Discovery and mechanism of action studies of 4,6-diphenylpyrimidine-2-carbohydrazides as utrophin modulators for the treatment of Duchenne muscular dystrophy

Article abstract of DOI:10.1016/j.ejmech.2021.113431

Duchenne muscular dystrophy is a fatal disease with no cure, caused by lack of the cytoskeletal protein dystrophin. Upregulation of utrophin, a dystrophin paralogue, offers a potential therapy independent of mutation type. The failure of first-in-class utrophin modulator ezutromid/SMT C1100 in Phase II clinical trials necessitates development of compounds with better efficacy, physicochemical and ADME properties and/or complementary mechanisms. We have discovered and performed a preliminary optimisation of a novel class of utrophin modulators using an improved phenotypic screen, where reporter expression is derived from the full genomic context of the utrophin promoter. We further demonstrate through target deconvolution studies, including expression analysis and chemical proteomics, that this compound series operates via a novel mechanism of action, distinct from that of ezutromid.

Full text of DOI:10.1016/j.ejmech.2021.113431

European Journal of Medicinal Chemistry 220 (2021) 113431
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Research paper
Discovery and mechanism of action studies of 4,6-
diphenylpyrimidine-2-carbohydrazides as utrophin modulators for
the treatment of Duchenne muscular dystrophy
,
,
Aini Vuorinen a 1, Isabel V.L. Wilkinson a 1, Maria Chatzopoulou a, Ben Edwards b,
Sarah E. Squire b, Rebecca J. Fairclough b 2, Noelia Araujo Bazan a, Josh A. Milner a,
,
Daniel Conole a, James R. Donald a, Nandini Shah b, Nicky J. Willis a,
R. Fernando Martínez a, Francis X. Wilson c, Graham M. Wynne a, Stephen G. Davies a,
Kay E. Davies b **, Angela J. Russell a
,
, d, *
a Department of Chemistry, University of Oxford, Chemistry Research Laboratory, Mansfield Road, Oxford, OX1 3TA, UK
b Department of Physiology, Anatomy and Genetics, University of Oxford, Sir Henry Wellcome Building of Gene Function, South Parks Road, Oxford, OX1 3PT,
UK
c Summit Therapeutics Plc, 136a Eastern Avenue, Milton Park, Abingdon, Oxfordshire, OX14 4SB, UK
d Department of Pharmacology, University of Oxford, Mansfield Road, Oxford, OX1 3PQ, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Duchenne muscular dystrophy is a fatal disease with no cure, caused by lack of the cytoskeletal protein
dystrophin. Upregulation of utrophin, a dystrophin paralogue, offers a potential therapy independent of
mutation type. The failure of first-in-class utrophin modulator ezutromid/SMT C1100 in Phase II clinical
trials necessitates development of compounds with better efficacy, physicochemical and ADME prop-
erties and/or complementary mechanisms. We have discovered and performed a preliminary optimi-
sation of a novel class of utrophin modulators using an improved phenotypic screen, where reporter
expression is derived from the full genomic context of the utrophin promoter. We further demonstrate
through target deconvolution studies, including expression analysis and chemical proteomics, that this
compound series operates via a novel mechanism of action, distinct from that of ezutromid.
© 2021 Elsevier Masson SAS. All rights reserved.
Received 1 December 2020
Received in revised form
25 March 2021
Accepted 27 March 2021
Available online 20 April 2021
Keywords:
Duchenne muscular dystrophy
Utrophin
Phenotypic drug discovery
Mechanism of action
Target deconvolution
Chemical proteomics
Photoaffinity labelling
1. Introduction
developed and are starting to be approved for clinical use in certain
patient cohorts [1]. Oral small molecules have been developed to
Duchenne muscular dystrophy (DMD) is an X-linked progressive
muscle-wasting disease caused by a lack of the cytoskeletal protein
dystrophin. There is currently no disease modifying treatment for
all patients, although various promising approaches (e.g. exon
skipping, readthrough of stop codons and gene therapy) are being
compensate for the missing dystrophin by replacing it with its
autosomal paralogue utrophin [2]. This therapy would be appli-
cable to all patients regardless of their dystrophin mutation and
would be systemically bioavailable.
* Corresponding author. Department of Chemistry, University of Oxford, Chemistry Research Laboratory, Mansfield Road, Oxford, OX1 3TA, UK.
** Corresponding author. Department of Physiology, Anatomy and Genetics, University of Oxford, Sir Henry Wellcome Building of Gene Function, South Parks Road, Oxford,
OX1 3PT, UK.
E-mail addresses: kay.davies@dpag.ox.ac.uk (K.E. Davies), angela.russell@chem.ox.ac.uk (A.J. Russell).
Joint first authors.
Current address: Emerging Innovations Unit, Discovery Sciences, BioPharmaceuticals R&D, AstraZeneca, Cambridge, CB2 0SL, UK.
1
2
https://doi.org/10.1016/j.ejmech.2021.113431
0223-5234/© 2021 Elsevier Masson SAS. All rights reserved.

A. Vuorinen, I.V.L. Wilkinson, M. Chatzopoulou et al.
European Journal of Medicinal Chemistry 220 (2021) 113431
The first-in-class utrophin modulator, ezutromid (formerly SMT
form disubstituted chlorides 4 was also high yielding (61e76%),
apart from some analogues bearing a free OH group (36 and 42% for
4k and 4r, respectively). Symmetrical analogues were synthesised
with 2 equivalents of boronic acid. DABCO-mediated displacement
of the 2-chloro substituent yielded nitriles 5 in mostly moderate to
good yields (52e91%), barring once more analogues bearing a free
OH group (50% and 46% for 5k and 5r, respectively). Methanolysis
of nitriles 5 provided the methyl esters 6 and treatment with hy-
drazine resulted in the desired hyrazides 7 in moderate to good
yields (44e96%).
