Synthesis and characterization of quinoline-carbaldehyde derivatives as novel inhibitors for leishmanial methionine aminopeptidase 1

Article abstract of DOI:10.1016/j.ejmech.2019.111860

Methionine aminopeptidase 1 of Leishmania donovani (LdMetAP1) is a novel antileishmanial target for its role in vital N-terminal methionine processing. After LdMetAP1 expression and purification, we employed a series of biochemical assays to determine opt

Full text of DOI:10.1016/j.ejmech.2019.111860

Journal Pre-proof
Synthesis of quinoline-carbaldehyde derivatives as novel inhibitors for leishmanial
methionine aminopeptidase 1
Saleem Yousuf Bhat, Peddapaka Jagruthi, Angapelly Srinivas, Mohammed Arifuddin,
Insaf Ahmed Qureshi
PII:
S0223-5234(19)31012-8
DOI:
https://doi.org/10.1016/j.ejmech.2019.111860
EJMECH 111860
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 12 July 2019
Revised Date: 30 October 2019
Accepted Date: 5 November 2019
Please cite this article as: S.Y. Bhat, P. Jagruthi, A. Srinivas, M. Arifuddin, I.A. Qureshi, Synthesis of
quinoline-carbaldehyde derivatives as novel inhibitors for leishmanial methionine aminopeptidase 1,
European Journal of Medicinal Chemistry (2019), doi: https://doi.org/10.1016/j.ejmech.2019.111860.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. 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.
©
2019 Published by Elsevier Masson SAS.

Synthesis of Quinoline-carbaldehyde derivatives as novel inhibitors for
Leishmanial Methionine aminopeptidase 1
a
b
b
b
Saleem Yousuf Bhat , Peddapaka Jagruthi , Angapelly Srinivas , Mohammed Arifuddin
a
and Insaf Ahmed Qureshi *
a
Department of Biotechnology & Bioinformatics, School of Life Sciences, University of
Hyderabad, Prof. C.R. Rao Road, Hyderabad 500046, India
b
Department of Medicinal Chemistry, National Institute of Pharmaceutical Education and
Research (NIPER), Hyderabad 500037, India.
*
E-mail: insaf@uohyd.ac.in
Abstract
Leishmanial methionine aminopeptidase 1 (LdMetAP1) is a novel antileishmanial target for
its role in vital N-terminal methionine processing. After LdMetAP1 expression and
purification, we employed a series of biochemical assays to determine optimal conditions for
catalysis, metal dependence and substrate preferences for this ubiquitous enzyme. Screening
of newly synthesized quinoline-carbaldehyde derivatives in inhibition assays led to the
identification of HQ14 and HQ15 as novel and specific inhibitors for LdMetAP1 which
compete with substrate for binding to the catalytic active site. Both leads bind LdMetAP1
with high affinity and possess druglikeness. Biochemical studies suggested HQ14 and HQ15
to be comparatively less effective against purified HsMetAP1 and showed no or less toxicity.
We further show selectivity and inhibition of lead inhibitors is sensed through a non-catalytic
Thr residue unique to LdMetAP1. Finally, structural studies highlight key differences in the
binding modes of HQ14 and HQ15 to LdMetAP1 and HsMetAP1 providing structural basis
1

for differences in inhibition. The study demonstrates the feasibility of deploying small drug
like molecules to selectively target the catalytic activity of LdMetAP1 which may provide an
effective treatment of leishmaniasis.
Keywords
Leishmania donovani; Methionine aminopeptidase 1; Quinoline-carbaldehyde derivatives;
Drug discovery
1
. Introduction
Leishmaniasis is a major parasitic burden after malaria and appears in three major forms -
visceral (also known as kala-azar and the most severe form), cutaneous (commonest form),
and mucocutaneous form. Visceral form is often referred to as the disease of the poor and is
associated with malnutrition, bad housing, population displacement and weak immune
system. This disease is also connected to environmental changes, rampant urbanization and
is caused by the dimorphic parasites of genus Leishmania which are disseminated into
mammalian
hosts
through
the
bite
of
sandfly
(http://www.who.int/mediacentre/factsheets/fs375/en/). The major symptoms of leishmaniasis
include ulcerative skin lesions, fever, low RBC count, enlargement of spleen and liver which
makes it life-threatening. Nearly one million new cases and about 30,000 deaths are reported
annually in over 80 countries around the globe [1]. Absence of any significant vaccination
accompanied with the reports of toxicity [2] and resistance against the current line of
antileishmanials [3], there is a pressing need to identify newer molecular targets for
therapeutic intervention against leishmaniasis.
Aminopeptidases are metal dependent exopeptidases that cleave amino acids from the N-
terminal end in proteins and peptides [4]. The critical roles played by these enzymes in
2

nutrient acquisition [5] and the maintenance of protein turnover [6] makes them vital.
Methionine aminopeptidases (MetAPs) remove initiator methionine from the N-terminus of
nascent polypeptides either co-translationally or post-translationally [7-8] to enable necessary
post-translational modifications [4,9]. MetAPs are ubiquitous, highly conserved and require a
divalent metal cofactor for activity [4,7,10]. These enzymes are often seen as Co(II) activated
metallo-enzymes with a dinuclear metal binding active site surrounded by residues DDHEE
[
7,10,11]. Apart from Co(II), other divalent metal ions like Fe(II), Mn(II) and Mg(II) are
major activators [12]. However, the physiological activator appears to be Fe(II) [13].
Methionine aminopeptidases are grouped into two categories: type I and type II MetAP based
on the presence of an additional amino acid insertion of nearly 60 amino acids residues
towards the catalytic domain in the C-terminus of type II MetAP [8,14]. Prokaryotes express
only MetAP1 whose deletion is reported to be lethal [15]. Eukaryotes express both types of
MetAPs and develop lethality if either one or both MetAP genes are deleted [16]. Therefore,
MetAPs have received attention as molecular targets for drug discovery against parasitic
diseases [17] and severe adversaries like cancer [18].
Our study reports the expression, purification and a detailed characterization of methionine
aminopeptidase 1 from L. donovani (LdMetAP1). We have also synthesized and screened a
library of quinoline-carbaldehyde derivatives to find novel inhibitors for LdMetAP1. Using
molecular, biochemical, biophysical and computational approaches, we show key differences
between LdMetAP1 and HsMetAP1 and explain their differential inhibition by quinoline-
carbaldehyde derivatives. We finally come up with two lead compounds with high
druglikeness that may guide the development of highly specific and potent inhibitors against
LdMetAP1 and find use to combat leishmaniasis.
3

2
. Results
2
.1 Identification of LdMetAP1gene, homology modelling and protein purification
To find genes responsible for methionine excision in the genome of L. donovani, we searched
for LdMetAP genes that were homologous to HsMetAP1. We found a copy of both classes of
methionine aminopeptidases (MetAPs) encoded by all species of Leishmania. Amongst the
two MetAPs, one belongs to MetAP2 class based on the presence of an extra C terminal
insertion near the catalytic domain while as, other MetAP belongs to the MetAP1 class of
methionine aminopeptidases. Both LdMetAP1 and LdMetAP2 show high homology to their
respective human counterparts and display a two domain organisation which is typical of
M24 aminopeptidases (Fig. 2A). The analysis of LdMetAP1 sequence revealed this protein to
be 392 residues with its approximate molecular weight poised at 43.20 kDa. Since full length
LdMetAP1 could not be purified, we removed the initial 60 amino acid residues from the
LdMetAP1 ORF from the N-terminal non-catalytic domain and expressed and purified a
truncated functional 38 kDa LdMetAP1. The truncation was done knowing that the neglected
N-terminal 60 residues are far-off from the C-terminal catalytic domain and have no role in
LdMetAP1 catalysis.
A homology model of 38 kDa LdMetAP1 was built and validated (Fig. 2B). RMSD
differences between LdMetAP1 and TbMetAP1 were 0.2 Å asserting the two structures to be
very similar due to high sequence homology and the same pita-bread fold organization.
Furthermore, overlay of the modelled LdMetAP1 and HsMetAP1 showed RMSD converge to
0
.51 Å suggesting high structural similarity between all MetAP1 enzymes. The secondary
structural composition of LdMetAP1 resembles more with TbMetAP1 than HsMetAP1. In
fact, secondary as well as tertiary structures of LdMetAP1 and TbMetAP1 are highly
redundant. Like TbMetAP1, the active site of LdMetAP1 is located in the pita-bread fold
bearing two metal ions and surrounded by two large α-helices and two β-sheets (Fig. 2C).
4

