Targeting the transmembrane domain 5 of latent membrane protein 1 using small molecule modulators

Article abstract of DOI:10.1016/j.ejmech.2021.113210

Protein-protein interactions (PPIs) play a critical role in living cells and represent promising targets for the drug discovery and life sciences communities. However, lateral transmembrane PPIs are difficult targets for small-molecule inhibitor developme

Full text of DOI:10.1016/j.ejmech.2021.113210

European Journal of Medicinal Chemistry 214 (2021) 113210
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Targeting the transmembrane domain 5 of latent membrane protein 1
using small molecule modulators
,
Bo Zhang a b, Yibo Wang a, Cong Lin a, Hongyuan Li a, Xiaojie Wang a, Yinghua Peng c,
,
,
,
, b, *
Konstatin S. Mineev d e, Andrew J. Wilson f g, Hongshuang Wang a **, Xiaohui Wang a
a Laboratory of Chemical Biology, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun, Jilin, 130022, China
b Department of Applied Chemistry and Engineering, University of Science and Technology of China, Hefei, Anhui, 230026, China
c State Key Laboratory for Molecular Biology of Special Wild Economic Animals, Institute of Special Animal and Plant Sciences, Chinese Academy of
Agricultural Sciences, Changchun, Jilin, 130112, China
d ShemyakinꢀOvchinnikov Institute of Bioorganic Chemistry, Russian Academy of Sciences, Moscow, 117997, Russia
e Department of Biological and Medicinal Physics, Moscow Institute of Physics and Technology, Dolgoprudny, 141701, Russia
f School of Chemistry, University of Leeds, Woodhouse Lane, Leeds, LS2 9JT, UK
g Astbury Centre for Structural Molecular Biology, University of Leeds, Woodhouse Lane, Leeds, LS2 9JT, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Protein-protein interactions (PPIs) play a critical role in living cells and represent promising targets for
the drug discovery and life sciences communities. However, lateral transmembrane PPIs are difficult
targets for small-molecule inhibitor development given less structural information is known and fewer
ligand discovery methods have been explored compared to soluble proteins. In this study, the in-
teractions of the transmembrane domain 5 (TMD-5) of latent membrane protein 1 (LMP-1) of Epstein-
Barr virus (EBV) were disrupted by pentamidine derivatives to curb the committed step of EBV infec-
tion. A pentamidine derivative 2 with a 7-atom di-amide linker had the best activity whilst switching the
amide regiochemistry in the linker influenced membrane permeability and abolished anti TMD-5 ac-
tivity. Molecular dynamics simulations were performed to understand the interaction between pent-
amidine derivatives and TMD-5, and to rationalise the observed structure-activity relationships. This
study explicitly demonstrated that the interaction of small molecule with lipid should be considered
alongside interaction with the protein target when designing small molecules targeting the PPIs of TMDs.
In all, this study provides proof of concept for the rational design of small molecules targeting trans-
membrane PPIs.
Received 21 November 2020
Received in revised form
14 January 2021
Accepted 14 January 2021
Available online 27 January 2021
Keywords:
Protein-protein interactions
Pentamidine analogues
Latent membrane protein 1
The fifth transmembrane domain (TMD-5)
Epstein-barr virus
Small molecule modulators
© 2021 Elsevier Masson SAS. All rights reserved.
1. Introduction
compared to soluble proteins [4]. Despite the hydrophobic nature
of the transmembrane helix, the presence of charged residues is
Protein-protein interactions (PPIs) are fundamental for living
cells [1,2]. However, until recently they were considered undrug-
gable given that PPI interfaces are typically large and flatter than
those usually targeted by small molecules [3]. Lateral trans-
membrane PPIs are even more difficult to probe because less
structural information is known and fewer ligand discovery
methods have been applied to the identification of inhibitors when
believed to be important for membrane protein functions [5e8].
For instance, with assistance of a charged aspartic acid (D150) in its
transmembrane domain 5 (TMD-5) [9], the oncogenic latent
membrane protein 1 (LMP-1) of Epstein-Barr virus (EBV) can form
homo-oligomeric complexes on lipid rafts to active downstream
signaling, leading to cell transformation [10e13]. As the first human
cancer virus to be discovered, it is highly infectious and about 90%
of the world’s population is infected [14]. Studies have shown that
EBV infection is highly correlated with multiple malignancies,
including Burkitt’s lymphoma, Hodgkin’s lymphoma, stomach
cancer, nasopharyngeal carcinoma, etc. [15e17] Unfortunately,
LMP-1 function is independent of ligand binding and it effectively
serves as a constitutively active receptor following oligomerization
* Corresponding author. Laboratory of Chemical Biology, Changchun Institute of
Applied Chemistry, Chinese Academy of Sciences, Changchun, Jilin, 130022, China.
** Corresponding author.
E-mail addresses: hongshuang.wang@ciac.ac.cn (H. Wang), xiaohui.wang@ciac.
ac.cn (X. Wang).
https://doi.org/10.1016/j.ejmech.2021.113210
0223-5234/© 2021 Elsevier Masson SAS. All rights reserved.

B. Zhang, Y. Wang, C. Lin et al.
European Journal of Medicinal Chemistry 214 (2021) 113210
[18]. Therefore, the lateral transmembrane PPIs between mono-
mers represent one of the limited current targets for drug discovery
against EBV. Recent studies showed that TMD-5 itself can also form
an oligomer [9,19], rendering it an ideal model for small molecule
screening and development targeting LMP-1 to curb the committed
step of EBV infection [20e22]. Pentamidine, which is a clinical
antifungal medication, has been shown to disrupt the oligomeri-
zation of LMP-1 (Fig. 1) [20,21]. However, its inhibition efficiency is
moderate and its molecular mechanism of action is not well un-
derstood. Herein pentamidine analogues 1e5 with linkers of
different lengths, varieties and orientations were designed and
their anti TMD-5 activities were tested. A pentamidine derivative
with a 7-atom di-amide linker, 2, had the best activity. Interestingly,
switching the regiochemistry of the amides in the linker to give a
pentanediamide derivative abolished anti TMD-5 activity. Further
molecular dynamics simulations were performed to understand
the interaction between the small molecules 1e5 and TMD-5, and
to rationalise the observed structure-activity relationships (SAR).
Fig. 2. Concentration-dependent inhibition curves of pentamidine analogues on TMD-
5 oligomerization was measured by chimeric TMD-5 ToxR assay.
membrane [23,24]. The chimeric TMD-5 was expressed with
maltose binding protein for localization on the membrane and
transcription factor ToxR to provide a reporter for changes in
oligomerization. As shown in Fig. 2 and Table 1, pentamidine de-
rivative 2, which was constructed by replacing the bis(oxy)-
pentane linker of pentamidine with pentanediamide and has the
same number of linker atoms as pentamidine, inhibited ToxR ac-
tivity in a concentration-dependent manner with an IC50 of
2. Results and discussion
The chemical structure of pentamidine (Fig. 1) is symmetrical
and contains two benzamidine motifs coupled by a bis(oxy)-
pentane linker. Previous limited SAR studies showed that the
positively charged benzamidine group is critical for anti TMD-5
activity [20,21]. However, how the linker affects its anti TMD-5
activity is not clear. Herein SAR analyses on the linker connecting
two benzamidine groups in pentamidine was explored. Consid-
ering that only hydrogen bond acceptors (oxygens) exist in the
linker of pentamidine, amide or secondary amine groups (both
hydrogen bond acceptors and donors) were incorporated into
pentamidine analogues to enhance protein-ligand interactions.
An E. coli FHK-12 cell-based ToxR assay was developed to study
0.6
0.3
mM, which is 15-fold higher than pentamidine
(9.1 0.6
mM) [21], suggesting it could disrupt the oligomerization
of TMD-5. By changing the linker length by one-atom with respect
to 2, 1 was formed by linking two benzamidine motifs with buta-
nediamide and 3 was designed by linking the benzamidine motifs
with hexanediamide. Compared to 2, 1 (IC50 ¼ 111.5 15.0
mM) and
3 (IC50 ¼ 42.0 14.8
m
M) showed much weaker anti TMD-5 po-
tencies, which indicates that the length of the linker is critical and a
linker of 7-atoms in length is optimal. However, switching the
the association of the transmembrane helix in
a biological
Fig. 1. Chemical structures of pentamidine and its derivatives 1e5. The linker is highlighted in red in pentamidine.
2

B. Zhang, Y. Wang, C. Lin et al.
European Journal of Medicinal Chemistry 214 (2021) 113210
Table 1
rationalise the observed SAR. The root-mean-square deviation
(RMSD) of TMD-5 over 50 ns was analyzed across the series of
pentamidine analogues (Fig. 4). The RMSD values of TMD-5 in the
protein-ligand complexes showed that all TMD5s can reach a stable
state in 50 ns.
Binding mode analysis showed that all compounds exhibited
similar interaction patterns with TMD-5: one amidine group
interacted with the charged D150 of TMD-5 to sterically encumber
oligomer formation whilst the other amidine group reached out to
the polar head group of the membrane lipids (Fig. 5). 1, 3 and 5
caused bending of TMD-5 in the membrane (the bending angles of
TMD-5 are shown in Fig. S1), which would incur an energy penalty
and therefore reduce the TMD-5 binding affinity and anti-TMD
activity. In contrast, 2 and 4, had much less influence on confor-
mational changes of TMD-5 than the other compounds, which
indicated that a lower energy penalty was required to bend TMD-5
when 2 or 4 binds to TMD-5.
Interactions of the linker nitrogen with the lipid head groups,
which appear to be a key driving force for bending of TMD-5 in the
membrane, were further investigated. Compared to other pent-
amidine derivatives, 2 and 4 showed formation of fewer H-bonds
between linker nitrogen and lipid head groups (Fig. 6, white col-
umns). Meanwhile, 4 is the only analogue with an electron with-
drawing group para to the amidine, decreasing the ability of
forming hydrogen bonds of amidines. As shown in Fig. 6 (grey
columns), amidines of 4 formed the fewest hydrogen bonds with
TMD-5 among all pentamidine analogues. These provide an
explanation as to why they did not substantially bend the TMD-5
helix. Together, these MD simulations in the membrane milieu
indicate that 2 and 4 should have potent TMD-5 binding affinity
and anti-TMD activity. However, experimental evidence indicated
that 2 was a potent TMD-5 disruptor while 4 had no apparent anti-
TMD-5 activity.
It should be noted that during the above MD simulations the
small molecules were first directly docked into TMD-5 in the
membrane milieu. Considering that pentamidine derivatives first
permeate into the membrane and then interact with TMD-5, it was
therefore speculated that 4 might have poor membrane perme-
ability. In order to test this hypothesis, the free energy changes for 2
and 4 crossing the membrane were calculated. As shown in Fig. 7A,
2 overcame an energy barrier of 1.5 kcal/mol to enter the mem-
brane and stayed in the middle of membrane with a free energy
change of ꢀ6.8 kcal/mol; 4 had to overcome an energy barrier with
~3.0 kcal/mol to permeate into the membrane and had a free en-
ergy minimum of ꢀ4.9 kcal/mol. Compared to 2, the parameters
IC50 for ToxR assay, cell viability and TMD-5 binding affinity (Kd) of pentamidine
analogues.
Compound
IC50 of ToxR (
m
M)
FHK 12 cell viability (
m
M)
Kd (mM)
1
2
3
4
5
111.5 15.0
0.6 0.3
42.0 14.8
>200
>200
56.4 2.1
>200
>200
>200
13.5 3.3
2.6 0.4
8.9 2.3
n.d
31.2 1.4
7.3 1.8
n.d., not determined because the binding affinity was too weak to fit; All experi-
ments were performed at least three times independently, and data are given as the
mean SD.
regiochemistry of the amide from 2 to give 4 resulted in the
abolition of anti TMD-5 activity (IC50 > 200 mM), which revealed a
critical importance of amide orientation on inhibitory potency.
Changing the linker from an ether group (pentamidine) or amide
group (2) to an amino group (5), the anti TMD-5 activity also
diminished (IC50 ¼ 31.2 1.4
mM). It should be noted that no de-
rivative showed significant cell proliferation inhibition (Table 1 for
cell viability). Together, the results showed the linker composition,
length and the functional group orientation profoundly influence
pentamidine derivative’s anti TMD-5 activity.
To further confirm the interaction between pentamidine de-
rivatives and the chimeric TMD-5 system happened in the mem-
brane domain, a direct TMD-5 biophysical binding experiment in
the membrane environment was performed [9]. Tryptophan (Trp)
emission fluorescence intensity of TMD-5 was recorded upon
titration with pentamidine derivatives; this led to quenching of the
TMD-5 intrinsic fluorescence (Fig. 3). The dissociation constants are
summarized in Table 1. A dissociation constant of 2.6 0.4
determined for 2. Compared to 2, 1 (Kd ¼ 13.5 3.3 M) and 3
M) had weaker binding affinity to TMD-5, which is
mM was
m
(Kd ¼ 8.9 2.3
m
consistent with the TMD-5 ToxR functional assays and confirmed
the optimal linker of 7-atoms for pentamidine analogues. Inter-
estingly, the binding affinity of 4, which was designed by reversing
the orientation of the amides in the linker of 2, was too weak to fit.
Finally, the TMD-5 binding affinity of 5 was 7.3
1.8 mM, again
weaker than that of 2. Collectively, the TMD-5 biophysical binding
data were consistent with the TMD-5 ToxR activity data and
confirmed the linker profoundly influences the anti-TMD-5 activity
of pentamidine derivatives.
Molecular dynamics (MD) simulations were employed to un-
derstand how pentamidine derivatives interact with TMD-5 and to
Fig. 3. Representative titration curves of TMD-5 with pentamidine analogues. 290 nm
was used as the excitation wavelength for intrinsic Trp protein fluorescence; emission
fluorescence at 337 nm was plotted against the pentamidine analogue concentration.
Fig. 4. Time evolution curves of RMSD of TMD-5 helix backbone during molecular
dynamics simulations of TMD-5/pentamidine analogue complexes.
