Structure-activity relationship study and optimisation of 2-aminopyrrole-1-benzyl-4,5-diphenyl-1H-pyrrole-3-carbonitrile as a broad spectrum metallo-β-lactamase inhibitor

Article abstract of DOI:10.1016/j.ejmech.2017.05.061

A SAR study on derivatives of 2-amino-1-benzyl-4,5-diphenyl-1H-pyrrole-3-carbonitrile 5a revealed that the 3-carbonitrile group, vicinal 4,5-diphenyl and N-benzyl side chains of the pyrrole are important for the inhibitory potencies of these compounds against members representing the three main subclasses of metallo-β-lactamases (MBLs), i.e. IMP-1 (representing the B1 subgroup), CphA (B2) and AIM-1 (B3). Coupling of 5a with a series of acyl chlorides and anhydrides led to the discovery of two N-acylamide derivatives, 10 and 11, as the two most potent IMP-1 inhibitors in this series. However, these compounds are less effective towards CphA and AIM-1. The N-benzoyl derivative of 5a retained potent in vitro activity against each of MBLs tested (with inhibition constants in the low μM range). Importantly, this compound also significantly enhanced the sensitivity of IMP-1, CphA- or AIM-1-producing cell cultures towards meropenem. This compound presents a promising starting point for the development of a universal MBL inhibitor, targeting members of each of the major subgroups of this family of enzymes.

Full text of DOI:10.1016/j.ejmech.2017.05.061

Accepted Manuscript
Structure-activity relationship study and optimisation of 2-aminopyrrole-1-benzyl-4,5-
diphenyl-1H-pyrrole-3-carbonitrile as a broad spectrum metallo-β-lactamase inhibitor
Ross P. McGeary, Daniel T.C. Tan, Christopher Selleck, Marcelo Monteiro Pedroso,
Hanna E. Sidjabat, Gerhard Schenk
PII:
S0223-5234(17)30432-4
10.1016/j.ejmech.2017.05.061
EJMECH 9491
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 3 March 2017
Revised Date: 16 April 2017
Accepted Date: 31 May 2017
Please cite this article as: R.P. McGeary, D.T.C. Tan, C. Selleck, M.M. Pedroso, H.E. Sidjabat,
G. Schenk, Structure-activity relationship study and optimisation of 2-aminopyrrole-1-benzyl-4,5-
diphenyl-1H-pyrrole-3-carbonitrile as a broad spectrum metallo-β-lactamase inhibitor, European Journal
of Medicinal Chemistry (2017), doi: 10.1016/j.ejmech.2017.05.061.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Structure-activity relationship study and optimisation of 2-aminopyrrole-1-
benzyl-4,5-diphenyl-1H-pyrrole-3-carbonitrile as a broad spectrum
metallo-β-lactamase inhibitor
a,
a
a
a
Ross P. McGeary *, Daniel T. C. Tan , Christopher Selleck , Marcelo Monteiro Pedroso ,
b
a
Hanna E. Sidjabat & Gerhard Schenk
a
The University of Queensland, School of Chemistry and Molecular Biosciences, St. Lucia,
QLD 4072, Australia.
b
The University of Queensland, UQ Centre for Clinical Research, Building 71/918, Royal
Brisbane & Women's Hospital Campus, Herston, QLD 4029, Australia.
*
Corresponding author
Tel. + 61-7-3365-3955
E-mail: r.mcgeary@uq.edu.au
1

ACCEPTED MANUSCRIPT
Abstract
A SAR study on derivatives of 2-amino-1-benzyl-4,5-diphenyl-1H-pyrrole-3-carbonitrile 5a
revealed that the 3-carbonitrile group, vicinal 4,5-diphenyl and N-benzyl side chains of the
pyrrole are important for the inhibitory potencies of these compounds against members
representing the three main subclasses of metallo-β-lactamases (MBLs), i.e. IMP-1
(representing the B1 subgroup), CphA (B2) and AIM-1 (B3). Coupling of 5a with a series of
acyl chlorides and anhydrides led to the discovery of two N-acylamide derivatives, 10 and 11,
as the two most potent IMP-1 inhibitors in this series. However, these compounds are less
effective towards CphA and AIM-1. The N-benzoyl derivative of 5a retained potent in vitro
activity against each of MBLs tested (with inhibition constants in the low µM range).
Importantly, this compound also significantly enhanced the sensitivity of IMP-1, CphA- or
AIM-1-producing cell cultures towards meropenem. This compound presents a promising
starting point for the development of a universal MBL inhibitor, targeting members of each
of the major subgroups of this family of enzymes.
Key words
Metallo-β-lactamase
Enzyme Inhibition
Antibiotics Resistance
2

ACCEPTED MANUSCRIPT
1
. Introduction
It is well accepted that the global over-use of antibiotics, such as β-lactam antibiotics [1] in
both humans and livestock is a major cause of the rapid emergence of antibiotic resistance in
a growing number of pathogens [2, 3]. As a result, there is a dire need for novel classes of
antibiotics [3, 4]. The continuous search for such compounds appears unavoidable due to the
ability of pathogens to respond to the evolutionary pressure exerted by these compounds by
evolving novel resistance strategies. A highly successful strategy involves the use of β-
lactamases, enzymes which hydrolytically degrade β-lactam antibiotics [5-7].
β-Lactamases are categorised into four classes: A, B, C and D [8]. The class A, C and D β-
lactamases utilise a nucleophilic serine residue to cleave the β-lactam ring, and hence are
known as serine-β-lactamases (SBLs) [5, 8]. Inhibitors for the class A and D SBLs have been
successfully developed and are clinically administered in combination with β-lactam
antibiotics [5]. On the other hand, the zinc-dependent class B β-lactamases, or metallo-β-
lactamases (MBLs), are not inhibited by any of the clinically available SBL inhibitors [5, 6,
8
, 9]. In addition, the MBLs are capable of hydrolysing a wide range of β-lactam antibiotics,
including carbapenems such as imipenem (1, Figure 1), which are regarded as ‘last good’ or
last-line’ antibiotics for the treatment of infections caused by antibiotic-resistant bacteria [5,
, 8-10]. Moreover, the resistant genes responsible for the expression of these metallo-
‘
6
enzymes are highly transferable on bacterial mobile genetic elements such as plasmids, and
have been widely disseminating among pathogenic Gram-negative bacteria [1, 2, 6, 8, 9].
Exacerbating the threat of MBLs to health care is the observation these enzymes can alter and
expand their substrate range easily via mutations, thus adapting to novel antibiotics and
potentially avoiding inhibitors [11-13]. All of these reasons therefore require the issue of
MBL-mediated bacterial resistance to be addressed urgently.
3

