1,2,4]triazolo[4,3-a]quinoxaline-8-carboxylate (7). To a
stirring mixture of ethyl 3ꢀchloroꢀ2ꢀmethylquinoxalineꢀ6ꢀcarboxylate (6; 3.0 g, 12 mmol, 1.0 eq,)
in nꢀBuOH (45 mL) in a sealed tube was added 3ꢀchlorobenzhydrazide (3.04 g, 17.9 mmol, 1.5
eq). The tube was sealed and heated to 160 °C. After heating at 160 °C for 3h, the reaction mixꢀ
ture was cooled to room temperature and diluted with EtOAc (50 mL) and H O (50 mL). The
2
organic layer was separated and concentrated in vacuo to afford a crude solid. This crude mixture
was purified by reversed phase HPLC (MeCN and 0.1% aqueous NH OAc as eluent) to afford 7
4
1
(
0.800 g, 18% yield). H NMR (300 MHz, CDCl ) δ 8.21 (d, J = 8.4 Hz, 1H), 8.09 (d, J = 8.4 Hz,
3
1
H), 7.96 (s, 1H), 7.70 (m, 3H), 7.60 (m, 1H), 4.27 (q, J = 6.9 Hz, 2H), 3.12 (s, 3H), 1.28 (s, J =
+
6.9 Hz, 3H). MS(ESI): m/z 367.23 [M + H] .
Ethyl 1-(2-chloro-5-fluorophenyl)-4-methyl-[1,2,4]triazolo[4,3-a]quinoxaline-8-carboxylate
(8).
ACS Paragon Plus Environment

Page 21 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Preparation of 2-chloro-5-fluorobenzohydrazide (48). To a stirring solution of methyl 2ꢀ
chloroꢀ5ꢀfluorobenzoate (1.0 g, 5.3 mmol, 1.0 eq) in EtOH (10 mL) was added hydrazine hydrate
(2.8 g, 53 mmol, 10.0 eq). The reaction mixture was heated to 50 °C for 1 h. The reaction mixꢀ
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ture was cooled to 0 °C and triturated with Et O (20 mL). The resultant precipitate was filtered
2
1
to afford 2ꢀchloroꢀ5ꢀfluorobenzohydrazide (48; 0.800 g, 80% yield) as white solid. H NMR
(300 MHz, DMSOꢀd6) δ 9.5 (brs, 1H), 7.38 (dd, J = 8.7, 4.8 Hz, 1H), 7.36ꢀ7.26 (m, 2H).
+
MS(ESI): m/z 189.06 [M + H] .
To a stirring solution of ethyl 3ꢀchloroꢀ2ꢀmethylquinoxalineꢀ6ꢀcarboxylate (6; 3.0 g, 12.7 mmol,
1
.0 eq.) in nꢀBuOH (45 mL) in a microwave vial was added hydrazide 48 (3.1 g, 16.5 mmol, 1.3
eq.). The vial was capped and heated at 160 °C in the microwave for 1 h. After heating at 160
C in the microwave for 1 h, the reaction mixture was cooled to room temperature and concenꢀ
°
trated in vacuo to afford crude compound. The crude compound was purified by reverse phase
column chromatography (MeCN and 0.1% aqueous NH OAc as eluent) to afford 6 (1.0 g, 21%
4
1
yield). H NMR (400 MHz, CDCl ) δ 8.21 (dd, J = 8.2, 1.6 Hz, 1H), 8.11 (d, J = 8.8 Hz, 1H),
3
7
.98 (d, J = 1.6 Hz, 1H), 7.67 (dd, J = 4.8, 8.8 Hz, 1H), 7.47 (m, 2H), 4.29 (q, J = 7.6 Hz, 2H),
+
3
.12 (s, 3H), 1.3 (t, J = 7.6 Hz, 3H). MS(ESI): m/z 370.92 [M + H] .
Ethyl
1-(2-chloro-5-methoxyphenyl)-4-methyl-[1,2,4]triazolo[4,3-a]quinoxaline-8-
carboxylate (9).