The metal ions are indispensable for catalysis because of their role in the generation of a
strong nucleophile to break the sessile peptide bond while acting on nascent polypeptides or
peptide substrates. Even though LdMetAP1 is somewhat symmetrical in nature, its catalytic
site is asymmetrically located in the highly conserved N-terminal domain. Besides, sequence
alignment highlighted the evolutionary conservation of LdMetAP1 active site and high
sequence conservation with all MetAP1 class enzymes including the HsMetAP1 (Fig. 2D).
High resemblance of LdMetAP1 to HsMetAP1 is also evident from the presence of zinc-
finger motif in the N-terminal extension of both enzymes.
Recombinant LdMetAP1 was amplified from the genomic DNA of L. donovani and cloned
into the expression vector pET28a. After expression in E. coli BL21 (DE3) cells, LdMetAP1
was purified to homogeneity by Ni-NTA affinity chromatography and gel filtration
chromatography (Fig. 2E). Other recombinantly expressed proteins T291A (lanes 3 and 4 in
Fig. 2E) and HsMetAP1 (Fig. 2F) were purified in a similar fashion. The quality of all
purified proteins was assessed by 10% (w/v) SDS-PAGE which demonstrated all proteins to
be highly homogenous. Furthermore, comparative analysis of the LdMetAP1 partition
coefficient (k ) with those of the standard set of proteins suggested this enzyme to be
av
monomeric in solution (Fig. 2G).
2
.2 Determination of secondary structure and thermal stability
CD spectroscopy was employed to determine the secondary structure content of LdMetAP1
and establish whether the average secondary structure of this enzyme changes after mutating
Thr291 to Ala. The results of Far-UV spectra showed all recombinant proteins expressed in
properly folded form with LdMetAP1 composed of 21% α-helices and 24% β-sheets and
T291A carrying 17% α-helices and 29% β-sheets (Fig. 3A). Therefore, T291A had slightly
lesser α-helices and more β-sheets than LdMetAP1. Heat denaturation assays in the
5

temperature range of 293 to 380 K indicated both LdMetAP1 and T291A to have high
thermostability. However, native LdMetAP1 showed higher thermostability (T 354.5 ± 0.3
M
K) than T291A (T 347.5 ± 0.3 K) possibly due to more helices (Fig. 3B).
M
2
.3 LdMetAP1 is a Fe(II) dependent methionine aminopeptidase
Methionine aminopeptidases are exopeptidases implicated in the excision of initiator
methionine residue from nascent polypeptides. Therefore, amidolytic activity of LdMetAP1
was assessed using a synthetic substrate L-methionine-4-methyl-coumaryl-7-amide in the
presence of different metal chlorides. The kinetic parameters determined with different
metals at the physiological pH (Supplementary Fig. 1A) revealed LdMetAP1 to be highly
activated by divalent cobalt (Fig. 3C). However, LdMetAP1 showed highest affinity for
fluorogenic substrate when it was supplemented with divalent iron (K = 40.4 ± 5.4 µM) in
m
the assay buffer. This points towards the LdMetAP1 being an iron-dependent aminopeptidase
3
−1
−1
(
k /K = 114.71 ± 16.28 × 10 M min ) than a cobalt-dependent aminopeptidase (k /K =
cat m cat m
3
−1
−1
6
7.62 ± 4.73 × 10 M min ). The other major activators of LdMetAP1 were divalent
3
−1
−1
manganese (k /K = 41.71 ± 3.85 × 10 M min ) and magnesium (k /K = 23.36 ± 2.92 ×
cat
m
cat
m
3
−1
−1
1
0 M min ), while zinc didn’t prove to be a good activator for the amidolytic activity of
LdMetAP1 (Table 1).
While assessing whether LdMetAP1 can cleave other residues from the synthetic substrates,
we tested other fluorogenic substrates in in vitro aminopeptidase assay. Unlike LdMetAP2,
the enzyme LdMetAP1 doesn’t prefer the cleavage of smaller amino acid residues like Ala
from Ala-AMC (Fig. 3D). In fact, LdMetAP1 didn’t cleave alanine at all. However,
LdMetAP1 hydrolyzed Leu from L-leucine-4-methyl-coumaryl-7-amide and showed minimal
activity for di- and tripeptides. The results suggested LdMetAP1 to be specific to methionine
hydrolysis and highlighted the bigger residue preference of this enzyme for activity.
6

2
.4 Identification of selective LdMetAP1 Quinoline-carbaldehyde core bearing
inhibitors
We synthesized 20 Quinoline-carbaldehyde derivatives (Fig. 1) and screened them in
aminopeptidase assays. Synthesis of the target compound 4 was carried out in 2 steps as
illustrated in Scheme-1. In the first step, the commercially available 8-Hydroxy-2-
methylquinoline (1) was subjected to oxidation with SeO to obtain the corresponding 8-
2
hydroxy-2-quinoline carbaldehyde (2) by using the reported method [19]. Subsequently, 8-
hydroxy-2-quinoline carbaldehyde (2) was subjected to cyclization reaction with substituted
anilines and thioglycolic acid as per the literature report [20] to afford the final product [4
(HQ1-HQ20)]. The purity analysis by HPLC has demonstrated all compounds to be highly
pure except HQ11 which possessed degradative tendency (Supplementary Table 1). The
newly synthesised compounds were confirmed and characterized by spectroscopic techniques
1
13
1
such as, H-NMR and C-NMR etc. The H-NMR spectra demonstrated characteristic peaks
around 3.8 ppm and 4.1 ppm region which corresponds to CH of the 4-thiazolidinone ring.
2
Whereas, the peak at 7.26 ppm as singlet corresponds to the CH attached to quinoline ring. In
1
3
C-NMR, peaks around 3.94 ppm and 65.53 ppm are those of 4-thiazolidinone carbons
while, a peak at 171 ppm corresponds to the carbonyl carbon of the 4-thiazolidinone. The
13
DEPT C-NMR studies were performed for the three most active compounds HQ13, HQ14
HQ15 to confirm the final products. DEPT spectra of compound 14 demonstrated a
&
characteristic negative peak at 33.13 ppm which corresponds to CH of 4-Thaizolidinone
2
ring, whereas the positive peak at 66.35 is due to CH of 4-Thiazolidinone and peak at 20.97 is
due to CH of aromatic rings (Supplementary material). The remaining peaks are that of the
3
aromatic ring. Four promising compounds showing encouraging inhibition of LdMetAP1
were identified and their relative potency for the specific inhibition of LdMetAP1 was
determined by enumerating the IC values in inhibition assay (table 2). The most potent
5
0
7