OX01914 analogues bearing a free phenol could act as important
key intermediates for further functionalisation. However, synthetic
Route A provided phenol analogues such as 7k (Scheme 1) with low
overall yields (6%, full synthesis details described in the Supporting
Information). Use of the O-benzyl protecting group significantly
improved the Suzuki coupling yield, however the subsequent
displacement with cyanide led to a mixture of unidentified prod-
ucts (data not shown). An alternative strategy was therefore pur-
sued, replacing the 2-chloro substituent with a 2-thiomethyl group
8 to be further oxidised (11) and displaced with cyanide 12 (Route
B, Scheme 1). Employing a N-Boc-protecting group on the hydra-
zide allowed functionalisation at the phenolic OH without
competing nucleophilic reaction of the hydrazide (Scheme S2).
Though longer, this synthetic sequence allowed the synthesis of key
intermediate N-Boc-7k on a 3 g scale in 37% yield over 8 steps. This
route was also used to synthesise OPh analogue 14c for the inves-
tigation of substituent effect on the pyrimidine ring. OMe analogue
14d was the adventitious product of nitrile methanolysis during the
synthesis of 14c, arising from displacement of the phenoxy group
with methanol. Adopting milder conditions with MeONa allowed
access to the desired OPh analogue 14c.
C1100), was developed using
a high throughput cell-based
phenotypic screen [3]. The screening cell line consisted of immor-
talised murine DMD disease model (mdx) myoblasts transfected
with 8.4 kb of the human utrophin A promoter upstream of firefly
luciferase (H2K-mdx utrnA-luc, Fig. 1a) [4]. Evaluation of ezutromid
in a Phase II clinical trial in DMD patients revealed encouraging
signs of on-target activity, including reduced muscle fibre damage
and increased levels of utrophin, at the trial midpoint [2]. However,
ezutromid is susceptible to CYP-mediated metabolism [5] and rapid
clearance in patients [6], and its therapeutic benefits were not
sustained through to the end of the trial, leading to discontinuation
of its clinical development. Therefore, utrophin modulators with
improved efficacy, physicochemical and ADME properties and/or
complementary mechanisms of action are now required. Here we
describe a novel class of small molecule utrophin modulators,
discovered in a new screening cell line, that operate via a previously
undescribed mode of action, as proven by expression analysis and
chemoproteomics.
2. Results and discussion
A cell-based reporter system which exploits the full utrophin
promoter in its endogenous genomic context and encompassing all
known utrophin isoforms and regulatory elements was developed
by knock-in of firefly luciferase into one UTRN exon 7 of a
dystrophin-null mouse (LUmdx; mdx,utrluc/þ, Fig. 1b, procedure
described in the Experimental Methods). 7000 drug-like small
molecules from our in house compound collection were screened in
immortalised myoblasts from the LUmdx mouse. After counter-
screening for firefly luciferase assay interference [7,8] to ensure
genuine activation of the utrophin promoter, this screen gave rise to
several novel hit series’ of utrophin modulators.
Interestingly, one of these compounds, an unusual heteroaryl
acyl hydrazide denoted OX01914 (compound 1, Fig. 1f), demon-
strated activity in the new LUmdx reporter cell line, but not the
original H2K-mdx utrnA-luc screening cell line [4] (Fig. 1c), with the
activity confirmed to be independent of luciferase interference
(Supplementary Fig. 1a and b). This finding suggested that OX01914
may operate via a mechanism of action which is distinct from that
of ezutromid, which we recently reported to be mediated by the
antagonism of the aryl hydrocarbon receptor [9].
In addition, a shorter synthesis (Route C, Scheme 1) was inves-
tigated to access cyclohexyl analogue 22 in 6 steps. First, 3-
phenylisoxazol-5-amine 16 was treated with methyl oxalyl chlo-
ride and the product cyclised to 4-hydroxypyrimidine 18. POCl3
displacement of the hydroxy group with chloride allowed for a
palladium coupling with cyclohexene to give 20. Olefin 20 was then
hydrogenated and treated with hydrazine hydrate to provide the
cyclohexyl analogue 22. Advantageously, this route could accelerate
SAR development on one aryl ring at a time.
2.2. Structure-activity relationships around OX01914
The ability of OX01914 to upregulate utrophin mRNA in LUmdx
and H2K mdx myoblasts was confirmed by qPCR, with a 2-fold in-
crease observed at the highest dose in LUmdx (Figs. 1d and 2). Next,
upregulation of utrophin expression by OX01914 was investigated
in human DMD muscle cells at the protein level. Excitingly,
OX01914 treatment led to a statistically significant 3-fold increase
in utrophin protein expression (Fig. 1e), far in excess of the 1.5 fold
increase along the membrane required to see a beneficial effect [2].
On the basis of these encouraging results in mdx and DMD muscle
cells, the physicochemical/ADME properties of OX01914 were
assessed to guide compound optimisation. OX01914 demonstrated
good aqueous solubility and permeability, however, was rapidly
metabolised by mouse hepatocytes (T1/2 3 min) e a profile that was
also sustained in human hepatocytes (T1/2 28 min, Fig. 1g).
