Discovery and optimization of withangulatin A derivatives as novel glutaminase 1 inhibitors for the treatment of triple-negative breast cancer

Article abstract of DOI:10.1016/j.ejmech.2020.112980

To develop novel GLS1 inhibitors as effective therapeutic agents for triple-negative breast cancer (TNBC), 25 derivatives were synthesized from the natural inhibitor withangulatin A (IC₅₀ = 18.2 μM). Bioassay optimization identified a novel and

Full text of DOI:10.1016/j.ejmech.2020.112980

Journal Pre-proof
Discovery and optimization of withangulatin A derivatives as novel glutaminase 1
inhibitors for the treatment of triple-negative breast cancer
Wu-Xi Zhou, Chen Chen, Xiao-Qin Liu, Ying Li, Yao-Lan Lin, Xiu-Tao Wu, Ling-Yi
Kong, Jian-Guang Luo
PII:
S0223-5234(20)30952-1
DOI:
https://doi.org/10.1016/j.ejmech.2020.112980
EJMECH 112980
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 24 September 2020
Revised Date: 25 October 2020
Accepted Date: 27 October 2020
Please cite this article as: W.-X. Zhou, C. Chen, X.-Q. Liu, Y. Li, Y.-L. Lin, X.-T. Wu, L.-Y. Kong, J.-
G. Luo, Discovery and optimization of withangulatin A derivatives as novel glutaminase 1 inhibitors
for the treatment of triple-negative breast cancer, European Journal of Medicinal Chemistry, https://
doi.org/10.1016/j.ejmech.2020.112980.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
©
2020 Elsevier Masson SAS. All rights reserved.

Graphical abstract

Discovery and optimization of withangulatin A derivatives as
novel glutaminase 1 inhibitors for the treatment of triple-negative
breast cancer
a, #
a, #
a
a
a
a
Wu-Xi Zhou , Chen Chen , Xiao-Qin Liu , Ying Li , Yao-Lan Lin , Xiu-Tao Wu ,
a,
a,
Ling-Yi Kong *, Jian-Guang Luo *
a
State Key Laboratory of Natural Medicines, Jiangsu Key Laboratory of Bioactive Natural
Product Research, School of Traditional Chinese Pharmacy, China Pharmaceutical
University, 24 Tong Jia Xiang, Nanjing, 210009, PR China
#
Authors contributed equally
*
Corresponding author: lykong@cpu.edu.cn (L.-Y. Kong), luojg@cpu.edu.cn (J.-G. Luo)
Abstract
To develop novel GLS1 inhibitors as effective therapeutic agents for triple-negative breast
cancer (TNBC), 25 derivatives were synthesized from the natural inhibitor withangulatin A
(
(
IC = 18.2 μM). Bioassay optimization identified a novel and selective GLS1 inhibitor 7
50
IC = 1.08 μM). In MDA-MB-231 cells, 7 diminished cellular glutamate levels by blocking
5
0
glutaminolysis pathway, further triggering the generation of reactive oxygen species to induce
caspase-dependent apoptosis. Molecular docking indicated that 7 interacted with a new
reacting site of allosteric binding pocket by forming various interactions in GLS1. The
intraperitoneal administration of 7 at a dose of 50 mg/kg exhibited remarkable therapeutic
effects and no apparent toxicity in the MDA-MB-231 xenograft model, indicating its
potential as a novel GLS1 inhibitor for treatment of TNBC.
Keywords: Withangulatin A derivatives; GLS1 inhibitors; Molecular docking; Antitumor;
1

TNBC
1
. Introduction
Glutaminolysis is one of the main metabolic pathways in cancer cells to produce energy
to meet the demands of their rapid and sustained growth [1-3]. Glutamine is hydrolyzed by
glutaminase 1 (GLS1) into glutamate, which is further converted into α-ketoglutarate by
glutamate dehydrogenase. The α-ketoglutarate produced in this process can enter the
tricarboxylic acid cycle to supply cellular nutrients [4-6]. Moreover, glutaminolysis pathway
is essential for cellular reactive oxygen species (ROS) homeostasis [5]. Owing to the
addiction of many cancer cells to glutamine, blocking glutaminolysis was considered to be an
attractive strategy for cancer therapy [7-9].
GLS1, as a key metabolic enzyme, controls the first step in the glutaminolysis pathway.
Suppression of the expressed form of GLS1 has antitumor activity across many tumor types,
including glioma, pancreatic, lymphoma, breast and renal cancers [10-14]. Notably, GLS1
expression level is associated with high grade and metastatic breast cancer [15]. In particular,
TNBC cells have high levels of GLS1 as well as an increased uptake of glutamine to supply
the cells with intermediates to meet the TCA cycle [16-18]. Currently, the treatment of TNBC
usually relies on chemotherapy. Several studies have aimed at the development of novel
chemotherapies [19-22], while there is still a lack of satisfactory therapeutic agents in the
market. In recent years, TNBC cells are reported to be used for screening antitumor activity
of GLS1 inhibitors in vitro, of which the most commonly used is MDA-MB-231 [23]. These
observations highlight the critical connection between breast carcinoma and GLS1 and
suggest that GLS1 is a potential therapeutic target for anti-TNBC drug discovery and
development [8, 24-26]. Although the potential benefits of GLS1 inhibition have been
recognized for many years, small molecule GLS1 inhibitors were relatively scarce and
2

merely chemically synthesized derivatives, DON, BPTES and CB-839, were available to date
27, 28]. Nevertheless, some shortcomings, such as nonspecificity, low solubility and
[
moderate potency, may limit their pharmacological applications [29, 30].
Natural products are important sources for the discovery and development of novel
anticancer drugs [31]. Recently, we found that withangulatin A (WA, Fig. 1) with a
withanolide scaffold isolated from Physalin species exhibited potent anticancer activities
[
32-35]. In several cancer models, WA has been proven to inhibit tumor metastasis, reduce
cell proliferation, and induce cell apoptosis, which makes WA a promising drug for cancer
therapy [36, 37]. More interestingly, WA was found to have a moderate GLS1 inhibitory
activity (IC = 18.2 μM). Structure-based optimizations based on WA lead to the discovery
5
0
of a potent GLS1 inhibitor 7 with IC = 1.08 μM and favorable physicochemical properties.
5
0
In particular, this compound exhibited a different binding site from traditional GLS1
inhibitors. Intraperitoneal administration of 7 exhibited remarkable therapeutic effects and no
apparent toxicity in the MDA-MB-231 xenograft model, indicating its potential as a novel
GLS1 inhibitor for treatment of TNBC.
Fig. 1. Structure of withangulatin A (WA)
2
. Results and discussion
2
.1. Design of GLS1 inhibitors based on WA
WA was found to have moderate GLS1 inhibitory activity (IC = 18.2 μM). Further
5
0
3