Preparation of 2-chloro-5-methoxybenzohydrazide (49). To a stirring solution of 2ꢀchloroꢀ5ꢀ
hydroxybenzoic acid (5.0 g, 29 mmol, 1.0 eq) in DMF (50 mL) at 0 °C was added K CO (20.0
2
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
g, 145 mmol, 5.0 eq) followed by methyl iodide (21.7 mL, 348 mmol, 12.0 eq). The reaction
mixture was heated to 40 °C. After heating at 40 °C for 12 h, the reaction mixture was cooled to
room temperature and diluted with preꢀcooled H O (50 mL). The aqueous layer was extracted
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
with ethyl acetate (3 X 50 mL). The combined organics were washed with H O (50 mL), dried
2
over Na SO , filtered, and concentrated in vacuo to afford methyl 2ꢀchloroꢀ5ꢀmethoxybenzoate
2
4
1
(
5.2 g, 89% yield) as a liquid that was used without further purification. H NMR (300 MHz,
CDCl ) δ 7.38ꢀ7.32 (m, 2H), 6.96 (dd, J = 9.0, 3.0 Hz, 1H), 3.93 (s, 3H), 3.82 (s, 3H). MS(ESI):
3
m/z 201.19 [M + H] +. To a solution of 2ꢀchloroꢀ5ꢀmethoxybenzoate (5.0 g, 24.9 mmol, 1.0 eq.)
in EtOH (50 mL) was added hydrazine hydrate (12.5 mL, 249 mmol, 10.0 eq). The reaction
mixture was heated to 50 °C. After heating for 3 h at 50°C, the reaction mixture was cooled to 0
°
C and triturated with Et O (100 mL). The formed precipitate was filtered to afford the title
2
compound (49; 5.2 g, 84% purity by LCMS) as a white solid that was used without further puriꢀ
1
fication. H NMR (300 MHz, DMSOꢀd6) δ 9.58 (brs, 1H), 7.38 (d, J = 8.7 Hz, 1H), 7.01 (dd, J =
+
9
.0, 3.3 Hz, 1H), 6.93 (d, J = 3.0 Hz, 1H), 3.77 (s, 3H). MS(ESI): m/z 201.06 [M + H] .
To a stirring solution of ethyl 3ꢀchloroꢀ2ꢀmethylquinoxalineꢀ6ꢀcarboxylate (6; 0.75 g, 3.0 mmol,
.0 eq) in nꢀBuOH (11.3 mL) in a microwave vial was added 2ꢀchloroꢀ5ꢀ
1
methoxybenzohydrazide (49; 0.78 g, 3.9 mmol 1.3 eq). The microwave vial was capped and
heated to 160 °C in the microwave. After heating at 160 °C in the microwave for 0.5 h, the reacꢀ
tion mixture was cooled to room temperature and concentrated in vacuo to afford crude solid.
The crude compound was purified by reverse phase column chromatography (MeCN and 0.1%
aqueous NH4OAc as eluent) to afford ethyl 1ꢀ(2ꢀchloroꢀ5ꢀmethoxyphenyl)ꢀ4ꢀmethylꢀ
1
[
1,2,4]triazolo[4,3ꢀa]quinoxalineꢀ8ꢀcarboxylate (9; 0.35 g, 29% yield). H NMR (400 MHz,
CDCl3) δ 8.22 (dd, J = 8.4, 1.6 Hz, 1H), 8.10 (d, J = 12.8 Hz, 1H), 8.05 (d, J = 1.6 Hz, 1H), 7.56
ACS Paragon Plus Environment

Page 23 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
dd, J = 8.0, 1.6 Hz, 1H), 7.21 (m, 2H), 4.29 (q, J = 7.6 Hz, 2H), 3.87 (s, 3H), 3.11 (s, 3H), 1.3 (t,
+
J = 7.6 Hz, 3H). MS(ESI): m/z 397.2 [M + H] .
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Ethyl
1-(2-chloro-5-isopropoxyphenyl)-4-methyl-[1,2,4]triazolo[4,3-a]quinoxaline-8-
carboxylate (10).
Preparation of 2-chloro-5-isopropoxybenzohydrazide (50). To a stirring solution of 2ꢀchloroꢀ
5
ꢀhydroxybenzoic acid (7.0 g, 41 mmol, 1.0 eq.) in DMF (70 mL) at 0 °C was added K CO3
2
(28.0 g, 203 mmol, 5.0 eq.) and isopropyl iodide (20.3 mL, 203 mmol, 5.0 eq.). The reaction
mixture was heated to 80 °C. After stirring at 80 °C for 12 h, the reaction mixture was cooled to
room temperature, diluted with preꢀcooled H O, and extracted with EtOAc (3 X 70 mL). The
2
combined organics were washed with H O (70 mL), dried over Na SO , filtered, and concentratꢀ
2
2
4
ed in vacuo to afford isopropyl 2ꢀchloroꢀ5ꢀisopropoxybenzoate (8.1 g, 77% yield) as a liquid that
1
was used without further purification. H NMR (300 MHz, CDCl ) δ 7.32–7.26 (m, 2H), 6.91
3
(dd, J = 9.3, 3.3 Hz, 1H), 5.28–5.24 (m, 1H), 4.58–4.49 (m, 1H), 1.38 (d, J = 6.0 Hz, 6H), 1.33
+
(d, J = 6.3 Hz, 6H). MS(ESI): m/z 256.12 [M+H] . To a stirring solution of isopropyl 2ꢀchloroꢀ
5
ꢀisopropoxybenzoate (8.0 g, 31 mmol, 1.0 eq.) in EtOH (80 mL) was added hydrazine hydrate
(15.6 mL, 310 mmol, 10.0 eq.). The reaction mixture was heated to 80 °C. After heating at 80 °C
for 12h, the reaction mixture was cooled to room temperature and concentrated in vacuo to obꢀ
tain a crude solid. The crude solid was purified by silica gel chromatography (3% MeOH in
DCM) to afford 2ꢀchloroꢀ5ꢀisopropoxybenzohydrazide (50; 7.1 g, 100%) as an offꢀwhite solid.
1
H NMR (300 MHz, DMSOꢀd6) δ 9.52 (brs, 1H), 7.35 (d, J = 9.0 Hz, 1H), 6.98 (dd, J = 9.0, 3.0
Hz, 1H), 6.88 (d, J = 3.0 Hz, 1H), 4.66–4.58 (m, 1H), 4.45 (brs, 2H), 1.26 (d, J = 6.0 Hz, 6H).
MS(ESI): m/z 229.03 [M+H]+.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
To a stirring solution of ethyl 3ꢀchloroꢀ2ꢀmethylquinoxalineꢀ6ꢀcarboxylate (6; 0.75 g, 3.0 mmol,
1
.0 eq) in nꢀBuOH (11.3 mL) in a microwave vial was added 2ꢀchloroꢀ5ꢀ
isopropoxybenzohydrazide (50; 0.89 g, 3.9 mmol, 1.3 eq). The vial was capped and the reaction
mixture was heated to 160 °C in the microwave. After heating for 30 min at 160 °C in the miꢀ
crowave, the reaction mixture was cooled to room temperature and concentrated in vacuo to obꢀ
tain a crude solid. The crude solid was purified by reverse phase HPLC (Reveleris Cꢀ18, 80%
MeCN in 0.1% NH4OAc) to afford the title compound (10; 0.50 g, 39% yield) as an offꢀwhite
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
solid. H NMR (300 MHz, DMSOꢀd6) δ 8.15 (m, 2H), 7.85 (brs, 1H), 7.73 (d, J = 8.4 Hz, 1H),
7
.41–7.35 (m, 2H), 4.72–4.64 (m, 1H), 4.21 (q, J = 6.6 Hz, 2H), 2.98 (s, 3H), 1.31 (d, J = 6.3 Hz,
+
6H), 1.24 (t, J = 6.3 Hz, 3H). MS(ESI): m/z 425.16 [M+H] .
Ethyl
1-(2-chloro-5-isobutoxyphenyl)-4-methyl-[1,2,4]triazolo[4,3-a]quinoxaline-8-
carboxylate (11).
Preparation of 2-chloro-5-isobutoxybenzohydrazide (51). To a stirring solution of 2ꢀchloroꢀ5ꢀ
hydroxybenzoic acid (7.0 g, 41 mmol, 1.0 eq.) in DMF (70 mL) at 0 °C was added K CO (28.0
2
3
g, 203 mmol, 5.0 eq.) and 1ꢀbromoꢀ2ꢀmethylpropane (21.7 mL, 203 mmol, 5.0 eq.). The reacꢀ
tion mixture was heated to 80 °C. After heating at 80 °C for 12h, the reaction mixture was
cooled to room temperature, diluted with precooled H O, and extracted with EtOAc (3 x 70 mL).
2
The combined organics were washed with H O (70 mL), dried over Na SO , filtered, and conꢀ
2
2
4
centrated in vacuo to afford isobutyl 2ꢀchloroꢀ5ꢀisobutoxybenzoate (10.0 g, 35.1 mmol) as a liqꢀ
1
uid that was used without further purification. H NMR (400 MHz, CDCl3) δ 7.33 (s, 1H), 7.31
(d, J = 6.0 Hz, 1H), 6.94 (dd, J = 8.8, 3.2 Hz, 1H), 4.12 (d, J = 6.4 Hz, 2H), 3.72 (d, J = 6.8 Hz,
2
H), 2.11 – 2.06 (m, 2H), 1.03 (d, J = 2.4 Hz, 6H), 1.02 (d, J = 2.4 Hz, 6H). To a stirring solution
ACS Paragon Plus Environment

Page 25 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of isobutyl 2ꢀchloroꢀ5ꢀisobutoxybenzoate (10.0 g, 35.1 mmol, 1.0 eq.) in EtOH (100 mL) was
added hydrazine hydrate (17.6 mL, 350 mmol, 10.0 eq). The reaction mixture was heated to 80
°C. After heating at 80 °C for 12 h, the reaction mixture was cooled to room temperature and
concentrated in vacuo to obtain a crude solid. The crude solid was purified by silica gel chromaꢀ
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
tography (3% MeOH in DCM) to afford 2ꢀchloroꢀ5ꢀisobutoxybenzohydrazide (51; 8.1 g, 95%
1
yield) as an offꢀwhite solid. H NMR (300 MHz, DMSOꢀd ) δ 9.52 (brs, 1H), 7.36 (d, J = 9.0
6
Hz, 1H), 7.00 (dd, J = 9.0, 3.0 Hz, 1H), 6.91 (d, J = 3.0 Hz, 1H), 4.45 (brs, 2H), 3.75 (d, J = 6.6
+
Hz, 2H), 2.05–1.96 (m, 1H), 0.96 (d, J = 6.6 Hz, 6H). MS(ESI): m/z 243.08 [M+H] .
To a stirring solution of ethyl 3ꢀchloroꢀ2ꢀmethylquinoxalineꢀ6ꢀcarboxylate (6; 0.75 g, 3.0 mmol,
1
.0 eq) in nꢀBuOH (11.3 mL) was added 2ꢀchloroꢀ5ꢀisobutoxybenzohydrazide (51; 0.94 g, 3.87
mmol, 1.3 eq.) in a microwave vial. The vial was capped and heated to 160 °C in the microwave.
After heating at 160 °C for 30 min in the microwave, the reaction mixture was cooled to room
temperature and concentrated in vacuo to obtain a crude solid. The crude solid was purified by
reversed phase chromatography (Reveleris Cꢀ18, 80% MeCN in 0.1% NH OAc(aq.)) to afford
4
ethyl 1ꢀ(2ꢀchloroꢀ5ꢀisobutoxyphenyl)ꢀ4ꢀmethylꢀ[1,2,4]triazolo[4,3ꢀa]quinoxalineꢀ8ꢀcarboxylate
1
(
11; 0.55 g, 42% yield) as an offꢀwhite solid. H NMR (400 MHz, DMSOꢀd ) δ 8.15–8.14 (m,
6
2
H), 7.85 (brs, 1H), 7.74 (d, J = 8.8 Hz, 1H), 7.43–7.39 (m, 2H), 4.21 (q, J = 7.2 Hz, 2H), 3.82 (t,
J = 6.4 Hz, 2H), 2.98 (s, 3H), 2.05–2.02 (m, 1H), 1.22 (t, J = 6.8 Hz, 3H), 0.97 (dd, J = 6.4, 2.8
+
Hz, 6H). MS(ESI): m/z 438.8 [M+H] .