ACCEPTED MANUSCRIPT
Figure 1. Chemical structure of Imipenem 1.
MBLs are subdivided based on their sequence homology, metal ion requirement for activity,
and substrate selectivity into as many as four sub-classes: B1, B2, B3 and B4 [6, 7, 14-17].
The B1 and B3 MBLs are catalytically active with two zinc ions, and exhibit a broad
substrate profile, whereas the mono-zinc B2 MBLs are solely carbapenemases [6, 14]. Little
is known about members of the B4 subgroup other than that they may form a catalytically
active di-metallic centre only in the presence of a substrate [17]. Since none of the B4
enzymes has yet been associated with antibiotic resistance they have been excluded from the
present study.
Since its discovery in 1988 in Japan, the plasmid-acquired B1 MBL Imipenemase 1 (IMP-1)
has been detected world-wide [9]. The global emergence of IMP-1, combined with extensive
structural and kinetics information of this enzyme reported in the literature, make it a suitable
starting point in the quest to develop novel, clinically useful MBL inhibitors [18-20]. CphA is
an MBL from the pathogenic Aeromonas hydrophilia. It belongs to the B2 subclass of MBL
and is of interest because it has a narrow specificity profile, efficiently hydrolysing only
carbapenems [21]. Another MBL of clinical relevance and interest to us is the recently
discovered Adelaide Imipenemase (AIM-1) [22], so named after its place of isolation in
Australia [23]. AIM-1 is currently the only known B3 MBL encoded on a mobile transferable
genetic element [23], and it can easily be mutated to become an efficient catalyst that
4

ACCEPTED MANUSCRIPT
hydrolyses members of all major β-lactam groups [11]. While there have been several
inhibitors of IMP-1 reported [18, 19, 24-29] and some CphA inhibitors [21, 30, 31], thus far
there are no known inhibitors of AIM-1.
As part of our program into the development of MBL inhibitors, we identified pyrrole 5a as a
competitive inhibitor of IMP-1, possessing an inhibition constant (K ) of 21 ± 10 µM [27].
i
Docking simulations of compound 5a bound in the active site of IMP-1 suggested
hydrophobic interactions of the 4,5-diphenyl groups of 5a with Val25, Trp28 and Val31,
along with hydrogen bonding stabilisations between (i) the amino functional group of 5a and
His139 (N-N distance 3.03 Å) and (ii) the cyano nitrogen of 5a and Lys161 (N-N distance
3
.10 Å). Curiously, no interaction between 5a and either of the zinc ions in the active site of
IMP-1 was predicted by the modelling program [27] (Figure 2).
Figure 2. Computational model of pyrrole 5a docked into the active site of IMP-1 as
predicted by MolDock [32]. Atom colours: carbon on ligand in olive green, carbon on IMP-1
2
+
in grey, nitrogen in blue, oxygen in red and hydrogen in white. The Zn ions are shown as
red spheres. The dashed, green lines represent hydrogen bonds. Adapted from [27].
5

ACCEPTED MANUSCRIPT
To further explore the potential mode of binding of 5a in the active site of IMP-1, to
determine SAR relationships, and to examine specificity of this inhibitor with respect to the
B2 and B3 MBLs CphA and AIM-1, respectively, we decided to examine the effects of (i)
substituting the vicinal 4,5-diphenyl rings with different aromatic and aliphatic groups; (ii)
acylating the 2-amino group with various aliphatic and aromatic groups; (iii) modifying the
3
-carbonitrile group, and (iv) varying the chain length of the phenyl group attached to the
pyrrole nitrogen.
2
. Chemistry
The synthetic routes to compounds 5-19 are shown in Schemes 1-5. Syntheses of the 2-
amino-3-cyanopyrroles were accomplished by treating α-hydroxyketones 2a-c with suitable
primary amines 3a-f in hot toluene in the presence of a catalytic amount of trichloroacetic
acid. After initial formation of the unisolated α-aminoketones 4, addition to the reaction
mixture of malononitrile led to ring-closure and heterocycle formation [33, 34]. Using this
approach, pyrroles 5a-g were prepared in good yields (Scheme 1). Pyrrole 5h could not be
prepared by this method. Instead, 5h was prepared in two steps as shown in Scheme 2.
Heating benzoin 2a with aniline 3f gave the α-aminoketone 6, which was isolated in 63%
yield. Subsequent treatment of 6 with malononitrile and sodium ethoxide in ethanol [34] gave
the desired N-phenylpyrrole 5h in 60% yield.
6

ACCEPTED MANUSCRIPT
O
R
CN
R
R
NH2
NH
R
NH2
O
n
n
N
a
b
R
n
R
OH
R1
R1
R1
2
2
2
a, R = Ph
b, R = Me
c, R = 2-furanyl 3c, R = H, n = 2
3a, R = H, n = 1
4
5a, R = Ph, R = H, n = 1
1
1
3b, R = OMe, n = 1
5b, R = Me, R = H, n = 1
1
1
5c, R = 2-furanyl, R = H, n = 1
1
1
3
3
3
d, R = H, n = 3
5d, R = Ph, R = OMe, n = 1
1
1
e, R = H, n = 4
5e, R = Ph, R = H, n = 2
1
1
f, R = H, n = 0
5f, R = Ph, R = H, n = 3
1
1
5
g, R = Ph, R = H, n = 4
1
5
h, R = Ph, R = H, n = 0
1
Scheme 1. Reagents and conditions: (a) Cl CCO H, toluene, ∆, 4-6 h; (b) CH (CN) , ∆,
3
2
2
2
overnight. 5a (67%); 5b (61%); 5c (59%); 5d (63%); 5e (78%); 5f (53%); 5g (78%); 5h
0%).
(
Scheme 2. Reagents and conditions: (a) Cl CCO H, toluene, ∆ (63%); (b) CH (CN) , NaOEt,
3
2
2
2
EtOH, ∆ (60%).
Initial attempts to acylate the poorly nucleophilic 2-amino group of 5a with one molar
equivalent of various aromatic and aliphatic acid chlorides led in some cases to poor yields of
the desired monoacylated derivatives, For example, attempted acylation of 5a with benzoyl
chloride following the method of Müller et al. [35] gave only 37% yield of the benzamide 8b,
7

ACCEPTED MANUSCRIPT
along with 52% recovered starting material. Repeating the reaction with an excess of acid
chloride or acid anhydride in refluxing toluene led to diacylation, resulting in the formation
of the imides 7a-d in good yields (Scheme 3). These imides could be converted to their
corresponding amides 8a-d by partial deacylation using methanolic potassium carbonate
(Scheme 3).
Scheme 3. Reagents and conditions: (a) (i) Ac O, ∆, 3 h, (ii) 25% NH , 7a (79%) or RCOCl,
2
3
Et N, toluene, ∆, overnight, 7b (76%); 7c (86%); 7d (93%); (b) K CO , MeOH, r.t., 1 h, 8a
3
2
3
(68%); 8b (56%); 8c (95%); 8d (87%).
Imides 7 were only formed in some cases from pyrrole 5a using the acylation conditions
described in Scheme 3. In other cases only the mono-acylated amides were obtained (i.e., 11,
1
4 and 15) and for some acylating agents a mixture of mono- (i.e., 10 and 13) and diacylated
products resulted (i.e., 9 and 12). The reaction of pyrrole 5a with phthalic anhydride gave the
expected imide 16 (Figure 3).
8