inhibitors (HQ11, HQ13, HQ14 and HQ15) were then again tested for selectivity using both
LdMetAP1 and HsMetAP1 enzymes. While as, HQ11 showed a similar potency for
LdMetAP1 and HsMetAP1 enzymes, inhibitors HQ13, HQ14 and HQ15 were highly specific
to LdMetAP1 (Fig. 4A-D and Supplementary Fig. 3A). Using time course assays of 30 min,
we determined the mode of inhibition of promising inhibitors of LdMetAP1. HQ11 which
inhibits LdMetAP1 (IC 8.8 µM) and HsMetAP1 (IC 8.9 µM) with almost equal potency
5
0
50
was found to inhibit LdMetAP1 in an uncompetitive manner as both K and V decreased
m
max
with soaring HQ11 (Fig. 4E). Other hit compounds (HQ13, HQ14 and HQ15) inhibited
LdMetAP1 through a competitive mode of inhibition as K showed an increase and V
m
max
stood steady with an increase in the concentration of these inhibitors (Fig. 4F-H). A
substitution at the C4 position of the six carbon ring appeared to be essential for potency as
other substitutions didn’t seem to contribute greatly to the inhibitor potency. Interestingly,
barring a –CN group substitution at C3, other substitutions made at this position didn’t seem
to add to inhibitor potency. In contrast, all substitutions made at C4 contributed to potency.
Hence, different analogs bearing different groups at C4 position were synthesized (Fig. 1).
Positioning of a chloride atom at C4 position as in HQ15 increased the inhibitor potency
greatly rendering HQ15 as the most potent inhibitor (IC 1.1 µM) of LdMetAP1 in this
5
0
library (Fig. 4D). Substitution of smaller electronegative atom amongst halogens like fluorine
IC 45.4 µM) or a bigger atom of the same group like bromine (IC 251.6 µM) at C4
(
5
0
50
position was found inefficient in inhibition assays (Supplementary Fig. 1B). Addition of a -
CH group as in HQ14 (IC 1.3 µM) showed good potency as well (Fig. 4C).
3
50
We next determined the toxicity of hit compounds with MTT assay and found all compounds
except HQ11 exhibiting low toxicity. While determining whether the identified hits have
druglikeness, we determined physiochemical properties along with the parameters like LogP,
8

LogS and GI absorption. We found all hits possess promising druglikeness as none of them
violated the Lipinski’s rule (Table 3).
2
.5 Lead molecules HQ11, HQ13, HQ14 and HQ15 bind to LdMetAP1 with high affinity
Physical interactions of enzymes LdMetAP1 and HsMetAP1 with HQ11, HQ13, HQ14 and
HQ15were monitored using Surface plasmon resonance (SPR) binding assays. Sensorgrams
showed a direct interaction between HQ compounds and recombinant enzymes with HQ11
-10
binding (Fig. 5A and B) with high affinity to both LdMetAP1 (K 1.61 × 10 M) and
D
-10
HsMetAP1 (K 1.52 × 10 M). Other leads HQ13, HQ14 and HQ15 had higher binding
D
affinities (Table 4) for LdMetAP1 (Fig. 5C, E and G) than HsMetAP1 (Fig. 5D, F and H).
Indeed, lead inhibitor molecule HQ15 showed comparatively lesser affinity for HsMetAP1
-
7
-8
(
K 3.61 × 10 M) and higher affinity for LdMetAP1 (K 1.50 × 10 M) which explains its
D
D
poor inhibition of HsMetAP1. Meanwhile, fluorescence spectroscopic plots showed all lead
molecules HQ11, HQ13, HQ14 and HQ15 quenching the fluorescence emission of
LdMetAP1 in a concentration dependent manner (Supplementary Fig. 2A-D). Determination
of the binding parameters using modified Stern-Volmer plots showed all lead molecules to
have high binding affinity for LdMetAP1 which corroborates with the surface plasmon
resonance results.
2
.6 Quinoline-carbaldehyde specificity and potency is majorly mediated through a Thr
residue unique to LdMetAP1
Quinoline-carbaldehyde core bearing compound library showed specificity to LdMetAP1 and
not to HsMetAP1 even though LdMetAP1 and HsMetAP1 have high structural similarity and
an identical domain and fold organization. However, amino acid sequences of LdMetAP1 and
HsMetAP1 carry key amino acid differences at S1 and S1’ sub-sites (Fig. 2D). Molecular
9

docking of hit compounds with LdMetAP1 suggests Thr291, His300 and Asn304 to be the
key residues involved in inhibitor binding. Since His300 and Asn304 are conserved in both
LdMetAP1 and HsMetAP1 enzymes, we reasoned they may not be the key residues
conferring inhibitory specificity. However, Thr291 is unique to LdMetAP1, and is replaced
by Cys in HsMetAP1. Mutation of Thr291 with either Ala or Val resulted in the loss of
affinity of specific inhibitor HQ15 to LdMetAP1 from -7.4 kcal/mol to -6.8 kcal/mol in
computational assays (Fig. 6E-G). We followed up computational assays by site directed
mutagenesis and mutated Thr291 to Ala. While determining whether there was a change in
the catalytic efficiency of LdMetAP1 after mutating Thr291 to Ala, we performed
aminopeptidase assay of mutazyme T291A and found it catalytically less efficient (Fig. 6A).
Moreover, there was a remarkable loss in affinity of T291A for substrate (K 969.7 ±
m
2
15.1µM), indicating Thr291 to be an important substrate binding residue for LdMetAP1.
Intriguingly, inhibition assays of LdMetAP1 and T291A with inhibitors HQ13, HQ14 and
HQ15 suggested T291A to be very less prone to inhibition by specific inhibitors than the
native LdMetAP1 (Fig. 6B-D and Supplementary Fig. 3B). This demonstrated Thr291 to be a
crucial specificity conferring residue for the inhibition of LdMetAP1 by Quinoline-
carbaldehyde core bearing competitive inhibitors.
2
.7 Structural basis of inhibition of LdMetAP1
In an attempt to unravel the atomic interactions of lead compounds with LdMetAP1, we
performed molecular docking studies. For analysis the conformation of ligands with protein
having the highest binding affinity was chosen. Firstly, the ligand binding site of HQ11
(uncompetitive inhibitor of LdMetAP1) was predicted. HQ11 was found interacting with
LdMetAP1 exclusively via hydrophobic interactions involving residues Arg107, Gly108,
Asp106, Pro240, Thr291, Gly292, His293, Thr301, Ala302 and Asn304 (Fig. 7A). HQ13 had
1
0

a lone hydrogen bond with Thr291 and hydrophobic interactions with residues Tyr185,
Tyr290, Gly292, His293, His300, Glu326 and Trp343 (Fig. 7B). Furthermore, both HQ14
and HQ15 interact with Thr291, His300 and Asn304 through hydrogen bonding and have
hydrophobic interactions with residues Arg107, Tyr185, Tyr186, Ser289 and Tyr290 (Fig 7C
and D). Therefore, lead molecules HQ14 and HQ15 bind LdMetAP1 via three hydrogen
bonds while as, HQ13 and HQ11 bind to LdMetAP1 via one and no hydrogen bonds,
respectively. The binding site and binding mode of HQ11 was contrastingly different from
those of HQ13, HQ14 and HQ15 which bind at the catalytic active site providing a structural
basis for the differences in the mode of inhibition. Meanwhile, all leads bound HsMetAP1
with different binding modes and lower affinity (except HQ11) which may be a major
contributing factor to the exceptionally lower potency of HQ13, HQ14 and HQ15 against
HsMetAP1 (Fig. 7E-H).
2
.8 Lead inhibitor HQ15 forms a stable complex with LdMetAP1
After molecular docking of all lead compounds, LdMetAP1 and T291A were simulated alone
and in complex with lead molecule HQ15 for 30 ns. RMSD of backbone atoms used to assess
the convergence of structure to equilibrium after HQ15 binding suggested LdMetAP1 and
T291A to be stable as there was no significant deviation in the RMSD. The binding of HQ-15
to both LdMetAP1 as well as T291A increased the RMSD by approximately 0.8 Å (0.08 nm)
suggesting a minor change induced in the protein structure after ligand binding (Fig. 8A).
This matches with the fluorescence spectroscopic studies where HQ15 binding quenched the
fluorescence but didn’t affect the protein folding greatly as neither a ‘blue shift’ nor ‘red
shift’ was observed (Supplementary Fig. 2D). Meanwhile, both LdMetAP1-HQ15 as well as
T291A-HQ15 complex reached equilibrium one after another with former assuming
precedence.
1
1