3

B. Zhang, Y. Wang, C. Lin et al.
European Journal of Medicinal Chemistry 214 (2021) 113210
Fig. 5. Conformations of TMD-5 with pentamidine analogues after equilibration. The grey lines represent the lipids and the orange spheres represent phosphorus atoms in the
lipids. TMD-5 is shown as a green ribbon. The key residue D150 is shown with spheres and the cyan sticks correspond to the compounds.
oligomerization which is most likely due to its poor membrane
permeability. In contrast, 2, which had better membrane perme-
ability, interacted with TMD-5 without apparent bending of the
TMD-5 helix, favoring TMD-5 binding and TMD-5 oligomerization
disruptor activity.
The free energy profiles for 2 binding to TMD-5 are shown in
Fig. 8. In the PMF profiles, compound 2 reached a global minimum
at z ¼ 5.5 Å with ꢀ4.6 kcal/mol. However, the minimum for 4 is at
z ¼ ꢀ0.5 Å with ꢀ3.6 kcal/mol. The conformations and locations
with lowest free energy of compounds 2 and 4 interacting with
TMD-5 are consistent with the equilibrium simulation results (see
Fig. 5). Significantly, the presence of TMD-5 decreased the energy
barrier for compound permeation into the membrane. However,
the barrier for permeation of 4 is still slightly higher than 2 whilst
binding of 4 is less favorable than that of 2 based on the free en-
ergies. Considering the broader permeation barrier for 4, it may be
kinetically unfavourable for 4 to enter the membrane even in the
presence of TMD-5.
Fig. 6. Number of H-bonds per frame (total number of H-bonds/number of frames)
3. Conclusions
formed between nitrogen in the linker and lipids (white columns) and between ni-
trogen in amidines and TMD-5 (grey columns).
We designed pentamidine analogues to disrupt the oligomeri-
zation of TMD-5 by targeting D150. 2 with a linker length of 7
governing partition of 4 into the membrane milieu were kinetically
and thermodynamically less favorable. To validate the calculation
results about pentamidine derivatives 2 and 4 crossing cell mem-
branes, the MDCK transport experiments were performed. The
apparent permeability coefficient (Papp) of compounds 2 and 4 are
shown in Fig. 7B. An apparent permeability coefficient (Papp) of
4.28 ꢁ 10ꢀ6 1.24 ꢁ 10ꢀ7 cm/s was determined for 2. Compared to
2, 4 (Papp ¼ 1.07 ꢁ 10ꢀ6 1.03 ꢁ 10ꢀ7 cm/s) had much weaker
membrane penetration ability, which is consistent with the above
free energy calculations. Therefore, 4, generated by switching the
position of the carbonyl group and amino group relative to the
linker in 2, showed weak ability to bind and disrupt TMD-5
atoms had the best activity whilst switching the amide regio-
chemistry in the linker influenced membrane permeability and
abrogated anti TMD-5 activity. Further examination of the effects of
pentamidine derivatives on the oligomerization and function of full
length of LMP-1 of EBV is worthy of investigation. In contrast to
traditional drug discovery, targeting the PPIs of TMD is a multi-
body problem; the interaction of small molecule with lipid
should be considered alongside small molecule interactions with
target proteins, and target protein-lipid interactions. The perme-
ability of the small molecule, which may be influenced by the
structural orientation of its functional groups, is critical because
local lipophilicity is likely to influence conformation, stability, and
4

B. Zhang, Y. Wang, C. Lin et al.
European Journal of Medicinal Chemistry 214 (2021) 113210
Fig. 7. (A) The free energy profiles (PMFs) along the permeation route of pentamidine analogues into a model bilayer; (B) The apparent permeability coefficient (Papp) of pent-
amidine analogues 2 and 4 was measured by MDCK transport experiments.
LB þ arabinose (0.0025%) and chloramphenicol (30
different concentrations of testing compound. Five replica Cultures
were incubated with shaking at 37 ꢂC for 20 h and
-galactosidase
activity was measured using a SYNERGY H1 Micro-plate Reader
(BioTek Instruments, Carlsbad, CA, USA). Briefly, 5 L of culture was
transferred to the wells of a Costar 3596 polystyrene 96-well plate
(Corning, NY, USA) containing 100 L Z buffer/chloroform (1%
mg/mL) with
b
m
m
b-
mercaptoethanol, 10% chloroform, 89% A buffer: 1 M sodium
phosphate, 10 mM KCl, 1 mM MgSO4 and pH 7.0). Cells were lysed
by addition of 50
in Z buffer) and shaking at 28 ꢂC for 10 min. 50
Nitrophenyl- -galactoside (ONPG, 0.4% w/v in Z buffer) was added
and
mL Z buffer/SDS (1.6% w/v sodium dodecyl sulfate
mL Z buffer/o-
b
b
-galactosidase activity was measured by monitoring the re-
action at 405 nm for a period of 20 min at 28 ꢂC. The effect of testing
compound on FHK12 E. coli. proliferation was also investigated by
measuring the OD at 600 nm (Table 1).
Fig. 8. The binding free energy profiles of compounds 2 and 4 interacting with TMD-5.
The reaction coordinate was defined as the distance along the z axis between the
center of mass of non-hydrogen atoms of the compound and the carbon atom of the
carboxyl group of D150 (see Fig. S2).
4.2. Peptide synthesis
TMD-5 peptide (KKKK-WQLLAFFLAFFLDLILLIIALYL-KKKK) was
synthesized and purified by
a CRO company (GL Biochem,
Shanghai, China). The identity and purity of this peptide was
confirmed by MALDI and HPLC (>98%) (Fig. S8 and S9). The con-
centration of TMD-5 was determined by absorbance at 280 nm
oligomerization in the transmembrane domain. In all, this study
provides proof of concept for the rational design of small molecules
targeting transmembrane PPIs.
using the extinction coefficient of 6990 Mꢀ1cmꢀ1
.
4. Materials and method
4.3. Biophysical binding titration
4.1. ToxR assay
The intrinsic Trp fluorescence of TMD-5 was used to investigate
the interaction of TMD-5 with pentamidine analogues. Fluores-
cence measurements were performed on a Cary Eclipse spectro-
fluorimeter (Agilent Technologies, Santa Clara, CA, USA). All
measurements were carried out under room temperature in a
2 ꢁ 10 mm quartz cell (Starna Cells, Atascadero, CA, USA). 290 nm
was chosen as the excitation wavelength of Trp fluorescence and
ToxR assay was performed and validated as described previously
(Scheme 1) [20,21]. Briefly, ToxR7-TMD-5 plasmid (200 ng) was
transformed into 200
42 ꢂC for 90 s and incubated on ice for 2 min, followed by addition
of 800
L SOC media and incubation with shaking at 37 ꢂC for 1 h.