efforts focus on structural modifications, and optimization of this scaffold are still needed
today. In the search for novel antitumor GLS1 inhibitors with higher potency, we designed a
series of WA derivatives in which a 4-hydroxyl group was modified with different types of
substituents derived from small molecule GLS1 inhibitors via different linkers (Fig. 2).
BPTES, as the first reported small molecule GLS1 allosteric inhibitor, showed potent
GLS1 inhibitory activity (IC = 3.3 µM) [38-41]. The structure of BPTES (Fig. 2) consists
5
0
of two exactly equivalent parts including thiadiazole, amide, and phenyl groups. This
structural characteristic allows it to easily interact with GLS1 allosteric pocket and effectively
trap GLS1 as an inactive tetramer [42]. On the basis of the scaffold of BPTES, some
derivatives were synthesized for the purposed of improving the GLS1 inhibitory activity and
structure-activity relationship (SAR) data were presented [43, 44] The amide groups play an
important role in binding, while the terminal thiadiazole can be replaced by a variety of
amide groups while retaining the inhibitory activity of GLS1. In addition, the phenyl groups
can be modified without losing the GLS1 inhibitory activity [44]. Thus, substituents
containing these binding groups or their analogues were first introduced to WA to improve
the enzymatic inhibition activities and cell inhibition activities.
WA showed the presence of two hydroxyl and an α, β-unsaturated ketone functions (Fig.
1
). For WA analogs, the OH group at the C-4 position was utilized to incorporate the ester
linkage, since the reactivity of the hydroxyl group at C-4 is much higher than that at C-14
45]. Previous studies found that the enone moiety of WA is essential for its anticancer
[
activity and that the acetylation of 4-hydroxyl could lead to improvement of activity [46].
However, the introduction of large steric hindered groups into 4-hydroxyl of WA resulted in a
loss of the potency. Thus, different linkers with varying structure and length were evaluated
as survey of the potential chemical space. In addition, these linkers contain two carbonyl
4

groups connected to the substituents via an amide linkage, which was expected to be
susceptible to binding with GLS1 allosteric pocket. Therefore, in the present study, different
types of substituents derived from BPTES were introduced to the 4-hydroxyl of WA by
different linkers leading to the discovery of a novel molecule 7 with enhanced GLS1
inhibitory activity as the candidate for TNBC therapy.
Fig. 2. Design of GLS1 inhibitors based on WA and BPTES binding groups.
2
.2. Chemistry
The synthetic routes to the intermediates a-d and the WA derivatives 1-25 are shown in
Schemes 1 and 2 and detailed in the experimental section. Initially, the 4-hydroxyl group in
WA was acylated with glutaric anhydride, succinic anhydride, maleic anhydride or phthalic
anhydride in dry DCM under reflux conditions to give intermediates a, b or c, respectively
(Scheme 1). Besides, 4-(2-fluoro-4-nitrophenyl)morpholine was prepared with
3
,4-difluoronitrobenzene and morpholine in DMSO, followed by treatment with 10% Pd-C in
the presence of hydrogen in ethyl acetate to produce intermediate d (Scheme 2).
Subsequently, the appropriate substituents were conjugated with a, b or c in the presence of
catalyst HOBt•H O, EDCI•HCl and pyridine in DMF at room temperature to obtain the target
2
5

compounds 1-21 (Scheme 1). To validate that the enone moiety of WA is essential for
activity, compounds 22-25 were identified by Michael addition (Scheme 1).
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
H
H
H
H
A, B, or C
(i)
R2
(ii)
H
H
OH
H
OH
H
OH
OAc
H
OH
OAc
OAc
OAc
R2
R1
O
O
O
O
HO
R1
R2
R1
R2
OH
O
O
O
O
O
O
22-25
WA
a-c
1-21
O
O
O
O
A
B
R1
O
O
O
O
N
S
H
N
S
N
S
H
N
N
H
N
R2
N
H
N
H
O
C
F
N
N
F
H
H
N
O
N
N
O
Scheme 1. Synthesis of WA derivatives 1-25. Reagents and conditions: (i) TEA, DCM,
reflux, 40-72 h. (ii) HOBt•H O, EDCI•HCl, pyridine, DMF, rt, 15-48 h.
2
Scheme 2. Synthesis of intermediate d. Reagents and conditions: (i) Morpholine, KH PO ,
2
4
DMSO, 60 °C, 24 h. (ii) 10% Pd-C/H , EtOAc, rt, 6 h.
2
2
.3. Anti-proliferative activity in vitro
MDA-MB-231 cells, a TNBC cells, have high expression of GLS1 [16-18]. These cells
are reported to be used for screening antitumor activity of GLS1 inhibitors in vitro [23].
Therefore, we primarily began our research by screening the inhibitory activities of WA
derivatives in MDA-MB-231. The cytotoxicity of WA derivatives against MDA-MB-231 cell
line was tested, and the IC values and the maximum inhibition ratio were summarized in
5
0
6

Table 1. Obviously, the current data suggest that the hydroxyl substitutions at the C-4 atom
are crucial; when hydroxyl at the C-4 atom is substituted (compounds 1-21), most derivatives
demonstrate better anti-proliferative activity than that of WA. Meanwhile, the derivatives,
with addition at the C-3 (compounds 22-25), revealed no anti-proliferative activity, which
demonstrated that enone moiety of WA was essential for inhibitory activity [46]. As shown in
Table 1, compounds 1, 2, 8, 9, 15 and 16 containing either a thiadiazole ring or a thiazole
ring did not improve activity, while the introduction of benzothiazole displayed the highest
inhibition (IC = 0.67±0.03 µM) for compound 3. In addition, the heterocyclic ring was
5
0
replaced by a phenyl ring that could be useful to improve the inhibitory activity (compounds
, 11 and 18). Therefore, the phenyl ring of the substituents was maintained, and WA
4
derivatives with phenyl ring were prioritized. Subsequently, the introduction of an N,
N-dimethylaniline in compounds 5, 12 and 19 improved inhibitory activity. This was also
observed with the electron-donating methoxy group in compounds 6, 13 and 20. Interestingly,
morpholine saturated ring substituent derivatives (7, 14 and 21) in which phenyl ring,
nitrogen and oxygen were included exhibited potent anti-proliferative activities.
To further validate our design strategy of improving inhibition activities, three linkers
were selected. When the chain length of the linker region was five carbon atoms, compound 7
exhibited strongest inhibitory activity. However, compounds 14 and 21, which contained four
carbon atoms or an olefinic bond in linker region, reduced inhibitory activities. These results
suggested that the flexibility of the linker determining the spatial orientation of substituent
portion was important for the inhibitory activities. Moreover, these linkers contain two
carbonyl groups connected to the substituents via an amide linkage, which might provide a
binding group for protein-ligand interactions, but the intermediates with a carboxyl
(compounds a-c) revealed no inhibitory activity. Based on that, we identified a derivative 7
with a substituent 3-fluoro-4-morpholinoaniline which dramatically improved the potency
7

(
IC = 0.32 ± 0.003 µM). Almost complete inhibition was achieved by 7 in a biochemical
50
assay compared with WA and BPTES (Table 1), which indicated good efficacy in a
cell-based assay.
2
.4. Inhibition of GLS1 activity
Based on potent inhibition effects in MDA-MB-231 cells, the inhibition effects on GLS1
were necessary to be evaluated to further characterize our products. The inhibition effects of
WA derivatives, WA and BPTES on the GLS1 activity was investigated using the GLS1
inhibitor screening assay kit. The GLS1 inhibitor screening assay kit is designed to measure
the hydrolase activity of GLS1 for screening and profiling applications. All the synthesized
derivatives were initially tested at a concentration of 10 μM for their potency to inhibit GLS1
(Table 1). Among the active compounds showing greater than 50% inhibition, the one with
the highest inhibition rate was selected for IC50 measurement under these screening
conditions and compared with WA and BPTES. The results are shown in Table 1. Excitingly,
all of the three compounds showed increased inhibitory activities on GLS1, especially 7 with
an IC of 1.08 μM with improved efficacy of about 20-fold than that of WA (IC , 18.2 μM).
5
0
50
It was more potent compared with BPTES with IC of 4.57 μM (Table 1). To date, 7 is the
50
best GLS1 inhibitor in our studies, which validated our structure modification strategy.
Taken together, when the linker is glutaric acid, substituent is a phenyl ring with fluorine,
nitrogen and oxygen, and hydroxyl substitutions at the C-4 atom of WA, the enzyme
inhibitory activity and cytotoxic activity are improved. The most potent compound 7
exhibited approximately 20 times stronger cytotoxicity and 20 times higher inhibitory activity
of GLS1 than those of WA, respectively. Since topological polar surface area (tPSA) and
LogP values have potential impact on the ability of cell penetration, we calculated those
values of 7. According to the calculated values, compound 7 had better LogP and other
8

physicochemical properties than WA itself (Table S1), and deserved further biological
activity assays and pharmacological mechanistic study as a new antitumor candidate agent.
Table 1. Inhibition of GLS1 enzyme and MDA-MB-231 cells by WA derivatives.
GLS1
Inhibition
MDA-MB-231
Comp
d
R1
R2
Max-IN
IC50
(μM)
IC50
(μM)
a
b
b
H
(%)
c
(%)
-
CH CH CH
2
2
2
a
b
c
-OH
-OH
-OH
<50
<50
<50
<50
<50
<50
<50
-
-
-
-
-
-
-
-
-
>50
>50
>50
-
-
-
-
-CH
CH -
2
2
-CH=CH-
-
-
-
-
-
-
CH CH CH
2
2
2
2
2
2
2
2
1
2
3
4
5
6
2.17±0.07 71.6±1.5
1.00±0.06 84.8±2.4
0.67±0.03 94.1±0.6
0.49±0.02 95.4±0.3
1.07±0.05 92.6±1.9
-
CH CH CH
2
2
-
CH CH CH
2
2
-
CH CH CH
2
2
-
CH CH CH
51.76±0.8
8
2
2
-
CH CH CH
56.30±2.7
8
0.44±0.00
92.9±0.5
6
2
2
-
9

-
CH CH CH
87.20±5.0
6
0.32±0.00
3
2
2
2
7
8
9
1.08±0.02
95.2±0.4
-
-CH
-CH
-CH
-CH
-CH
-CH
CH
CH
CH
CH
CH
CH
-
-
-
-
-
-
<50
<50
<50
<50
<50
<50
-
-
-
-
-
-
8.07±0.35 77.5±1.4
2.06±0.09 79.6±2.5
3.50±0.16 79.5±1.2
1.87±0.11 80.4±3.4
1.62±0.09 83.3±1.2
1.02±0.04 82.9±3.8
2
2
2
2
2
2
2
2
2
2
2
2
1
0
1
2
3
1
1
1
5
3.45±3.3
8
1
4
-CH
2
CH
2
-
-
1.66±0.07 80.9±0.9
15
16
17
18
19
20
-CH=CH-
-CH=CH-
-CH=CH-
-CH=CH-
-CH=CH-
-CH=CH-
<50
<50
<50
<50
<50
-
-
-
-
-
-
8.45±0.45 65.4±2.5
1.83±0.09 88.4±1.1
>50
0
1.32±0.03 88.9±1.6
1.56±0.05 72.1±2.6
5
1.58±1.8
9
0.79±0.00
94.1±0.5
9
5
2.44±3.8
9
1.38±0.00
90.8±1.2
5
2
1
-CH=CH-
-
-
CH CH CH
2
2
2
2
2
3
<50
<50
-
-
>50
>50
0
0
-
-CH CH CH
2
2
2
2
-
1
0

2
4
5
-CH
2
CH
2
-
<50
<50
-
-
>50
>50
0
0
2
-CH=CH-
4
1.40±0.9 18.20±0.5
WA
6.34±0.17 74.7±1.6
9.08±0.27 68.5±1.3
9
6
BPTE
52.35±1.6
2
d
4.57±0.09
S
a
b
In vitro inhibition of compound concentration of 10 μM in GLS1. The concentration of an inhibitor
c
where the response is reduced by half. In vitro maximal inhibition in MDA-MB-231 cell lines. All values
d
are the means ± SD of four independent experiments. Positive control.
2
.5. Cell viability and clonogenicity
To investigate the ability of 7 to induce cell death, MDA-MB-231 cells were cultured
with different concentrations of 7, and the cell viability was measured. The results are shown
in Fig. 3A. The results suggest that 7 significantly decreased the viability of MDA-MB-231
cells in a dose-dependent manner, which is more potent than that of WA. We also examined
the long-term effect on the proliferation of MDA-MB-231 cells under treatment of 7 by
colony formation assay (Fig. 3B and C). Consistently, the proliferation of MDA-MB-231
cells was notably inhibited by 7, and compared with WA, it is more evident.
11

Fig. 3. Compound 7 inhibits the proliferation of MDA-MB-231 cells. (A) MDA-MB-231 cells were treated
with the different concentrations of 7 or WA for 24 hours and cell viability was detected by MTT assay. (B)
Compound 7 inhibited colony formation of MDA-MB-231 cells. MDA-MB-231 cells were treated with the
different concentration of 7 or WA for 24 hours. After treatment, the cells were seeded into 6-well plates
for two weeks. (C) The clonogenicity inhibition was presented as percentage relative to the vehicle
treatment. Data represent the means ± SD (n = 4), *P < 0.05, **P < 0.01, ***P < 0.001, compared with
control group.
2
.6. Compound 7 induced apoptosis in MDA-MB-231 cells
To validate the decreased viability of 7-treated cells is attributed to apoptosis,
MDA-MB-231 cells were treated with 7 and apoptotic changes were tested. Annexin V/PI
double-stained assay measured with flow cytometry showed that 7 significantly elevated the
total number of Annexin V/ PI double-stained (early apoptosis and late apoptosis) cells (Fig.
4
A and B). Furthermore, western blotting analysis suggested that 7 caused a dose-dependent
increase in the expression of cleaved caspase-3 and cleaved caspase-9 (Fig. 4C and D),
which are widely used apoptotic marker [47]. These data indicated that 7 induced apoptosis
1
2

and that it was via a caspase-dependent pathway in GLS1 high-expressing MDA-MB-231
cell line.
Fig. 4. Compound 7 induces apoptosis of MDA-MB-231 cells. (A) MDA-MB-231 cells were treated with
different concentrations of 7 for 12 h, and then fixed and stained with Annexin V-FITC/PI and analyzed by
flow cytometry; Acquisition of Annexin V-FITC and PI data are presented as a percentage (%) in each
quadrant. The early and late apoptotic cells are presented in the lower and upper right quadrant,
respectively. (B) And the apoptosis rate was quantitated. Data represent the means ± SD (n = 4), ***P <
0
.001, compared with control group. (C) Western blotting analysis of 7 in MDA-MB-231 cells.
MDA-MB-231 cells were treated with the different concentrations of 7 for 12 h; then the protein
expression of caspase-3, cleaved caspase-3, caspase-9, and cleaved caspase-9 were evaluated by western
blotting; GAPDH was used as an internal control. (D) Quantitative analysis. Data represent the means ±
SD (n = 4), ***P < 0.001, compared with control group.
2
.7. Compound 7 inhibited cellular glutamate production
Cellular Glutamate was detected using Amplex® Red Glutamic Acid Assay Kit
(
Invitrogen). Experiments are carried out following the manufacturer's instructions. In the
assay, compound 7 was found to diminish glutamate production in a dose-dependent manner
1
3