ACCEPTED MANUSCRIPT
Ph
Ph
CN
Ph
CN
Ph
CN
O
O
O
NO2
N
Ph
N
Ph
N
N
N
N
H
H
O
NO2
Ph
Ph
Ph
O
NO2
9
(26%)
10 (36%)
11 (47%)
Ph
Ph
CN
Ph
Ph
CN
O
Ph
Ph
CN
O
O
NO2
N
N
H
N
N
N
N
H
O
Ph
Ph
Ph
1
2 (59%)
13 (10%)
14 (46%)
Ph
Ph
CN
Ph
Ph
CN
O
O
N
H
N
N
N
Ph
Ph
O
15 (66%)
16 (43%)
Figure 3. Amide and imide products arising from the acylation of pyrrole 5a. Reaction
conditions used were the same as those described in Scheme 3.
To investigate the effect that the nitrile group of 5a has on its inhibitory activity, two
transformations of this group were performed. The first, involving the hydrolysis of the nitrile
to the carboxamide proceeded uneventfully using hot polyphosphoric acid (Scheme 4) [36].
The second derivatisation of the nitrile group of 5a, a 1,3-dipolar cycloadddtion with azide to
give a tetrazole, proved to be more troublesome. Attempted reaction of 5a with sodium azide
in the presence of either triethylammonium hydrogensulfate in hot toluene [37], or with
sodium azide and ammonium chloride in hot N,N-dimethylformamide [38] failed. The
desired transformation was finally achieved, albeit in low yield, by reacting pyrrole 5a with
aluminium triazide, formed in situ using trimethylsilyl azide and trimethylaluminium, in hot
toluene [39] (Scheme 4).
9

ACCEPTED MANUSCRIPT
Scheme 4. Reagents and conditions: (a) (i) P O , H PO , ∆, 5 h; (ii) 25% NH (79%); (b)
2
5
3
4
3
TMSN , Me Al, toluene, ∆, 28 h (26%).
3
3
The final desired transformation of 5a was removal of the N-benzyl group, to determine its
role in the inhibitory activity of this compound. This reaction also proved to be problematic:
attempted hydrogenolysis of the N-benzyl group of 5a using hydrogen gas over palladium-
carbon [40] failed, as did attempted debenzylation with aluminium trichloride [41]. In both
cases only starting material was recovered. The required product 19 was ultimately obtained
by debenzylation of the N-(4-methoxybenzyl) derivative 5d using aluminium trichloride in
anisole [42].
Scheme 5. Reagents and conditions: (a) AlCl , PhOMe, ∆, 17 h (36%).
3
3
. Results and Discussion
Initial studies of the structure-activity relationship of the aminopyrroles 5a-h focused on
IMP-1 inhibition. Table 1 shows the inhibitory activity of compounds 5a-h, all measured at
1
0

ACCEPTED MANUSCRIPT
1
0 µM, against IMP-1. It can be seen that replacement of the geminal diphenyl groups of the
starting molecule 5a with methyl groups (5b) leads to significant loss of potency. Substitution
of the phenyl groups with furan rings does not increase potency (5c), nor does the addition of
a 4-methoxy group to the N-benzyl moiety (5d). Extending the N-benzyl group of 5a by one,
two or three methylenes (compounds 5e, 5f and 5g, respectively) results in marked loss of
inhibitory activity, and activity is abolished when the N-benzyl group of 5a is replaced with a
an N-phenyl group (5h). These results indicate the crucial role that the benzylic group (benzyl
or 4-methoxybenzyl) on the pyrrole nitrogen plays in maintaining inhibitory activity against
IMP-1, and the importance of the two aromatic groups on the 4- and 5-positions of the
pyrrole.
Table 1. Inhibition data for 2-amino-3-cyanopyrroles 5a-h (10 µM) against IMP-1 activity.
Inhibitor
% Inhibition
26.0 ± 2.7
2.3 ± 3.3
Inhibitor
% Inhibition
4.8 ± 0.8
4.1 ± 0.8
8.7 ± 1.1
0
5a
5e
5f
5b
5c
23.2 ± 3.9
20.1 ± 4.3
5g
5h
5d
For the imide (7a-d) and amide (8a-d) derivatives of 5a, the data in Table 2 show that the
aliphatic amides and imides were less active against IMP-1 than 5a. The aromatic imide (7b)
and aromatic amide (8b), however, were both more potent than 5a. The dibenzoyl imide (7b)
was later shown to exhibit a mixed inhibition mode against IMP-1, with a less potent K value
i
(see below). To test where the imide (7b) was hydrolysing during the course of the assay to
the amide (8b), 7b was incubated with IMP-1 overnight. No trace of 8b was detected in the
1
1

ACCEPTED MANUSCRIPT
mixture using electrospray ionisation mass spectrometry (ESI-MS), suggesting that the imide
7
b is stable towards hydrolysis by the enzyme (data not shown).
Other than the dibenzoyl imide 7b and the amide 8b, the 4-nitrobenzamide (10), the 3-
nitrobenzamide (14), the 4-benzoylbenzamide (11) and the 3,5-dimethylbenzamide derivative
(13) all showed inhibitory activity greater than 5a. Additional kinetic data for these
compounds are discussed below.
Table 2. Inhibition data for N-acyl amides and N,N-diacyl imides 7-16 (10 µM) against IMP-
1
activity. For comparison, compound 5a showed inhibition of 26% at 10 µM (Table 1).
Inhibitor
% Inhibition
18.6 ± 4.8
48.0 ± 5.9
17.1 ± 3.4
14.3 ± 2.4
16.1 ± 6.6
50.9 ± 2.5
21.0 ± 2.5
8.2 ± 4.5
Inhibitor
% Inhibition
20.2 ± 3.4
36.9 ± 2.4
44.3 ± 3.3
13.3 ± 2.4
32.3 ± 4.7
48.7 ± 1.9
23.3 ± 3.7
20.3 ± 3.7
7a
9
7b
10
11
12
13
14
15
16
7c
7d
8a
8b
8c
8d
Derivatives with modifications to the 3-cyano group (compounds 17-18) and deletion of the
N-benzyl group (19) were also examined (Table 3). Conversion of the cyano group of 5a to
the primary amide gave 17, which was devoid of activity. The tetrazole 18, formed by
cycloaddition of azide to the nitrile group of 5a, was poorly active, as was the de-N-
benzylated compound 19. These results indicate that the 3-nitrile group of 5a is necessary for
1
2

ACCEPTED MANUSCRIPT
potent activity of these inhibitors. Similarly, the N-benzyl group on the pyrrole is optimal
(compare 19 with 5d-h)
Table 3. Inhibition data for pyrrole derivatives 17-19 (10 µM) against IMP-1 activity.
Inhibitor
% Inhibition
0
Inhibitor
% Inhibition
8.7 ± 1.3
Inhibitor
% Inhibition
12.0 ± 2.6
17
18
19
The compounds that showed inhibitory activity higher than that of 5a were further examined
to determine their inhibition constants and their modes of inhibition (competitive or
uncompetitive) (Table 4). The most potent competitive inhibitor of the series is the 4-
nitrobenzamide 10, with a K value of 1.24 µM. The 4-benzoylbenzamide 11 is also a potent
ic
inhibitor, but its mode of inhibition is uncompetitive (Kiuc 1.52 µM). Benzamides 8b, 13 and
1
4 are slightly more potent than 5a, but each has a mixed mode of inhibitory activity, with
their uncompetitive inhibition constants being smaller than their competitive counterparts.
Imide 7b showed a modest K value of 55 µM.
iuc
These results suggest that the presence of an N-aroyl group appending the 3-amino group of
5
a can in some cases improve these compounds’ potencies against IMP-1. Notable is the
effect of these compounds on the mode of inhibition. For example, the 3-nitrobenzamide
derivative 14 inhibits potently and mainly uncompetitively, whereas the isomeric 4-nitro
analogue is exclusively a competitive inhibitor, as is the 4-benzoylbenzamide derivative 11.
The absence of a substituent (8b) or the presence of a 3,5-dimethylbenzamide group (13)
leads to mixed mode inhibition. Thus, relatively minor modifications have a significant
influence on the mode of their binding interactions. This observation, while previously noted
with other MBL inhibitors [17-22] may also be a reflection of conformational dynamics and
fitness landscape of MBLs [11-13, 43, 44].
1
3