50 L of the transformation mixture was used to inoculate 5 mL
mL FHK12 competent cells with heat shock at
m
m
5

B. Zhang, Y. Wang, C. Lin et al.
European Journal of Medicinal Chemistry 214 (2021) 113210
Scheme 1. ToxR assay. TMD-5 can self-associate in an E. coli Membrane to activate the transcriptional activator ToxR, promoting binding to the ctx promoter and initiating lacZ
transcription. It will further regulate the production of -galactosidase. A colorimetric readout is obtained by addition of ONPG to lysed cells.
b
emission at 310e450 nm was measured. Appropriate controls were
subtracted from spectra obtained on the samples. 1 М TMD-5 in
50 mM HEPES (pH ¼ 7.4) buffer contained 150 M C14 betaine (3-
range electrostatic interactions [35]. Non-bonded interactions were
switched off at 10e12 Å and periodic boundary conditions were
applied in all directions. The timestep was set as 2 fs and each
system was run for 50 ns. Only the last 30 ns were used for
hydrogen bond analysis.
m
m
(N,N-dimethyl- myristylammonio)propanesulfonate) micelle was
titrated with different concentrations of testing compound. The
fluorescence intensity at 337 nm was plotted against compound
concentration. The raw data was fitted by non-linear least square
method using the equation:
The compound permeation simulations were set with the same
conditions. To explore the free energy changes when compounds
permeate into the lipid bilayer, umbrella sampling simulations
were performed. A harmonic biasing potential with a force constant
of 10 kcal/(mol$Å) along the perpendicular direction of lipid bilayer
was applied on the center of mass (COM) of non-hydrogen atoms of
compounds. The sampling windows were spaced every 0.5 Å
from ꢀ40 to 40 Å resulting in 161 windows. Each window was run
for 3 ns as only the last 2 ns was used for analysis. The free energy
profiles were rebuilt with the Weighted Histogram Analysis
Method (WHAM) and the tolerance was set to 10ꢀ6 [36,37]. The
statistical uncertainties were estimated as described previously
[38].
F ¼ 0.5 ꢁ (2 ꢁ F0-FPL ꢁ (KDþ[LT]þ[PT]-((KDþ[LT]þ[PT])2-
4 ꢁ [LT] ꢁ [PT])0.5))
where F, the observed fluorescence; F0, initial fluorescence of pro-
tein in the absence of ligand; FPL, adjustable parameter for protein-
ligand complex molar fluorescence; KD, dissociation constant; [LT],
total concentration of the ligand; [PT], total protein concentration.
4.4. Docking
Free energy profiles of compounds binding to TMD-5 were also
calculated by umbrella sampling methods. Same simulation con-
ditions described above for compound permeation simulations
were applied. The reaction coordination was defined as the dis-
tance along z axis between the COM of non-hydrogen atoms of
compound and the carbon atom of carboxyl group of D150 (see
Fig. S2). Each window was run for 11 ns as only the last 10 ns was
used for analysis.
The structure for the TMD-5 was built as described previously
[22]. The state of D150 was set to be charged and the compounds
were prepared by Maestro [25]. The charge state of compounds
were generated at pH ¼ 7.0 2.0 (all compounds are protonated
with the charge of þ2). AutoDock Vina 1.1.2 [26] was applied with
default setting options to dock and score the compounds. The best
scored conformations were used for molecular dynamics
simulations.
4.6. MDCK transport experiment
4.5. Molecular simulations
Dulbecco’s Modified Eagle’s Medium (DMEM) and TrypsineEDTA
were purchased from HyClone (Uath, USA). Penicillin, streptomycin
The TMD-5 helix and compound were embedded in a 1,2-
dimyristoyl-sn-glycero-3-phosphocholine (DMPC) bilayer and sol-
vated with TIP3P water molecules in a 60 ꢁ 60 ꢁ 120 Å3 box via the
CHARMM-GUI membrane builder protocol [27]. 150 mM KCl was
added to neutralize the system and mimic the physiological con-
centration. The protein, lipids, and ions were described by the
CHARMM36 force field [28,29] and compounds 1, 2, 3, and 4 were
parameterized by the CGenFF program [30]. Considering that the
penalties of compound 5 was too high, it was re-parameterized
with GAAMP program [31]. All MD simulations were performed
by the NAMD2.12 program package [32].
and G418 (þ10000 Units/mL penicilin, þ10000
mg/mL streptomycin)
were purchased from Grand Island (NY, USA). Foetal bovine serum
(FBS) and Hank’s Balanced Salt Solution (HBSS) were obtained from
BI (Biological Industries Israel Beit Haemek LTD., CT, USA). Trans-
well™ plates of 12 wells (12 mm, pore size 0.4 mm) were purchased
from Corning Costar (Cambridge, MA, USA).
4.6.1. Cell culture and evaluate the viability of cells
MDCK cells were cultured in DMEM, supplemented with 10% (v/
v) FBS, 1% penicillin-streptomycin solution and 0.5 mL G418 solu-
The models with ~40,000 atoms were simulated under the NPaT
tion (600 m
g/mL) at 37 ꢂC in an incubator containing 5% CO2.
ꢀ
ensemble with a pressure of 1 atm at 310.15 K. The Nose-Hoover
The passage number of the cells was 20e35. Viability of cells
was directly measured using MTT assay to evaluate the maximum
nontoxic concentration to the cells (Fig. S3). For the MTT assay,
MDCK cells were seeded onto a 96-well culture plate with a density
Langevin piston method [33,34] was applied for the pressure con-
trol and the Langevin thermostat for the temperature coupling. The
particle mesh Ewald (PME) algorithm was employed to treat long-
6

B. Zhang, Y. Wang, C. Lin et al.
European Journal of Medicinal Chemistry 214 (2021) 113210
of 4000 cells/well and incubated at 37 ꢂC for 24 h. Then the medium
was replaced with 100 mM pentamidine analogues 2 and 4 with
(Hz), and integration. Detailed procedures with synthetic rout were
exhibited in Scheme 2.
different concentrations, which diluted in DMEM containing 10%
FBS. The control group was treated with the same amount solvent
without drugs. Meanwhile, the wells without cells were used as
4.7.1. Synthetic and 1H and 13C NMR and HRMS characterization of
compound 1 [40]
blank control. After 4 h, the medium was replaced with 100
mL MTT
N1,N4-bis(4-carbamimidoylphenyl)succinamide (1). To a solu-
tion of 4-aminobenzamidine dihydrochloride (2.08 g, 10 mmol) in
pyridine (8 mL) and DMF (40 mL) was added succinyl dichloride
(0.56 mL, 5 mmol). The resulting mixture was stirred at 150 ꢂC for
2 h, and a large amount of grey solid was generated. Then the re-
action mixture was cooled to room temperature and filtrated. The
filter cake was washed with water (20 mL) and acetone (20 mL) and
dried under high vacuum for 3 h to give the crude compound 1,
which was purified by flash column chromatography to afford 1
(1.70 g, 48%) as a white solid. M.p. 218.3e220.9 ꢂC (lit [40]. >300 ꢂC
at the concentration of 1 mg/mL to form formazan crystals. And
then the MTT solution was removed and the formazan crystals were
dissolved into 150 mL DMSO. The absorbance at 570 nm was showed
on a microplate reader (Tecan, Austria). The percentage of cell
viability was calculated as follows:
Cell viability % ¼ (model-blank) / (control-blank) ꢁ 100%
in a dihydrochloride salt). 1H NMR (300 MHz, DMSO‑d6)
J ¼ 8.7 Hz, 4H), 7.61 (d, J ¼ 8.7 Hz, 4H), 4.65e2.99 (m, 8H) 2.69 (s,
4H) ppm; 13C NMR (75 MHz, DMSO‑d6):
d
¼ 7.71 (d,
4.6.2. Evaluation of MDCK cell monolayers
The formation of functional epithelial layers was monitored by
measuring the transepithelial electrical resistance (TEER) with an
epithelial volt-ohm meter in order to evaluate the integrity of the
MDCK cell monolayers before each experiment [39]. Only cell
d
¼ 171.1, 171.1, 162.6, 162.6,
141.2, 141.2, 130.2, 130.2, 130.2, 130.2, 127.6, 127.6, 118.5, 118.5, 118.5,
118.5, 31.6, 31.6 ppm; HRMS (ESI-Orbitrap) m/z: [M þ H]þ calcd for
C
18H21N6Oþ2 353.1721; found 353.1719.
monolayers with a TEER of above 200
U
/cm2 were used for the
transport assays. The TEER values were also measured after the
experiment to check monolayer integrity.
