H. Cheng, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127302
introduced as R3, but only low to moderate inhibition potency was
obtained (entries 10–15). Among all the tested groups, the H atom
demonstrated the best activity and thus was identified as the optimal
group for R3 (entries 2–15 vs. entry 1). If other diphenyl ether structures
were employed, comparable or reduced activity was attained (entries
16–20 vs. entry 2). Interestingly, a pyridine (Y = N) instead of a phenyl
group (Y = CH) led to improved activity (IC50 = 1.21
0.10 μmol/L
0.11 μmol/L for entry 1). However,
for entry 21 vs. IC50 = 1.52
employment of both Ring A and Ring C with pyridine-derived groups
(X = Y = N) resulted in an inhibition ratio of only 14% (entry 22).
From the above results, X and Y were optimized as CH and N, respec-
tively. Furthermore, replacement of the 2,4-Cl2 group with 2-Cl-4-CF3
at the R1 position gave rise to slightly higher activity (entry 23 vs. entry
21). Unfortunately, poor levels of inhibition were produced when
substituents other than H were introduced into both Ring A and Ring B
(entries 24–28). Moreover, a pyrazine-containing target compound
(3c’) was also synthesized, but it merely exhibited moderate inhibitory
Fig. 1. The design strategy.
the amino group of Series A with an amide functionality (as shown in
Fig. 1, right column). The resulting target molecules, namely N-(4-
SCR as well as complex II and/or complex III. To our delight, N-(4-(2-
chloro-4-(trifluoromethyl)phenoxy)phenyl)picolinamide (3w), one of
these target compounds, displayed good activity against both SCR and
complex III, leading to an active inhibitor of complex III which has
never been reported before. Furthermore, computational simulations
provided more information about its possible binding mode.
activity against SCR (entry 29, IC50 = 8.72
0.16 μmol/L). Replacing
the electron-withdrawing group (CF3) of compound 3w with an elec-
tron-donating group (CH3) resulted in compound 3d’, which showed
slightly reduced potency than 3w (entry 30 vs. entry 23). Besides,
changing the Cl group with a bulkier Br group gave rise to compound
3e’, which also exhibited marginally lower activity than 3w (entry 31
vs. entry 23). It appeared that both electronic and steric properties af-
fected the inhibitory activity of the resulting target compounds against
SCR, with the Cl and CF3 groups as the optimal groups on the diphenyl
ether skeleton. Finally, compound 3w was identified as the most active
SCR inhibitor, which manifested comparable potency with the two
commercial fungicides (entry 23 vs. entries 32–33).
First of all, the designed target compounds were synthesized using
the traditional amide synthesis from activated carboxylic acids and
amines. In detail, the amide bond was constructed from aromatic car-
boxylic acids and diphenyl ether-containing anilines using 1-ethyl-3-(3-
dimethylaminopropyl) carbodiimide hydrochloride (EDCI) combined
with 1-hydroxybenzotriazole (HOBt) as the activating reagents (as de-
picted in Scheme 1). Using this procedure, total 31 target compounds
were prepared and characterized by NMR and HRMS. To better present
our experimental results, the three phenyl rings were assigned as Ring
A, Ring B and Ring C, respectively (Scheme 1).
Aiming to confirm the inhibiting target of 3w (complex II, complex
III or dual–target), we further tested its activity against SCR, complex II
and complex III (as listed in Table 2). As expected, penthiopyrad se-
monstrated outstanding inhibitory activity exclusively against complex
III (entries 2–3). Similar with azoxystrobin, 3w showed an inhibition
rate of only 31% against complex II at the concentration of 10 μmol/L,
but considerably higher activity against complex III (entry 1). There-
fore, compound 3w was identified as a potent inhibitor of complex III.
With the discovery of 3w as an active complex III inhibitor, its
possible binding mode was then elucidated using computational simu-
lations (shown in Fig. 2). Complex III is composed of two binding sites
(Qo site and Qi site), so it’s of great significance to dock 3w into both
sites. It was worth noting that the calculation of the binding free en-
ergies was based on the docking scores in AutoDock 4.2. The docking of
3w into the Qo site imparted the calculated binding energy of
−7.7 kcal/mol for the best binding conformation, while the binding
energy was calculated as −6.7 kcal/mol when it was docked into the Qi
site. These results indicated that 3w would presumably bind to the Qo
site, which inspired us to investigate the binding mode of 3w into the
Qo site of complex III (shown in Fig. 2A). It appeared that crucial non-
polar interactions were formed between this inhibitor and the target
enzyme (PDB ID: 1SQB). The pyridine ring of 3w could enter the hy-
drophobic pocket surrounded by residues P270, F274, Y131 and F128.
Additionally, it could also form π-π stacking interactions with residues
Y131 and F274. Meanwhile, hydrophobic interactions were observed
between the diaryl ether scaffold and residues L294, I298, L121, I146,
M124, F274. Notably, the Cl and CF3 groups on the diphenyl ether
skeleton favored the enhancement of the hydrophobic interactions. As
we know, amide-π stacking interaction is one of the most frequently
observed interactions between proteins and ligands, which can largely
affect the biological activity.47 The amide bond in 3w showed good
amide-π stacking interactions with residues F274 and Y278. In terms of
was then assayed at a concentration of 10 μmol/L, and those with the
inhibitory rates of above 50% were further selected to determine their
IC50 values (as listed in Table 1). Two commercial SCR-inhibiting fun-
trols. Initially, various aromatic groups were attempted while retaining
the optimized diphenyl ether fragment (2,4-Cl2 as R1, X = CH, R2 = H)
from our previous work (entries 1–15).33 If an unsubstituted phenyl
ring (R3 = H) was introduced as Ring C, a high inhibition rate and an
IC50 value of 1.52
0.11 μmol/L were obtained (entry 1). It seemed
that the substituent positions on Ring C had a significant impact on the
inhibitory activity (entries 2–7). A methyl group employed at different
positions triggered varied inhibition potency, with 3-Me as R3 ex-
hibiting higher performance than the 2-Me and 4-Me counterparts
(entry 3 vs. entries 2 and 4). If 2-Cl, 3-Cl or 4-Cl was introduced as R3, a
similar trend was observed (entries 5–7). Polysubstituted groups were
also employed as R3, but only moderate inhibition rates of 31–46%
were obtained (entries 8–9). Other 3-substituted groups were
Scheme 1. The synthesis of target compounds 3a–3e′.
2