in MDA-MB-231 cells (Fig. 5). In addition, compared with WA, 7 was more potent in
reducing cellular glutamate production. These results demonstrated that 7 could effectively
inhibit GLS1 and block the glutamine hydrolysis pathway, which could lead to cell death.
Fig. 5. Effect of compound 7 and WA on the cellular glutamate levels of MDA-MB-231 cells.
2
.8. Compound 7 induced reactive oxygen species (ROS) generation
Cell signals mediated by reactive oxygen species (ROS) can promote tumor development
at a certain physiological level [48], but when the level is too high, reactive oxygen can cause
great damage to macromolecules in cells [49]. However, tumor cells can directly control ROS
levels through molecules produced by the glutamine metabolism pathway, preventing high
levels of ROS from causing chromosomal instability [50]. Hence, glutamine metabolism is
vital for cellular ROS homeostasis. The ROS level was tested using a fluorogenic dye
(DCFH-DA) by fluorescence microscopy and flow cytometry. As shown in Fig. 6A and B, 7
could significantly enhance intracellular ROS levels in a dose-dependent manner in
MDA-MB-231 cells, which was more potent than WA. In particular, as shown in Fig. 6C, it
was clear that significant fluorescence intensity of MDA-MB-231 cells confirmed that 7 was
able to achieve superior properties on cellular internalization with the highest level of
fluorescence intensity detected at 640 nM in comparison with that of WA. These results
1
4

proved that 7 as a GLS1 inhibitor could effectively block the glutamine hydrolysis pathway,
further triggering the generation of ROS.
Fig. 6. ROS activity. 10000/well A549 cells treated or not treated (Ctrl) with different concentration of 7,
or WA (0, 160, 320, or 640 nM) for 3 h. (A) Fluorescence microscopic images of intracellular ROS
production by DCFH-DA staining (green) in MDA-MB-231 cells (B) Quantitative analysis of ROS
generation by the flow cytometry. (C) At a concentration of 640 nM, 7 showed improved ROS activity in
comparison with control group and WA.
2
.9. Molecular docking studies and binding pattern analysis
The structure of GLS1 has been determined, revealing the presence of four units in the
asymmetric tetramer, which has been shown to be necessary for the catalytic activity of GLS1
1
5

[
51]. In the crystal structure of GLS1, there were two main different binding pockets, a
substrate binding pocket and an allosteric binding pocket for ligand to occupy inhibiting the
enzyme activity of GLS1.
In order to investigate the possible binding pattern of compound 7 with GLS1, the
docking study was conducted based on the GLS1 crystal structure (PDB code 3UO9).
According to binding model, 7 was found to interact with a different reacting site of the
allosteric pocket in GLS1 (Fig. 7A and B). The acetoxyl in inhibitor 7 could form hydrogen
bonding contact with the protein backbone amide groups of LYS-320. The two carbonyl
groups of the linker made hydrogen bonds with ARG-317 and another LYS-320. These
H-bond interactions allowed this linker to protrude into the narrow active-site channel formed
by the tetramer of GLS1, thus orientating the 4-morpholinoaniline moiety into another
binding site region. Then the morpholine ring of 7 linked to phenyl ring bound to the inner
pocket and made a hydrogen bond with the residue of LYS-320. The phenyl ring formed
strong π- π stacked interactions with the residues of ARG-317. All interactions helped
stabilize the conformation and combination of the ligand lying in the new site of the allosteric
pocket of GLS1. Furthermore, different from the docking result of WA which only made
hydrogen bonds with LYS-320 (Fig. 7C and D), 7 had a stronger binding capacity. Moreover,
7
took a different binding mode from that of BPTES and its analogs of which the thiadiazole
group and the aliphatic linker occupy the allosteric pocket by forming hydrogen bonds with
Lys320, Leu323 and Leu321 [40]. Collectively, 7 exploited the GLS1 allosteric binding site
via three distinct parts of the molecule including WA, linker, and the substituent.
1
6

Fig. 6. Docking studies results of 7 in GLS1 (PDB code 3UO9). (A) 7 shown with the protein surface
representation of GLS1; (B) Key interactions of 7 with GLS1; (C) WA shown with the protein surface
representation of GLS1; (D) Key interactions of WA with GLS1. 7 and WA are rendered as a stick and
colored by atom type, with yellow carbons, blue nitrogens, and red oxygens. The key residue atoms in
GLS1 which interact with compound 7 are shown as sticks and colored by atom type, with light blue
carbon, red oxygen, and blue nitrogen, and their surface is colored according to the coulombic potential.
The yellow dashed lines represent the π- π interactions. The red dashed lines represent the hydrogen bonds.
2
.10. Compound 7 suppressed tumor growth in vivo
To investigate the antitumor activity of compound 7 in vivo, we established a xenograft
model by subcutaneously injecting MDA-MB-231 cells in nude mice. After four weeks of
treatment by 7 (i.p., 25 and 50 mg/kg per 3 days) and BPTES (i.p., 25 mg/kg per 3 days), the
mice were euthanized, and their organs were harvested for histopathologic analysis. As
1
7

shown in Fig. 8A, the growths of tumors were effectively inhibited in the 7-treated groups
compared with the vehicle group. Tumor volume and tumor weight were observably reduced
after treatment with 7 (Fig. 8B and C). Moreover, compound 7 had no significant effect on
the body weight of mice or obvious signs of adverse effects during the experimental
treatment period (Fig. 8D). To investigate the effect of compound 7 on organ damage, the
H&E staining of heart, liver, spleen, lung, and kidney tissues were analyzed. The
histopathological morphology changes of organs were not detected in compound 7-treated
tumor-bearing mouse group, which showed that there was no obvious toxicity under the
current treatment paradigm (Fig. S31). These results clearly indicated that compound 7
effectively inhibited the tumor proliferation in a preclinical mouse model without causing
apparent toxicity.
Fig. 7. Compound 7 inhibited TNBC growth in vivo. (A) Images of MDA-MB-231 tumor-bearing mice
1
8

after treated with 7 or BPTES. Changes of tumor volume (B), tumor weight (C), and body weight (D) were
shown. Data represent the means ± SD (n = 6), *P < 0.05, ***P < 0.001, compared with control group.
Scale bar = 5 mm.
3
. Conclusion
A series of WA derivatives in which a 4-hydroxyl group was modified with different
types of substituents derived from BPTES via different linkers were designed, synthesized,
and evaluated as potential GLS1 inhibitors. These WA-based core compounds showed strong
antiproliferative activity against MDA-MB-231 cells bearing overexpressed GLS1. Among
them, 7 exhibited stronger inhibition against GLS1 compared with WA and BPTES in
enzyme assay. Compound 7 could diminish cellular glutamate levels by blocking
glutaminolysis pathway, further triggering the generation of reactive oxygen species, thus
inducing caspase-dependent apoptosis. According to binding model, 7 exhibited a different
binding pattern from WA and BPTES, which exploited the GLS1 allosteric binding site via
three characteristic parts of the molecule including WA, linker, and the substituent. Moreover,
compound 7 effectively inhibited the MDA-MB-231 xenograft tumor growth without
apparent toxicity. As a result, 7 showed promising application as a GLS1 inhibitor from a
natural product for the treatment of TNBC that deserves further development.
4
4
4
. Experimental
.1. Chemistry
.1.1. General methods
All solvents were reagent grade or HPLC grade. Unless otherwise specified, all materials
were obtained from commercial suppliers (Sigma-Aldrich, Ark-Pharm, TCI, Aladdin or
Energy Chemical) and used as supplied without further purification. WA, used as starting
1
9

material, was isolated from Physalis angulata var. villosa by our group, and its structure was
identified by ESIMS and NMR. All the reactions were monitored by TLC analysis. Analytical
TLC was carried out on silica gel plates HSGF254. Chromatographic purification was
conducted on commercial silica gel column (200-300 meshes, Qingdao Haiyang Chemical
Co., Ltd., China). Preparative HPLC purification was performed on Shimadzu apparatus with
a 1200 series multi-wavelength detector using a Shim-pack RP-C18 column (20 × 200 mm,
i.d.) at a flow rate of 10 mL/min. All final compounds have purity ≥ 95% as determined by an
1
Agilent1260 instrument equipped with multiple wavelength diode array detector. H (500
13
MHz) and C (126 MHz) nuclear magnetic resonance (NMR) spectra were measured on a
Bruker ACF-500 NMR instrument (Bruker, Karlsruhe, Germany) at ambient temperature
using TMS as the internal standard. Solutions were prepared in CDCl with chemical shifts (δ)
3
which are given in ppm referenced to deuterated solvent as an internal standard.
High-resolution ESI mass spectrometry (HRMS) data were acquired on an Agilent 6520B
Q-TOF mass instrument (Agilent, America).
4
.1.2. Procedure for synthesis of intermediate a
The preparation of a was performed following the route illustrated in Schemes 1 and 2.
WA (1 eq., 20 mg, 0.038 mmol) was dissolved in pre-dried DCM (0.5 mL). Then TEA (0.5
mL) and glutaric anhydride (1.5 eq., 6.5 mg, 0.057 mmol) was added at room temperature.
The reaction mixture was stirred for 72 h in refluxing. The mixture was washed with water (1
mL × 3) and brine (1 mL × 3), respectively. The organic layers were dried over Na SO and
2
4
concentrated in vacuo. The residue was purified by silica gel column chromatography
hexanes/EtOAc = 1/30) to get the desired pure product a.
(
4
2
.1.2.1.
(20S,
2R)-15α-acetoxy-5β,6β-epoxy-14β-dihydroxy-1-oxowitha-2,16,24-trienolide-4β-yl
2
0

1
hemiglutarate (a). Pale yellow amorphous powder. Yield: 21.9 mg (90.2%). H NMR (500
MHz, CDCl ), δ 7.05 (1H, dd, J = 9.8, 6.0 Hz, H-3), 6.24 (1H, d, J = 9.8 Hz, H-2), 5.68 (1H,
3
d, J = 2.5 Hz, H-16), 5.22 (1H, d, J = 2.6 Hz, H-15), 4.75 (1H , d, J = 6.0 Hz, H-4), 4.23 (1H,
ddd, J = 12.3, 6.6, 3.4 Hz, H-22), 3.35 (1H, s, H-6), 1.94 (3H, s, -COCH ), 1.93 (3H, s,
3
Me-28), 1.90 (3H, s, Me-27), 1.39 (3H, s, Me-19), 1.11 (3H, d, J = 7.0 Hz, Me-21), 1.08 (3H,
1
3
s, Me-18). C NMR (126 MHz, CDCl ), δ 201.5 (C-1), 176.4 (-COOH), 172.1 (-COO-),
3
1
69.7 (-COCH3), 166.3 (C-26), 162.5 (C-17), 148.6 (C-24), 140.5 (C-3), 133.3 (C-2), 122.3
(
C-25), 121.1 (C-16), 83.7 (C-15), 81.6 (C-14), 79.4 (C-22), 72.0 (C-4), 61.3 (C-5), 60.4
C-6), 52.2 (C-13), 48.1 (C-10), 39.6 (C-9), 37.4 (C-12), 35.5 (C-20), 34.6 (C-8), 33.1 (C-23),
(
3
3.1 (-OCO-CH -CH -CH COOH), 32.6 (-OCO-CH -CH -CH -COOH), 24.6 (C-7), 21.5
2 2 2- 2 2 2
(C-11), 21.1 (-COCH ), 20.6 (C-28), 19.9 (-OCO-CH -CH -CH -COOH), 17.7 (C-21), 16.0
3 2 2 2
(
C-19), 15.8 (C-18), 12.6 (C-27). HRMS m/z calculated for C H O , 640.2884; found [M +
35
44 11
+
H] , 641.2856.
4
.1.3. Procedure for synthesis of intermediate b
The preparation of b was performed following the route illustrated in Schemes 1 and 2.
WA (1 eq., 20 mg, 0.038 mmol) was dissolved in pre-dried DCM (0.5 mL). Then TEA (0.5
mL) and succinic anhydride (1.5 eq., 5.7 mg, 0.057 mmol) was added at room temperature.
The reaction mixture was stirred for 50 h in refluxing. The mixture was washed with water (1
mL × 3) and brine (1 mL × 3), respectively. The organic layers were dried over Na SO and
2
4
concentrated in vacuo. The residue was purified by silica gel column chromatography
EtOAc) to get the desired pure product b.
(
4
2
.1.3.1.
(20S,
2R)-15α-acetoxy-5β,6β-epoxy-14β-dihydroxy-1-oxowitha-2,16,24-trienolide-4β-yl
1
hemisuccinate (b). Pale yellow amorphous powder. Yield: 23.3 mg (98.0%). H NMR (500
2
1

