ACS Medicinal Chemistry Letters
Letter
physicochemical property. Several lyngbyastatin 7-based
analogues were designed by changing the length of the
terminal lipid chain, replacing the side chain amide of Gln with
other polar functional groups, and attaching hydrophilicity/
lipophilicity-modulating groups at the terminus of the pendant
side chain. Molecular docking using Autodock Vina31 was
applied to virtually evaluate the interaction between the
designed analogues and the HNE binding site (Figure S1),
using the cocrystal structure of lyngbyastatin 7 and porcine
pancreatic elastase (PPE)12 as a starting point and to validate
the docking approach. Several hits with improved affinity were
obtained, and we first selected two analogues to corroborate
our strategy to attach various side chain at late stage and to
generate diverse analogues in a few steps. In addition to
binding affinity of the designed molecules, we also took their
cLogD into account to probe the nature of the HNE binding
pocket. Moreover, if this molecule were to be developed for
inhalation, hydrophilicity/lipophilicity would also affect the
absorption rate and bioavailability.32 Therefore, analogues 17
and 18 characterized by a cyclohexyl and morpholine group,
respectively (Figures 3), were prioritized for synthesis due to
their differential molecular properties. The synthetic strategy to
generate these two molecules was similar to the previously
reported one for obtaining lyngbyastatin 7 (1) (see Supporting
The HNE inhibitory activities of 15, 16, 17, and 18 were
evaluated in parallel with lyngbyastatin 7 (1) at the enzyme
level (see Figure 3 for IC50 values and dose−response curves in
Figure 4A for the most active compounds), and some critical
information could be obtained from the structure−activity
relationship (SAR) analysis. The first important message is that
the side chain is actually essential to afford the elastase
inhibitory activity. As the side chain was shortened to an acetyl
group (compound 16), the activity decreased over 300-fold. It
is the first time that the minimal cyclodepsipeptide structure of
effective elastase inhibition was defined. Although 3-amino-6-
hydroxy-2-piperidone (Ahp) was the known pharmacophore
and the macrocyclic conformation was known to be essential,
the macrocycle alone (in acetylated form) was insufficient to
substantially inhibit the enzyme. At least a minimal side chain
was required to achieve recognition and binding toward HNE.
Second, the elastase binding pocket could not tolerate a bulky
functional group, especially at the conjunction site of the side
chain and the macrocycle. As the side chain was switched to a
Boc group (compound 15), the activity was completely lost;
according to the result derived from molecular docking, this
structure also displayed a decreased binding affinity to HNE
(−6.7 kcal/mol) compared with lyngbyastatin 7 (−7.2 kcal/
mol). When bulky groups were attached at the terminal site
(compounds 17 and 18), the potency only decreased slightly.
The next message is, the HNE binding pocket prefers a more
lipophilic terminal chain, as the analogue with higher
lipophilicity (compound 17) retained more activity than the
one with lower lipophilicity (compound 18). It was previously
known that the Abu unit adjacent to Ahp contributes to the
selectivity for elastase.6,11,12,33,34
Figure 4. In vitro biological evaluation of lyngbyastatin 7 (1),
analogue 17, and analogue 18. (A) In the HNE enzyme assay, HNE
was first incubated with lyngbyastatin 7 (1), analogue 17, or analogue
18, respectively, for 15 min in 0.1 M Tris−0.5 M NaCl (pH 7.5), and
then N-(OMe-succinyl)-Ala-Ala-Pro-Val-p-nitroanilide was used as
substrate to monitor the enzyme activity. The IC50 value for each
compound is 29 2, 103 6, and 197 15 nM, respectively. Data
are presented as mean
SD (n = 3), relative to 0.5% DMSO
treatment + vehicle. (B) In the cell viability assay, BEAS-2B cells were
cotreated with the tested compound and 100 nM HNE for 24 h. Cell
viability was monitored using MTT reagent. The EC50 values for
lyngbyastatin 7 (1), analogue 17, and analogue 18 were determined as
20 4, 91 17, and 185 33 nM, respectively. Data are presented
as mean SD (n = 3), relative to 0.5% DMSO treatment + vehicle.
(C) In order to monitor the changes in transcript level of IL1B in the
presence of different compounds at various concentrations, total RNA
was extracted after BEAS-2B cells were cotreated with lyngbyastatin 7
(1), analogue 17 and analogue 18 or solvent control and 100 nM
HNE or vehicle for 3 h. After cDNA synthesis, qPCR was carried out
while using GAPDH as endogenous control. Data are presented as
mean SD, *P < 0.05, ****P < 0.0001 compared to HNE-treated
cells using ANOVA, Dunnett’s t test (n = 3).
In addition, the potency of 17 and 18 was further assessed at
the cellular and transcriptional level side by side with
lyngbyastatin 7 (1) (Figures 4B and 4C). Compared with 1,
17 displayed similar potency in protecting bronchial epithelial
cells against elastase-induced antiproliferation, as measured by
MTT assay, and in upregulating transcript levels of pro-
inflammatory cytokine IL1B, as measured by reverse tran-
E
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX