Bridged tetrahydroisoquinolines as selective NADPH oxidase 2 (Nox 2) inhibitors

Article abstract of DOI:10.1039/c3md00061c

(1SR,4RS)-3,3-Dimethyl-1,2,3,4-tetrahydro-1,4-(epiminomethano)naphthalenes were synthesized in 2-3 steps from commercially available materials and assessed for specificity and effectiveness across a range of Nox isoforms. The N-pentyl and N-methylenethiophene substituted analogs 11g and 11h emerged as selective Nox₂ inhibitors with cellular IC₅₀ values of 20 and 32 μM, respectively.

Full text of DOI:10.1039/c3md00061c

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
MedChemComm
View Article Online
View Journal
CONCISE ARTICLE
Bridged tetrahydroisoquinolines as selective NADPH
oxidase 2 (Nox2) inhibitors†
Cite this: DOI: 10.1039/c3md00061c
a
a
Eugenia Cifuentes-Pagano,a Jaideep Saha,ab Gabor Csanyi, Imad Al Ghouleh,
´
´
a
a
bc
a
*
´
´
Sanghamitra Sahoo, Andres Rodrıguez, Peter Wipf, Patrick J. Pagano
b
*
and Erin M. Skoda
Received 25th February 2013
Accepted 22nd May 2013
(1SR,4RS)-3,3-Dimethyl-1,2,3,4-tetrahydro-1,4-(epiminomethano)naphthalenes were synthesized in 2–3
steps from commercially available materials and assessed for specificity and effectiveness across a range
of Nox isoforms. The N-pentyl and N-methylenethiophene substituted analogs 11g and 11h emerged as
selective Nox2 inhibitors with cellular IC50 values of 20 and 32 mM, respectively.
DOI: 10.1039/c3md00061c
www.rsc.org/medchemcomm
cytosolic subunits but is regulated by Poldip2.21,22 The result of
activation of these enzymes is the generation of ROS in the form
Introduction
of O2_ꢀ (Nox1, 2, 5) and H2O2 (Nox4).23,24 ROS production is
mediated by electron transfer from NADPH in the cytosol to FAD
to form FADH2. Single electron transfer to heme groups on the
transmembrane domains and subsequent transfer to molecular
oxygen on the opposite side of the membrane forms O2_ꢀ, which
can be converted to H2O2 by superoxide dismutase (SOD).19,25
Downstream effects of this ROS generation include changes in
gene expression, cellular signaling, host defense and inam-
mation, and cell growth regulation.19 The inability of currently
available agents to specically inhibit a particular NADPH
oxidase along with the combinative and varied expression of
these isozymes in cells and tissue has made it difficult to assess
their individual contributions to disease.26
Reactive oxygen species (ROS) play a pivotal role in the devel-
opment of cardiovascular diseases, cancer, neurological disor-
ders, and other pathologies.1–4 A major source of ROS is a family
of enzymes, NADPH oxidases (Nox), that catalyze electron
transfer from NADPH to molecular oxygen to give superoxide
(O2_ꢀ) and/or hydrogen peroxide (H2O2). Nox plays a crucial role
in signaling cascades initiated by pro-inammatory stimuli
including hormones, vasoactive agents, and cytokines as well as
mechanical stress.5–9 Members of this family include Nox1–5 as
well as Duox1 and 2; in the human cardiovascular system, Nox1,
2, 4, and 5 isoforms are prevalent.
The major catalytic subunit of these Nox isozymes possesses
six transmembrane domains with a cytosolic C-terminus con-
taining NADPH- and FAD-binding domains. Specically, Nox1, 2,
and 4 are constitutively associated with membrane-bound
p22phox, the complex of which forms cytochrome b558.10–13 On
the other hand, Nox5 does not require p22phox or cytosolic
subunits but uniquely contains calcium-activating EF domains
at its N-terminus.14–18 Furthermore, the Nox isozymes differ in
requirements for specic cytosolic subunits for activation and
organization.19,20 Nox1 associates with GTPase Rac1, cytosolic
activator NoxA1, and cytosolic organizer NoxO1. Nox2 (Fig. 1)
associates with Rac1 or Rac2 as well as cytosolic activator p67phox
and cytosolic organizer p47phox while Nox4 requires no classical
Among the isoforms, Nox2 (aka gp91phox, the rst Nox iso-
form discovered) has been implicated in cardiovascular disease
(CVD) processes including atherosclerosis,27 hypertension,28,29
ischemia reperfusion,30 cardiac hypertrophy,31 stroke,32 and
restenosis.33,34 In addition to CVDs, Nox2 has more recently
aVascular Medicine Institute, Department of Pharmacology and Chemical Biology,
University of Pittsburgh, Pittsburgh, PA 15260, USA. E-mail: pagano@pitt.edu
bCenter for Chemical Methodologies and Library Development, Department of
Chemistry, University of Pittsburgh, Pittsburgh, PA 15261, USA. E-mail: emf53@
pitt.edu
cDepartment of Pharmaceutical Sciences, University of Pittsburgh, Pittsburgh, PA
15260, USA
† Electronic supplementary information (ESI) available: Experimental procedures
and complete spectroscopic data for new compounds, protocols for cell and
cell-free assays. See DOI: 10.1039/c3md00061c
Fig. 1 Representation of Nox2 structure and activation.
This journal is ª The Royal Society of Chemistry 2013
Med. Chem. Commun.