Seeking to establish the minimum structural moieties required
for activity (Fig. 1f), we investigated (i) a small range of head group
replacements (Table 1), (ii) alterations of the pyrimidine core
(Table 2), (iii) the role of the substituents on the 4,6-positions of the
pyrimidine ring (Table 3) and (iv) the influence of regiochemistry
on the phenyl rings (Table 4).
Leveraging chloride intermediate 4a from Route A to OX01914
allowed for rapid synthesis of head group variants including hy-
drazine analogue 23 and methylene linked hydrazide analogue 25
(Table 1, Scheme S3). Meanwhile, the methyl ester intermediate 6a
allowed facile synthesis of amide 26, carboxylic acid 27 and a
hydroxamic acid analogue 28 (Table 1, Scheme S3). Finally, imide
analogue 30 was synthesised from 6a in three steps: reduction to
an alcohol (29, Scheme S3), Dess-Martin periodinane-mediated
oxidation to an aldehyde then reaction with hydrazine.
2.1. Chemistry
Intriguingly, these alterations on the head group were not
tolerated, with hydrazine analogue 23, methylene linked analogue
25, methyl ester intermediate 6a and imide analogue 30 all inactive
in the LUmdx reporter gene assay when tested at concentrations up
to 100 mM. Even more subtle alterations including to amide, car-
boxylic acid and hydroxamic acid moieties led to inactive analogues
(Table 1).
As part of the hit confirmation process, and to establish a syn-
thetic route to further analogues, resynthesis of OX01914 was un-
dertaken (Route A, Scheme 1, synthesis of analogues using Route A
detailed in Scheme S1). In an initial Suzuki reaction using an excess
of 2,4,6-trichloropyrimidine 2, selective monosubstitution to afford
3 was achieved in good yield (82%). The second Suzuki coupling to
2

A. Vuorinen, I.V.L. Wilkinson, M. Chatzopoulou et al.
European Journal of Medicinal Chemistry 220 (2021) 113431
Fig. 1. Schematic of firefly luciferase reporter gene assay phenotypic screens for utrophin gene upregulation: a) H2K-mdx utrnA-luc cell line, b) LUmdx cell line; c) OX01914 (1)
activates luciferase in the LUmdx cell line but not in H2K-mdx utrnA-luc line, N ¼ 3, representative graph shown; d) OX01914 increases utrophin mRNA levels in LUmdx myoblasts,
n ¼ 2, see also Fig. 2; e) OX01914 increases utrophin protein levels by approximately 3-fold, N ¼ 2; f) structure of OX01914 and areas of interest for structure-activity relationships;
g) luciferase inhibition, physicochemical and ADME properties of OX01914. Solubility: aqueous thermodynamic solubility, Caco2 permeability efflux ratio (ER) and apical to
basolateral (A/B) transport, mHep T1/2: half-life in mouse hepatocytes and hHep T1/2: half-life in human hepatocytes.
To investigate the SAR of the core scaffold, pyridine analogues 33
and 40, as well as regioisomeric pyrimidine 34 were synthesised
from the respective 4,6-dichloro-2-carboxylate esters in two steps:
Suzuki coupling followed by treatment with hydrazine hydrate
(Scheme S4). Triazine analogue 38 was synthesised from 2,4,6-
trichlorotriazine in a similar sequence to the OX01914 analogues
in Scheme S1 (Scheme S4).
However, alterations to the structure of OX01914 led to a loss of
activity, except in the case of triazine analogue 38, which was found
to give comparable or somewhat improved activity compared to
the original hit 1 (Table 2). In this study we continued to focus on
optimisation of the pyrimidine core. The large effect on activity of
relatively subtle structural alterations was interesting to note, for
example, the introduction of a methyl substituent at the 5 position
of the pyrimidine ring which led to inactive analogue 14b. As such,
5-methylated pyrimidine acyl hydrazides can be usefully employed
as negative controls in assay development and target deconvolu-
tion studies.
structure, by replacing one or both aromatic substituents with
methyl substituents, oxygen-linkers and a cyclohexyl substituent
(Table 3, synthesis described in Scheme 1). All of these modifica-
tions led to inactive analogues.
Finally, several analogues bearing substituted phenyl rings
(Table 4, 7b-7j) were synthesised to delineate the effect of pe-
ripheral regiochemistry on activity. Symmetrical analogues were
chosen for this investigation due to their facile synthesis (Route A).
Small substituents like fluorine maintained activity regardless of
regiochemistry, while meta- and para-methoxy groups were also
tolerated. Interestingly, ortho- and meta-CF3 substitution both led
to loss of activity, while the para position tolerated the widest range
of substituents.
2.3. Mechanism of action studies
To confirm that OX01914 (1) upregulates utrophin expression
via a mechanism distinct from that of the AhR antagonist ezu-
tromid, qPCR of utrophin and AhR was undertaken in H2K mdx
Next, recognising the desirability of steering away from flat
polyaromatic molecules [10] we investigated introducing more 3D
myoblasts (Fig. 2). Both ezutromid (3 mM) and OX01914 (30 mM)
3

A. Vuorinen, I.V.L. Wilkinson, M. Chatzopoulou et al.
European Journal of Medicinal Chemistry 220 (2021) 113431
oxygen and its position affects the activity. Additionally, com-
pounds 7p, and 7u were prepared to explore steric effects.
