A.C.A. de Souza, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127350
further corroborate this binding hypothesis, an intermolecular re-
cognition experiment was carried out by means of extended molecular
dynamics (MD) simulations. Briefly, YopH and 3j were included in a
simulation box of TIP3P-type water molecules at a distance higher than
40 Å each other. Following energy minimization, heating, and density
equilibration, two replicas of unrestrained MD simulations were run
and trajectories were produced for 500 ns in each replica for a total
simulation time of 1 µs. Notably, 3j quickly approaches YopH, most
likely driven by electrostatic interactions that led the molecule to in-
teract within the same positively charged cleft al-ready identified by
molecular docking. Cluster analysis of MD frames showed that 3j spend
more than 75% of its time in that site. In the most representative pose
extrapolated from MD trajectories, 3j establishes direct H-bond inter-
actions with R278, and S345, whereas a water-bridged H-bond inter-
action is established with K379. The chalcone core is inserted in a hy-
drophobic cleft bounded by I344, M382, Y383, and A390, whose
conformation is adapted to the presence of the ligand (Fig. 2B).
Overall, computational modeling results are in line with experimental
evidence and further substantiate the non-competitive allosteric mode
of action of 3j at a molecular level.
Further, 3j efficacy in inhibiting the growth of Y. enterocolitica was
evaluated in vitro. Interestingly a direct effect of 3j in bacterial growth
was not observed, the maximum inhibition was nearly 20% after 48 h at
the evaluated concentrations (8 to 250 µM) (Fig. S90). YopH is essential
for virulence but not for bacterial growth, thus we assumed that 3j
inhibition would impair the intracellular survival of Y. enterocolitica
rather than its viability.
In summary, we report herein the synthesis and PTP inhibitory ac-
tivity of chalcones derivatives, while the inhibitory mechanism of the
derivative 3j toward a small collection of PTPs was characterized. We
found that compound 3j is a low-micromolar (Ki < 2 µM) PTPs in-
hibitor, although having different inhibition mechanisms on some PTPs.
Compound 3j is indeed a competitive inhibitor of LYP, PTP-PEST and
PtpB, and a non-competitive inhibitor of PTP1B, PtpA and YopH.
Remarkably, the mechanism of YopH inhibition by 3j was proven to be
irreversible and non-competitive. Further, 3j inhibition mechanism was
corroborated by the result of limited proteolysis. In addition, we
showed that CA is an irreversible mixed type inhibitor of YopH. Our
limited proteolysis analysis indicated that 3j binding alters the cleavage
pattern of YopH helix α2, suggesting that the binding site of 3j might be
in proximity of the well-known second substrate binding site of YopH.
Finally, molecular docking studies revealed that 3j bind within the
second binding site of YopH, and in the catalytic site for LYP, in line
with experimental evidence. Thus, we propose that 3j binds closely to
the second binding site of YopH, and irreversibly inhibits its activity.
Accordingly, the second binding site of YopH and compound 3j present
potential for the development of new allosteric inhibitors with a higher
potency and specificity for this enzyme.
Fig. 2. Binding mode of 3j to YopH predicted by molecular docking (A) or MD
simulations (B). A) the crystallo-graphic structure of YopH coded by PDB:
4YAA, after reverting the W345Y mutation, is shown as green cartoon. The
catalytic site is colored magenta, the catalytic cysteine is shown as sticks.
Residues of the second substrate binding site are colored orange; 3j binding
posed by GOLD is shown as yellow stick, that from AutoDock is shown as cyan
sticks. B) Magnification of the binding mode of 3j within the allosteric site of
YopH such as predicted by MD simulations. The molecule is shown as yellow
sticks, the protein as green cartoon. Residues contacted by 3j and described in
the text are showed as green sticks and are labeled. Polar contacts are high-
lighted by black dashed lines.
binding sites of YopH, a behavior that could be justified by the large
change in cleavage profiles presented herein and the mixed type in-
hibition displayed by CA. The MS results found for 3j and CA (Figs. S80
and S88) show a reduction in intensity of the fragment 2022 m/z,
suggesting that helix α2 may be strongly affected by a non-competitive
inhibitor.
Given the irreversible non-competitive inhibition and change in
helix α2 cleavage pattern, we next sought to determine the binding
mode of 3j in the three-dimensional structure of YopH using compu-
tational modeling tools. To this aim, we performed a blind docking
experiment with GOLD and AutoDock to scan the whole surface of the
crystallographic structure of YopH17 with 3j. Notably, both programs
converged to dock the molecule in a highly-basic groove (colored or-
ange in Fig. 2A) that is far from the catalytic cavity (colored magenta in
by anions in multiple X-ray crystallography structures of YopH, as well
as it has been identified in other works as putative second binding site
for substrates or inhibitors.18,38 Although the docking poses of GOLD
and AutoDock were slightly different, none of the programs placed 3j
within the catalytic site, in agreement with experimental evidences
described above. The same docking protocol was used to monitor the
interaction of 3j to LYP (Fig. S89), clearly suggesting that the molecule
binds exclusively within the catalytic site of this protein, in agreement
with the competitive mechanism of action observed experimentally. To
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
Acknowledgments
We dedicate this work to the memory of Professor Maurizio Botta, a
friend and mentor who loved to share knowledge. The authors would
like to thank the following facilities for their contributions to this work,
the Chemistry Department, Central Laboratory of Structural Biology
and Clinical Analysis Laboratory Unit (Sector Microbiology, University
Hospital) of the Federal University of Santa Catarina. We gratefully
acknowledge financial support from CNPq and CAPES (Brazil).
4