Ethyl
1-(5-(tert-butoxy)-2-chlorophenyl)-4-methyl-[1,2,4]triazolo[4,3-a]quinoxaline-8-
carboxylate (12).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Preparation of 5-(tert-butoxy)-2-chlorobenzohydrazide (52). To a stirred solution of methyl
2
ꢀchloroꢀ5ꢀhydroxybenzoate (5.0 g, 26.8 mmol, 1.0 eq) in toluene (50 mL) was added N,Nꢀ
dimethylformamide diꢀtertꢀbutyl acetal (12.8 mL, 53.6 mmol, 2.0 eq). The reaction mixture was
heated to 110 °C. After heating at 110 °C for 16 h, the reaction mixture was cooled to room temꢀ
perature and additional N,Nꢀdimethylformamide diꢀtertꢀbutyl acetal (6.4 mL, 26.8 mmol, 1.0 eq)
was added. The reaction mixture was heated back to 110 °C. After heating an additional 16h at
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
110°C, the reaction was cooled to room temperature, and concentrated in vacuo to obtain a crude
solid. The crude solid was purified by silica gel chromatography (10% EtOAc in hexanes) to
afford methyl 5ꢀ(tertꢀbutoxy)ꢀ2ꢀchlorobenzoate (2.2 g, 34% yield) as a colorless liquid and reꢀ
1
covered starting material (2.5 g, 50%). H NMR (300 MHz, DMSOꢀd ) δ 7.47 (d, J = 8.7 Hz,
6
1
H), 7.35 (d, J = 2.7 Hz, 1H), 7.20 (dd, J = 9.0, 3.0 Hz, 1H), 3.85 (s, 3H), 1.31 (s, 9H). MS(ESI):
+
m/z 243.10 [M+H] . To a stirring solution of methyl 5ꢀ(tertꢀbutoxy)ꢀ2ꢀchlorobenzoate (3.30 g,
13.6 mmol, 1.0 eq.) in EtOH (66 mL) was added hydrazine hydrate (11.3 mL, 136 mmol, 10.0
eq). The reaction mixture was heated to 80 °C. After heating at 80 °C for 16 h, the reaction mixꢀ
ture was cooled to room temperature and concentrated in vacuo to obtain crude solid. The crude
solid was purified by silica gel chromatography (3% EtOH in EtOAc) to afford the title comꢀ
1
pound (52; 2.4 g, 73% yield) as an offꢀwhite solid. H NMR (400 MHz, DMSOꢀd ) δ 9.54 (brs,
6
1
2
H), 7.37 (d, J = 8.4 Hz, 1H), 7.04 (dd, J = 8.8, 3.2 Hz, 1H), 6.93 (d, J = 2.8 Hz, 1H), 4.47 (brs,
+
H), 1.30 (s, 9H). MS(ESI): m/z 243.12 [M+H] .
To a stirring solution of ethyl 3ꢀchloroꢀ2ꢀmethylquinoxalineꢀ6ꢀcarboxylate (6; 0.12 g, 0.48
mmol, 1.0 eq.) in 2,6ꢀlutidine (0.6 mL) was added 5ꢀ(tertꢀbutoxy)ꢀ2ꢀchlorobenzohydrazide (52;
0
.12 g, 0.48 mmol, 1.0 eq.) The reaction mixture was concentrated in vacuo to afford a crude
solid. The crude solid was purified by silica gel chromatography (30% EtOAc in heptanes) to
ACS Paragon Plus Environment

Page 27 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
afford the title compound (12; 0.15 g, 71% yield). H NMR (400MHz, CDCl ) δ 8.28 (dd, J =
3
1
.6, 8.4 Hz, 1H), 8.17 (d, J = 8.4 Hz, 1H), 8.05 ꢀ 7.99 (m, 1H), 7.63 ꢀ 7.56 (m, 1H), 7.38 ꢀ 7.31
13
(m, 2H), 3.16 (s, 3H), 1.43 (s, 9H). C NMR (101MHz, CDCl ) δ 168.8, 155.6, 155.1, 147.4,
3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
44.7, 139.5, 130.8, 130.1, 129.1, 128.8, 128.7, 127.6,127.4, 125.4, 117.5, 80.3, 29.7, 28.8, 21.5.
+
MS(ESI): m/z 410.9 [M+H] .