4.7.2. Synthetic and 1H and 13C NMR and HRMS characterization of
compound 2 [41]
N1,N5-bis(4-carbamimidoylphenyl)glutaramide (2). Following
the procedure described above for the synthesis of compound 1,
4.6.3. Preparation of transport samples
MDCK cells were seeded at the density of 5 ꢁ 104 cells/well on
Transwell. Culture medium was replaced with fresh medium every
two days, and the monolayers were grown for 4e5 days to reach
confluence and differentiation after seeding. Before the experi-
ments, the MDCK cell monolayers were washed three times with
pre-warmed HBSS medium (pH 7.4), the monolayers were incu-
bated in HBSS for 30 min at the finality. Then the drug solutions
were added to either the apical (AP, 0.5 mL) or basolateral side (BL,
bisbenzamidine
2
(1.84 g, 50%) was prepared from 4-
aminobenzamidine dihydrochloride and glutaryl dichloride. M.p.
213.5e214.6 ꢂC (lit [41]. >300 ꢂC in a dihydrochloride salt). 1H NMR
(300 MHz, DMSO‑d6):
d
¼ 7.72 (d, J ¼ 8.6 Hz, 4H), 7.63 (d, J ¼ 8.6 Hz,
4H), 2.41 (t, J ¼ 7.1 Hz, 4H), 1.99e1.84 (m, 2H) ppm; 13C NMR
(75 MHz, DMSO‑d6):
d
¼ 171.5, 171.5, 162.7, 162.7, 141.2, 141.2, 131.1,
131.1, 131.1, 131.1, 127.2, 127.2, 127.2, 127.2, 118.7, 118.7, 36.0, 36.0,
21.2 ppm; HRMS (ESI-Orbitrap) m/z: [M þ H]þ calcd for C19H23N6Oþ2
367.1877; found 367.1873.
1.5 mL). 500 mL of the sample was collected from the opposite side
at 150 min and then immediately frozen, lyophilized and preserved
below ꢀ20 ꢂC. Before quantitative analysis, the samples were dis-
4.7.3. Synthetic and 1H and 13C NMR and HRMS characterization of
compound 3 [41]
solved in 100 mL water.
N1,N6-bis(4-carbamimidoylphenyl)adipamide (3). Following the
procedure described above for the synthesis of compound 1, bis-
The apparent permeability coefficients (Papp) were calculated
as:
benzamidine
3
(2.09 g, 55%) was prepared from 4-
Papp¼dQ/dt ꢁ (1/C0A)
aminobenzamidine dihydrochloride and adipoyl dichloride. M.p.
P
app: apparent permeability coefficient (cm/s), dQ/dt: perme-
211.4e212.9 ꢂC (lit [41]. >300 ꢂC in a dihydrochloride salt). 1H NMR
ability rate of the compound on the receiver.
(300 MHz, DMSO‑d6):
d
¼ 7.71 (d, J ¼ 8.1 Hz, 4H), 7.63 (d, J ¼ 8.1 Hz,
4H), 2.37 (s, 4H), 1.63 (s, 4H) ppm; 13C NMR (75 MHz, DMSO‑d6):
4.7. Chemistry methods
d
¼ 171.9, 171.9, 162.9, 162.9, 141.5, 141.5, 130.2, 130.2, 130.2, 130.2,
127.8, 127.8, 118.6, 118.6, 118.6, 118.6, 36.7, 36.7, 25.2, 25.2 ppm;
HRMS (ESI-Orbitrap) m/z: [M þ H]þ calcd for C20H25N6O2þ 381.2034;
found 381.2032.
Common reagents and materials were purchased from com-
mercial sources and used as received without further purification
unless otherwise stated. Anhydrous ethanol (EtOH) was refluxed
with magnesium powder and distilled under nitrogen before use.
Anhydrous methanol (MeOH) and dichloromethane (DCM) were
freshly distilled from CaH2. Anhydrous dioxane was freshly distilled
from anhydrous CaCl2 prior to use. TLC plates were visualized by
exposure to ultra violet light (UV). High resolution mass spec-
trometer (HRMS) data were acquired in positive ion mode using a
Thermo Fisher LTQ ORBITRAP XL with an electrospray ionization
(ESI) source. Nuclear magnetic resonance (NMR) spectra were ac-
quired on a Bruker AV-300 spectrometer (300 MHz 1H, 75 MHz
4.7.4. Synthetic and 1H and 13C NMR and HRMS characterization of
compound 4 [42]
N,N’-(propane-1,3-diyl)bis(4-cyanobenzamide) (9). To a solu-
tion of propane-1,3-diamine (0.83 mL, 10 mmol) in Et3N (2.8 mL)
and DCM (30 mL) was added dropwise 4-cyanobenzoyl chloride
(3.31 g, 20 mmol) in DCM at 0 ꢂC by a constant pressure titration
funnel, then warmed to ambient temperature. After stirring for
24 h, the solvent was removed under vacuum and the residue was
washed with saturated aqueous NaHCO3 (30 mL), hydrochloride
acid aqueous (30 mL, 1 N) and water (30 mL) to give crude com-
pound 9 (3.20 g). The crude product was recrystallized by (50e90)%
aqueous DMSO to yield compound 9 (2.96 g, 89%) as a white solid.
M.p. 215.2e216.1 ꢂC (lit [42]. 212.5e213.5 ꢂC). 1H NMR (300 MHz,
13C). Chemical shifts (
d) are expressed in ppm downfield from tet-
ramethylsilane (TMS) using non-deuterated solvent present in the
bulk deuterated solvent (CDCl3: 1H 7.26 ppm, 13C 77.16 ppm;
DMSO‑d6: 1H 2.50 ppm, 13C 39.52 ppm). Data are represented as
follows: chemical shift, multiplicity (s ¼ singlet, d ¼ doublet,
t ¼ triplet, m ¼ multiplet, br ¼ broad), coupling constant in Hertz
DMSO‑d6):
d
¼ 8.77 (t, J ¼ 5.4 Hz, 2H), 8.03e7.91 (m, 8H), 3.35 (dd,
J ¼ 12.7, 6.6 Hz, 4H), 1.81 (p, J ¼ 6.9 Hz, 2H) ppm. 13C NMR (75 MHz,
7

B. Zhang, Y. Wang, C. Lin et al.
European Journal of Medicinal Chemistry 214 (2021) 113210
Scheme 2. Synthetic route for pentamidine analogues. Reagents and conditions: (a) pyridine, DMF, 150 ꢂC; (b) DCM, Et3N, ambient temperature; (c) EtOH, HCl (gas); (d) EtOH,
NH3 (gas), ambient temperature; (e) DMSO, Et3N, 150 ꢂC; (f) dioxane, MeOH, HCl (gas), ambient temperature.
DMSO‑d6):
d
¼ 165.3, 165.3, 139.0, 139.0, 132.9, 132.9, 132.9, 132.9,
freezer to form precipitate, which was filtrated, washed with water
(10 mL) and dried under high vacuum to give compound 4 (0.90 g,
41% from compound 10) as a white solid. M.p. 230.4e231.6 ꢂC (lit
[42]. 302.5e304.5 ꢂC in a dihydrochloride salt). 1H NMR (300 MHz,
128.4, 128.4, 128.4, 128.4, 118.8, 118.8, 114.0, 114.0, 37.8, 37.8,
29.3 ppm; HRMS (ESI-Orbitrap) m/z: [M þ H]þ calcd for C19H17N4Oþ2
322.1273; found 322.1277.
Diethyl-4,4’-((propane-1,3-diylbis(azanediyl))bis(carbonyl))
dibenzimidate (10). To a solution of dry EtOH (50 mL) was bubbled
fresh prepared HCl (gas) for 8 h under ice bath. Then compound 9
(2.0 g, 6.02 mmol) was added to the fresh ethanol solution of HCl
(50 mL) and the reaction mixture was slowly warmed to ambient
temperature. After vigorously stirred for 4 d, the reaction mixture
was cooled in a freezer to form precipitate. The precipitate was
filtered and the filter cake was washed with diethyl ether
(3 ꢁ 100 mL), dried under high vacuum to yield crude compound 10
(1.76 g), which was unstable and used for the next step without
further characterization and purification.
N,N’-(propane-1,3-diyl)bis(4-carbamimidoylbenzamide) (4). To
a solution of dry EtOH (50 mL) was bubbled fresh prepared NH3
(gas) for 8 h under ice bath. The compound 10 (1.76 g) was added to
the saturated ethanolic ammonia solution (50 mL). The reaction
mixture was sealed and slowly warmed to ambient temperature
and vigorously stirred for 4 d. Then, the mixture was cooled in a
DMSO‑d6):
d
¼ 8.59 (s, 2H), 7.85 (s, 8H), 6.48 (s, 6H), 3.34 (d,
J ¼ 5.2 Hz, 4H), 1.77 (m, 2H) ppm; 13C NMR (75 MHz, DMSO‑d6):
d
¼ 166.2, 166.2, 162.3, 162.3, 139.0, 139.0, 137.3, 137.3, 136.9, 136.9,
136.9, 136.9, 136.0, 136.0, 136.0, 136.0, 37.6, 37.6, 29.7, 29.7 ppm.
HRMS (ESI-Orbitrap) m/z: [M þ H]þ calcd for C19H23N6O2þ 367.1877;
found 367.1875.
4.7.5. Synthetic and 1H and 13C NMR and HRMS characterization of
compound 5 [43]
4,4’-(pentane-1,5-diylbis(azanediyl))dibenzonitrile (11) [44]. To
a solution of 4-fluorobenzonitrile (11 g, 90 mmol) in Et3N (14 mL)
and DMSO (60 mL) was added pentane-1,5-diamine (2.3 g,
23 mmol). The reaction mixture was stirred at 150 ꢂC for 3 h. Then
the reaction mixture was poured into ice-water mixture (400 mL)
and a large amount of pale-yellow solid was generated. The solid
was filtered and recrystallized with DMSO/H2O (8:1) to yield a pale
yellow solid (5.3 g, 75%). M.p. 141.1e142.7 ꢂC. 1H NMR (300 MHz,
8

B. Zhang, Y. Wang, C. Lin et al.
European Journal of Medicinal Chemistry 214 (2021) 113210
DMSO‑d6):
d
¼ 7.42 (d, J ¼ 8.7 Hz, 4H), 6.68 (t, J ¼ 5.1 Hz, 2H), 6.60
References
(d, J ¼ 8.8 Hz, 4H), 3.04 (dd, J ¼ 12.2, 6.4 Hz, 4H), 1.56 (dt, J ¼ 13.7,
7.0 Hz, 4H), 1.42 (dd, J ¼ 14.3, 7.8 Hz, 2H) ppm. 13C NMR (75 MHz,
[1] V. Azzarito, K. Long, N.S. Murphy, A.J. Wilson, Inhibition of alpha-helix-
mediated protein-protein interactions using designed molecules, Nat. Chem.
5 (2013) 161e173, https://doi.org/10.1038/Nchem.1568.
[2] M.P. Stumpf, T. Thorne, E. de Silva, R. Stewart, H.J. An, M. Lappe, C. Wiuf,
Estimating the size of the human interactome, Proc. Natl. Acad. Sci. U. S. A. 105
(2008) 6959e6964, https://doi.org/10.1073/pnas.0708078105.
[3] S. Surade, T.L. Blundell, Structural biology and drug discovery of difficult
targets: the limits of ligandability, Chem. Biol. 19 (2012) 42e50, https://
doi.org/10.1016/j.chembiol.2011.12.013.
[4] T.A. Stone, C.M. Deber, Therapeutic design of peptide modulators of protein-
protein interactions in membranes, Biochim. Biophys. Acta. Biomembr. 1859
(2017) 577e585, https://doi.org/10.1016/j.bbamem.2016.08.013.
[5] G.A. Caputo, E. London, Cumulative effects of amino acid substitutions and
hydrophobic mismatch upon the transmembrane stability and conformation
of hydrophobic alpha-helices, Biochemistry-US. 42 (2003) 3275e3285,
https://doi.org/10.1021/bi026697d.
[6] N.J. Gleason, V.V. Vostrikov, D.V. Greathouse, R.E. Koeppe, Buried lysine, but
not arginine, titrates and alters transmembrane helix tilt, Proc. Natl. Acad. Sci.
U. S. A. 110 (2013) 1692e1695, https://doi.org/10.1073/pnas.1215400110.
[7] V.V. Vostrikov, B.A. Hall, D.V. Greathouse, R.E. Koeppe II, M.S. Sansom, Changes
in transmembrane helix alignment by arginine residues revealed by solid-
state NMR experiments and coarse-grained MD simulations, J. Am. Chem.
Soc. 132 (2010) 5803e5811, https://doi.org/10.1021/ja100598e.
DMSO‑d6):
d
¼ 152.7, 152.7, 133.8, 133.8, 133.8, 133.8, 121.2, 121.2,
112.1, 112.1, 112.1, 112.1, 95.7, 95.7, 42.5, 42.5, 28.6, 28.6, 24.6 ppm.
HRMS (ESI-Orbitrap) m/z: [M þ H]þ calcd for C19H21N4þ 304.3970;
found 304.3975.
Dimethyl-4,4’-(pentane-1,5-diylbis(azanediyl))dibenzimidate-
dihydrochloride (12). To a solution of dry dioxane (50 mL) and dry
MeOH (20 mL) was bubbled fresh prepared HCl (gas) for 8 h at ice
bath. Then compound 11 (5.3 g) was added to the prepared solution
(50 mL) and slowly warmed to ambient temperature and stirred for
1 h, then another portion of mixed solution (20 mL) was added to
reaction. After stirring for 2 d, a large amount of precipitate was
generated. The reaction mixture was filtered and the filter cake was
washed with diethyl ether (100 mL), dried under reduced pressure
to achieve crude compound 12 (4.7 g), which was unstable and
used for the next step without further characterization and
purification.