In the alkoxy-substituted series (7l-7o) activity is maintained
until the n-propoxy analogue 7n, though somewhat diminished,
but the bulkier n-butoxy and phenoxy (7o and 7p, respectively)
were both inactive. Interestingly, the ethylene glycol analogue 7s,
was active (EC50 ~ 9uM), suggesting the incorporation of a polar
atom within the chain is favourable for activity, when the length of
the chain is increased.
Compounds 7k and 7r bearing a hydroxy group gave promising
results with the lowest EC50 values (1.57
tively) and the methoxymethyl variant 7t gave an EC50 of 4.65
m
M and 1.90
m
M, respec-
M.
m
These results show that the position of the oxygen atom along the
chain can be changed with no loss in activity. Synthesis of PEG-
chain variants 49 and the alkyne variant 50 proceeded smoothly
from intermediate N-Boc-7k via an etherification reaction with the
respective mesylated alcohols [18e20] and subsequent N-Boc-
deprotection (Scheme S5). Importantly, 49 and 50 were both active
Fig. 2. Both ezutromid (3
mM) and OX01914 (30 mM) upregulate utrophin mRNA
in the low micromolar range (EC50s of 12.7 and 14.9
respectively).
mM,
expression in H2K mdx myoblasts, but only ezutromid upregulates AhR expression (N
biological replicates ¼ 2, n ¼ 3 technical replicates). ** and *** indicate p-values of
<0.005 and < 0.0005 respectively.
Ultimately, combination of the PEG alkyne and diazirine pho-
toaffinity groups in dual-tagged probe 64 (Fig. 3a) resulted in ac-
tivity (EC50 ¼ 20.0
2.29 mM, Table 5) comparable to that of
significantly increased utrophin mRNA levels compared to DMSO
control (1.7 and 1.6 fold respectively). However, while ezutromid
increased AhR mRNA levels approximately 2 fold, consistent with
AhR antagonism and in accordance with previous reports [9],
OX01914 treatment did not result in any significant difference in
AhR expression. This result supports the hypothesis that ezutromid
and OX01914 upregulate utrophin through different mechanisms.
Target identification can greatly assist lead optimisation by
enabling structure activity relationship (SAR) studies to be con-
ducted more efficiently and rationally [11,12]. Furthermore, iden-
tification of on- and off-targets can allow optimisation to reduce
side effects, increase efficacy and find new applications for the
compounds. Here, a chemical proteomics strategy using cell-
permeable photoaffinity-labelled clickable probes [13] was
employed to identify proteins targeted by the OX01914 chemotype.
Use of a photoaffinity label can allow identification of even
weakly bound proteins, by enabling UV-mediated covalent binding
of the probe to its protein target [14]. A trifluoromethylphenyl
diazirine (TPD) was selected as the photoaffinity label, with its
placement ‘nested’ within the structure of OX01914 to minimise
structural perturbation [15]. Substitution of the diazirine in the
para position was informed by the SAR for the series (Table 5).
Synthesis of trifluoromethylphenyl diazirines from the corre-
sponding ketones is well-established [15] and was performed over
4 steps for the synthesis of diazirine-labelled probes 63e65
(Scheme S6). N-Boc-protection was employed for the installation of
the acyl hydrazine after diazirine synthesis as without it hydrazine
was found to reduce the diazirine ring back to the diaziridine.
Gratifyingly, photoaffinity probe 63 was found to have good
OX01914. Noting the inactivity of 5-methylpyrimidine analogues of
OX01914 such as 14b, dual-tagged probe 65 (Fig. 3a, Table 5) was
also synthesised to serve as a negative control (synthesis of both
probes reported in Scheme S6).
With both active and negative control probes in hand, their
protein-interaction profiles were assessed with gel-based labelling
experiments using live dystrophic mouse myoblasts. Both probes
64 and 65 (6.25e50 mM), were treated alongside DMSO vehicle for
2 h. After irradiation (365 nm) and cell lysis, the probe-labelled
lysate was subjected to a CuAAC-mediated click reaction with
TAMRA-azide, then SDS-PAGE. In-gel fluorescence showed dose-
dependent labelling of proteins (Fig. 3b), with no labelling apparent
in the vehicle control.
Protein target identification was next performed in live human
DMD myoblasts using the two probes (30
samples were pre-treated with an excess concentration of the un-
modified compound OX01914 (50 M) to occupy the target binding
mM). In addition, control
m
sites and help distinguish non-specifically binding proteins. Probe-
labelled proteins were clicked to azide-biotin, then enriched using
streptavidin beads. Enriched proteins underwent on-bead tryptic
digest and the peptides analysed by nano liquid chromatography-
tandem mass spectrometry (nanoLC-MS/MS). Evaluation of pro-
teins enriched by active probe 64, but not enriched after pre-
treatment with excess parent compound OX01914, delivered a
shortlist of potential OX01914 targets (red proteins, Fig. 3c).
Meanwhile, proteins enriched by the active probe 64 but not the
structurally similar inactive probe 65 are likely to mediate the
therapeutic effects of the OX01914 chemotype (Fig. 3d, full list of
proteins identified available in Supplementary Information). As
anticipated, unlike as was previously seen for ezutromid-based
photoaffinity probes [9], the aryl hydrocarbon receptor was not
detected by these acyl hydrazide probes, supporting the hypothesis
that this series of compounds operates via a distinct mechanism of
action to that of ezutromid. Interestingly, ATP5F1 (subunit b of
mitochondrial ATP synthase) was identified as a target by the active
probe 64 against both inactive and competitor controls. Oxidative
metabolism and ATP synthesis are known to be severely compro-
mised in DMD pathology [21e24], while induction of slower
oxidative muscle phenotypes through small molecule regulation of
oxidative respiration and metabolism have been shown in pre-
clinical trials to upregulate utrophin and mitigate dystrophic pa-
thology [25e27]. A protein network database search [28,29]
activity (Table 5, EC50 ¼ 7.22 0.72
mM) in the utrophin promoter
reporter gene assay.
