Y. Yang, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127508
Table 2
When the oxadiazole ring in PI-1840 was replaced by meta- (2a) or
para-phenyl (2e), only a slight decrease in the inhibitory activity was
observed. Further substitution of the 3-pyridinyl at Ar3 with other
aromatic rings, the inhibitory activity diminished in various extent (2b-
2d and 2f). Compounds with meta- or para-phenyl introduced at Ar2
(2a-2h and 2k-2n) were generally less active than their analogs with an
oxadiazole ring at this area. The substitution of the oxadiazole ring in
PI-1840 with a pyridinyl (2i) was evidently unfavored, and the adverse
effect of a pyridinyl at Ar2 was further supported by the activity data of
compounds 2j, 2q and 2r. Interestingly, although 2e and 2g were
comparably active, the presence of a five-membered thienyl at Ar3
obviously made the compound more sensitive to the substitution of n-
butyl with phenoxyl and results in significant loss of activity (2f vs 2h),
which is consistent with data obtained for series 1.
Inhibitory activities of selected compounds against CT-L, T-L and C-L activities
of human 20S proteasome.
Compd
CT-L
T-L
C-L
1c
0.12
0.18
1.08
0.63
0.04
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
1d
0.05
0.05
0.15
2a
PI-1840
1
The IC50 values are calculated from two independent measures.
pyridinyl was varied (compounds 1a-1d). It looks like that the Ar3
moiety in area C is tolerable to structural variations, while heteroaro-
matic groups seem to be favored (1a vs 1b, 1c and 1d), especially those
with a hetero-atom vicinal to the oxadiazole ring (PI-1840 vs 1b, 1c and
1d). The n-butyl on the phenyl in area A of PI-1840 was then replaced
with phenoxyl group (1e), which caused a slight decrease in the in-
hibition against the CT-L activity. However, the effects of such a sub-
stitution on inhibitory activities appeared to vary with different Ar3
moieties. When Ar3 changed to furyl (1c), substitution of n-butyl with
phenoxyl (1f) resulted in a significant decline in the inhibitory activity.
When benzoheterocyclic rings were incorporated in area A (1 g-1l), the
inhibitory activities were generally weaker. Again, the discrepancy in
activity loss for compounds with different Ar3 was observed. The re-
placement of area A in PT-1840 with 2-benzofuryl only led to marginal
decrease in the inhibitory activity when five-membered furyl or thienyl
presented at Ar3 (1c, 1d vs 1i, 1j), while for compounds with naphthyl
or 2-quinolyl at Ar3, a dramatic drop in activity was detected (1a, 1b vs
1k, 1l). The different extent in activity loss for compounds with dif-
ferent Ar3 implicates the involvement of alternative binding poses,
which might be sensitive to the Ar3 substitution.
Compounds (1c, 1d and 2a) with the highest potency against CT-L
activity in each series were further evaluated for their effects on T-L and
C-L activities of human 20S proteasome. PI-1840 was again taken as a
reference compound. As illustrated in Table 2, all the four compounds
The constitutive proteasome (cCP) and the immunoproteasome
(iCP) are two different types of eukaryotic proteasomes. The former is
expressed in all eukaryotic cells, whereas the latter is mainly expressed
in cells of hematopoietic origin. PI-1840 was selective for β5c over
β5i,22 and consistently, compounds 1c and 1d showed no significant
inhibition against the β5i subunit (with percentage inhibition of 7.3%
and 25.3% at 10 μM, respectively).
Molecular docking of compound lc was performed to explore the
molecular basis for the selective inhibition against different 20S pro-
teasome subunits. As shown in Fig. 2, compound lc forms hydrogen
bonds with residues Thr1 and Thr21 in the active site of the β5c sub-
unit. Additional hydrophobic interactions with Ala20, Met45-Ala49,
Gly129 and Tyr169 of the β5c subunit were also observed (Fig. 2A). For
Fig. 2. The interaction modes of compound lc with (A) β5c, (B) β2c, (C) β1c and (D) β5i subunits as proposed by molecular docking. For cCP and iCP, the structures
of 5LF3 and 6E5B were applied respectively.
5