4,4’-(pentane-1,5-diylbis(azanediyl))dibenzimidamide (5). To a
solution of dry EtOH (30 mL) was bubbled fresh prepared NH3 (gas)
for 8 h at ꢀ20 ꢂC. Then compound 12 (1.0 g) was added to the
saturated ethanolic ammonia solution (30 mL). The reaction
mixture was sealed and slowly warmed to ambient temperature.
After vigorously stirred for 3 d, a large amount of white solid was
generated. The reaction mixture was filtered and the filter cake was
dispersed to a solution of NaOH aqueous (30 mL, 1 M). After stirring
for 40 min, the solid was filtered and the filter cake was washed
with water (10 mL) and then dried under high vacuum to yield
compound 5 (0.63 g, 23% from compound 11) as a white solid. M.p.
161.6e163.2 ꢂC (lit [45]. 215 ꢂC in a tetrahydrochloride salt). 1H
[8] G.A. Caputo, E. London, Position and ionization state of Asp in the core of
membrane-inserted alpha helices control both the equilibrium between
transmembrane and nontransmembrane helix topography and trans-
membrane helix positioning, Biochemistry-US. 43 (2004) 8794e8806, https://
doi.org/10.1021/bi049696p.
[9] D.W. Sammond, C. Joce, R. Takeshita, S.E. McQuate, N. Ghosh, J.M. Martin,
H. Yin, Transmembrane peptides used to investigate the homo-oligomeric
interface and binding hotspot of latent membrane protein 1, Biopolymers
95 (2011) 772e784, https://doi.org/10.1002/bip.21672.
[10] B. Sugden, An intricate route to immortality, Cell. 57 (1989) 5e7, https://
doi.org/10.1016/0092-8674(89)90165-7.
[11] D. Wang, D. Liebowitz, E. Kieff, An EBV membrane protein expressed in
immortalized lymphocytes transforms established rodent cells, Cell. 43 (1985)
831e840, https://doi.org/10.1016/0092-8674(85)90256-9.
[12] K.M. Kaye, K.M. Izumi, E. Kieff, Epstein-barr-virus latent membrane protein-1
is essential for B-lymphocyte growth transformation, Proc. Natl. Acad. Sci. U. S.
A. 90 (1993) 9150e9154, https://doi.org/10.1073/pnas.90.19.9150.
[13] U. Dirmeier, B. Neuhierl, E. Kilger, G. Reisbach, M.L. Sandberg,
W. Hammerschmidt, Latent membrane protein 1 is critical for efficient growth
transformation of human B cells by Epstein-Barr virus, Cancer. Res. 63 (2003)
2982e2989.
NMR (300 MHz, DMSO‑d6):
J ¼ 8.5 Hz, 4H), 6.20 (s, br, 4H), 5.98 (m, 2H), 3.03 (m, 4H), 1.57 (m,
4H), 1.45 (m, 2H) ppm; 13C NMR (75 MHz, DMSO‑d6):
d
¼ 7.52 (d, J ¼ 8.4 Hz, 4H), 6.51 (d,
d
¼ 163.0,
163.0, 150.9, 150.9, 128.1, 128.1, 128.1, 128.1, 122.9, 122.9, 111.1, 111.1,
111.1, 111.1, 43.0, 43.0, 28.9, 28.9, 24.8 ppm; HRMS (ESI-Orbitrap) m/
z: [M þ H]þ calcd for C19H27Nþ6 339.2292; found 339.2294.
[14] J.I. Cohen, Epstein-Barr virus infection, N. Engl. J. Med. 343 (2000) 481e492,
https://doi.org/10.1056/NEJM200008173430707.
[15] K.J. Flavell, P.G. Murray, Hodgkin’s disease and the Epstein-Barr virus, Mol.
Pathol. 53 (2000) 262e269, https://doi.org/10.1136/mp.53.5.262.
[16] G. Henle, W. Henle, V. Diehl, Relation of Burkitt’s tumor-associated herpes-
ytpe virus to infectious mononucleosis, Proc. Natl. Acad. Sci. U. S. A. 59 (1968)
94e101, https://doi.org/10.1073/pnas.59.1.94.
Declaration of competing interest
[17] G. Brady, G.J. MacArthur, P.J. Farrell, Epstein-Barr virus and Burkitt lymphoma,
The authors declare that they have no known competing
financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
J.
Clin.
Pathol.
60
(2007)
1397e1402,
https://doi.org/10.1136/
jcp.2007.047977.
[18] A. Kieser, K.R. Sterz, The latent membrane protein 1 (LMP1), Curr. Top.
Microbiol. Immunol. 391 (2015) 119e149, https://doi.org/10.1007/978-3-
319-22834-1_4.
[19] E.F. Kot, Y. Wang, S.A. Goncharuk, B. Zhang, A.S. Arseniev, X. Wang,
K.S. Mineev, Oligomerization analysis as a tool to elucidate the mechanism of
EBV latent membrane protein 1 inhibition by pentamidine, Biochim. Biophys.
Acknowledgements
This work was supported by the National Natural Science
Foundation of China [21602216, 21877106, 21807098]; Chinese
Academy of Sciences (CAS) Pioneer Hundred Talents Program; the
Young Talents Program of Chinese Academy of Agricultural Sci-
ences; and the Royal Society-Newton Advanced Fellowship
(NA170152). Computing time was supported by the Network and
Computing Center, Changchun Institute of Applied Chemistry, CAS,
the China Science and Technology Cloud, the National Super-
computing Center in Shenzhen and the Beijing Super Cloud
Computing Center (BSCC).
Acta.
Biomembr.
1862
(2020),
183380,
https://doi.org/10.1016/
j.bbamem.2020.183380.
[20] X. Wang, J.P. Saludes, T.X. Zhao, A. Csakai, Z. Fiorini, S.A. Chavez, J. Li, G.I. Lee,
K. Varga, H. Yin, Targeting the lateral interactions of transmembrane domain 5
of Epstein-Barr virus latent membrane protein 1, Biochim. Biophys. Acta.
Biomembr.
1818
(2012)
2282e2289,
https://doi.org/10.1016/
j.bbamem.2012.05.013.
[21] X. Wang, Z. Fiorini, C. Smith, Y. Zhang, J. Li, L.R. Watkins, H. Yin, Repositioning
antimicrobial agent pentamidine as a disruptor of the lateral interactions of
transmembrane domain 5 of EBV latent membrane protein 1, PloS. One. 7
(2012), e47703, https://doi.org/10.1371/journal.pone.0047703.
[22] Y. Wang, Y. Peng, B. Zhang, X. Zhang, H. Li, A.J. Wilson, K.S. Mineev, X. Wang,
Targeting trimeric transmembrane domain 5 of oncogenic latent membrane
protein 1 using a computationally designed peptide, Chem. Sci. 10 (2019)
7584e7590, https://doi.org/10.1039/c9sc02474c.
[23] E. Lindner, D. Langosch, A ToxR-based dominant-negative system to investi-
gate heterotypic transmembrane domain interactions, Proteins. 65 (2006)
803e807, https://doi.org/10.1002/prot.21226.
Appendix. Supplementary data
Supplementary data associated with this article can be found in
the online version, at https://doi.org/10.1016/j.ejmech.2021.113210.