1
-(2-Chlorophenyl)-4-methyl-[1,2,4]triazolo[4,3-a]quinoxaline-8-carboxylic acid (13). To a
stirring solution of ethyl 1ꢀ(2ꢀchlorophenyl)ꢀ4ꢀmethylꢀ[1,2,4]triazolo[4,3ꢀa]quinoxalineꢀ8ꢀ
o
carboxylate (0.600 g, 1.77 mmol) in THF/H O (2:1, 16.5 mL) at 0 C was added LiOH (0.064 g,
2
2.7 mmol, 1.5 eq.). The reaction mixture was warmed to room temperature. After stirring 2h,
the organics were concentrated in vacuo and the remaining aqueous layer was acidified with
KHSO . The aqueous layer was extracted with 5% MeOH in DCM (3 x 20 mL), dried over
4
Na SO filtered, and concentrated in vacuo to afford crude solid. The crude solid was triturated
2
4,
1
with Et O (6 mL) to afford the title compound (13; 0.500 g, 84% yield). H NMR (400 MHz,
2
DMSOꢀd ): δ 13.20 (brs, 1H), 8.12 (s, 2H), 7.83 (m, 3H), 7.78 (s, 1H), 7.71 (dt, J = 11.6, 2.0 Hz,
6
13
1
H), 2.98 (s, 3H). C NMR (101 MHz, DMSOꢀd ) δ 166.1, 155.0, 147.4, 144.8, 138.9, 134.1,
6
1
33.9, 133.3, 130.8, 130.5, 130.2, 128.8, 128.4,127.8, 125.6, 116.8, 21.4. EIꢀHR calcd for
+
C H ClN O (M ) 339.0643, found 339.0642.
17
12
4
2
1-(2-Chloro-5-fluorophenyl)-4-methyl-[1,2,4]triazolo[4,3-a]quinoxaline-8-carboxylic
acid
(
14). To a stirring solution of ethyl 1ꢀ(2ꢀchloroꢀ5ꢀfluorophenyl)ꢀ4ꢀmethylꢀ[1,2,4]triazolo[4,3ꢀ
a]quinoxalineꢀ8ꢀcarboxylate (1.8 g, 4.9 mmol) in THF/H O (2:1, 49 mL) was added LiOH (0.30
2
g, 7.4 mmol, 1.5 eq.). After 3h, the organics were concentrated in vacuo. The resultant aqueous
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
layer was acidified with NaHSO , extracted with EtOAc, dried over Na SO , filtered, and conꢀ
4
2
4
centrated in vacuo to afford crude compound. The crude compound was triturated with Et O (16
2
1
mL) to afford the title compound (14; 1.6 g, 92% yield) as a white solid. H NMR (400 MHz,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
DMSOꢀd6) δ 13.20 (brs, 1H), 8.15 (s, 2H), 7.94 (dd, J = 9.2, 4.8 Hz, 1H), 7.84ꢀ7.83 (m, 3H),
13
2
.99 (s, 3H). C NMR (101 MHz, DMSOꢀd ) δ 166.1, 162.81 ꢀ 159.73 (m, 1C), 155.0, 146.3 (d,
6
J = 1.5 Hz, 1C), 144.8, 138.8, 132.6 (d, J = 8.8 Hz, 1C),130.9, 130.3, 129.9 (d, J = 3.7 Hz, 1C),
1
2
29.4 (d, J = 9.5 Hz, 1C), 128.5, 125.4, 121.28 ꢀ 120.76 (m, 1C), 120.67 ꢀ 120.22 (m, 1C), 116.7,
+
1.4. EIꢀHR calcd for C H ClFN O (M ) 357.0549, found 357.0548.
1
7
11
4
2
1
-(2-chloro-5-methoxyphenyl)-4-methyl-[1,2,4]triazolo[4,3-a]quinoxaline-8-carboxylic acid
(15). To a stirring solution of ethyl 1ꢀ(2ꢀchloroꢀ5ꢀmethoxyphenyl)ꢀ4ꢀmethylꢀ[1,2,4]triazolo[4,3ꢀ
a]quinoxalineꢀ8ꢀcarboxylate (9; 1.3 g, 3.3 mmol, 1.0 eq.) in THFꢀH O (2:1, 36 mL) was added
2
LiOH (0.21 g, 4.9 mmol, 1.5 eq.). After stirring at room temperature for 3h, the organics were
concentrated in vacuo. The aqueous layer was acidified with NaHSO , extracted with EtOAc,
4
dried over Na SO , filtered, and concentrated in vacuo to afford the crude solid. This crude mixꢀ
2
4
ture was triturated with Et O (5 mL) to afford the title compound (15; 0.71 g, 58% yield) as a
2
1
white solid. H NMR (400 MHz, DMSOꢀd6) δ 13.2 (brs, 1H), 8.13 (d, J = 0.8 Hz, 2H), 7.86 (s,
1
3
1
H), 7.74 (d, J = 8.8 Hz, 1H), 7.42 (m, 2H), 3.83 (s, 3H), 2.97 (s, 3H). C NMR (101 MHz,
DMSOꢀd6) δ 166.1, 159.1, 155.0, 147.3, 144.7, 138.8, 131.5, 130.8, 130.1, 128.4, 128.4, 125.5,
+
1
3
25.2, 119.6, 118.3, 117.0, 56.5, 21.4. EIꢀHR calcd for C H ClN O (M ) 369.0749, found
1
8
14
4
3
69.0749.