MHz, CDCl ), δ 7.05 (1H, dd, J = 9.7, 6.1 Hz, H-3), 6.19 (1H, d, J = 9.7 Hz, H-2), 5.66 (1H,
3
d, J = 2.3 Hz, H-16), 5.20 (1H, d, J = 2.5 Hz, H-15), 4.72 (1H , d, J = 6.1 Hz, H-4), 4.23 (1H,
ddd, J = 10.7, 6.7, 3.4 Hz, H-22), 3.31 (1H, s, H-6), 1.92 (3H, s, -COCH ), 1.92 (3H, s,
3
Me-28), 1.84 (3H, s, Me-27), 1.39 (3H, s, Me-19), 1.10 (3H, d, J = 7.0 Hz, Me-21), 1.07 (3H,
1
3
s, Me-18). C NMR (126 MHz, CDCl ), δ 201.6 (C-1), 177.7 (-COOH), 172.9 (-COO-),
3
1
69.7 (-COCH3), 166.4 (C-26), 162.3 (C-17), 148.8 (C-24), 141.0 (C-3), 133.0 (C-2), 122.2
(
C-25), 122.0 (C-16), 83.8 (C-15), 81.6 (C-14), 79.3 (C-22), 71.6 (C-4), 61.1 (C-5), 60.4
C-6), 52.2 (C-13), 48.0 (C-10), 39.7 (C-9), 37.5 (C-12), 35.5 (C-20), 34.7 (C-8), 33.1 (C-23),
(
3
1.7 (-OCO-CH -CH -COOH), 30.8 (-OCO-CH -CH -COOH), 24.6 (C-7), 21.4 (C-11), 21.1
2 2 2 2
(
-COCH ), 20.5 (C-28), 17.7 (C-21), 16.0 (C-19), 15.8 (C-18), 12.5 (C-27). HRMS m/z
3
+
calculated for C H O , 626.2727; found [M + H] , 627.2788.
3
4
42 11
4
.1.4. Procedure for synthesis of intermediate c
The preparation of c was performed following the route illustrated in Schemes 1 and 2.
WA (1 eq., 20 mg, 0.038 mmol) was dissolved in pre-dried DCM (0.5 mL). Then TEA (0.5
mL) and maleic anhydride (1.5 eq., 5.6 mg, 0.057 mmol) was added at room temperature.
The reaction mixture was stirred for 40 h in refluxing. The mixture was washed with water (1
mL × 3) and brine (1 mL × 3), respectively. The organic layers were dried over Na SO and
2
4
concentrated in vacuo. The residue was purified by silica gel column chromatography
hexanes/EtOAc = 1/20) to get the desired pure product c.
(
4
2
.1.4.1.
(20S,
2R)-15α-acetoxy-5β,6β-epoxy-14β-dihydroxy-1-oxowitha-2,16,24-trienolide-4β-yl
1
hemimaleate (c). Pale yellow amorphous powder. Yield: 22.9 mg (96.5%). H NMR (500
MHz, CDCl ), δ 7.06 (1H, dd, J = 9.8, 6.0 Hz, H-3), 6.57 (1H, d, J = 12.0 Hz, -C=C-), 6.20
3
(1H, d, J = 9.8 Hz, H-2), 5.69 (1H, d, J = 12.0 Hz, -C=C-), 5.66 (1H, d, J = 2.2 Hz, H-16),
2
2

5
.21 (1H, d, J = 2.5 Hz, H-15), 4.79 (1H , d, J = 6.0 Hz, H-4), 4.24 (1H, ddd, J = 10.3, 6.4,
3
.4 Hz, H-22), 3.32 (1H, s, H-6), 1.92 (3H, s, -COCH ), 1.92 (3H, s, Me-28), 1.84 (3H, s,
3
1
3
Me-27), 1.37 (3H, s, Me-19), 1.10 (3H, d, J = 7.0 Hz, Me-21), 1.06 (3H, s, Me-18). C NMR
(
(
(
126 MHz, CDCl ), δ 201.7 (C-1), 172.3 (-COOH), 169.7 (-COCH3), 166.4 (C-26), 165.1
3
-COO-), 162.3 (C-17), 148.8 (C-24), 143.8 (-CH=CH-), 141.1 (C-3), 133.1 (C-2), 122.2
C-25), 121.0 (C-16), 116.9 (-CH=CH-), 83.8 (C-15), 81.6 (C-14), 79.3 (C-22), 71.4 (C-4),
6
1.1 (C-5), 60.5 (C-6), 52.2 (C-13), 48.0 (C-10), 39.8 (C-9), 37.5 (C-12), 35.5 (C-20), 34.7
(
C-8), 33.0 (C-23), 24.6 (C-7), 21.5 (C-11), 21.1 (-COCH ), 20.6 (C-28), 17.7 (C-21), 16.0
3
(
C-19), 15.8 (C-18), 12.5 (C-27). HRMS m/z calculated for C H O , 624.2571; found [M +
3
4
40 11
+
H] , 625.2605.
4
.1.5. Procedure for synthesis of intermediate d
To a mixture of morpholine (2 eq., 2.4 g, 27.1 mmol) and KH PO (2 eq., 4.7 g, 27.1
2
4
mmol) in DMSO (24 mL) was added 3, 4 - difluoronitrobenzene (1 eq., 2.2 g, 13.6 mmol),
and the resulting mixture was stirred at 60 ºC for 24 h. The resulting solution was added to
cool water, and extracted with ethyl acetate (50 mL × 3). The organic layer was dried over
Na SO , filtered and concentrated under vacuum to give a yellow solid (yield: 2.9 g, 94.2%).
2
4
The solid (2.9 g, 12.8 mmol) was dissolved in ethyl acetate (60 mL), and 10% Pd-C (10%,
.3 g) was added. The resulting mixture was stirred under H for 6 h. The mixture was
0
2
filtered and the filtrate was concentrated under vacuum to give d [52].
1
4
.1.5.1. 3-fluoro-4-morpholinoaniline (d). Pale brown solid. Yield: 2.3 g (87.6%). H NMR
(
500 MHz, CDCl ), The hydrogen signals of the phenyl ring [δ 6.78 (1H, t, J = 8.8), 6.39 (2H,
3
1
3
m)], 3.83 (4H, t, J = 4.2 Hz, -(CH -) O), 2.95 (4H, t, J = 3.9 Hz, -N(-CH -) ). C NMR (126
2
2
2 2
MHz, CDCl ), The carbon signals of the phenyl ring [δ 157.8, 143.0, 131.8, 120.4, 110.7,
3
1
04.0], 67.2 (-(CH -) O), 67.2 (-(CH -) O), 51.8 (-N(-CH -) ), 51.8 (-N(-CH -) ).
2 2 2 2 2 2 2 2
2
3

4
.1.6. General procedure for synthesis of compounds 1-21
EDCI·HCl (2.2 eq.), HOBt·H O (1.5 eq.), and pyridine (3.0 eq.) were added to a solution
2
of Acid (the intermediates a, b, or c; 1.0 eq.) in DMF at room temperature under N2
atmosphere. After 30 min, a solution of R -H (Scheme 1; 1.2 eq.) in DMF was added
2
dropwise. The mixture was then stirred for 15 h at room temperature and, upon completion of
the reaction, water was added and extracted with EtOAc. The organic layers were then
washed with brine, dried over Na SO , filtered and concentrated in vacuo. The residue was
2
4
purified via preparative HPLC using acetonitrile/water (1/1) as the solvent system to obtain
the target compounds.
4
2
5
1
9
.1.6.1.
(20S,
2R)-15α-acetoxy-5β,6β-epoxy-14β-dihydroxy-1-oxowitha-2,16,24-trienolide-4β-yl
-((1,3,4-thiadiazol-2-yl)amino)-5-oxopentanoate (1). Pale yellow amorphous powder. Yield:
1
5.4 mg (56.2%). H NMR (500 MHz, CDCl ), δ 8.80 (1H, s, -SCH=N-), 6.98 (1H, dd, J =
3
.9, 5.8 Hz, H-3), 6.25 (1H, d, J = 9.9 Hz, H-2), 5.69 (1H, d, J = 2.6 Hz, H-16), 5.23 (1H, d, J
=
2.7 Hz, H-15), 4.90 (1H , d, J = 5.8 Hz, H-4), 4.27 (1H, ddd, J = 12.7, 6.0, 3.5 Hz, H-22),
3
.45 (1H, s, H-6), 1.95 (3H, s, -COCH ), 1.93 (3H, s, Me-28), 1.86 (3H, s, Me-27), 1.38 (3H,
3
1
3
s, Me-19), 1.13 (3H, d, J = 7.0 Hz, Me-21), 1.11 (3H, s, Me-18). C NMR (126 MHz,
CDCl ), δ 201.7 (C-1), 171.9 (-CONH-), 170.8 (-COO-), 169.7 (-COCH ), 166.4 (C-26),
3
3
1
1
6
61.9 (C-17), 159.6 (-SCN-NH-), 148.8 (C-24), 147.7 (-SCHN-), 140.2 (C-3), 133.2 (C-2),
22.3 (C-25), 121.4 (C-16), 83.8 (C-15), 81.6 (C-14), 79.0 (C-22), 71.4 (C-4), 62.8 (C-5),
1.2 (C-6), 52.3 (C-13), 48.1 (C-10), 39.8 (C-9), 37.3 (C-12), 35.4 (C-20), 35.1 (C-8), 34.9
(C-23), 33.6 (-NHCOCH -), 32.6 (-OCOCH -), 24.7 (C-7), 21.6 (C-11), 21.5 (-COCH ), 21.1
2 2 3
(
-CH CH CH -), 20.6 (C-28), 17.4 (C-21), 16.5 (C-19), 16.0 (C-18), 12.6 (C-27). HRMS m/z
2
2
2
+
calculated for C H N O S, 723.2826; found [M + H] , 724.2857.
3
7
45
3
10
2
4