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
View Article Online
MedChemComm
Concise Article
been implicated in neurodegenerative diseases such as Hun- in a complex whole cell or in vivo environment. Heterologous
tington's,35 Alzheimer's,36 and Parkinson's diseases.37 Many of Nox1, 2, 4, and 5 cell systems were used as a testing paradigm
the insights into the involvement of Nox2 in these processes for these studies.
have until now been obtained by using a peptidic isoform-
specic inhibitor of Nox2 developed in our laboratory.38–40
Nevertheless, small molecules are still the preferred therapeutic
Results and discussion
strategy for clinical use and thus the quest for such isoform-
specic inhibitors of Nox has intensied in recent years.
However, due to the complex assembly and the high degree of
homology among the various members of the Nox family, the
development of isoform-specic inhibitors has proven
challenging.
In order to accelerate the understanding of the role of Nox2 in
disease etiology, we set out to identify a probe molecule that
inhibits Nox2 selectively over Nox1, 4, and 5. An acceptable
probe would also show a lack of activity against XO, which also
produces ROS. The latter screen would serve two purposes: (1)
to eliminate compounds that inhibit an oxidase of distinct
composition; and (2) to preclude agents that directly scavenge
ROS. Accordingly, we optimized a cell-based primary assay
using stable Nox2-transfected COS cells56 with L-012 chem-
iluminescence57 as a detection system for O2_ꢀ generation. A
screen of a subset of small organic molecules from the
University of Pittsburgh Center for Chemical Methodologies
and Library Development (UPCMLD) at concentrations of 100,
50, 25, and 12.5 mM led to the identication of hit compound 7
(Fig. 3). Structurally related compounds 8 and 9 were also
present in the screening library, which allowed for some initial
structure–activity relationship (SAR) information to be gained.
Compound 8 was inactive, suggesting that substitution at the
5-position of the heterocycle was required for activity. In
contrast, 9 was active, suggesting that substitution could also be
tolerated at C-8 of the bridged tetrahydroisoquinoline motif.
A series of secondary assays was used to conrm activity and
rule out undesirable mechanisms of action. First, the concen-
tration-dependent effect of 7 was studied using the same
conditions as the primary screen (L-012 chemiluminescence in
COS-Nox2 cells), revealing that 7 had an IC50 of 45 mM (Fig. 4A,
closed squares). Cell-free activity in lysed COS-Nox2 cells was
promising, since 7 was also able to inhibit O2_ꢀ production in a
system in which assembly of active Nox2 subunits is achieved by
treatment with the anionic amphiphile LiDS and the reaction is
initiated by cofactor NADPH.58 Under these conditions, 7
appeared to be effective at inhibiting Nox2 activity since 6.3 mM
ameliorated LiDS-stimulated O2_ꢀ generation to almost non-
stimulated levels (Fig. 4B). Next, because selectivity over the
closely related isoform Nox4 was one of our most stringent
benchmarks, the concentration–response activity of 7 against
this isoform was tested. Since Nox4 is generally accepted to
directly form H2O2 rather than O2_ꢀ, the more appropriate
Amplexꢀ Red uorescence was used as the detection reporter in
Nox4 transfected COS cells;13 and it was compared to similar
assay conditions using the COS-Nox2 cells with addition of SOD
To date, several small molecule dual or multi-target Nox
inhibitors have been identied (Fig. 2) and have been used as
positive controls in phenotypic assays.41–44 Diphenylene iodo-
nium (DPI, 1) is a commonly used positive control for Nox
testing; however, it is also an irreversible and non-selective
inhibitor of avin-dependent enzymes such as xanthine oxidase
(XO) and nitric oxide synthase.19,42 Apocynin (2) was originally
thought to inhibit Nox; however, recent reports attribute activity
solely to its role as an nonspecic inhibitor and antioxidant.45,46
Similarly, S17834 is an inhibitor of O2_ꢀ formation by a number
of enzymes;47 and AEBSF (3) is a serine protease inhibitor that
also inhibits binding of cytochrome b558 to p47phox
. Ebselen
48
(4), a known ROS scavenger and mimic of glutathione peroxi-
dase, and its analogs inhibit both Nox1 and Nox2 activity as well
as other Nox isoforms to a lesser extent.49 Pan-inhibitors of Nox
include VAS2870 (ref. 50) and its derivative VAS3947 (ref. 51)
shown to inhibit Nox1, Nox2, and Nox4 but not XO. Dual Nox
inhibitors include fulvene-5 (5) that inhibits Nox2 and Nox4,52
GKT136901 that inhibits Nox1 and Nox4,53 and celastrol (6),
which preferentially inhibits Nox1 and Nox2 over Nox4.54
Perhaps the only inhibitor to date reported to display isoform
specicity is ML171, which inhibits Nox1.55 Due to the wide
distribution of the Nox enzymes in a variety of cells in the body
as well as their benecial role in signalling, nonspecic Nox
inhibitors are likely to cause undesired effects in vivo. A selective
inhibitor of Nox2 would help to differentiate the Nox isoforms
Fig.
3
Tetrahydroisoquinoline hit 7 (CID3323417) and related analogs 8
(CID3160422) and 9 (CID4005560) from a screen of a subset of compounds from
Fig. 2 Nonspecific or indirect inhibitors of Nox2 activity.
the UPCMLD library.
Med. Chem. Commun.
This journal is ª The Royal Society of Chemistry 2013