ACCEPTED MANUSCRIPT
Table 4. Competitive (K ) and uncompetitive (K ) inhibition constants of inhibitors 5a, 7b,
ic
iuc
8
b, 10, 11, 13 and 14 against IMP-1.
Inhibitor
Structure
K (µM)
ic
K (µM)
iuc
a
5
a
21 ± 10
-
7
8
b
b
468 ± 6
10.7 ± 9.3
1.24 ± 0.59
-
55 ± 19
4.3 ± 1.8
-
1
1
0
1
1.52 ± 0.36
1
1
3
4
20.2 ± 10.9
23.7 ± 25.2
7.8 ± 2.5
3.3 ± 0.8
a
Data from [27]
1
4

ACCEPTED MANUSCRIPT
The ultimate goal from a clinical point of view is the development of an inhibitor that is
effective against MBLs from each of the subgroups, a so-called universal MBL inhibitor. The
structural and functional differences between the three subclasses of MBLs have made the
search for such broad-spectrum inhibitors challenging, and efforts towards this goal have met
with limited success. Several compounds have been shown to inhibit more than one subclass
of MBL. Schofield’s group has reported that thiomandelic acid 20 is a potent inhibitor of
several MBLs of the B1 and B3 subclasses, but not of CphA, a B2 MBL [45]. Further
development of inhibitors containing a thiol group led to the di(mercaptomethyl)benzoic acid
2
1 and the mercaptoacetophenone 22 as compounds with submicromolar inhibition against
all classes of MBL [31]. The same group later reported cyclic boronates as broad-spectrum
MBL inhibitors, e.g. 23, which are able to inhibit various MBLs and SBLs. Inhibitor 23
showed low- or submicromolar activity against a number of B1 subclass MBLs, as well as the
B2 MBL CphA. However, no activity for the cyclic boronates was reported for B3 MBLs
[
46].
Here, compounds 5a, 8b, 10 and 11 were additionally tested for inhibitory activity against the
B2 subclass MBL CphA and the B3 subclass MBL AIM-1. Results are summarised in Table
5
. The 4-nitrobenzamide compound 10 showed good activity against the B3 enzyme AIM-1,
but was devoid of activity against the B2 enzyme CphA. On the other hand compound 11,
while a potent IMP-1 inhibitor, was only weakly active against CphA, and not at all active
against AIM-1. In contrast, compound 5a was shown to be a potent inhibitor of both CphA
1
5

ACCEPTED MANUSCRIPT
and AIM-1, binding in a predominantly uncompetitive manner to the former and
competitively to the latter. Similarly, the benzamide derivative 8b was a potent mixed mode
inhibitor of CphA, and a potent competitive inhibitor of AIM-1. The observed differences in
binding modes reflect both structural differences among MBLs from different subgroups and
conformational dynamics that may facilitate alternative binding modes. The modes of binding
of the thiol inhibitors 20-22 in the active site of MBLs involves direct binding of the thiolate
group to the zinc ion(s). For the cyclic boronate 23, the oxygens attached to boron bind to
zinc, and the group acts as a mimic of the tetrahedral intermediate involved in β-lactam
hydrolysis [31, 46]. The modes of binding of compounds 5a and 8b have not yet been
visualised but the structure of these molecules suggests that a direct coordination to the metal
ion(s) in the active site may not be essential for a potent inhibitor. This interpretation is in
agreement with recent in vitro evolution studies that demonstrated that residues essential for
substrate, and by analogy, inhibitor recognition may not be close to the immediate
coordination sphere of the catalytically essential metal ions [11-13, 43, 44, 47-49].
Identifying and visualising such remote inhibitor binding sites may thus provide an attractive
target for the design of universal MBL inhibitors, and compounds such as 5a and 8b may
play a pivotal in this endeavour. Efforts towards these goals are currently in progress.
1
6

ACCEPTED MANUSCRIPT
Table 5. Comparison of inhibitory activities of compounds 5a, 8b, 10 and 11 against
representatives of each subclass of MBL.
Inhibitor
IMP-1 (B1)
K (µM)
CphA (B2)
AIM-1 (B3)
K (µM)
K (µM)
K (µM)
iuc
K (µM)
K (µM)
iuc
ic
iuc
ic
ic
5
a
21 ± 10
10.7 ± 9.3
1.24 ± 0.59
-
-
52 ± 7
11 ± 4
-
5.5 ± 0.8
9 ± 2
-
2.2 ± 0.3
1.6 ± 0.4
10.9 ± 2.7
-
12.5 ± 2.2
13 ± 13
47 ± 17
-
8
b
4.3 ± 1.8
-
1
1
0
1
1.52 ± 0.36 872 ± 50 161 ± 32
While the design of potent broad-spectrum inhibitors of MBLs is a crucial step in the fight
against the spread of antibiotic resistance a promising compound needs to display both
effective in vitro inhibition of MBLs and in vivo enhancement of the sensitivity of MBL-
producing pathogens towards β-lactam antibiotics. We thus measured the MIC values of
IMP-1-, CphA- and AIM-1-producing E. coli cultures in the absence and presence of
compounds 5a, 8b, 10 and 11. The results are summarised in Table 6. While none of these
compounds inhibit bacterial growth in the absence of meropenem (Supporting Information,
Figure S1), each of them inhibited cell growth in the presence of this antibiotic, consistent
with a synergetic effect between the compounds tested and the antibiotic (Supporting
Information, Figure S2). Among the four compounds tested 8b displayed the strongest effect
among the representatives of the three major MBL subclasses. Overall, a reasonable
agreement between inhibition constants (Table 5) and MIC values (Table 6) is observed. For
instance, compounds 10 and 11, which are either poor or no inhibitors for CphA and AIM-1
(Table 5), have also very modest effect on corresponding MIC values (Table 6). In contrast,
compounds 5a and 8b, the only compounds identified here as “universal” MBL inhibitors
1
7

ACCEPTED MANUSCRIPT
(Table 5), also have a significant effect on the MIC values of cell cultures expressing any of
the three MBLs tested here (Table 6). This correlation is most significant for compound 8b.
Relative to the blanks the sensitivity towards meropenem (measured in terms of the reduction
of the MIC values) is enhanced ~20-fold for IMP-1- and AIM-1-producing cells and over 40-
fold for the CphA-expressing strain.
Table 6. MIC values (µg/mL) for meropenem, measured in the absence (Blank) and presence
of compounds 5a, 8b, 10 or 11. Data were measured for E. coli BL21 expressing either IMP-
1
, CphA or AIM-1.
E. coli BL21 MICs of
MICs of meropenem measured in the presence of
inhibitor (µg/mL)
expressing
MBL
meropenem
without
5a
8b
10
11
inhibitor
(µg/mL)
IMP-1
CphA
AIM-1
1
1
2
0.5
0.047
0.25
0.047
0.023
0.094
0.25
0.75
0.5
0.5
0.75
0.75
The inhibitors were used in the concentration of 30 µM.
4
. Conclusion
Pyrrole 5a is a broad spectrum inhibitor of members of all three subclasses of MBLs,
although its mode of action varies from exclusively competitive (IMP-1) to mixed mode of
1
8

