Structure-aided optimization of 3-O-β-chacotriosyl epiursolic acid derivatives as novel H5N1 virus entry inhibitors

Article abstract of DOI:10.1016/j.bmcl.2020.127518

It is urgent to develop new antiviral agents due to the continuous emergence of drug-resistant strains of influenza virus. Our earlier studies have identified that certain pentacyclic triterpene saponins with 3-O-β-chacotriosyl residue are novel H5N1 virus entry inhibitors. In the present study, a series of C-28 modified 3-O-β-chacotriosyl epiursolic acid derivatives via conjugation with different kinds of sides were synthesized, of which anti-H5N1 activities in A549 cells were evaluated in vitro. Among them, 10 exhibited strongest anti-H5N1 potency at the low-micromole level without cytotoxicity, surpassing the potency of ribavirin. Further mechanism studies of the lead compound 10 based on HI, SPR and molecular modeling revealed that these new 3-epiursolic acid saponins could bind tightly to the viral envelope HA protein, thus blocking the invasion of H5N1 viruses into host cells.

Full text of DOI:10.1016/j.bmcl.2020.127518

Bioorganic&MedicinalChemistryLetters30(2020)127518
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Structure-aided optimization of 3-O-β-chacotriosyl epiursolic acid
derivatives as novel H5N1 virus entry inhibitors
Sumei Lia,e, Xiuhua Jiac,e, Hui Lib, Yilu Yed, Xuesha Zhangb, Yongfeng Gaob, Guoqing Guoa,
⁎
⁎
Shuwen Liud, , Gaopeng Songb,
a Department of Human Anatomy, School of Medicine, Jinan University, Guangzhou 510632, China
b College of Materials and Energy, South China Agricultural University, Guangzhou 510642, China
c Department of Ophthalmology, The Third Affiliated Hospital, Sun Yat-sen University, Guangzhou 510630, China
d School of Pharmaceutical Sciences, Southern Medical University, Guangzhou 510515, China
A R T I C L E I N F O
A B S T R A C T
Keywords:
It is urgent to develop new antiviral agents due to the continuous emergence of drug-resistant strains of influenza
virus. Our earlier studies have identified that certain pentacyclic triterpene saponins with 3-O-β-chacotriosyl
residue are novel H5N1 virus entry inhibitors. In the present study, a series of C-28 modified 3-O-β-chacotriosyl
epiursolic acid derivatives via conjugation with different kinds of sides were synthesized, of which anti-H5N1
activities in A549 cells were evaluated in vitro. Among them, 10 exhibited strongest anti-H5N1 potency at the
low-micromole level without cytotoxicity, surpassing the potency of ribavirin. Further mechanism studies of the
lead compound 10 based on HI, SPR and molecular modeling revealed that these new 3-epiursolic acid saponins
could bind tightly to the viral envelope HA protein, thus blocking the invasion of H5N1 viruses into host cells.
3-Epiursolic acid saponins
Semi-synthesis
H5N1 entry inhibitors
Structure–activity relationships
Highly pathogenic H5N1 influenza A virus (IAV) has resulted in a
world-wide devastating effect on public health with a substantial
mortality and morbidity since 1997.1,2 However, there is no effective
vaccine to prevent H5N1 virus up to now. Currently, the treatment of
H5N1 influenza A infection only depends on a limited number of an-
tiviral drugs including amantadine, rimantadine, zanamivir, oseltamivir
and so on, which are able to inhibit two viral targets, namely M2 ion
channel and neuraminidase (NA).3,4 It is worth noting that prevention
of H5N1 influenza infection by antiviral drugs presents a formidable
challenge as the high variability within each segment of H5N1 virus
leads to the virus developing resistance to the above existing antiviral
drugs in the current clinical use.5–7 Therefore, there is an urgent need to
develop new antiviral agents with different mechanisms of action.
Hemagglutinin (HA) that is a mushroom-like glycoprotein on the
surface of IAVs, plays a key role in the viral lifecycle.8,9 It is well known
that HA can bind to sialylglyconjugates on the surface of the host cells,
leading to endocytosis of the virus and subsequently membrane fusion
of the virus and the host cells.10–12 Considering HA is essential for both
host cell recognition and membrane fusion, it can be regarded as a
potential target for antiviral therapy development. Encouragingly,
human monoclonal antibodies can target HA and show broad spectrum
antiviral activities against H1N1, H3N2, H5N1 and so on.13,14 Besides
antibodies, more and more IAVs entry inhibitors have been identified
and developed as new antiviral agents15–17, which are expected to have
enormous value for the treatment of H5N1 influenza infections.