-(2-Chloro-5-isopropoxyphenyl)-4-methyl-[1,2,4]triazolo[4,3-a]quinoxaline-8-carboxylic
To a stirring solution of ethyl 1ꢀ(2ꢀchloroꢀ5ꢀisopropoxyphenyl)ꢀ4ꢀmethylꢀ
1
acid (16).
[1,2,4]triazolo[4,3ꢀa]quinoxalineꢀ8ꢀcarboxylate (10; 3.0 g, 7.1 mmol, 1.0 eq) in 1:1 mixture of
ACS Paragon Plus Environment

Page 29 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
THF/H O (120 mL) was added lithium hydroxide (0.42 g, 18 mmol, 2.5 eq.). After stirring for 4
2
h, the reaction mixture was concentrated in vacuo to obtain a crude solid. The crude solid was
dissolved in H O (40 mL) and washed with EtOAc (3 x 40 mL). The aqueous layer was acidiꢀ
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
fied with NaHSO (aq.) and extracted with EtOAc (3 x 40 mL). The combined organic layers
4
were dried over Na SO , filtered and concentrated in vacuo to afford the title compound (16;
2
4
1
2
7
4
1
.15 g, 76% yield). H NMR (400 MHz, DMSOꢀd ) δ 13.21 (brs, 1H), 8.12 (d, J = 0.8 Hz, 2H),
6
.85 (brs, 1H), 7.71 (d, J = 8.8 Hz, 1H), 7.40 (d, J = 3.2 Hz, 1H), 7.36 (dd, J = 8.8, 3.2 Hz, 1H),
13
.68 (m, 1H), 2.98 (s, 3H), 1.28 (dd, J = 14.8, 6.0 Hz, 6H). C NMR (101 MHz, DMSOꢀd6) δ
66.1, 157.3, 154.9, 147.2, 144.7, 138.8, 131.6, 130.9, 130.1, 128.5, 128.4, 125.6, 124.8, 121.2,
+
119.9, 117.0, 70.9, 22.0, 21.9, 21.4. EIꢀHR calcd for C H ClN O (M ) 397.1062, found
2
0
18
4
3
3
97.1061.
1-(2-Chloro-5-isobutoxyphenyl)-4-methyl-[1,2,4]triazolo[4,3-a]quinoxaline-8-carboxylic ac-
id (17). To a stirring solution of ethyl 1ꢀ(2ꢀchloroꢀ5ꢀisobutoxyphenyl)ꢀ4ꢀmethylꢀ
1,2,4]triazolo[4,3ꢀa]quinoxalineꢀ8ꢀcarboxylate (11; 2.35 g, 5.35 mmol, 1.0 eq) in 1:1 mixture of
[
THF/H O (94 mL) was added lithium hydroxide (0.32 g, 13 mmol, 2.5 eq). After stirring for 3h,
2
the reaction mixture was concentrated in vacuo to obtain a crude solid. The crude solid was disꢀ
solved in H2O (35 mL) and washed with EtOAc (3 x 35 mL). The aqueous layer was acidified
with NaHSO4 (aq.) and extracted with EtOAc (3 x 40 mL). The combined organic layers were
dried over Na2SO4, filtered and concentrated in vacuo to afford the title compound (17; 1.73 g,
1
7
9% yield). H NMR (300 MHz, DMSOꢀd6) δ 13.21 (brs, 1H), 8.13 (d, J = 0.9 Hz, 2H), 7.87
(brs, 1H), 7.72 (d, J = 9.0 Hz, 1H), 7.42–7.36 (m, 2H), 3.85–3.80 (m, 2H), 2.98 (s, 3H), 2.06–
13
1.99 (m, 1H), 0.96 (dd, J = 6.9, 3.3 Hz, 6H). C NMR (101 MHz, DMSOꢀd ) δ 166.1, 158.6,
6
ACS Paragon Plus Environment