A click handle was desired for attachment of a purification tag
such as biotin, and an alkyne was chosen for use in the well-
precedented biorthogonal copper-catalysed azide-alkyne cycload-
dition (CuAAC) reaction [16]. A polyethylene glycol (PEG) chain
linker was also desired to improve accessibility of the alkyne for
efficient CuAAC [17]. Different functionalities were introduced to
see how steric effects and hydrophilic/hydrophobic groups were
tolerated (Table 5). Initially the length of an n-alkyloxy chain was
increased systematically from methoxy to n-butoxy (7l-7o) and it
was compared to its para-hydroxy substituted counterpart 7k.
Compounds 7q-7t were then prepared to see how the presence of
4

A. Vuorinen, I.V.L. Wilkinson, M. Chatzopoulou et al.
European Journal of Medicinal Chemistry 220 (2021) 113431
Scheme 1. Routes employed towards synthesis of hit compound OX01914 and analogues.
Reagents and conditions: i) R1-arylboronic acid (1 eq), Pd(OAc)2, PPh3, Na2CO3, DME/H2O, 85 ꢀC, 24 h; ii) R2-arylboronic acid, Pd(OAc)2, PPh3, Na2CO3, DME/H2O, 85 ꢀC, 24 h; iii) KCN,
DABCO, DMSO/H2O, 80 ꢀC, 1e3 h; iv) AcCl, MeOH, 65 ꢀC, 2e3 h; v) NH2NH2$H2O, PhMe/EtOH, rt, 24 h; vi) Phenylboronic acid (1 or 2.5 eq), Pd(PPh3)4, 2 M K2CO3, PhMe, 95 ꢀC, 24 h;
vii) arylboronic acid, Pd(OAc)2, PPh3, Na2CO3, DME/H2O, 85 ꢀC, 24 h; viii) PhOH, CsCO3, MeCN, rt, 60 h; ix) Oxone®, CH3CN/PhMe/H2O, 60 ꢀC, 3 d; x) KCN, DMSO, rt, 24 h; xi) MeONa,
MeOH/PhCH3, 10 ꢀC to rt, 3 h; xii) ClCOCOOMe, pyridine, 0 ꢀC, 1 h; xiii) PtO2, H2, EtOH, rt, 3 h; xiv) POCl3, PhNMe2, 100 ꢀC, 3 h; xv) cyclohex-1-en-1-ylboronic acid, Pd(PPh3)4, 2 M
K2CO3, PhMe, 80 ꢀC, 3 h; xvi) H2, Pd/C (10% wt), EtOAc, rt, 3 h.
5

A. Vuorinen, I.V.L. Wilkinson, M. Chatzopoulou et al.
European Journal of Medicinal Chemistry 220 (2021) 113431
Table 1
Table 4
Preliminary structure e activity relationships on the head group, N ꢁ 2 biological
Investigation of phenyl ring regiochemistry to determine the optimal position for
substitution on hit compound OX01914. N ꢁ 1 biological replicates, n ¼ 3 technical
replicates. IA ¼ inactive, R2 indicates goodness of fit where only technical replicates
were performed.
replicates, n ¼ 3 technical replicates, IA ¼ inactive.
#
R
EC50 SEM (mM)
1
CONHNH2
NHNH2
CH2CONHNH2
COOMe
CONH2
COOH
20.5 1.79
23
25
6a
26
27
28
30
IA
IA
IA
IA
IA
IA
IA
#
R
EC50 SEM (
m
M)
R2
1
Ph
2-F
3-F
4-F
2-OMe
3-OMe
4-OMe
2-CF3
3-CF3
4-CF3
20.5 1.79
11.0
>10
12.8
IA
9.22
7.14
IA
7b
7c
7d
7e
7f
7g
7h
7i
0.94
0.88
0.96
CONHOH
CH¼NNH2
0.56
0.95
IA
>10
Table 2
7j
0.95
Preliminary structure e activity relationships on the core, N ꢁ 2 biological replicates,
n ¼ 3 technical replicates, IA ¼ inactive.
3. Conclusions
A cell-based phenotypic reporter gene assay exploiting the full
utrophin promoter was developed by knock-in of firefly luciferase
into one utrophin exon 7 allele in a dystrophin-null mouse
(LUmdx). Screening of compounds in an immortalised myoblast
#
X
Y
Z
Scaffold
EC50 SEM (mM)
Table 5
Structure e activity relationships for the synthesis of photoaffinity click probes,
N ꢁ 2 biological replicates, n ¼ 3 technical replicates, IA ¼ inactive.
1
N
N
N
N
e
N
N
CH
CH
N
N
e
A
A
A
A
B
A
20.5 1.79
IA
33
34
38
40
14b
CH
CH
N
IA
8.98a
IA
e
N
CeMe
IA
a
R2 ¼ 0.91 (goodness of fit where only technical replicates were performed).
suggested that ATP5F1 expression and activity are not mediated by
AhR; the effects of targeting ATP5F1 on mitochondrial function and
utrophin regulation in DMD cells will form the basis of future
studies.