Pentacyclic triterpenes characterized by hydrophobic pentacyclic
scaffolds, possess valuable inhibitory activities against human im-
munodeficiency virus (HIV), Ebola virus (EBOV), IAVs and hepatitis C
virus (HCV) infection.18–20 Further investigations indicate that these
pentacyclic triterpenes are able to block viral entry into the host cells to
inhibit viral infection.21–24 Particularly, a number of pentacyclic tri-
terpenes saponins with different saccharide chains have been obtained
as novel IAVs entry inhibitors25–27, which can inhibit IAVs entry into
cells by either blocking virus attachment to host cell receptor or virus-
cell membrane fusion. For example, Y3 (Fig. 1), an analog of oleanic
acid (OA), conjugated with acetylated galactose at C-17-COOH in OA,
exhibited remarkable anti-IAVs potency in vitro and in vivo by inter-
acting with HA.25 Rollinger reported that several new oleanane-type
triterpene saponins Rg-1 and Rg-2 (Fig. 1) comprising a branched or
linear trisaccharide moiety at C-3-OH, were identified as potent anti-
IAVs agents with IC50 values in the low nanomolar range.26
In our previous study, a new series of methyl 3-O-β-chacotriosyl
ursolic acid (1, Fig. 2) and its derivatives displayed excellent inhibitory
activity against H5N1 IAV entry based on an efficient HIV-based
⁎ Corresponding authors.
E-mail addresses: liusw@smu.edu.cn (S. Liu), songgp1021@scau.edu.cn (G. Song).
e These authors contributed equally to this work.
https://doi.org/10.1016/j.bmcl.2020.127518
Received 15 July 2020; Received in revised form 21 August 2020; Accepted 21 August 2020
Availableonline31August2020
0960-894X/©2020ElsevierLtd.Allrightsreserved.

S. Li, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127518
O
OH
O
O
COOH
AcO
HO
HO
OAc
AcO
COOH
HN
HO
O
OH
O
O
O
O
AcO
HO
HO
O
O
HO
O
HO
O
O
Rg-2
Rg-1
O
Y3
O
OH
OH
O
HO
HO
HO
HO
OH
HO
Fig. 1. Representative oleanane-type triterpene saponins Y3, Rg-1 and Rg-2 as IAV entry inhibitors.
SARs
4 R=H;
5 R=Et;
6 R=n-pentyl;
7 R=cyclopentyl;
8 R=Bn;
O
OR
SO
alteration of the
C-3 configuration
O
O
R1
OMe
enhance SI
SO
1 R1=OMe;
2 R1=NMe2;
SO
3
9 R1=H; R2=Me;
10 R1=R2=Me;
11R1=H; R2=cyclopentyl;
12 R1=H; R2=Ph;
13 R1=H; R2=Bn;
R1
N
bioisosteric
R2
HO
O
HO
HO
O
O
O
HO
SO
O
O
S=
replacement
HO
HO
OH
HO
Fig 2. Structures of the lead compounds 1–3 and title saponins 4–13.
pseudotyping system to screen a semisynthetic saponin library.27 On
the basis of SPR studies and docking predictions, 1 and its analogs were
claimed to bind tightly to HA, disrupting the interaction of HA with the
sialic acid receptor. Further structure–activity relationships (SARs) re-
vealed that not only the chacotriosyl residue but also the ursolic acid
(UA) aglycone moiety played a profound effect on anti-H5N1 activity,
though several modifications were tolerated at certain positions. 27 Our
prior studies on compound 1 suggested that the 17-position (COOCH3
group) was preferred for further replacement with its bioisoster di-
methylamide (2, Fig. 2), which was beneficial to improving the selec-
tive index. However, 2 still exhibited moderate cytotoxicity against
MDCK cells with the CC50 value of 47 µM, which restricted its phar-
maceutical use as a potential anti-H5N1 entry inhibitor. Moreover, both
inhibition against H5N1 virus and mechanisms of action of 2 are still
unclear despite increased selective index.