ACCEPTED MANUSCRIPT
inhibition (CphA and AIM-1). SAR studies show that the N-benzyl group is optimal for
activity. Removal of this group altogether, or increasing or decreasing the distance between
the nitrogen atom and the phenyl group decreases potency. The 4,5-geminal diphenyl groups
may be replaced with furans with no significant loss of activity, but replacement with methyl
groups leads to almost complete loss of activity. Derivatisation of the nitrile group of 5a to a
tetrazole led to a substantial loss of activity, and hydrolysis of the nitrile group to the primary
amide resulted in complete loss of activity. Benzoylation of 5a gave inhibitor 8b, which is
also a broad spectrum MBL inhibitor, more potent than 5a. Importantly, 8b also significantly
enhances the sensitivity of cell cultures expressing representatives from the three major
subgroups of MBLs between 20- to 40-fold. The 4-nitrobenzoyl inhibitor 10 is a potent
competitive inhibitor of IMP-1 and shows mixed mode inhibition against AIM-1. It is,
however, inactive towards CphA. The 4-(benzoyl)benzoyl derivative 11 is also a potent
inhibitor of IMP-1, although now in a purely uncompetitive mode. It is only weakly active
against CphA, and not at all active against AIM-1.
The enzyme inhibition and SAR data described in this study should aid in the development of
compounds that are more potent and active against all three subclasses of MBLs. In particular
compound 8b has emerged as a promising universal lead for MBL inhibitors. Further studies
will focus on investigating the inhibitory effects of other substituents on the 3-amino group of
5
a, and visualising the binding mode(s) of the 5a and 8b as they may provide insight into
sites remote from the metal binding centre that are effective for the development of universal
MBL inhibitors. Further studies will focus on diminishing the lipophilicity of these
compounds, investigating the inhibitory effects of other substituents on the 3-amino group of
5
a, and visualising the binding mode(s) of the 5a and 8b as they may provide insight into
sites remote from the metal binding centre that are effective for the development of universal
MBL inhibitors.
1
9

ACCEPTED MANUSCRIPT
5
. Experimental
5
.1. General experimental
Thionyl chloride and petroleum ether, PE (b.p. 40-60 °C) were distilled before use. Other
chemicals and reagents were used without prior purification. NMR spectra were recorded on
Bruker AVANCE 500 or 400 MHz spectrometers. Chemical shifts are reported in parts per
1
13
million (ppm) on a scale relative to the solvent peak CDCl ( H, 7.24 ppm; C, 77.0 ppm)
3
1
13
and DMSO-d ( H, 2.49 ppm; C, 39.5 ppm) Coupling constants (J) are reported in Hz and
6
.
peak multiplicities described as singlet (s), doublet (d), triplet (t), doublet of doublets (dd),
quintet (quin), triplet of triplets (tt), multiplet (m), or broad (br). Low-resolution ESI-MS
measurements were carried out on a Bruker Esquire HCT (High Capacity 3D ion trap)
instrument with a Bruker ESI source. High-resolution electrospray ionisation accurate mass
measurements (HRMS) were recorded on a Bruker MicrOTOF-Q (quadrupole–Time of
Flight) instrument with a Bruker ESI source. Accurate mass measurements were carried out
with external calibration using sodium formate as reference calibrant. Low resolution and
accurate mass EI were carried out on a Finnigan MAT 900XL Trap instrument at 70eV
ionising energy. Accurate mass samples were carried out at 5000 Resolving Power as electric
sector scan measurements using high boiling PFK (perfluorokerosene) as the reference
material. Infrared (IR) spectroscopy samples were analysed as neat on a Perkin Elmer
Frontier FT-IR/NIR Spectrometer.
5
.2. 2-Amino-1-benzyl-4,5-diphenyl-1H-pyrrole-3-carbonitrile (5a) [33, 50]
2
0

ACCEPTED MANUSCRIPT
A mixture of benzoin 2a (4.2 g, 20 mmol), benzylamine 3a (2.1 mL, 2.1 g, 20 mmol) and
trichloroacetic acid, TCA (0.17 g, 1.0 mmol) in dry toluene (20 mL) was heated under reflux
using a Dean-Stark apparatus for 2 h. A yellow solution was obtained. Malononitrile (0.66 g,
1
0 mmol) was then added to the reaction mixture, followed by 3 drops of pyridine. The
mixture was further heated under reflux under argon for 17 h. A colour change from yellow
to olive green and finally dark brown was observed. The excess solvent was removed in
vacuo to give a brown semi solid residue (7.33 g). The residue was purified by silica gel
column chromatography [EtOAc/PE (2:23) → EtOAc/PE (7:43)], followed by
recrystallisation from diethyl ether to afford the title compound 5a as a light brown solid
(
4.67 g, 67%), m.p.:173.5-175.9 °C (lit.[34] m.p.: 173-175 °C). R : 0.42 (30% EtOAc in PE,
f
1
visible as a blue spot under UV light). H NMR (500 MHz, CDCl ) δ (ppm): 3.85 (br s, 2H,
3
NH , D O exchangeable), 4.90 (s, 2H, NCH Ph), 7.06-7.08 (m, 2H, Ar-H), 7.13-7.15 (m, 3H,
2
2
2
1
3
Ar-H), 7.18-7.32 (m, 8H, Ar-H), 7.35-7.38 (m, 2H, Ar-H). C NMR (125 MHz, CDCl ) δ
3
(
ppm): 47.0 (NCH Ph), 75.9 (C-CN), 117.5 (C≡N), 121.0 (Ar-C), 125.65 (Ar-C), 125.63
2
(Ar-C), 126.4 (Ar-CH), 128.0 (Ar-CH), 128.1 (Ar-CH), 128.2 (Ar-CH), 128.67 (Ar-CH),
1
28.72 (Ar-CH), 129.3 (Ar-CH), 130.8 (Ar-CH), 131.0 (Ar-CH), 133.2 (Ar-C), 136.1 (Ar-
+
-1
C), 145.9 (C-NH ). ESI-MS: m/z 350.1 [M + H] . IR ν
(cm ): 3406.0 (N-H stretch),
2
max
3
1
333.0 (N-H stretch), 2199.7 (C≡N stretch), 1624.5, 1601.7, 1541.8, 1496.8, 1470.9, 1438.1,
1
365.4, 1339.3. H NMR and ESI-MS data are in agreement with the literature [33].
5
.3. 2-Amino-1-benzyl-4,5-dimethyl-1H-pyrrole-3-carbonitrile 5b
Acetoin 2b (0.88 g, 10 mmol) and benzylamine 3a (1.1 mL, 1.1 g, 10 mmol) were used for
the synthesis of the title compound 5b, which was obtained as a dark brown solid after silica
gel column chromatography of the crude product [EtOAc/PE (1:4) → EtOAc/PE (3:7)] (2.3
2
1