The flexibility of unbound and HQ15 bound enzymes during the 30 ns MDS run was
assessed by plotting RMSF. Other than the major fluctuations occurring in the C-terminal
region (residues 310-315) and with non-catalytic residues Cys221 and Ser250 (of truncated
LdMetAP1) which are far off from the catalytic pocket, all systems were highly rigid (Fig.
8
B). Specific fluctuations in the loop region towards C-terminal end possibly indicate the
occurrence of slight conformational changes after HQ15 binding.
Unbound and HQ15 bound enzyme structures were also analyzed using Rg which describes
compactness by measuring the mass of atoms relative to the mass of whole complex.
LdMetAP1 showed relatively more compactness than T291A. Furthermore, LdMetAP1-
HQ15 complex had higher compactness than T291A-HQ15 complex. Therefore, the
compactness and stability of LdMetAP1 increased dramatically after complex formation with
HQ-15 (Fig. 8C). The MD simulations of LdMetAP1 and T291A and its complexes with
HQ15 assert the complexes to be stabilized by conformational rearrangements and that HQ15
like other lead molecules is held with LdMetAP1 by hydrophobic interactions and hydrogen
bonding. Similar results were observed through molecular docking.
3
. Discussion
The removal of initiator methionine from nascent polypeptides by MetAPs is an evolutionary
conserved process essential for growth, proliferation and survival of both prokaryotes and
eukaryotes. Like human and yeast genome, the leishmanial genome encodes both type I and
type II forms of MetAPs. Deletion of MetAPs from the yeast or other eukaryotic genomes
causes growth and developmental abnormalities and leads to lethality [16,21]. Moreover,
specific inhibition of MetAPs with small molecule inhibitors leads to immune suppression
[
22] and blockade of angiogenesis in humans [23]. Specific and potent inhibition of MetAPs
has also been used to evade and eliminate parasitic infections [24] suggesting specific
1
2

inhibition or inactivation of leishmanial MetAPs could become a mainstream strategy to
block the growth and proliferation of L. donovani. We cloned, expressed and purified both
LdMetAP1 and HsMetAP1 and established that both of these recombinantly expressed
enzymes are active under physiological like conditions against a synthetic fluorogenic
substrate L-methionine-4-methyl-coumaryl-7-amide widely used to determine activity of
MetAPs and inhibitor screening.LdMetAP1 showed high catalytic efficiency than HsMetAP1
[
25] and TbMetAP1 [26] at physiological pH. While screening other synthetic substrates
1
carrying a smaller or bigger P residue (such as, Ala or Leu), we found LdMetAP1 either
inactive or marginally active suggesting its high specificity for the hydrolysis of Met. Like
other type I MetAPs, LdMetAP1 was highly activated by divalent cobalt which can be
attributed to the ability of this metal to act as a hard Lewis acid [27] as a result of which it
attains high affinity for hard bases like water or hydroxide ions that are involved in the
catalytic mechanism [28] of MetAPs. Furthermore, cobalt has a partially filled 3d orbital and
thence preferred oxidation states of +2 and +3 which allow it to form low-spin trigonal
bipyramidal, high-spin octahedral and low-spin octahedral complexes [26]. Because of the
higher stability of cobalt’s low-spin and high coordination number complexes, the metal ion
charge intensifies which leads to higher substrate binding and the stability of reaction
intermediates. Such chemical properties result in higher catalytic efficiency of cobalt
activated MetAPs. Although highly activated by Co(II), LdMetAP1 had highest affinity for
substrate when activated by Fe(II) much like EcMetAP1 [13]. In fact, LdMetAP1 seemed
specific to Fe(II) as the substrate binding increased dramatically. Interestingly, the
concentration of cytosolic Fe(II) increases while the concentration of Co(II) stays steady
when EcMetAP1 is over-expressed in E. coli which further supports the narrative that iron is
the physiological activator of type I MetAPs. Like TbMetAP1 [26], LdMetAP1 didn’t show a
noticeable activity with Zn(II) against L-methionine-4-methyl-coumaryl-7-amide. Whether or
1
3

not LdMetAP1 can be activated by Zn(II) like TbMetAP1 [26] against other substrates that
resemble physiological substrates of MetAPs like Met-Gly-Met-Met remains to be seen. Zinc
is the second most plenteous metal in human body [29] and is resistant to redox changes at
biological potential [30]. Therefore, it is difficult to rule out the possibility of zinc being one
of the major physiological activators of MetAPs. Structurally, MetAPs are very similar;
therefore, different metal preferences can be due to amino acid substitutions near the catalytic
site allowing slight rearrangement of active site residues which alter metal preferences and
metal dependence. As such, HsMetAP1 could be highly activated by Co(II) due to a smaller
active site [31]. All type 1 MetAPs including LdMetAP1 are monomeric in solution and have
a two domain organization i) N-terminal domain which reportedly binds to ribosome [32] to
take up the N-terminal methionine processing of nascent polypeptides and ii) C-terminal
catalytic domain which carries catalytic residues DDHEE embedded with two divalent metal
ions indispensable for catalysis [7,28]. Modelled LdMetAP1, like the crystal structure of
TbMetAP1 carries more β-sheets than α-helices. This in turn matches with the Far-UV CD
spectroscopic data whose analysis suggested that LdMetAP1 carries 21% α-helices and 24%
β-sheets.
While screening inhibitor libraries of distinct structural classes, we identified quinoline-
carbaldehyde core bearing inhibitors HQ14 and HQ15 which inhibited LdMetAP1
specifically. Most inhibitors of this library except HQ11 showed remarkably high specificity
towards LdMetAP1 than HsMetAP1 possibly due to crucial amino acid substitutions near the
catalytic site. The key structural element necessary for inhibitors to carry potency was the C4
position of the 6-carbon ring linked to the quinoline-carbaldehyde core by a Nitrogen atom of
pentameric ring. Multiple substitutions with different groups revealed –Cl substitution to be
most efficient for potency while –CH substitution was equally efficient. Thr291 of
3
LdMetAP1 appeared to be directly interacting with leads and crucial to mediate inhibitor
1
4

potency as mutazyme (T291A) carrying Ala at the place of Thr was less susceptible to
inhibition by competitive inhibitors HQ13, HQ14 and HQ15. Thr291 is an important
substrate binding residue whose mutation to Ala or Val caused rotameric flipping of active
site residues and the residues surrounding active site like Arg107, Tyr185, Tyr290, His293
and Asn304 which potentially alters inhibitor potency. Interestingly, Thr is replaced by Cys
in HsMetAP1 and is not prone to inhibition by HQ13, HQ14 and HQ15. Meanwhile, T291A
was comparatively less thermostable than LdMetAP1 possibly due to a slight decline in the
percentage of α-helices. Decline in α-helices tends to destabilize protein structures and
renders them vulnerable to denaturation [33]. Furthermore, time course biochemical
experiments suggested inhibitor HQ11 which carries a fluorine substitution at the C3 position
of the 6-carbon ring linked to the quinoline-carbaldehyde core binding LdMetAP1 at a site
other than the active site. Similar observations were made with molecular docking. This
could be due to the highly electronegative nature of fluorine [34] that enables the inhibitor
(HQ13) to either bind at multiple sites to the enzyme or bind away from the active site rich in
highly reactive residues as depicted by molecular docking studies. Interestingly, HQ11 was
equally potent against LdMetAP1 and HsMetAP1 with similar binding affinities implying it
could have a common binding site in both enzymes. This is indeed possible as LdMetAP1
and HsMetAP1 carry a high sequence homology in the C-terminal catalytic domain and have
almost identical structures. In contrast, inhibitors HQ13, HQ14 and HQ15 seemed specific to
LdMetAP1 and inhibited it competitively with similar binding modes and encouraging IC50
values. Surface plasmon resonance and fluorescence spectroscopic measurements suggested
all lead inhibitors binding LdMetAP1 with higher affinity than HsMetAP1 except HQ11
which binds to both enzymes with high affinity. HQ11 also carries toxicity possibly because
of off-target binding due to the presence of reactive fluorine in the hexameric ring. Other lead
molecules specific to LdMetAP1 viz HQ13, HQ14 and HQ15 showed comparatively low
1
5

toxicity and drug likeness. Importantly, lead molecule HQ15 also formed a stable complex
with LdMetAP1 in a 30 ns MDS run suggesting this to be an important lead for inhibitor
development against LdMetAP1. The molecular basis of remarkably high potency and
specificity of HQ14 and HQ15 towards LdMetAP1 and not HsMetAP1 remains largely
unknown. However, specificity may manifest due to crucial amino acid differences between
LdMetAP1 and HsMetAP1 within 6 Å distance from the catalytic site. The differing residues:
(LdMetAP1 vs. HsMetAP1) Val220 vs. Ile221, Ser221 vs. Thr222, Thr291 vs. Cys292 render
HsMetAP1 catalytic site compact due to bulkier side chains which possibly makes it
inaccessible to quinoline-carbaldehyde core containing leads and thus, less vulnerable to
inhibition.
Specific inhibitors of LdMetAP1 have a potential to circumvent the mayhem of drug
resistance to current regimen of antileishmanials which has also been a major roadblock for
the effective elimination of leishmaniasis. Identification of specific LdMetAP1 inhibitors
have a potential to aid the development of new antileishmanial agents. Moreover, specific
inhibitors of LdMetAP1 like HQ14 and HQ15 may also be used to inhibit the growth and
proliferation of both promastigotes and amastigotes as LdMetAP1 is a vital house-keeping
enzyme expressed in both life forms of Leishmania species thereby extending the importance
and efficacy of specific LdMetAP1 inhibitors. Improvement to lead molecules like HQ14 and
HQ15 along with other quinoline-carbaldehyde core bearing inhibitors in their selectivity and
potency toward LdMetAP1 over HsMetAP1 may thus lead to the development of a novel
class of antileishmanial agents.
1
6

4
. Experiments
4
.1 Sequence analysis and Homology modelling
LdMetAP1
https://www.ebi.ac.uk/Tools/hmmer/) which also described this enzyme to be a member of
M24 aminopeptidases. The sequence was further analysed with web based SMART server
35] for motif analysis, domain organization and presence of signal peptide and internal
repeats. For physico-chemical properties ProtParam tool was used
https://web.expasy.org/protparam/). Homology model of LdMetAP1 was generated with the
sequence
was
identified
through
HMMER
web
server
(
[
(
program Modeller 9.15 version [36] using the structural coordinates of T. brucei methionine
aminopeptidase 1 (PDB code: 4FUK) as a template (sequence identity, 70%). The loops and
Ramachandran outliers were fixed and the modelled structure was energy minimized with
Gromacs 4.6.3 [37]. The geometry of model was validated with RAMPAGE [38] and the
structure was visualized in PyMOL [39].
4
.2 Cloning and site directed mutagenesis
Gene sequence encoding for functional LdMetAP1 was amplified from the genomic DNA of
Leishmania donovani DD8 strain using primers specific for LdMetAP1 (5'-
ATTGGATCCATGGTGGACGAGCGGCTCTTCAAC-3'
and
5'-
AGCAAGCTTTCAGATTTTGATTTCGCTGGGGTC-3'). Likewise, ORF of HsMetAP1
containing 1161 nucleotides was amplified from the cDNA (generated from the mRNA
isolated
from
human
fibroblasts)
using
primers
(5’-
5’-
TTAGGATCCATGGCGGCCGTGGAGACG-3’and
AGGCTCGAGTTAAAATTGAGACATGAAGTGAG-3’). The PCR products of both genes
were cloned in the expression vector pET28a in frame with an N-terminal 6x-Histidine tag
between the sites BamHI and HindIII, and BamHI and XhoI, respectively. Both clones were
1
7

confirmed by restriction digestion and DNA sequencing. Meanwhile, Thr291 positioned at
the active site of LdMetAP1 structure was mutated to Ala in a PCR reaction using primers
containing a mutation (5'-GATACCATGCCCGGCGTAGCTGCGCACTA-3' and 5'-
TAGTGCGCAGCTACGCCGGGCATGGTATC-3'). The PCR was performed as follows: 98
℃
for 30 sec, 98 ℃ for 30 sec, 60 ℃ for 1 min, 72 ℃ for 6 min 30 sec and 72 ℃ for 10
minutes. Methylated template plasmid DNA was digested with DpnI for 1 hr at 37 ℃ and the
reaction was transformed into XL1-Blue cells. The transformants observed on plate were
grown in LB media for plasmid extraction to confirm the mutation with DNA sequencing.
4
.3 Expression and purification of recombinant enzymes
For LdMetAP1, T291A and HsMetAP1 expression, E. coli BL21 (DE3) cells were used and
the protein expression induced with 0.5 mM IPTG for 20 hours at 18°C in a shaking
incubator. The IPTG induced E. coli BL21 (DE3) cultures expressing LdMetAP1, T291Aand
HsMetAP1 were harvested by centrifugation at 4°C and resuspended in the lysis buffer (50
mM Tris-HCl pH 8.0, 250 mM NaCl and 25mM imidazole). Cells were lysed by sonication
and the cell lysate centrifuged at 14000 rpm at 4°C for 45 minutes. The resultant supernatant
was loaded onto a pre-packed His-Trap column (GE Healthcare) equilibrated with the lysis
buffer for protein binding on an AKTA purification system (GE Healthcare). After extensive
washes, all enzymes were eluted out with a buffer bearing 50 mM Tris-HCl (pH 8.0), 250
mM NaCl & 250 mM imidazole. The protein size and quality was checked on 10% (w/v)
SDS-PAGE. The purest fractions of LdMetAP1 were concentrated and loaded onto a pre-
calibrated Superdex 200 column (GE Healthcare) for size-exclusion chromatography. The
protein concentration was estimated photometrically on Nano drop (Thermo Scientific) using
−
1
−1
molar absorption coefficient and molecular weight of 45380 mol cm and 41480 Da for
−
1
−1
LdMetAP1 and 55350 mol cm and 46252Da for HsMetAP1, respectively.
1
8