#
R1
R2
R3
EC50 SEM (mM)
1
7k
7l
H
OH
OMe
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
20.5 1.79
1.57 0.24
2.28 0.09
2.71 0.30
5.62 0.09
IA
7m
7n
7o
7p
7q
7r
7t
7s
7u
49
50
63
OEt
Table 3
OnPr
Preliminary structure e activity relationships on the substituents, N ꢁ 2 biological
OnBu
replicates, n ¼ 3 technical replicates, IA ¼ inactive.
OPh
IA
nPr
4.54 0.57
1.90 0.34
4.65 0.50
8.95 1.02
IA
12.7 1.57
14.9 1.17
7.22 0.72
CH2OH
CH2OMe
OCH2CH2OMe
OCH2CH2OBn
OCH2CH2OCH2CH2NHAc
OCH2CH2OCH2C^CH
H
#
R1
R2
EC50 SEM (mM)
1
Ph
H
Me
OPh
OMe
cyclohexyl
Ph
H
Me
Ph
Ph
Ph
20.5 1.79
64
65
OCH2CH2OCH2C^CH
OCH2CH2OCH2C^CH
H
20.0 2.29
IA
41a
42a
14c
14d
22
IA
IA
IA
IA
IA
Me
a
Obtained commercially.
6

A. Vuorinen, I.V.L. Wilkinson, M. Chatzopoulou et al.
European Journal of Medicinal Chemistry 220 (2021) 113431
Fig. 3. Target identification via chemical proteomics for the OX01914 chemotype. a) Structures of active 64 and inactive 65 photoaffinity-labelled clickable probes of OX01914; b) In-
gel fluorescence of active and inactive probe-labelled H2K mdx proteome; c) Volcano plot of proteins enriched by active probe 64 in the presence and absence of competitor
OX01914, top hits in red; d) Volcano plot of proteins enriched by active probe 64 compared to inactive probe 65, top hits in red.
LUmdx cell line led to discovery of a novel class of utrophin mod-
ulators based on the original hit molecule 4,6-diphenylpyrimidine-
2-carbohydrazide (OX01914, 1). OX01914 was further found to
upregulate utrophin mRNA in LUmdx myoblasts and protein in
human DMD myoblasts. The inactivity of OX01914 in the H2K-mdx
utrnA-luc phenotypic screen in which ezutromid was discovered,
along with its null effect on AhR-mediated gene expression indicate
that this class of compounds operate via a distinct mechanism of
action.
Synthesis of 43 analogues with systematic variations of the head
group, core and peripheral substituents led to a fuller under-
standing of the structure-activity relationship of OX01914. While
the majority of alterations diminished activity, many para-sub-
stituents on the phenyl rings led to more potent analogues. The SAR
of the series inspired the design and synthesis of active and inactive
cell-permeable photoaffinity probes to deconvolute the protein
targets of OX01914. Identification of probe-binding proteins in live
human DMD myoblasts by LC-MS/MS confirmed that AhR is not a
target of OX01914, while highlighting ATP5F1 as a target to be
further explored. Taken together, these findings reveal 4,6-
diphenylpyrimidine-2-carbohydrazides as novel utrophin modu-
lators worthy of further investigation for the treatment of
Duchenne muscular dystrophy.
relationships which may be considered as potential competing
interests: R.J.F., S.G.D., A.J.R. and K.E.D are minor shareholders of
Summit Therapeutics plc.
Acknowledgements
The authors wish to thank Summit Therapeutics plc (M.C., D.C.,
N.S.), the Medical Research Council (EP/L016044/1, I.V.L.W.;
1501AV003/CA2, B.E., S.E.S, K.E.D.), the Engineering and Physical
Sciences Research Council (EP/L016044/1, I.V.L.W.), Muscular Dys-
trophy Association (MDA212606, J.R.D., N.A.B.), Duchenne UK
(N.A.B.) and Muscular Dystrophy UK (RA4/3013/4, A.V.) for financial
support. We thank G. E. Morris (Oswestry, UK) for the gift of the
utrophin A MANCHO3 antibody. The authors gratefully acknowl-
edge Aurelia Bioscience (UK) for performing the firefly luciferase
inhibition and dual luciferase (Firefly/Renilla) CMV reporter assays;
Cyprotex Discovery Ltd (U.K.) for the physicochemical and ADME
evaluation; Key Organics (particularly Steve Brough and Becky Gill)
for analogue synthesis and helpful input to the route design.
Appendix A. Supplementary data
Declaration of competing interest
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.ejmech.2021.113431.
The authors declare the following financial interests/personal
7

A. Vuorinen, I.V.L. Wilkinson, M. Chatzopoulou et al.
European Journal of Medicinal Chemistry 220 (2021) 113431
6752e6756, https://doi.org/10.1021/jm901241e.
[11] K.M. Comess, S.M. McLoughlin, J.A. Oyer, P.L. Richardson, H. Stockmann,
Abbreviations
€
A. Vasudevan, S.E. Warder, Emerging approaches for the identification of
protein targets of small molecules - a practitioners’ perspective, J. Med. Chem.
61 (2018) 8504e8535, https://doi.org/10.1021/acs.jmedchem.7b01921.
[12] I.V.L. Wilkinson, G.C. Terstappen, A.J. Russell, Combining experimental stra-
tegies for successful target deconvolution, Drug Discov. Today (2020), https://
doi.org/10.1016/j.drudis.2020.09.016.
[13] M.H. Wright, S.A. Sieber, Chemical proteomics approaches for identifying the
cellular targets of natural products, Nat. Prod. Rep. 33 (2016) 681e708,
https://doi.org/10.1039/c6np00001k.