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
View Article Online
Concise Article
MedChemComm
Scheme 1 Synthesis of hit 7 and related analog 8, and reductive amination of
these building blocks to form tertiary amines 11a–h.
Scheme 2 Synthesis of biaryl analogs 12a–d via Suzuki cross-coupling with aryl
bromides.
Fig. 4 Concentration–response testing of 7 in (A) COS-Nox2 cells, (B) lysed COS-
Nox2 cells (cell-free preparation), and (C) COS-Nox4 cells. (A) Effect of 7 on Nox2
activity was measured in whole COS-Nox2 cells stimulated by 5 mM PMA using
L-012 and Amplexꢁ Red (AR). Data are expressed as % of vehicle control and
represent the mean ꢁ SEM of 6 independent experiments. (B) COS-Nox2 cell
lysate was preincubated with various concentrations of 7 for 15 min at 25 ꢂC. After
the addition of 130 mM LiDS, O2_ꢀ production was initiated by the addition of
180 mM NADPH and measured by the initial linear rate of SOD-inhibitable cyto-
chrome C reduction. O2_ꢀ production is expressed as nmol O2_ꢀ per min per 107
cell equivalents. (C) COS-Nox4 cells were pretreated with various concentrations
of 7 for 15 min. Initial rate of H2O2 production was measured using Amplexꢁ Red.
Data are expressed as the mean ꢁ SEM of 7 independent experiments.
synthesized to explore the SAR of this observed selectivity for
Nox2 over Nox4. Bridged tetrahydroisoquinoline derivatives
were synthesized in 2–3 steps from commercially available iso-
quinolines. Compound 8 was accessed through a double
allylation and carbometallation–cyclization sequence from iso-
quinoline as reported previously,59,60 and an analogous method
to convert O2_ꢀ to H2O2. Compound 7 had an IC50 of 40 mM
against Nox2-expressing cells when ROS inhibition was
assessed using Amplexꢀ Red (Fig. 4A, open circles), thus vali-
dating our results using two distinct ROS-detection assays.
However, 7 was found to be inactive against Nox4 (Fig. 4C).
Finally, the structure of compound 7 was conrmed by resyn-
thesis, and the freshly prepared sample retained its biological
activity.
Aer the activity and selectivity of 7 for Nox2 was conrmed
Scheme 3 Reduction of alkene 7 to alkane 13 and sulfonamide formation to
in both whole cell and cell-free assays, analogs of 7 were form analogs 14a and b.
This journal is ª The Royal Society of Chemistry 2013
Med. Chem. Commun.

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
View Article Online
MedChemComm
Concise Article
Table 1 IC50 values of inhibitors of ROS generation in Nox enzymes
Entry
1
Compound
Structure
Nox2a (mM)
Nox2b (mM)
46 ꢁ 11
Nox1b (mM)
31 ꢁ 0.7
Nox4a (mM)
Nox5b (mM)
20 ꢁ 3.1
7, CID3323417
40 ꢁ 9.7
>100
2
9, CID4005560
25 ꢁ 10
40 ꢁ 2.8
60 ꢁ 36
NI
36
3
4
8, CID3160422
>100
>100
>100
NI
>100
13
78 ꢁ 19
31 ꢁ 3.8
32 ꢁ 18
>100
18 ꢁ 1
5
6
11a
11b
30 ꢁ 7.5
68 ꢁ 33
>100
>100
>100
NI
50 ꢁ 9.0
>100
>100
>100
7
11c
>100
>100
>100
NI
>100
8
9
11d
11e
40
88 ꢁ 22
48 ꢁ 23
>100
>100
>100
>100
>100
12 ꢁ 0.5
31 ꢁ 11
10
11
12
11f
51
68 ꢁ 22
20 ꢁ 4.9
38 ꢁ 11
>100
>100
>100
>100
>100
NI
79 ꢁ 30
>100
11g
11h
20 ꢁ 1.9
32 ꢁ 1.9
>100
13
14
12a
12d
64 ꢁ 35
9.0 ꢁ 0.9
>100
ND
>100
ND
14 ꢁ 0.6
NI
NI
ND
15
16
12b
12c
>100
79
>100
>100
>100
>100
NI
>100
>100
>100
Med. Chem. Commun.
This journal is ª The Royal Society of Chemistry 2013