4
.4 Circular Dichroism (CD) measurements
Far-UV CD spectra of LdMetAP1 and T291A (3 µM each in 10 mM Potassium phosphate
buffer, pH 7.2) were recorded in a quartz cell of 0.2 cm path length on JASCO-J1500 CD
spectropolarimeter. For thermo-stability assays, spectral changes of LdMetAP1 and T291A
were analyzed in the temperature range of 293 to 380 K and the data fitted into the two-state
equilibrium unfolding model to enumerate thermodynamic parameters in Sigma plot 12.0
(San Jose, USA). CD spectral data was analyzed with the program DICHROWEB to predict
secondary structures of LdMetAP1 and T291A.
4
.5 Aminopeptidase activity assay
The LdMetAP1, T291A and HsMetAP1 activity was estimated by measuring the excision of
L-methionine from fluorogenic substrate L-methionine-4-methyl-coumaryl-7-amide (Sigma
Aldrich, USA) using InfiniteM200 Pro spectrofluorometer (TECAN, Switzerland) with a
wavelength pair of 355 and 460 nm used for excitation and emission, respectively. The
aminopeptidase activity assay buffer was supplemented with a metal chloride and 0.5 µM
LdMetAP1. After incubation at 37°C for 30 minutes, the substrate was added up to a final
concentration of 200 µM and the activity recorded in a 30 minute kinetic cycle with a break
of 1 min between measurements. To test whether LdMetAP1 cleaved other amino acid
residues, fluorogenic substrates L-alanine--4-methyl-coumaryl-7-amide, L-leucine-4-methyl-
coumaryl-7-amide,
L-arginine-4-methyl-coumaryl-7-amide,
L-lysine-alanine-4-methyl-
coumaryl-7-amide, L-alanine-alanine-phenylalanine-4-methyl-coumaryl-7-amide, L-alanine-
leucine-lysine-4-methyl-coumaryl-7-amide and L-alanine-alanine-phenylalanine-4-methyl-
coumaryl-7-amide were tested in aminopeptidase assays. Fluorescence emission due to
peptide bond hydrolysis was converted into product formation (AMC) and the steady state
kinetic parameters determined in Graphpad Prism.
1
9

4
.6 Chemistry
General procedure for the Synthesis of compound 2: (8-Hydroxxy-2-
quinolinecarbaldehyde)
8
-hydroxy-2-methylquinoline (1) (2g, 12.4 mmol.), selenium dioxide (1.74g, 15.8 mmol),
3
00 ml of degassed 1, 4-dioxane, and 1.5 ml of water were mixed and stirred in a 1L round
bottom flask. The resulting solution was then refluxed for 24 hours and the reaction was
monitored until completion using TLC method. The reaction mixture was then filtered off,
and the selenium metal was washed with dichloromethane, and the combined filtrates were
then evaporated off under reduced pressure. The crude product was purified by sublimation
under reduced pressure or by silica gel column chromatography to yield pure yellow needle
1
crystal. Yellow solid; Yield = 74.5%; Melting point = 97.0-98℃; H NMR (500 MHz,
CDCl ) δ 10.22 (s, 1H), 8.32 (d, J = 8.5 Hz, 1H), 8.17 (s, 1H), 8.06 (d, J = 8.5 Hz, 1H), 7.63
3
1
3
(
t, J = 8.0 Hz, 1H), 7.44 (dd, J = 8.3, 0.8 Hz, 1H), 7.29 (dd, J = 7.8, 1.0 Hz, 1H). C NMR
(
125 MHz, CDCl ) δ 192.72 (s), 153.02 (s), 150.17 (s), 137.81 (s), 137.49 (s), 130.96 (s),
3
1
30.44 (s), 118.04 (d, J = 11.5 Hz), 111.26 (s).
General procedure for one pot synthesis of 8-Hydroxyquinoline-substituted 4-
Thiazolidinones [4 (HQ1-HQ20)]
To a solution of substituted anilines (0.867 mmol, 1eq) in Dichloromethane at 0°C, 8-
Hydroxy-2-Quinolinecarbaldehyde (2) (1.300 mmol, 1.5eq) was added and the mixture
stirred for 5 min. Thioglycolic acid (1.734 mmol, 2eq) was then added to the above reaction
and stirring was continued for another 5 min at 0°C followed by the addition of silica gel (0.5
g, 100-200 mesh) as shown in Scheme 1. The reaction was stirred for 4 hr at room
temperature and the reaction progress monitored by TLC. After completion of the reactions
as indicated by TLC, the reaction mixture was extracted with DCM and the extract was
washed with sodium bicarbonate solution. The DCM layer was collected and concentrated to
2
0

give the crude product which was purified by using silica gel (100-200) column
chromatography by using 5% ethylacetate-hexane as eluent to get the pure product [4 (HQ1-
HQ20)] with 60-75% yield (Table 5).
Ar-NH (3)
2
SeO2
thioglycolic acid, silica(100-200)
O
rt 12 h, (60-75%)
Dioxane, reflux
DCM,
S
N
CH3
N
N
R
2
4 h, 70%
OH
OH
OH
N
8
-hydroxy-2-
methylquinoline
8-hydroxy-2-
O
quinolinecarbaldehyde
1
2
4
(HQ1-20)
Scheme 1: General synthesis scheme of new target molecules (4)
For purity analysis, all HQ molecules were subjected to HPLC by employing Isocratic
method and Phenomenex C8 150 X 4.6 column on Waters e2695 module with quaternary
solvent manager system using mobile phase [0.1 % Formic acid and Acetonitrile (30:70)] for
a run time of 10 min at 25°C. All the compounds were characterised by spectroscopic
methods and the spectral data is given in supplementary material. Spectral data for the
selected compounds HQ13, HQ14 and HQ15 are given below:
HQ13 (3-(2-(8-hydroxyquinolin-2-yl)-4-oxothiazolidin-3-yl)benzonitrile): Yield;70% ,
white solid, MP;175-178°C; IR(ATR); 3416.79, 2971.13, 2889.70, 2233.18, 1681.40,
-1 1
1
601.18, 1393.28, 1046.46 cm ; H NMR (500 MHz, CDCl ) δ 8.21 (d, J = 8.5 Hz, 1H), 7.75
3
(d, J = 6.9 Hz, 2H), 7.57 (ddd, J = 8.1, 2.1, 1.3 Hz, 1H), 7.40 (dddd, J = 18.8, 12.1, 9.0, 4.4
Hz, 5H), 7.21 (dd, J = 7.6, 0.9 Hz, 1H), 6.34 (s, 1H), 4.06 (dd, J = 16.0, 1.1 Hz, 1H), 3.87 (d,
1
3
J = 16.0 Hz, 1H); C NMR (125 MHz, CDCl ) δ 171.38 (s), 156.06 (s), 151.92 (s), 138.66
3
(
d, J = 9.0 Hz), 137.13 (s), 130.06 (d, J = 11.0 Hz), 128.72 (s), 128.06 (s), 127.32 (s), 118.43
s), 117.95 (d, J = 14.9 Hz), 113.40 (s), 111.42 (s), 65.53 (s), 32.94 (s); HRMS; m/z
(
13
calculated for C H N O S [M+H]+ :348.0807; found : 348.0804.; C NMR (DEPT) (126
1
9
13
3
2
MHz, CDCl ) δ 138.68 (s), 130.06 (d, J = 10.8 Hz), 128.71 (s), 128.08 (s), 127.34 (s), 118.45
3
(s), 118.01 (s), 111.42 (s), 65.54 (s), 32.94 (s).
2
1

HQ14 (2-(8-hydroxyquinolin-2-yl)-3-(p-tolyl)thiazolidin-4-one): Yield; 72%, white solid,
-1 1
MP;123-125°C; IR(ATR); 3422.63, 2917.71, 1683.64, 1502.03, 1329.42 cm ; H NMR (500
MHz, CDCl ) δ 8.16 (d, J = 8.6 Hz, 1H), 7.91 (s, 1H), 7.49 (s, 1H), 7.47 (d, J = 4.8 Hz, 1H),
3
7
.44 (d, J = 8.0 Hz, 1H), 7.31 (t, J = 8.6 Hz, 1H), 7.26 (s, 1H), 7.18 (d, J = 8.4 Hz, 3H), 7.07
(d, J = 8.3 Hz, 2H), 6.33 (s, 1H), 4.06 (dd, J = 15.8, 1.5 Hz, 1H), 3.88 (d, J = 15.8 Hz, 1H),
1
3
2
1
1
.23 (s, 3H); C NMR (125 MHz, CDCl ) δ 171.21 (s), 157.22 (s), 151.92 (s), 138.33 (s),
3
37.18 (s), 134.81 (s), 129.93 (s), 128.44 (s), 128.01 (s), 124.90 (s), 118.80 (s), 117.88 (s),
+
11.23 (s), 66.16 (s), 33.13 (s), 20.97 (s); HRMS; m/z calculated for C H N O S [M+H]
19
16
2
2
:
337.1011; found : 337.1014.
HQ15 (2-(8-hydroxyquinolin-2-yl)-3-(4-chlorophenyl)thiazolidin-4-one): Yield; 80%,
pale yellow crystals, MP;133-135°C; IR(ATR); 3470.07, 3203.07, 2900.07, 1689.91,
-
1 1
1
1
6
596.44, 1490.38, 1328.08, 769.62 cm ; H NMR (500 MHz, CDCl ) δ 8.17 (d, J = 8.6 Hz,
3
H), 7.84 (s, 1H), 7.46 (dd, J = 17.7, 8.2 Hz, 2H), 7.33 – 7.26 (m, 3H), 7.25 – 7.18 (m, 3H),
13
.32 (s, 1H), 4.05 (dd, J = 15.9, 1.3 Hz, 1H), 3.87 (d, J = 15.9 Hz, 1H); C NMR (125 MHz,
CDCl ) ) δ 171.22 (s), 156.65 (s), 151.92 (s), 138.46 (s), 136.99 (s), 136.08 (s), 132.52 (s),
3
1
3
29.40 (s), 128.59 (s), 128.02 (s), 125.81 (s), 118.62 (s), 117.94 (s), 111.32 (s), 65.91 (s),
3.05 (s); HRMS; m/z calculated for C H ClN O S [M+H]+ :357.0465; found : 357.0463.;
1
8
13
2
2
1
3
C NMR (DEPT) (126 MHz, CDCl ) δ 138.35 (s), 129.39 (s), 128.54 (s), 125.83 (s), 118.66
3
(s), 117.95 (s), 111.20 (s), 66.00 (s), 33.06 (s).
4
.7 LdMetAP1 in vitro inhibition assay
Twenty Quinoline-carbaldehyde derivatives (Fig. 1), dissolved in 100% DMSO, were
screened as inhibitors for LdMetAP1 and HsMetAP1 (as control) in 96-well plates to identify
hits. Initially, the enzyme present in the assay buffer along with the metal supplement was
incubated for 30 minutes. Post-incubation, substrate L-methionine-4-methyl-coumaryl-7-
2
2

amide (125 µM) was added and the fluorescence recorded with InfiniteM200 Pro
spectrofluorometer (TECAN, Switzerland). The net percentage decrease in fluorescence in
comparison to the control enzymatic reactions was taken as percentage inhibition for every
compound tested. The IC obtained for each compound with different concentrations was
50
then used to identify and rank hits in the order of potency. For mode of inhibition, hit
compounds were incubated with enzyme (0.5 µM) in Fe (II) supplemented assay buffer for
3
0 minutes at 37 °C. Subsequently, the substrate was added from 0-200 µM and the activity
measurements recorded in a time course reaction for 30 minutes. The mode of inhibition was
determined by Lineweaver-Burk plots.
4
.8 Cell viability assay and assessment of druglikeness
For cell viability using MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide)
assay, mouse embryonic fibroblast cells were incubated in media containing 0, 50, 100 and
2
00 µM concentration of compounds HQ11, HQ13, HQ14 and HQ15 for 48 h in 96-well
plates. This was followed by the addition of MTT solution (5mg/ml) to each well and an
incubation of 4 h at 37°C. Post incubation, the formazan crystals developed were dissolved
by adding DMSO and gentle shaking at 37°C. Cell viability was determined
spectrophotometrically by absorbance measurements at λ 570 nm. Furthermore, program
SwissADME [40] was used for assessing physicochemical properties and druglikeness of all
leads.
4
.9 Surface Plasmon Resonance (SPR) binding assays and Fluorescence Spectroscopy
Enzymes (ligands) LdMetAP1 and HsMetAP1 were coupled to CM-5 dextran chip (GE
Healthcare) with amine coupling. Binding assays were performed using common running and
injection buffers composed of 1X PBS (pH 7.4) and 0.05% P-20. Analytes (HQ compounds)
2
3

dissolved in 2% DMSO were then passed over the immobilized enzymes to measure response
units (RU) as a function of concentration of analytes. After solvent correction, kinetic
analysis was carried out following a 1:1 Langmuir binding model in Biacore T200 evaluation
software 2.0.
For Fluorescence measurements, 1 µM final concentration of LdMetAP1 was taken in the
buffer containing 20 mM Tris-HCl (pH: 7.5) and 50 mM NaCl at 25°C. Using five
tryptophan residues of LdMetAP1 as intrinsic fluorescent probes, the conformational changes
of LdMetAP1 with hit compounds were recorded (λEX ∼ 295 nm, λ ∼ 300-400 nm) using
EM
the Jasco ETC J-815 fluorescence spectrophotometer (Tokyo, Japan). After gentle mixing,
LdMetAP1 was incubated with hit compounds for 30 min at 25°C prior to fluorescence
measurements. The relative fluorescence intensity obtained by (F /F ), where F and F are
0
1
0
1
the fluorescence intensities in the absence and presence of hits, was plotted against the hit
concentration and the Stern-Volmer quenching constant (K ) calculated by linear regression
SV
in Graphpad Prism. Binding parameters of all compounds were calculated with modified
Stern-Volmer plot as reported previously [41].
4
.10 Molecular docking and Molecular Dynamics Simulations
For molecular docking with AutoDock Vina 1.1.2 [42] version using Lamarckian genetic
algorithm, inhibitor coordinates were drawn manually and energy minimized. Docking
studies were conducted following a grid based procedure as reported previously [7] with
modifications in parameters and wherever necessary.
Molecular dynamics simulations consider receptor flexibility; therefore, this powerful method
was used to assess the stability and rigidity of LdMetAP1 and T291A separately and in
complex with HQ15 using group parameters root mean square deviation (RMSD), root mean
square fluctuation (RMSF), radius of gyration (Rg) and free energy calculations with
2
4