A/B
ADME
AhR
CuAAC
DMD
ER
apical to basolateral transport
absorption, distribution, metabolism, and excretion
aryl hydrocarbon receptor
copper-catalysed azide-alkyne cycloaddition
Duchenne muscular dystrophy
efflux ratio
hHep
IA
human hepatocytes
inactive, where EC50 > 100 mM
[14] E. Smith, I. Collins, Photoaffinity labeling in target- and binding-site identifi-
cation, Future Med. Chem.
fmc.14.152.
7 (2015) 159e183, https://doi.org/10.4155/
LC-MS/MS liquid chromatography-tandem mass spectrometry
mHep
PEG
[15] J.R. Hill, A.A.B. Robertson, Fishing for drug targets: a focus on diazirine pho-
toaffinity probe synthesis, J. Med. Chem. 61 (2018) 6945e6963, https://
doi.org/10.1021/acs.jmedchem.7b01561.
mouse hepatocytes
polyethylene glycol
[16] R.D. Row, J.A. Prescher, Constructing new bioorthogonal reagents and re-
actions, Acc. Chem. Res. 51 (2018) 1073e1081, https://doi.org/10.1021/
acs.accounts.7b00606.
[17] C.J. Pickens, S.N. Johnson, M.M. Pressnall, M.A. Leon, C.J. Berkland, Practical
considerations, challenges, and limitations of bioconjugation via azide-alkyne
cycloaddition, Bioconjugate Chem. 29 (2018) 686e701, https://doi.org/
10.1021/acs.bioconjchem.7b00633.
SAR
structure-activity relationship
SDS-PAGE sodium dodecyl sulfate polyacrylamide gel
electrophoresis
SEM
TAMRA
TPD
standard error of the mean
tetramethylrhodamine
trifluoromethylphenyl diazirine
[18] C.M. Timmers, H.J.J. Loozen, H.T. Stock, Ring-annulated Dihydropyrrolo[2,1-
Alpha]isoquinolines, US 2011 0028450 A1, 2011.
[19] J. Bucher, T. Wurm, K.S. Nalivela, M. Rudolph, F. Rominger, A.S.K. Hashmi,
Cyclization of gold acetylides: synthesis of vinyl sulfonates via gold vinylidene
complexes, Angew. Chem. Int. Ed. 53 (2014) 3854e3858, https://doi.org/
10.1002/anie.201310280.
[20] R. Mori, A. Kato, K. Komenoi, H. Kurasaki, T. Iijima, M. Kawagoshi, Y.B. Kiran,
S. Takeda, N. Sakai, T. Konakahara, Synthesis and in vitro antitumor activity of
novel 2-alkyl-5- methoxycarbonyl-11-methyl-6H-pyrido[4,3-b]carbazol-2-
ium and 2-alkylellipticin-2-ium chloride derivatives, Eur. J. Med. Chem. 82
(2014) 16e35, https://doi.org/10.1016/j.ejmech.2014.05.032.
[21] A.V. Kuznetsov, K. Winkler, F.R. Wiedemann, P. Von Bossanyi, K. Dietzmann,
W.S. Kunz, Impaired mitochondrial oxidative phosphorylation in skeletal
muscle of the dystrophin-deficient mdx mouse, Mol. Cell. Biochem. 183
(1998) 87e96, https://doi.org/10.1023/A:1006868130002.
[22] J.M. Percival, M.P. Siegel, G. Knowels, D.J. Marcinek, Defects in mitochondrial
localization and ATP synthesis in the mdx mouse model of duchenne
muscular dystrophy are not alleviated by PDE5 inhibition, Hum. Mol. Genet.
22 (2013) 153e167, https://doi.org/10.1093/hmg/dds415.
References
[1] S. Guiraud, K.E. Davies, Pharmacological advances for treatment in Duchenne
muscular dystrophy, Curr. Opin. Pharmacol. 34 (2017) 36e48, https://doi.org/
10.1016/J.COPH.2017.04.002.
[2] S. Guiraud, D. Roblin, K.E. Davies, The potential of utrophin modulators for the
treatment of Duchenne muscular dystrophy, Expert Opin. Orphan Drugs. 6
(2018) 179e192, https://doi.org/10.1080/21678707.2018.1438261.
[3] D.R. Chancellor, K.E. Davies, O. De Moor, C.R. Dorgan, P.D. Johnson,
A.G. Lambert, D. Lawrence, C. Lecci, C. Maillol, P.J. Middleton, G. Nugent,
S.D. Poignant, A.C. Potter, P.D. Price, R.J. Pye, R. Storer, J.M. Tinsley, R. van Well,
R. Vickers, J. Vile, F.J. Wilkes, F.X. Wilson, S.P. Wren, G.M. Wynne, Discovery of
2-arylbenzoxazoles as upregulators of utrophin production for the treatment
of duchenne muscular dystrophy, J. Med. Chem. 54 (2011) 3241e3250,
https://doi.org/10.1021/jm200135z.
[4] J.M. Tinsley, R.J. Fairclough, R. Storer, F.J. Wilkes, A.C. Potter, S.E. Squire,
D.S. Powell, A. Cozzoli, R.F. Capogrosso, A. Lambert, F.X. Wilson, S.P. Wren,
[23] E. Rybalka, C.A. Timpani, M.B. Cooke, A.D. Williams, A. Hayes, Defects in
mitochondrial ATP synthesis in dystrophin-deficient Mdx skeletal muscles
may be caused by complex I insufficiency, PLoS One 9 (2014), e115763,
https://doi.org/10.1371/journal.pone.0115763.