4
2
5
.1.6.2.
(20S,
2R)-15α-acetoxy-5β,6β-epoxy-14β-dihydroxy-1-oxowitha-2,16,24-trienolide-4β-yl
-oxo-5-(thiazol-2-ylamino)pentanoate (2). Pale yellow amorphous powder. Yield: 13.6 mg
1
(
49.5%). H NMR (500 MHz, CDCl ), δ 7.43 (1H, d, J = 7.2 Hz, =NCH=CHS-),7.00 (1H, d,
3
J=6.0 Hz, -SCH=CHN=), 6.96 (1H, dd, J = 9.9, 5.8 Hz, H-3), 6.25 (1H, d, J=9.9 Hz, H-2),
5
4
.70 (1H, d, J=2.6 Hz, H-16), 5.22 (1H, d, J=2.7 Hz, H-15), 4.94 (1H , d, J =5.8 Hz, H-4),
.26 (1H, ddd, J=12.7, 6.1, 3.5 Hz, H-22), 3.46 (1H, s, H-6), 1.95 (3H, s, -COCH ), 1.93 (3H,
3
s, Me-28), 1.86 (3H, s, Me-27), 1.38 (3H, s, Me-19), 1.13 (3H, d, J = 7.0 Hz, Me-21), 1.08
1
3
(
(
(
3H, s, Me-18). C NMR (126 MHz, CDCl ), δ 201.8 (C-1), 172.1 (-CONH-), 170.4
3
-COO-), 169.8 (-COCH ), 166.4 (C-26), 162.0 (C-17), 159.2 (-N=C-S-), 148.8 (C-24), 140.2
3
C-3), 136.1 (=NCH=CHS-), 133.0 (C-2), 122.3 (C-25), 121.3 (C-16), 113.7 (-SCH=CHN=),
8
3.8 (C-15), 81.5 (C-14), 79.1 (C-22), 71.0 (C-4), 63.0 (C-5), 61.3 (C-6), 52.2 (C-13), 48.0
C-10), 39.7 (C-9), 37.3 (C-12), 35.3 (C-20), 35.0 (C-8), 34.9 (C-23), 33.6 (-NHCOCH -),
(
2
3
2.6 (-OCOCH -), 24.7 (C-7), 21.8 (C-11), 21.5 (-COCH ), 21.0 (C-28), 20.6
2 3
(
-CH CH CH -), 17.4 (C-21), 16.7 (C-19), 16.1 (C-18), 12.6 (C-27). HRMS m/z calculated
2 2 2
+
for C H N O S, 722.2873; found [M + H] , 723.2906.
3
8
46
2
10
4
2
5
2
7
.1.6.3.
(20S,
2R)-15α-acetoxy-5β,6β-epoxy-14β-dihydroxy-1-oxowitha-2,16,24-trienolide-4β-yl
-(benzo[d]thiazol-2-ylamino)-5-oxopentanoate (3). Pale yellow amorphous powder. Yield:
1
0.2 mg (68.7%). H NMR (600 MHz, CDCl ), The hydrogen signals of the benzene ring [δ
3
.82 (1H, d, J = 7.8 Hz), 7.76 (1H, d, J = 8.1 Hz), 7.42 (1H, t, J = 7.6 Hz), 7.32 (1H, t, J = 7.6
Hz)], 6.99 (1H, dd, J = 9.9, 5.8 Hz, H-3), 6.26 (1H, d, J=9.9 Hz, H-2), 5.71 (1H, d, J=2.5 Hz,
H-16), 5.24 (1H, d, J=2.6 Hz, H-15), 4.89 (1H , d, J = 5.8 Hz, H-4), 4.25 (1H, ddd, J=12.6,
6
.1, 3.4 Hz, H-22), 3.46 (1H, s, H-6), 1.95 (3H, s, -COCH ), 1.94 (3H, s, Me-28), 1.87 (3H, s,
3
1
3
Me-27), 1.38 (3H, s, Me-19), 1.13 (3H, d, J = 7.0 Hz, Me-21), 1.04 (3H, s, Me-18). C NMR
2
5

(151 MHz, CDCl ), δ 201.8 (C-1), 172.1 (-CONH-), 170.9 (-COO-), 169.7 (-COCH ), 166.4
3 3
(C-26), 162.1 (C-17), 158.6 (-SC=N-), 148.7 (C-24), 140.3 (C-3), 133.1 (C-2), 122.3 (C-25),
1
1
4
3
21.3 (C-16), The carbon signals of the benzene ring [147.4, 131.8, 126.5, 124.3, 121.7,
20.7], 83.7 (C-15), 81.5 (C-14), 79.1 (C-22), 71.4 (C-4), 62.5 (C-5), 61.1 (C-6), 52.2 (C-13),
8.0 (C-10), 39.6 (C-9), 37.3 (C-12), 35.3 (C-20), 35.1 (-NHCOCH -), 34.9 (-OCOCH -),
2
2
3.4 (C-8), 32.7 (C-23), 24.6 (C-7), 21.5 (C-11), 21.5 (-CH CH CH -), 20.8 (-COCH ), 20.6
2
2
2
3
(C-28), 17.4 (C-21), 16.5 (C-19), 16.0 (C-18), 12.6 (C-27). HRMS m/z calculated for
+
C H N O S, 772.3030; found [M + H] , 773. 3102.
42
48
2
10
4
2
5
.1.6.4.
2R)-15α-acetoxy-5β,6β-epoxy-14β-dihydroxy-1-oxowitha-2,16,24-trienolide-4β-yl
(20S,
-((4-fluorophenyl)amino)-5-oxopentanoate (4). Pale yellow amorphous powder. Yield: 12.3
1
mg (44.1%). H NMR (500 MHz, CDCl ), δ 7.84 (1H, s, -CONH-), The hydrogen signals of
3
the phenyl ring [7.55 (2H, dd, J = 9.0, 4.8 Hz), 7.00 (2H, t, J = 8.7 Hz)], 7.07 (1H, dd, J = 9.8,
6
2
3
.0 Hz, H-3), 6.26 (1H, d, J = 9.8 Hz, H-2), 5.68 (1H, d, J = 2.6 Hz, H-16), 5.23 (1H, d, J =
.7 Hz, H-15), 4.77 (1H , d, J = 6.0 Hz, H-4), 4.23 (1H, ddd, J = 12.3, 7.1, 3.5 Hz, H-22),
.41 (1H, s, H-6), 1.94 (3H, s, -COCH ), 1.93 (3H, s, Me-28), 1.85 (3H, s, Me-27), 1.33 (3H,
3
1
3
s, Me-19), 1.11 (3H, d, J = 7.0 Hz, Me-21), 1.06 (3H, s, Me-18). C NMR (126 MHz,
CDCl ), δ 202.0 (C-1), 172.3 (-CONH-), 170.4 (-COO-), 169.7 (-COCH ), 166.2 (C-26),
3
3
1
62.6 (C-17), 148.6 (C-24), 140.6 (C-3), 133.3 (C-2), 122.3 (C-25), 120.9 (C-16), The carbon
signals of the phenyl ring [158.4, 134.4, 121.6, 121.6, 115.8, 115.6], 83.7 (C-15), 81.5 (C-14),
9.5 (C-22), 72.0 (C-4), 61.8 (C-5), 60.7 (C-6), 52.1 (C-13), 48.2 (C-10), 39.6 (C-9), 37.4
C-12), 36.2 (-NHCOCH -), 35.5 (C-20), 34.4 (C-8), 33.5 (C-23), 33.3 (-OCOCH -), 24.6
7
(
(
(
[
2
2
C-7), 21.5 (C-11), 21.4 (-COCH ), 21.1 (-CH -CH -CH -), 20.6 (C-28), 17.9 (C-21), 16.0
3
2
2
2
C-19), 15.9 (C-18), 12.6 (C-27). HRMS m/z calculated for C H FNO , 733.3262; found
4
1
48
10
+
M + Na] , 756.3125.
2
6