These data include MOL files and InChiKeys of the most important
compounds described in this article.
[24] W.P. Russ, D.M. Engelman, TOXCAT: a measure of transmembrane helix as-
sociation in a biological membrane, Proc. Natl. Acad. Sci. U. S. A. 96 (1999)
863e868, https://doi.org/10.1073/pnas.96.3.863.
€
€
[25] Schrodinger Release 2015-2: Maestro, Schrodinger, LLC, New York, NY, 2015.
9

B. Zhang, Y. Wang, C. Lin et al.
European Journal of Medicinal Chemistry 214 (2021) 113210
[26] O. Trott, A.J. Olson, Software news and update AutoDock Vina: improving the
speed and accuracy of docking with a new scoring function, efficient opti-
mization, and multithreading, J. Comput. Chem. 31 (2010) 455e461, https://
doi.org/10.1002/jcc.21334.
[27] S. Jo, T. Kim, V.G. Iyer, W. Im, CHARMM-GUI: a web-based graphical user
interface for CHARMM, J. Comput. Chem. 29 (2008) 1859e1865, https://
doi.org/10.1002/jcc.20945.
[28] R.B. Best, X. Zhu, J. Shim, P.E. Lopes, J. Mittal, M. Feig, A.D. Mackerell Jr.,
Optimization of the additive CHARMM all-atom protein force field targeting
improved sampling of the backbone phi, psi and side-chain chi(1) and chi(2)
weighted histogram analysis method for free-energy calculations on bio-
molecules. I. The method, J. Comput. Chem. 13 (1992) 1011e1021, https://
doi.org/10.1002/jcc.540130812.
[37] A. Grossfield, WHAM: the weighted histogram analysis method”, version
2.0.9. http://membrane.urmc.rochester.edu/wordpress/?page_id¼126.
[38] Y. Wang, R.K. Finol-Urdaneta, V.A. Ngo, R.J. French, S.Y. Noskov, Bases of
bacterial sodium channel selectivity among organic cations, Sci. Rep. 9 (2019),
15260, https://doi.org/10.1038/s41598-019-51605-y.
[39] F. Ingels, B. Beck, M. Oth, P. Augustijns, Effect of simulated intestinal fluid on
drug permeability estimation across Caco-2 monolayers, Int. J. Pharm. 274
(2004) 221e232, https://doi.org/10.1016/j.ijpharm.2004.01.014.
dihedral angles, J. Chem. Theory. Comput.
doi.org/10.1021/ct300400x.
8 (2012) 3257e3273, https://
[40] J.J. Vanden Eynde, A. Mayence, M. Mottamal, C.J. Bacchi, N. Yarlett, M. Kaiser,
R. Brun, T.L. Huang, Alkanediamide-Linked bisbenzamidines are promising
[29] J.B. Klauda, R.M. Venable, J.A. Freites, J.W. O’Connor, D.J. Tobias, C. Mondragon-
Ramirez, I. Vorobyov, A.D. MacKerell Jr., R.W. Pastor, Update of the CHARMM
all-atom additive force field for lipids: validation on six lipid types, J. Phys.
Chem. B. 114 (2010) 7830e7843, https://doi.org/10.1021/jp101759q.
[30] K. Vanommeslaeghe, E. Hatcher, C. Acharya, S. Kundu, S. Zhong, J. Shim,
E. Darian, O. Guvench, P. Lopes, I. Vorobyov, A.D. Mackerell Jr., CHARMM
general force field: a force field for drug-like molecules compatible with the
CHARMM all-atom additive biological force fields, J. Comput. Chem. 31 (2010)
671e690, https://doi.org/10.1002/jcc.21367.
[31] L. Huang, B. Roux, Automated force field parameterization for nonpolarizable
and polarizable atomic models based on ab initio target data, J. Chem. Theory.
Comput. 9 (2013) 3543e3556, https://doi.org/10.1021/ct4003477.
[32] J.C. Phillips, R. Braun, W. Wang, J. Gumbart, E. Tajkhorshid, E. Villa, C. Chipot,
R.D. Skeel, L. Kale, K. Schulten, Scalable molecular dynamics with NAMD,
J. Comput. Chem. 26 (2005) 1781e1802, https://doi.org/10.1002/jcc.20289.
[33] G.J. Martyna, D.J. Tobias, M.L. Klein, Constant-pressure molecular-dynamics
algorithms, J. Chem. Phys. 101 (1994) 4177e4189, https://doi.org/10.1063/
1.467468.
antiparasitic agents, Pharmaceuticals (Basel).
10.3390/ph9020020.
9 (2016), https://doi.org/
[41] T.L. Huang, J.J. Vanden Eynde, A. Mayence, M.S. Collins, M.T. Cushion,
D. Rattendi, I. Londono, L. Mazumder, C.J. Bacchi, N. Yarlett, Synthesis and SAR
of alkanediamide-linked bisbenzamidines with anti-trypanosomal and anti-
pneumocystis activity, Bioorg. Med. Chem. Lett. 19 (2009) 5884e5886,
https://doi.org/10.1016/j.bmcl.2009.08.073.
[42] D. Maciejewska, J. Zabinski, P. Kazmierczak, M. Rezler, B. Krassowska-Swie-
bocka, M.S. Collins, M.T. Cushion, Analogs of pentamidine as potential anti-
Pneumocystis chemotherapeutics, Eur. J. Med. Chem. 48 (2012) 164e173,
https://doi.org/10.1016/j.ejmech.2011.12.010.
[43] J.N. Ashley, H.J. Barber, A.J. Ewins, G. Newbery, A.D.H. Self, A chemotherapeutic
comparison of the trypanocidal action of some aromatic diamidines, J. Chem.
Soc. (1942) 103e116, https://doi.org/10.1039/jr9420000103.
[44] R.R. Tidwell, S.K. Jones, J.D. Geratz, K.A. Ohemeng, M. Cory, J.E. Hall, Analogs of
1,5-bis(4-Amidinophenoxy)Pentane (pentamidine) in the treatment of
experimental pneumocystis-carinii pneumonia, J. Med. Chem. 33 (1990)
1252e1257, https://doi.org/10.1021/jm00166a026.
[34] S.E. Feller, Y.H. Zhang, R.W. Pastor, B.R. Brooks, Constant-pressure molecular-
dynamics simulation - the Langevin piston method, J. Chem. Phys. 103 (1995)
4613e4621, https://doi.org/10.1063/1.470648.
[35] U. Essmann, L. Perera, M.L. Berkowitz, T. Darden, H. Lee, L.G. Pedersen,
A smooth particle mesh Ewald method, J. Chem. Phys. 103 (1995) 8577e8593,
https://doi.org/10.1063/1.470117.
[45] S.M. Bakunova, S.A. Bakunov, D.A. Patrick, E.V.K.S. Kumar, K.A. Ohemeng,
A.S. Bridges, T. Wenzler, T. Barszcz, S.K. Jones, K.A. Werbovetz, R. Brun,
R.R. Tidwell, Structure-activity study of pentamidine analogues as anti-
protozoal agents, J. Med. Chem. 52 (2009) 2016e2035, https://doi.org/
10.1021/jm801547t.
[36] S. Kumar, D. Bouzida, R.H. Swendsen, P.A. Kollman, J.M. Rosenberg, The
10