natural ursolic acid (UA) following the literature procedures28, of which
benzyl esterification furnished U228. The resulting glycosylation of U2
with the donor U3 proceeded to yield U4 in the presence of a strong
Lewis acid TMSOTf. This process took advantage of the participation of
the neighboring group in a stereoselective reaction at C-2 position in
Glu to produce the complete β-stereoselectivity. Then all the Bz groups
in Glu were deprotected with NaOMe in MeOH to give U5. Re-
gioselective protection of the C-6-OH and C-3-OH in U5 with pivaloyl
(Piv) group was performed by treatment with pivaloyl chloride in
pyridine at low temperature (−15 °C), giving rise to U6. Subsequent
coupling of U6 and the L-rhamnose donor U7 by means of a similar
TMSOTf-catalyzed glycosylation procedure afforded the trisaccharide
glycoside U8 with exclusive α-configuration at the new glycosidic
linkage. Then both piv and Ac groups in U8 were simultaneously de-
protected under the basic conditions to yield the title saponin 8. The
free C-17 carboxyl group in 4 was prepared by hydrogenolysis of Bn
group over 10% palladium/carbon. Subsequently, acetylation of 4 with
acetic anhydride was performed to generate the important intermediate
U9. Synthesis of the title saponins 5–7 and 9–13 was performed via an
acyl chlorination with oxalyl chloride and subsequent coupling reac-
tion, followed by removal of all the Ac groups using MeONa in MeOH.
The saponins 3–13 were tested for their inhibitory potency against
wild-type A/ Duck/Guangdong/212/2004 H5N1 viruses using the cy-
topathic effect (CPE) reduction assay in A549 cells while the MTT
screening was utilized to screen and exclude compounds with strong
toxicity toward A549 cells.9 As expected, little to no toxicity against
A549 cells was observed when 3β-OH of UA was converted to 3α-OH,
indicating the importance of the C-3 configuration of aglycone for the
selectivity index. Among these title saponins, the example of the effect
of compound 10 with strongest antiviral activity on CPE reduction was
Encouragingly, we unexpectedly found that alteration of the C-3
configuration of methyl ursolate from 3β-OH to 3α-OH leading to
methyl 3-O-β-chacotriosyl epiursolic acid (3, Fig. 2), could dramatically
reduce cytotoxicity against MDCK cells. In view of the above results, as
part of our persistent efforts toward the development of potential an-
tiviral entry inhibitors derived from natural products9–11,27, we de-
signed and synthesized a series of 3-O-β-chacotriosyl epiursolic acid
analogues 4–13 (Fig. 2) in cooperation with other chemical modifica-
tions at 17-COOH group, aiming to identify strongly potent H5N1 entry
inhibitors with high selective index. And then, the present study de-
termined whether the above 3-epiursolic acid saponins exert anti-H5N1
activity evaluated in cell-based assays and how they map out the po-
tential mechanism of action for inhibition against H5N1 IAV virus.
The synthesis of title saponins 4–13 was outlined in Scheme 1. The
important aglycone 3-epiursolic acid (U1) was obtained from the
2

S. Li, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127518
Scheme 1. Reagents and conditions: (a) BnCl,
K2CO3, DMF, 96%; (b) TMSOTf, CH2Cl2, 85%;
(c) MeONa, CH3OH, 93%; (d) PivCl, pyridine,
76%; (e) TMSOTf, CH2Cl2, 84% (f) NaOH,
CH3OH-THF-H2O, 93%; (g) 10% Pd/C, H2,
CH3OH-CH2Cl2, 84%; (h) Ac2O, DMAP, pyr-
idine, 87%; (h) (i) (COCl)2, CH2Cl2 (ii) R-Br,
K2CO3, DMF or R1R2NH, Et3N, CH2Cl2; (iii)
CH3ONa, CH3OH, 94% for 5, 90% for 6, 91%
for 7, 93% for 9, 85% for 10, 88% for 11, 90%
for 12, 90% for 13.
Fig 3. (A) Validation of the protection of A549 cells from influenza A/ Duck/Guangdong/212/2004 H5N1 by compound 10. (B) Dose-response curve for potential
compound 10 in the antiviral and cytotoxicity assay.
shown in Fig. 3A, demonstrating that these compounds could protect
Table 1
A549 cells from influenza virus H5N1-induced CPE compared with
Antiviral activity of compounds 3–13 against H5N1influenza A virus in A549
DMSO. In this manner, 10 displayed strong inhibition against H5N1 in
cells.
compound
A549 cells with an IC50 of 8.42 μM while exhibiting no cytotoxicity in
IC50 (µM)
CC50 (µM)
SI
vitro (Fig. 3B).