ACCEPTED MANUSCRIPT
g, 61%), m.p.: 110.3-115.9 °C. R : 0.52 (30% EtOAc in PE, visible as a blue spot under 254
f
1
nm). H NMR (400 MHz, CDCl ) δ (ppm): 1.99 (d, 3H, J = 0.5 Hz, CH ), 2.03 (d, 3H, J =
3
3
0
.6 Hz, CH ), 3.59 (br s, 2H, NH , exchangeable with D O), 4.90 (s, 2H, NCH Ph), 6.96-
3 2 2 2
13
6
.99 (m, 2H, Ar-H), 7.26-7.34 (m, 3H, Ar-H). C NMR (100 MHz, CDCl ) δ (ppm): 9.5
3
(
CCH ), 10.0 (CCH ), 45.9 (NCH Ph), 113.5 (C), 117.7 (C≡N), 119.2 (C), 125.7 (Ar-CH),
3 3 2
+
1
27.8 (Ar-CH), 129.1 (Ar-CH), 136.3 (Ar-C) 143.2 (C-NH ); LRMS: m/z 248.1 [M + Na] .
2
-1
IR νmax (cm ): 3418.5 (N-H stretch), 3330.8 (N-H stretch), 3233.3, 2202.1 (C≡N stretch),
1
1
650.9, 1603.3, 1564.2, 1509.2, 1496.2, 1455.0, 1441.9, 1354.6, 1303.7, 1244.4, 1176.6,
146.9, 1016.4.
5
.4. 2-Amino-1-benzyl-4,5-di(furan-2-yl)-1H-pyrrole-3-carbonitrile (5c)
Furoin 2c (1.92 g, 10.0 mmol) and benzylamine 3a (1.1 mL, 1.1 g, 10 mmol) were used for
the synthesis of the title compound 5c, which was obtained as a dark brown solid after silica
gel column chromatography of the crude product [EtOAc/PE (1:4) → EtOAc/PE (3:7)] (1.9
g, 59%), m.p.: 138.0-145.8 °C. R : 0.54 (30% EtOAc in PE, visible as a blue spot under UV
f
1
light). H NMR (400 MHz, CDCl ) δ (ppm): 3.91 (br s, 2H, NH , D O exchangeable), 4.88
3
2
2
(
s, 2H, NCH Ph), 6.34 (d, 2H, J = 1.2 Hz, furanyl-H), 6.43-6.46 (m, 2H, furanyl-H), 7.09-
2
7
.11 (m, 2H, Ar-H), 7.35-7.36 (m, 4H, Ar-H), 7.47 (dd, 1H, J = 0.8 Hz, J = 1.8 Hz, Ar-H).
1 2
1
3
C NMR (100 MHz, CDCl ) δ (ppm): 47.4 (NCH Ph), 74.1 (C-CN), 106.6 (Ar-CH), 111.0
3
2
(Ar-CH), 111.1 (C), 113.1 (Ar-CH), 114.2 (Ar-C), 115.2 (C), 116.7 (C≡N), 126.3 (Ar-CH),
1
28.2 (Ar-CH), 129.2 (Ar-CH), 135.5 (C) 141.4 (Ar-CH), 143.3 (Ar-CH), 143.5 (C), 146.4
-
-1
(
C-NH ), 147.3 (C); LRMS: m/z 328.1 [M + H] . IR ν
(cm ): 3454.2 (N-H stretch),
2
max
3
327.5 (N-H stretch), 3269.0, 3236.6, 3210.6, 3152.1, 3116.4, 2200.1 (C≡N stretch), 1673.7,
1
633.9, 1562.7, 1497.0, 1456.0, 1346.0, 1203.0, 1156.9, 1013.0.
2
2

ACCEPTED MANUSCRIPT
5
.5. 2-Amino-1-(4-methoxybenzyl)-4,5-diphenyl-1H-pyrrole-3-carbonitrile 5d
Benzoin 2a (2.12 g, 10.0 mmol) and 4-methoxybenzylamine 3b (1.3 mL, 1.4 g, 10 mmol)
were used for the synthesis of the title pyrrole 5d, which was obtained as a golden solid (2.4
g, 63%), m.p.: 148.0-150.6 °C. R : 0.40 (30% EtOAc in PE, visible as a blue spot under UV
f
1
light). H NMR (400 MHz, CDCl ) δ (ppm): 3.79 (s, 3H, OCH ), 3.88 (br s, 2H, NH , D O
3
3
2
2
exchangeable), 4.84 (s, 2H, NCH PhOCH ), 6.88 (d, 2H, J = 8.8 Hz, Ar-H), 6.99 (d, 2H, J =
2
3
1
3
8
.8 Hz, Ar-H), 7.14-7.27 (m, 10H, Ar-H). C NMR (100 MHz, CDCl ) δ (ppm): 46.5
3
(
NCH PhOCH ), 55.3 (OCH ), 75.7 (C-C≡N), 114.7 (Ar-CH), 117.5 (C≡N), 120.9 (Ar-C),
2 3 3
1
25.3 (Ar-CH), 125.6 (Ar-C), 126.4 (Ar-CH), 127.2 (Ar-CH), 127.8 (Ar-C), 128.1 (Ar-CH),
1
28.2 (Ar-CH), 128.68 (Ar-CH), 128.72 (Ar-CH), 129.0 (Ar-CH), 130.9 (Ar-C) 131.0 (Ar-
-
CH), 133.2 (Ar-C), 137.8 (Ar-C), 145.9 (Ar-C), 159.3 (Ar-C); LRMS: m/z 378.2 [M - H] .
-1
HRMS, found: m/z 378.1599, C H N O requires 378.1612. IR ν (cm ): 3416.3 (N-H
max
25
20
3
stretch), 3328.3 (N-H stretch), 3233.7, 2202.9 (C≡N stretch), 1651.5, 1603.3, 1564.4, 1508.8,
1
495.9, 1455.4, 1441.9, 1354.6, 1244.0, 1176.6, 1016.5.
5
.6. 2-Amino-1-phenethyl-4,5-diphenyl-1H-pyrrole-3-carbonitrile (5e)
Benzoin 2a (2.12 g, 10.0 mmol) and phenethylamine 3c (1.3 mL, 1.2 g, 10 mmol) were used
for the synthesis of the title compound. Silica gel column chromatography [EtOAc/PE (1:9)
→
EtOAc/PE (23:77)] of the crude product afforded the title pyrrole 5e as a brown solid (2.8
o
g, 78%), m.p.: 186.2-190.7 C. R : 0.33 (30% EtOAc in PE, visible as a blue spot under UV
f
1
light). H NMR (400 MHz, CDCl ) δ (ppm): 2.71 (t, 2H, J = 7.0 Hz, CH Ph), 3.46 (br s, 2H,
3
2
NH , D O exchangeable), 3.90 (t, 2H, J = 7.0 Hz, NCH ), 6.91-6.93 (m, 2H, Ar-H), 7.11-
2
2
2
2
3