MM/PBSA method. The 30 ns production run was performed with Gromacs 4.6.3 [35] using
GROMOS 96 43a1 as a force field. For complexes, HQ15 topology was built with PRODRG
server [43] and the docking conformation with highest affinity was taken. Before production
run, the topology of HQ15 was merged with both LdMetAP1 and T291A separately and the
simulations conducted following SPC model in a cubic box with water as a solvent. All other
procedures used for 30 ns production run were adapted from our previous study [7].
Author Contributions
SYB performed the experiments, analyzed the data, and drafted the manuscript. PJ and AS
synthesized the compounds used in this study under the guidance of MA. IAQ conceived,
designed and supervised the study and revised the manuscript.
Acknowledgements
The authors are grateful to Council of Scientific and Industrial Research, India [CSIR, project
no. 37(1686)/17/EMR-II], Government of India for funding. The authors acknowledge the
instruments facilities supported by DST-FIST and UGC-SAP to the Department of
Biotechnology & Bioinformatics, and Bioinformatics Infrastructure Facility, School of Life
Sciences, University of Hyderabad, Hyderabad for computational resources. SYB thanks to
Indian Council of Medical Research (ICMR) for Senior Research Fellowship. The authors
gratefully acknowledge Prof. Swati Saha for providing genomic DNA of Leishmania
donovani for this work.
Conflict of interest
The authors declare to have no conflict of interest whatsoever.
2
5

References
[
[
[
[
1] S. Burza, S.L. Croft, M. Boelaert, Leishmaniasis, The Lancet. 392 (2018) 951-970,
doi:10.1016/S0140-6736(18)31204-2.
2] S. Sundar, J. Chakravarty, Antimony toxicity, International Journal of Environmental
Research and Public Health. 7 (2010) 4267-77, doi:10.3390/ijerph7124267.
3] S. Mohapatra, Drug resistance in leishmaniasis: Newer developments, Tropical
Parasitology. 4 (2014) 4-9, doi:10.4103/2229-5070.129142.
4]
A. Taylor, Aminopeptidases: structure and function., The FASEB Journalꢀ: Official
Publication of the Federation of American Societies for Experimental Biology. 7
(1993) 290–298, doi:10.1186/1750-1326-5-48.
[
5] J. Timm, M. Valente, D. García-Caballero, K.S. Wilson, D. González-Pacanowska,
Structural characterization of acidic M17 leucine aminopeptidases from the TriTryps
and evaluation of their role in nutrient starvation in Trypanosoma brucei, mSphere. 2
(2017) e00226-17, doi:10.1128/mSphere.00226-17.
[
[
6] U. Jankiewicz, W. Bielawski, The properties and functions of bacterial
aminopeptidases, Acta Microbiologica Polonica. 52 (2003) 217-31.
7] S.Y. Bhat, A. Dey, I.A. Qureshi, Structural and functional highlights of methionine
aminopeptidase 2 from Leishmania donovani, International Journal of Biological
Macromolecules. 115 (2018) 940–954, doi:10.1016/j.ijbiomac.2018.04.090.
[
8] A. Addlagatta, X. Hu, J.O. Liu, B.W. Matthews, Structural basis for the functional
differences between type I and type II human methionine aminopeptidases,
Biochemistry. 44 (2005) 14741–14749, doi:10.1021/bi051691k.
[
[
9] C. Giglione, A. Boularot, T. Meinnel, Protein N-terminal methionine excision, Cellular
and Molecular Life Sciences. 61 (2004) 1455-74, doi:10.1007/s00018-004-3466-8.
10] K.W. Walker, R.A. Bradshaw, Yeast methionine aminopeptidase I can utilize either
2
6

Zn2+ or Co2+ as a cofactor: a case of mistaken identity?, Protein Scienceꢀ: A
Publication of the Protein Society. 7 (1998) 2684–7, doi:10.1002/pro.5560071224.
11] G. Yang, R.B. Kirkpatrick, T. Ho, G.F. Zhang, P.H. Liang, K.O. Johanson, D.J.
Casper, M.L. Doyle, J.P. Marino, S.K. Thompson, W. Chen, D.G. Tew, T.D. Meek,
Steady-state kinetic characterization of substrates and metal-ion specificities of the
full-length and N-terminally truncated recombinant human methionine
[
aminopeptidases
(Type
2),
Biochemistry.
40
(2001)
10645–10654,
doi:10.1021/bi010806r.
[
12] L. Meng, S. Ruebush, V.M. D’souza, A.J. Copik, S. Tsunasawa, R.C. Holz,
Overexpression and divalent metal binding properties of the methionyl aminopeptidase
from
Pyrococcus
furiosus,
Biochemistry.
41
(2002)
7199–7208,
doi:10.1021/bi020138p.
[
[
[
[
13] V.M. D’souza, R.C. Holz, The methionyl aminopeptidase from Escherichia coli can
function as an iron(II) enzyme?, Biochemistry. 38 (1999) 11079-85,
doi:10.1021/bi990872h.
14] S. Liu, J. Widom, C.W. Kemp, C.M. Crews, J. Clardy, Structure of human methionine
aminopeptidase-2 complexed with fumagillin., Science (New York, NY). 282 (1998)
1
324–7, doi:10.1126/science.282.5392.1324.
15] S.Y.P. Chang, E.C. McGary, S. Chang, Methionine aminopeptidase gene of
Escherichia coli is essential for cell growth, Journal of Bacteriology. 171 (1989) 4071-
2
, doi:10.1128/jb.171.7.4071-4072.1989.
16] X. Li, Y.H. Chang, Amino-terminal protein processing in Saccharomyces cerevisiae is
an essential function that requires two distinct methionine aminopeptidases.,
Proceedings of the National Academy of Sciences. 92 (2006) 12357-61,
doi:10.1073/pnas.92.26.12357.
2
7

[
[
17] P.J. Rosenthal, Proteases of malaria parasites: New targets for chemotherapy,
Emerging Infectious Diseases. 4 (1998) 49–57, doi:10.3201/eid0401.980107.
18] F. Frottin, W. V. Bienvenut, J. Bignon, E. Jacquet, A. Sebastian Vaca Jacome, A. Van
Dorsselaer, S. Cianferani, C. Carapito, T. Meinnel, C. Giglione, MetAP1 and MetAP2
drive cell selectivity for a potent anti-cancer agent in synergy, by controlling
glutathione
redox
state,
Oncotarget.
7
(2016)
63306-63323,
doi:10.18632/oncotarget.11216.
[
19] S.H. Chan, C.H. Chui, S.W. Chan, S.H.L. Kok, D. Chan, M.Y.T. Tsoi, P.H.M. Leung,
A.K.Y. Lam, A.S.C. Chan, K.H. Lam, J.C.O. Tang, Synthesis of 8-hydroxyquinoline
derivatives as novel antitumor agents, ACS Medicinal Chemistry Letters. 4 (2013)
1
70-4, doi:10.1021/ml300238z.
[
[
20] M.P. Thakare, P. Kumar, N. Kumar, S.K. Pandey, Silica gel promoted environment-
friendly synthesis of 2,3-disubstituted 4-thiazolidinones, Tetrahedron Letters. 55
(2014) 2463-66, doi:10.1016/j.tetlet.2014.03.007.
21] X. Hu, A. Addlagatta, J. Lu, B.W. Matthews, J.O. Liu, Elucidation of the function of
type 1 human methionine aminopeptidase during cell cycle progression, Proceedings
of
the
National
Academy
of
Sciences.
103
(2006)
18148-53,
doi:10.1073/pnas.0608389103.
[
[
22] R.C. Priest, J. Spaull, J. Buckton, R.L. Grimley, M. Sims, M. Binks, R. Malhotra,
Immunomodulatory activity of a methionine aminopeptidase-2 inhibitor on B cell
differentiation, Clinical and Experimental Immunology. 155 (2009) 514-22,
doi:10.1111/j.1365-2249.2008.03843.x.
23] N. Sin, L. Meng, M.Q.W. Wang, J.J. Wen, W.G. Bornmann, C.M. Crews, The anti-
angiogenic agent fumagillin covalently binds and inhibits the methionine
aminopeptidase, MetAP-2, Proceedings of the National Academy of Sciences. 94
2
8