As shown in Table 1, the goal of this preliminary SARs was to in-
vestigate the significance of substitutions on the C-17-COOH group with
respect to both anti-H5N1 activity and selective index. To confirm the
advantage of substituents attached to the 17-COOH position of 3-
epiursolic acid, 4 was synthesized to represent the minimal scaffold
with no substituents attached. Saponins 5–8 were designed to replace
the relatively small CH3 moiety (3) for a much bulkier alkyl or aryl
substituent to probe the effect of both length and size of side chains on
antiviral activity. By observing the antiviral potential of 3–8, it was
noted that all the 3-epiursolic acid esters-chacotriose conjugates 3 and
5–8 exhibited substantially improved anti-H5N1 activity relative to 4,
suggesting that the introduced alkyl or aryl group might be directly
involved in anti-IAV activity. Moreover, our rationale was that the
lengthened side chains in C-28 position (5 and 6) could potentially
3
16.45
48.72
18.06
29.32
40.52
> 50
12.24
0.58
1.83
0.52
1.45
3.26
363.42
282.75
374.63
488.60
523.72
NT
2.61
1.88
3.47
2.36
3.85
22.1
4
5.8
5
20.7
6
16.7
7
12.9
8
NT
9
0.60
> 1000
> 1000
> 1000
> 1000
NT
> 81.7
> 118.8
> 54.1
> 29.6
NT
10
8.42
0.36
11
18.48
33.83
> 50
47.22
1.22
1.61
12
13
ribavirin
1.05
129.65
1.38
2.7
IC50: 50% inhibitory concentration.
CC50: 50% cellular cytotoxicity concentration.
SI: selectivity index as CC50/IC50
.
3

S. Li, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127518
extend the CH3 moiety deeper into the unexplored side cavity, gen-
erating the additional hydrophobic interactions to enhance ligand af-
finity. However, this increase in steric bulk resulted in similar or a little
loss of efficacy in spite of reduced cytotoxicity against A549 cells
compared to 3. Likewise, decoration of C-17-COOH with enlarged side
chains (7 and 8) to exploit potential hydrophobic interactions also led
to a little and even significant decreased antiviral activity. These results
suggest that larger substituents sterically clashed inside that portion of
the active site, of which introduction should be avoided. A similar trend
was found for the bioisosteric surrogates 9–13 of 3-epiursolic acid ester
since the methanamide product 9 was more active than other three
monoamide analogs 11–13. It was interesting to note that all the amide
analogs showed marginal cytotoxicity toward A549 cells. These results
demonstrate that amidation at the 17-COOH position was beneficial for
improving the safety of these saponins while keeping their antiviral
activity. With the failure of positively charged monoamide groups with
different large substituent to produce antiviral agents, we therefore
proceeded with exploration of small neutral N, N-disubstitution groups.
Thus, the impact of N, N-disubstitution was determined with 10, a
disubstituted amide structure from 9. Even more encouraging was that
10 had an increased both anti-H5N1 potency and selective index,
possibly owing to optimal size and hydrophobic nature of dimethyla-
mino group in C-28 position. These results highlight the significance of
the short disubstituted amide residue in potential ligand affinity. Be-
cause of its good compromise between potency and selective index, 10
was considered for further studies of mechanism of action in vitro,
which was chosen as the lead compound.
Fig. 5. Comparisons of the behaviors of 10 vs anti-HA antibody in inhibition of
H5N1virus-induced aggregation of chicken erythrocytes.
activity of N1-typed neuraminidase (Fig. 4C).
It is influenza virus HA that plays a crucial role during viral invasion
into host cells, which can bind to sialic acid on the surface of red blood
cells (RBCs), leading to agglutination. Thus, a hemagglutination in-
hibition (HI) assay was performed to investigate whether the designed
3-O-β-chacotriosyl epiursolic acid derivatives blocked the ability of
viral particles to target host cell receptors. It was observed that 10
could inhibit the binding of influenza virus H5N1 to RBCs in a con-
centration-dependent manner (Fig. 5) as anti-HA antibody do, in-
dicating that these title saponins may target HA and block the inter-
action of H5 HA with sialic acid receptor in host cells.
Next, we examined 10 could block H5N1 virus production most
effectively at the virus entry step of cell infection based on a HA/HIV
pseudotyped virus model while a VSVG/HIV pseudoviral was used as a
specificity control to exclude inhibitory effect on post-entry infection of
HIV backbone. As shown in Fig. 4A, 10 exhibited potential inhibitory
activity against H5N1 (A/Tailand/Kan353/2004) with an IC50 of
6.29
0.23 uM, but no cytotoxic activity against MDCK cells
To further investigate the interaction of the lead 10 with HA, we
then characterized the affinity between HA and 10 based on a surface
plasmon resonance (SPR) assay. A recombinant HA protein from virus
strain A/Vietnam/1203/2004 (H5N1) was used as representative HA of
(Fig. 4B). Furthermore, we speculated that 10 could target H5N1 HA
due to the fact that it not only had negligible effect on VSV-G enveloped
pseudovirus (Fig. 4A) but also exhibited no significant inhibition on the
Fig. 4. (A) Compound 10 could inhibit H5N1 (A/Tailand/Kan353/2004) dose-dependently and did not inhibit VSVG pseudovirus. (B) 10 showed no cytotoxic
activity against MDCK cells. (C) 10 did not inhibit neuraminidase activity.
4

S. Li, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127518
Fig. 8. Evaluation of the efficacy of 10 against influenza infection via intranasal
administration. Balb/C mice, five in each group, were inoculated with 5LD50 of
virus and intranasally administrated with 10 at dose 0, 40, 80 and 160 mg/Kg/
d one daily for 5 days, starting 1 day before infection. The survival of infected
mice was monitored daily for 14 days. Infected mice treated with ribavirin were
used as positive control.
Fig. 6. Characterization of the affinity between 10 and H5N1 HA protein using
sequence of A/ Duck/Guangdong/212/2004 H5N1 was identical to that
of A/Vietnam/1194/2004, of which the neutral-pH crystal structure of
the HA has been well defined recently.9–11 Hence, 10 was docked into
the influenza HA protein of H5N1 (Influenza strain A/Vietnam/1194/
2004, PDB: 3ZP0).29
surface plasmon resonance (SPR) analysis.
IAV H5N1 strains, due to the fact that the amino acid sequence identity
of the full-length HA proteins between A/Vietnam/1203/2004 strain
and A/Duck/Guangdong/99 (H5N1) strain used in above antiviral
assay was over 97.0%.11 As shown in Fig. 6, 10 could favorably interact
with HA in a dose-dependent manner, and the equilibrium dissociation
constant (KD) of the interaction was 1.64 nM, indicating there was a
strong affinity between 10 with H5N1 HA.
As shown in Fig. 7, the surflex-dock study indicated that 10 could
occupy the binding pocket for sialic acid receptor, which shared a si-
milar binding mode as other pentacyclic triterpenes analogs.9 Specifi-
cally, the methyl moiety of 17-CON(CH3)2 and the 3-epiursolic acid
scaffold formed multiple hydrophobic interactions with Leu133,
Lys156, Lys193 and Leu194. More importantly, the chacotriosyl residue
played a pivotal role in occupying sialyllactose core binding site. It is
Here, to add further insights into the inhibition mechanism, we
attempted to perform docking study of the selected compound 10 with
the Surflex-Dock program of SYBYL 7.3. In this simulation, the HA1
Fig. 7. Structural representative of compound 10
binding within hemagglutinin (Protein Data Bank:
3ZP0) according to surflex-dock calculation. (A)
Overview of hemagglutinin. The binding pocket of
sialic acid is highlighted in blue square. Protein and
10 were shown as white surface and yellow spheres,
respectively. (B) Closer view of the binding pocket.
Compound 10 was shown as yellow sticks. (C) 3D
interaction plot green dashes indicate hydrogen
bonds.
5

S. Li, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127518
clear that residues in 130 loop, 190 helix and 220 loop were important
for sialic acid binding. 25 It was found that C-6-OH of D-glucose formed
hydrogen bond with Lys193 within 190 helix. Moreover, the C-2-OH
and C-4-OH of the L-rhamnose moiety linked to C-2-OH of D-glucose
formed stable hydrogen bonds with Ser136 within 130 loop and Gln226
Acknowledgements
This project was supported by the National Nature Science
Foundation of China (31941019, 31872521); the Natural Science
Foundation of Guangdong Province, China (2018A030313414); Science
&Technology Project of Guangzhou, China (201804010457) and the
special Project for prevention and control of SARS-CoV-2 from
Educational Commission of Guangdong Province of China
(2020KZDZX1039).
linked to C-4-OH of D-glucose also exhibited great bLin-rdhianmgnaoffisenitmyowieitthy
within 220 loop, respectively. Interestingly, the
those critical residues in 220 loop, where the C-2-OH and C-4-OH of the
L-rhamnose moiety formed stable hydrogen bonds with Lys222 and
Gln226, respectively.
Finally, compound 10 was advanced to a mouse influenza A (Duck/
Guangdong/212/2004 H5N1) model. As shown in Fig. 8, there was
remarkable inhibition of the mortality for infected mice treated with 10
compared with the untreated mice. The protective effect of 10 on mice
survival rate, was improved in a dose-dependent manner, though less
effective than that provided by ribavirin. Taken together, these data
indicate that 10 can protect mice from influenza virus H5N1 infection
in vivo.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.bmcl.2020.127518.
References
1. Meng WY, Yang QQ, Vrancken B, et al. Emerg Microbes Infect. 2019;8:823.
2. Zhong GX, Fan SF, Lopes TJS, et al. Front Microbiol. 2019;10:1411.
3. Zhu ZB, Li RM, Xiao GK, et al. Eur J Med Chem. 2012;57:211.
4. Zhao X, Li RF, Zhou Y, et al. J Med Chem. 2018;61:5187.
In the present study, we reported the synthesis and characterization
of a series of 3-epiursolic acid saponins with 3-O-β-chacotriosyl residue
generated from semisynthesis, and tested their anti-influenza A/Duck/
Guangdong/99 (H5N1) virus activities in A549 cells. All the title pen-
tacyclic triterpene saponins 4–13 showed no cytotoxicity to A549 cells,
but showed certain antiviral activity with a potency ranging from
moderate to potent. The subsequent SARs studies indicated that 10
showed the highest anti-H5N1 activity in vitro with IC50 at micromole
level and significant antiviral potency in vivo. In particular, 10 not only
inhibited virus-induced hemagglutination but also exhibited a strong
affinity to HA with KD value of 1.64 nM. These results suggested that
the 3-epiursolic acid saponins might bind region of HA and interfere
with influenza invasion by its interaction with the sialic acid receptor.
Our study might establish the importance of 3-epiursolic acid deriva-
tives for development of entry inhibitors of influenza viruses.
5. Bloom JD, Gong LI, Baltimore D. Science. 2010;328:1272.
6. Moscona A. N Engl J Med. 2005;353:2633.
7. Schade D, Kotthaus J, Riebling L, et al. J Med Chem. 2014;57:759.
8. Ekiert DC, Friesen RH, Bhabha G, et al. Science. 2011;333:843.
9. Liao YX, Chen LZ, Li SM, et al. Bioorg Med Chem. 2019;27:4048.
10. Li H, Chen LZ, Li SM, et al. Bioorg Med Chem Lett, 2019; 29: 2675.
11. Ye MD, Liao YX, Wu L, et al. Viruses. 2020;12:225.
12. Loregian A, Mercorelli B, Nannetti G, Compagnin C, Palù G. Cell Mol Life Sci.
2014;71:3659.
13. Whittle JRR, Zhang R, Khurana S, et al. Proc Natl Acad Sci USA. 2011;108:14216.
14. Laursen NS, Friesen RHE, Zhu XY, et al. Science. 2018;362:598.
15. Kadam RU, Wilson IA. Proc Natl Acad Sci USA. 2017;114:206.
16. Song GP, Shen XT, Li SM, et al. Eur J Med Chem. 2016;119:109.
17. van Dongen MJP, Kadam RU, Juraszek J, et al. Science. 2019;363:eaar6221.
18. Grishko VV, Galaiko NV, Tolmacheva IA, et al. Eur J Med Chem. 2014;83:601.
19. Wang C, Lu L, Na HY, et al. J Med Chem. 2014;54:7342.
20. Chen YY, Wang XC, Zhu YB, et al. Mol Pharm. 2020;17:2564.
21. Si LL, Meng K, Tian ZY, et al. Sci Adv. 2018;4:eaau8408.
22. Li HW, Li M, Xu RY, et al. Eur J Med Chem. 2019;163:560–568.
23. Shi YY, Si LL, Han X, et al. Med Chem Commun. 2017;8:1531.
24. Xiao SL, Tian ZY, Wang YF, Si LL, Zhang LH, Zhou DM. Med Res Rev. 2018;38:951.
25. Yu MR, Si LL, Wang YF, et al. J Med Chem. 2014;57:10058.
26. Mair CE, Grienke U, Wilhelm A, et al. J Nat Prod. 2018;81:515.
27. Song GP, Shen XT, Li SM, et al. Eur J Med Chem. 2015;93:431.
28. Wen XA, Sun HB, Liu J, et al. J Med Chem. 2008;51:3540.
29. Li SM, Jia XH, Shen XT, et al. Bioorg Med Chem. 2017;25:4384.
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.
6