ACCEPTED MANUSCRIPT
1
3
7
.25 (m, 10H, Ar-H), 7.34-7.36 (m, 3H, Ar-H). C NMR (100 MHz, CDCl ) δ (ppm): 36.2
3
(
CH ), 45.0 (N-CH ), 76.1 (C-CN), 117.7 (C≡N), 121.1 (Ar-C), 125.2 (Ar-C), 126.3 (Ar-
2 2
CH), 127.1 (Ar-CH), 128.1 (Ar-CH), 128.3 (Ar-CH), 128.5 (Ar-CH), 128.7 (Ar-CH), 128.8
Ar-CH), 128.9 (Ar-CH), 131.5 (Ar-CH), 133.1 (Ar-C), 137.6 (Ar-C), 145.5 (Ar-C); LRMS:
(
-
-
m/z 362.0 [M - H] . HRMS, found: m/z 362.1663, C H N requires 362.1663. IR ν (cm
25
20
3
max
1
)
: 3428.2 (N-H stretch), 3349.0 (N-H stretch), 3059.0, 3025.8, 2937.0, 2194.5 (C≡N stretch),
624.0, 1601.2, 1545.9, 1496.9, 1481.7, 1453.8, 1442.1, 1350.8, 1224.0, 1141.2, 1029.5,
013.7.
1
1
5
.7. 2-Amino-4,5-diphenyl-1-(3-phenylpropyl)-1H-pyrrole-3-carbonitrile 5f
Benzoin 2a (2.12 g, 10.0 mmol) and 3-phenyl-1-propylamine 3d (1.4 mL, 1.4 g, 10 mmol)
were used for the synthesis of the title compound. Filtration of the reaction mixture and
washing of the precipitate with cold toluene (2 x 30 mL) afforded the pyrrole 5f as a light
o
brown solid (2.0 g, 53%), m.p.: 159.7-169.0 C. R : 0.38 (30% EtOAc in PE, visible as a blue
f
1
spot under UV light). H NMR (400 MHz, CDCl ) δ (ppm): 1.85 (quin, 2H, J = 7.6 Hz,
3
CH CH Ph), 2.46 (t, 2H, J = 7.4 Hz, CH Ph), 3.65 (t, 2H, J = 7.8 Hz, NCH ), 7.00-7.02 (m,
2
2
2
2
13
2
H, Ar-H), 7.09-7.26 (m, 10H, Ar-H), 7.30-7.39 (m, 3H, Ar-H). C NMR (100 MHz,
CDCl ) δ (ppm): 30.7 (CH ), 32.5 (CH ), 42.6 (N-CH ), 75.7 (C-CN), 117.7 (C≡N), 120.9
3
2
2
2
(Ar-C), 125.3 (Ar-C), 126.3 (Ar-CH), 128.09 (Ar-CH), 128.13 (Ar-CH), 128.56 (Ar-CH),
1
28.59 (Ar-CH), 128.7 (Ar-CH), 128.91 (Ar-CH), 128.94 (Ar-CH), 131.2 (Ar-CH), 133.1
-
(
Ar-C), 140.2 (Ar-C), 145.2 (Ar-C); LRMS: m/z 376.1 [M - H] . HRMS, found: m/z
-1
3
76.1815, C H N requires 376.1819. IR ν
(cm ): 3420.0 (N-H stretch), 3338.3 (N-H
2
6
22
3
max
stretch), 3243.6, 3059.3, 3024.3, 2934.1, 2856.0, 2191.5 (C≡N stretch), 1947.4, 1632.5,
2
4

ACCEPTED MANUSCRIPT
1
1
601.3, 1558.4, 1505.4, 1495.3, 1462.9, 1453.1, 1444.1, 1372.4, 1356.8, 1320.3, 1227.9,
211.7, 1179.3, 1151.3, 1110.2, 1079.8, 1029.4.
5
.8. 2-Amino-4,5-diphenyl-1-(4-phenylbutyl)-1H-pyrrole-3-carbonitrile 5g
Benzoin 2a (2.12 g, 10.0 mmol) and 4-phenylbutylamine 3e (1.6 mL, 1.5 g, 10 mmol) were
used for the synthesis of the title compound. Silica gel column chromatography [EtOAc/PE
(1:4) → EtOAc/PE (23:77)] of the crude product afforded the pyrrole 5g as a brown solid
o
(
2.8 g, 78%), m.p.: 150.3-153.1 C. R : 0.40 (30% EtOAc in PE, visible as a blue spot under
f
1
UV light). H NMR (400 MHz, CDCl ) δ (ppm): 1.42-1.57 (m, 4H 2 x CH ), 2.47 (t, 2H, J =
3
2
7
1
2
.3 Hz, CH Ph), 3.64 (t, 2H, J = 7.0 Hz, NCH ), 7.02-7.05 (m, 2H, Ar-H), 7.16-7.26 (m,
2 2
1
3
0H, Ar-H), 7.30-7.31 (m, 3H, Ar-H). C NMR (100 MHz, CDCl ) δ (ppm): 27.8 (CH ),
3
2
8.9 (CH ), 34.9 (CH ), 43.2 (N-CH ), 75.8 (C-CN), 117.7 (C≡N), 120.9 (Ar-C), 125.2 (Ar-
2
2
2
C), 126.0 (Ar-CH), 126.2 (Ar-CH), 128.08 (Ar-CH), 128.12 (Ar-CH), 128.3 (Ar-CH), 128.4
(Ar-CH), 128.59 (Ar-CH), 128.64 (Ar-CH), 131.3 (Ar-CH), 133.2 (Ar-C), 141.3 (Ar-C),
-
1
3
2
1
7
45.1 (Ar-C); LRMS: m/z 390.1 [M - H] . HRMS, found: m/z 390.1970, C H N requires
27
24
3
-1
90.1976. IR νmax (cm ): 3418.6 (N-H stretch), 3341.1 (N-H stretch), 3243.8, 3061.6, 3024.6,
934.2, 2857.4, 2191.2 (C≡N stretch), 1947.6, 1631.6, 1601.0, 1558.2, 1495.4, 1462.7,
453.2, 1443.5, 1357.3, 1319.7, 1225.9, 1212.0, 1179.6, 1150.0, 1110.5, 1074.7, 1029.3,
33.5.
5
.9. 2-Amino-1,4,5-triphenyl-1H-pyrrole-3-carbonitrile 5h [34]
Malononitrile (0.18 g, 2.7 mmol) was added to a hot solution of α–aminoketone 6 (0.78 g, 2.7
mmol) in absolute EtOH (6 mL). A solution of sodium (72 mg, 3.1 mmol) in absolute EtOH
2
5