A. De Luca, K.E. Davies, Daily treatment with SMTC1100,
a novel small
molecule utrophin upregulator, dramatically reduces the dystrophic symp-
toms in the mdx mouse, PLoS One 6 (2011), e19189, https://doi.org/10.1371/
journal.pone.0019189.
ꢀ
[24] D. Capitanio, M. Moriggi, E. Torretta, P. Barbacini, S. De Palma, A. Vigano,
[5] M. Chatzopoulou, T.D.W. Claridge, K.E. Davies, S.G. Davies, D.J. Elsey, E. Emer,
A.M. Fletcher, S. Harriman, N. Robinson, J.A. Rowley, A.J. Russell, J.M. Tinsley,
R. Weaver, I.V.L. Wilkinson, N.J. Willis, F.X. Wilson, G.M. Wynne, Isolation,
structural identification, synthesis, and pharmacological profiling of 1,2-trans-
Dihydro-1,2-diol metabolites of the utrophin modulator ezutromid, J. Med.
Chem. 63 (2019) 2547e2556, https://doi.org/10.1021/acs.jmedchem.9b01547.
[6] V. Ricotti, S. Spinty, H. Roper, I. Hughes, B. Tejura, N. Robinson, G. Layton,
K. Davies, F. Muntoni, J. Tinsley, Safety, tolerability, and pharmacokinetics of
SMT C1100, a 2-arylbenzoxazole utrophin modulator, following single- and
multiple-dose administration to pediatric patients with duchenne muscular
dystrophy, PLoS One 11 (2016), e0152840, https://doi.org/10.1371/
journal.pone.0152840.
[7] A. Braeuning, Firefly luciferase inhibition: a widely neglected problem, Arch.
Toxicol. 89 (2015) 141e142, https://doi.org/10.1007/s00204-014-1423-3.
[8] I.V.L. Wilkinson, J.K. Reynolds, S.R.G. Galan, A. Vuorinen, A.J. Sills, E. Pires,
G.M. Wynne, F.X. Wilson, A.J. Russell, Characterisation of utrophin modulator
SMT C1100 as a non-competitive inhibitor of firefly luciferase, Bioorg. Chem.
94 (2020) 103395, https://doi.org/10.1016/j.bioorg.2019.103395.
H. Lochmüller, F. Muntoni, A. Ferlini, M. Mora, C. Gelfi, Comparative proteomic
analyses of Duchenne muscular dystrophy and Becker muscular dystrophy
muscles: changes contributing to preserve muscle function in Becker
muscular dystrophy patients, J. Cachexia. Sarcopenia Muscle. 11 (2020)
547e563, https://doi.org/10.1002/jcsm.12527.
[25] V. Ljubicic, P. Miura, M. Burt, L. Boudreault, S. Khogali, J.A. Lunde, J.-M. Renaud,
B.J. Jasmin, Chronic AMPK activation evokes the slow, oxidative myogenic
program and triggers beneficial adaptations in mdx mouse skeletal muscle,
Hum. Mol. Genet. 20 (2011) 3478e3493, https://doi.org/10.1093/hmg/
ddr265.
ꢁ
[26] C. Peladeau, A. Ahmed, A. Amirouche, T.E. Crawford Parks, L.M. Bronicki,
V. Ljubicic, J.-M. Renaud, B.J. Jasmin, Combinatorial therapeutic activation
with heparin and AICAR stimulates additive effects on utrophin A expression
in dystrophic muscles, Hum. Mol. Genet. 25 (2016) 24e43, https://doi.org/
10.1093/hmg/ddv444.
[27] P. Miura, J.V. Chakkalakal, L. Boudreault, G. Belanger, R.L. Hebert, J.-M. Renaud,
B.J. Jasmin, Pharmacological activation of PPARb/d stimulates utrophin A
expression in skeletal muscle fibers and restores sarcolemmal integrity in
mature mdx mice, Hum. Mol. Genet. 18 (2009) 4640e4649, https://doi.org/
10.1093/hmg/ddp431.
[9] I.V.L. Wilkinson, K.J. Perkins, H. Dugdale, L. Moir, A. Vuorinen,
M. Chatzopoulou, S.E. Squire, S. Monecke, A. Lomow, M. Geese, P.D. Charles,
P. Burch, J.M. Tinsley, G.M. Wynne, S.G. Davies, F.X. Wilson, F. Rastinejad,
S. Mohammed, K.E. Davies, A.J. Russell, Chemical proteomics and phenotypic
profiling identifies the aryl hydrocarbon receptor as a molecular target of the
utrophin modulator ezutromid, Angew. Chem. Int. Ed. 59 (2020) 2420e2428,
https://doi.org/10.1002/anie.201912392.
[28] STRING: Functional Protein Association Networks, (n.d.). https://string-db.org/
(accessed April 15, 2020).
[29] A.D. Rouillard, G.W. Gundersen, N.F. Fernandez, Z. Wang, C.D. Monteiro,
M.G. McDermott, A. Ma’ayan, The harmonizome: a collection of processed
datasets gathered to serve and mine knowledge about genes and proteins,
Database (2016), https://doi.org/10.1093/database/baw100, 2016.
[10] F. Lovering, J. Bikker, C. Humblet, Escape from flatland: increasing saturation
as an approach to improving clinical success, J. Med. Chem. 52 (2009)
8