4
V. Mehta et al. / Bioorg. Med. Chem. Lett. xxx (2015) xxx–xxx
flexibly the ligands into binding site with incrementally build up
algorithm.28 However, in general not all algorithm are relevant
for metalloenzyme (ME) docking and addressing the issues like
coordination geometry, atomic charge variability and proton trans-
fer as they are poorly and indirectly parameterized for ME.29
Lead-it uses empirical scoring function (leadit score) which cal-
Phe198 (3.05 Å) whereas linker amide groups with Gly140 and
Tyr297. The electrostatic interactions were characterized by N
and
Phe198 and anion–
over, stacking was also observed between oxacalix[4]arene
aromatic ring and Phe198 peptide. In addition, Van der Waals con-
tacts and weak –alkyl, –sigma interactions were observed with
O
of nitro group employing cation–p interaction with
p
interaction with Tyr91, respectively. More-
p–p
culate free binding energy
D
G þ
G of the protein ligand complex by
p
p
R
R
P
D
G ¼
D
G0 þ
D
Grot ꢁ Nrot
¼
D
DGlipo
ꢂð
DR;
DaÞ here;
ꢂ
peptides like Pro22, Lys24, Phe338, Tyr297, His21, Gly295, Phe141,
Cys142, Gly294, Met130 and Leu23. Similarly the other representa-
tive compound 5 binds via H-bond strongly with Glu140 (2.07 Å)
and also with Tyr297 and Phe198 residues. Here, the substituent
methyl benzothiazole was observed to penetrate into the binding
tunnel via significant hydrophobic interactions employing
Met130, Gly129, Gly294, Gly295, Leu23, Phe141, Tyr91, Phe198,
His132 and Cys142. As reflected from Table 3, the LeadIt score
was highest for the compounds 5 and 7 due to significant hydrogen
bonding and other non-covalent contacts which were exercised in
scoring function as explained above.
The scores of the top-ranked conformations were dominated by
the H-bonding term. In some complexes (5, 8) they were domi-
nated by H-bonds and less lipophilic contacts, while in complexes
(5, 7, 2, and 4), combination of hydrogen and lipophilic contacts
were observed (e.g., the largest lipophilic contribution to the score
is seen in complex 7). The comparison of contributions to score of
best-ranked pose in all complexes illustrates that polar interac-
tions outweigh nonpolar interactions.
lipocount
ðD
R;
Da
Þ is a scaling function penalizing deviation from ideal
geometry. Nrot is the number of free rotatable bonds that are
immobilized in the complex.30 The terms
DG0 and DGlipo are
adjustable parameter and lipophilic contact energy, respectively.
Docking was articulated by preparing the ligand and protein for
eradicating possible ambiguities. The compounds were docked
around the constructed active site, that is, 10 Å surrounding the
co-crystallized ligand trichostatin with the following considera-
tions (i) base placement using enthalpy and entropy hybrid
approach, (ii) scoring with threshold for full score contribution
and no score contribution of 0.30 and 0.70, respectively, (iii)
parameters of clash handling values for protein ligand clashes with
maximum allowed overlap volume of 2.9 Å and intra-ligand
clashes with clash factor of 0.6, and (iv) maximum number of solu-
tion per iteration and fragmentation of 200.
As it was possible to observe, only seven compounds were
docked into the defined binding pocket with the given docking
input constraints. The best docked poses were chosen on the basis
of scoring function and their binding characteristics comparing
with the reference ligand. However, the nature of extended geom-
etry of calix[4]arenes and anticipated exposed binding pocket of
enzymes makes the binding motifs and ligand’s orientation more
significant. In general, the docked poses were in 1, 3 alternate con-
formations with enhanced rigidity which further signals for better
inhibitory activity.31
Furthermore, to gauze all the key information of ligand–recep-
tor complex, the docked poses were imported for the Hyde32
assessment. Hyde considers desolvation effects and estimate atom
based binding free energy based on dehydration and hydrogen
bonding interactions by employing Eq. 1.
1
X
D
GHyde
¼
D
Gdesolvation
þ
D
GH-bond
ð1Þ
The co-factor Zn2+ chelates with amide functions of the com-
pounds 2 and 5 only. This chelation was probably decisive as the
similar compounds were also ranked highest in experimental
activities. This deciphers for the critical importance of the cofactor
in posing the bioactivity. Interestingly, the nitro group (Oꢃ)
attached to calix ring system showed hydrogen bonding with
Tyr91 and Lys267 in compounds 1 and 4, with Phe198 in com-
pounds 2, 5 and 7 and with Tyr91, Lys267, His179 and Gln192 in
compound 8.
atomi
Greater Hyde score of compound 11 (
D
G = ꢃ41.0) implies that it
display weak desolvation energies and can easily ligated with
receptor through substantial H-bonding using residues Phe198,
Glu140, Phe141, Phe338, Tyr264, Lys24 and His170. However,
compounds 2, 5 and 7 also possess the negative
DG score. In con-
trast, the ligands 4 and 8 are strongly solvated (NO2ꢃ-H2O543),
hence possess desolvation penalties, which also reflected in their
D
G values. In addition, these ligands form imperfect hydrogen
Simplifying the docking result for compound 2, we describe
here the interaction map in Figure 3, the Zn2+ coordinate to the
oxygen of amide linker. The nitro group shows H-bonding with
bond of deviated length.
Comprehensively, docking study inferred that compounds were
bound well within the binding pocket of HDAC protein and the
Figure 3. (a) 2D plot of the ligand showing target protein amino acid and their binding to the specified atoms of the ligand 2. (b) Orientation of the ligand into the binding
pocket of the HDAC protein.