ACCEPTED MANUSCRIPT
(3 mL) was then added dropwise to the mixture, and the mixture was heated under reflux for
1
8 h. The orange solid which formed was filtered off and washed with EtOH (2 x 10 mL).
Recrystallisation from MeOH afforded the pyrrole 5h as an orange solid (0.55 g, 60%), m.p.:
1
2
2
7
1
1
1
42.7-247.0°C (lit. [34] m.p. 248-250°C). H NMR (400 MHz, CDCl ) δ (ppm): 4.15 (br s,
3
H, NH ), 6.87-6.90 (m, 2H, Ar-H), 7.02-7.08 (m, 3H, Ar-H), 7.13-7.29 (m, 7H, Ar-H),
2
1
3
.33-7.39 (m, 3H, Ar-H). C NMR (100 MHz, CDCl ) δ (ppm): 74.2 (C-CN), 117.4 (C≡N),
3
21.3 (Ar-C), 124.7 (Ar-C), 126.6 (Ar-CH), 127.0 (Ar-CH), 128.0 (Ar-CH), 128.2 (Ar-CH),
28.3 (Ar-CH), 128.8 (Ar-CH), 129.0 (Ar-CH), 129.7 (Ar-C), 130.6 (Ar-C), 130.7 (Ar-CH),
+
33.1 (Ar-C), 135.3 (Ar-C), 146.5 (C-NH ); LRMS: m/z 358.0 [M + Na] . HRMS, found:
2
-1
m/z 358.1316, C H N Na requires 358.1315. IR (ν , cm ): 3473.5 (N-H stretch), 3384.4
23
17
3
max
(N-H stretch), 3221.5, 3053.5, 2199.6 (C≡N stretch), 1614.2, 1596.0, 1559.0, 1548.2, 1494.1,
1
474.2, 1453.6, 1439.3, 1355.6, 1268.9, 1221.5, 1154.2.
5
.10. 1,2-Diphenyl-2-(phenylamino)ethanone 6
A mixture of benzoin 2a (2.12 g, 10.0 mmol), aniline 3f (1.0 mL, 1.1 g, 10 mmol) and
trichloroacetic acid (0.08 g, 0.5 mmol) in dry toluene (20 mL) was refluxed using a Dean-
Stark apparatus for 3 h (LRMS monitoring). The yellow solution was evaporated in vacuo to
give a brown solid residue (3.3 g). Silica gel column chromatography [EtOAc/PE (1:19) →
EtOAc/PE (7:93)] afforded the α–aminoketone 6 as a brown solid (1.81 g, 63%), m.p.: 93.8-
1
9
5.7 °C (lit. [51] m.p. 98-99 °C). R : 0.58 (30% EtOAc in PE). H NMR (300 MHz, CDCl ) δ
f
3
(
(
(
ppm): 6.03 (s, 1H, CHNH), 6.68-6.71 (m, 3H, Ar-H), 7.09-7.29 (m, 5H, Ar-H), 7.39-7.44
1
3
m, 4H, Ar-H), 7.52 (tt, 1H, J = 2.4 Hz, J = 7.5 Hz), 7.95-7.98 (m, 2H, Ar-H). C NMR
1
2
75 MHz, CDCl ) δ (ppm): 63.3 (CHNH), 114.3 (Ar-CH), 118.6 (Ar-C), 128.2 (Ar-CH),
3
1
28.3 (Ar-CH), 128.7 (Ar-CH), 128.9 (Ar-CH), 129.1 (Ar-CH), 129.2 (Ar-CH), 133.5 (Ar-
2
6

ACCEPTED MANUSCRIPT
+
CH), 134.9 (Ar-CH), 137.1 (Ar-C), 145.2 (Ar-C), 196.7 (C=O); LRMS: m/z 288.0 [M + H] .
-1
IR νmax (cm ): 3408.8 (N-H stretch), 3386.2, 3050.3, 1679.3 (C=O stretch), 1596.4, 1579.1,
1
502.6, 1447.3, 1427.6, 1320.3, 1246.0, 1173.7.
5
.11. General procedure for the preparation of acyl chlorides [52]
A mixture of thionyl chloride (2.0 mL, 3.6 g, 30 mmol) and the corresponding carboxylic
acid (3 mmol) was refluxed for 1.5 h. Excess thionyl chloride was removed by distillation.
The residue obtained was co-evaporated with toluene (3 x 5 mL) in vacuo to ensure the
complete removal of thionyl chloride. The acyl chloride product was used without further
purification in the next step.
5
.12. General procedure for the condensation of pyrrole 5a with acyl chlorides
A solution of the appropriate acyl chloride (2.4 mmol) in toluene (2.0 mL) was added
dropwise to a stirred mixture of pyrrole 5a (0.21 g, 0.60 mmol) and triethylamine (0.50 mL,
0
.36 g, 3.6 mmol) in dry toluene (8 mL). The reaction mixture was refluxed overnight under
argon. The excess solvent was removed in vacuo and the residue obtained was dissolved in
EtOAc (20 mL) and washed successively with 5% HCl (10 mL), saturated NaHCO (10 mL)
3
and brine (10 mL). The organic layer was dried over Na SO and concentrated in vacuo. The
2
4
crude product was purified by silica gel column chromatography.
5
.13. General procedure for the partial deacylation of N,N-diacyl imides 7 to N-acyl amides
8
2
7

ACCEPTED MANUSCRIPT
Equimolar quantities of potassium carbonate (28 mg, 0.20 mmol) and the appropriate N,N-
diacyl imide 7 (0.20 mmol) were stirred at room temperature for 1 h in either MeOH or
acetone/MeOH (1:1) (2.5 mL). The excess solvent was removed in vacuo and the residue was
dissolved in CHCl (15 mL) and partitioned with water (15 mL). The aqueous layer was
3
extracted with CHCl (2 x 5 mL). The organic layers were combined, washed with brine (20
3
mL), dried over Na SO and concentrated in vacuo. The crude product 8 was purified by
2
4
recrystallisation or silica gel column chromatography.
5
.14. N-Acetyl-N-(1-benzyl-3-cyano-4,5-diphenyl-1H-pyrrol-2-yl)acetamide 7a
A mixture of pyrrole 5a (0.20 g, 0.58 mmol) and acetic anhydride (2.3 mL, 2.5 g, 24 mmol)
was heated under reflux for 3 h under argon. The reaction mixture was allowed to cool and
then poured into a mixture of ice and water. Ammonia (25%, 10 mL) was added to the
mixture until a red precipitate was formed. The precipitate was vacuum-filtered and
recrystallised from diethyl ether to afford the title compound 7a as a brown solid (0.20 g,
1
7
9%), m.p.: 173.0-177.8 °C. R : 0.45 (30% EtOAc in PE). H NMR (300 MHz, CDCl ) δ
f
3
(
(
(
(
ppm): 2.15 (s, 6H, 2 x COCH ), 4.85 (s, 2H, NCH Ph), 6.95-6.99 (m, 2H, Ar-H), 7.22-7.38
3 2
1
3
m, 13H, Ar-H). C NMR (75 MHz, CDCl ) δ (ppm): 25.6 (COCH ), 48.6 (NCH Ph), 92.5
3
3
2
C-CN), 114.4 (C≡N), 124.2 (Ar-C), 127.1 (Ar-CH), 127.5 (Ar-CH), 128.3 (Ar-CH), 128.4
Ar-CH), 128.8 (Ar-CH), 129.95 (Ar-CH), 129.01 (Ar-CH), 129.1 (Ar-CH), 130.0 (Ar-C),
1
31.4 (Ar-CH), 131.8 (Ar-C), 133.6 (Ar-C), 135.2 (C-NH ), 172.1 (C=O). ESI-MS: m/z
2
+
4
56.2 [M + Na] . HRMS, found: m/z 456.1683, C H N NaO requires 456.1682. IR ν
28
23
3
2
max
-1
(
cm ): 3033.8, 2223.8 (C≡N stretch), 1743.3 (C=O stretch), 1705.0 (C=O stretch), 1605.2,
1
533.7, 1497.2, 1479.3, 1455.9, 1446.4, 1429.1, 1364.1, 1349.7, 1255.8, 1202.2, 1016.6.
2
8