4
2
5
1
7
.1.6.5.
(20S,
2R)-15α-acetoxy-5β,6β-epoxy-14β-dihydroxy-1-oxowitha-2,16,24-trienolide-4β-yl
-((4-(dimethylamino)phenyl)amino)-5-oxopentanoate (5). Yellow amorphous powder. Yield:
1
2.2 mg (42.5%). H NMR (500 MHz, CDCl ), The hydrogen signals of the phenyl ring [δ
3
.41 (2H, d, J = 8.9 Hz), 6.71 (2H, d, J = 7.8 Hz)], δ 7.05 (1H, dd, J = 9.8, 6.0 Hz, H-3), 6.25
(
(
(
1H, d, J = 9.8 Hz, H-2), 5.68 (1H, d, J = 2.1 Hz, H-16), 5.22 (1H, d, J = 2.6 Hz, H-15), 4.78
1H , d, J = 5.9 Hz, H-4), 4.23 (1H, ddd, J = 12.4, 7.0, 3.5 Hz, H-22), 3.47 (1H, s, H-6), 2.90
6H, s, -N(-CH )), 1.94 (3H, s, -COCH ), 1.93 (3H, s, Me-28), 1.85 (3H, s, Me-27), 1.35 (3H,
3
3
1
3
s, Me-19), 1.11 (3H, d, J = 7.0 Hz, Me-21), 1.06 (3H, s, Me-18). C NMR (126 MHz,
CDCl ), δ 201.9 (C-1), 172.4 (-CONH-), 170.1 (-COO-), 169.7 (-COCH ), 166.3 (C-26),
3
3
1
62.5 (C-17), 148.6 (C-24), 140.6 (C-3), 133.2 (C-2), 122.3 (C-25), 120.9 (C-16), The carbon
signals of the phenyl ring [162.4, 128.3, 121.7, 121.7, 113.4, 113.4], 83.7 (C-15), 81.5 (C-14),
7
9.4 (C-22), 71.8 (C-4), 61.8 (C-5), 60.7 (C-6), 52.3 (C-13), 48.1 (C-10), 41.2 (-N(-CH ) ),
3 2
4
1.2 (-N(-CH ) ), 39.6 (C-9), 37.4 (C-12), 36.2 (-NHCOCH -), 35.5 (C-20), 34.6 (C-8), 33.5
3
2
2
(C-23), 33.2 (-OCOCH -), 24.6 (C-7), 21.5 (-(-CH )-), 21.2 (C-11), 20.6 (-COCH ), 19.5
2 2 3
(C-28), 17.8 (C-21), 16.0 (C-19), 16.0 (C-18), 12.6 (C-27). HRMS m/z calculated for
+
C H N O , 758.3778; found [M + H] , 759.3809.
43
54
2
10
4
2
5
1
.1.6.6.
2R)-15α-acetoxy-5β,6β-epoxy-14β-dihydroxy-1-oxowitha-2,16,24-trienolide-4β-yl
(20S,
-((4-methoxyphenyl)amino)-5-oxopentanoate (6). Pale yellow amorphous powder. Yield:
1
6.5 mg (58.2%). H NMR (500 MHz, CDCl ), δ 7.70 (1H, s, -CONH-), The hydrogen
3
signals of the phenyl ring [7.47 (2H, d, J = 9.0 Hz), 6.84 (2H, d, J = 9.0 Hz)], 7.06 (1H, dd, J
9.8, 6.0 Hz, H-3), 6.25 (1H, d, J = 9.8 Hz, H-2), 5.68 (1H, d, J = 2.6 Hz, H-16), 5.22 (1H, d,
J = 2.7 Hz, H-15), 4.78 (1H , d, J = 5.9 Hz, H-4), 4.22 (1H, ddd, J = 12.3, 6.9, 3.5 Hz, H-22),
.40 (1H, s, H-6), 1.94 (3H, s, -COCH ), 1.92 (3H, s, Me-28), 1.85 (3H, s, Me-27), 1.34 (3H,
=
3
3
2
7

1
3
s, Me-19), 1.11 (3H, d, J = 7.0 Hz, Me-21), 1.06 (3H, s, Me-18). C NMR (126 MHz,
CDCl ), δ 201.9 (C-1), 172.3 (-CONH-), 170.2 (-COO-), 169.7 (-COCH ), 166.2 (C-26),
3
3
1
62.6 (C-17), 148.6 (C-24), 140.6 (C-3), 133.2 (C-2), 122.3 (C-25), 120.9 (C-16), The carbon
signals of the phenyl ring [156.4, 131.4, 122.1, 121.7, 121.6, 114.2], 83.7 (C-15), 81.5 (C-14),
9.5 (C-22), 71.9 (C-4), 61.8 (C-5), 60.7 (C-6), 55.6 (-OCH ), 52.1 (C-13), 48.1 (C-10), 39.6
7
3
(
C-9), 37.4 (C-12), 36.1 (-NHCOCH -), 35.5 (C-20), 34.5 (C-8), 33.5 (C-23), 33.2
2
(
-OCOCH -), 24.6 (C-7), 21.5 (C-11), 21.4 (-COCH ), 21.1 (-CH -CH -CH -), 20.6 (C-28),
2
3
2
2
2
1
7
7.8 (C-21), 16.0 (C-19), 15.9 (C-18), 12.6 (C-27). HRMS m/z calculated for C H NO ,
4
2
51
11
+
45.3462; found [M + H] , 746.3485.
4
2
5
.1.6.7.
(20S,
2R)-15α-acetoxy-5β,6β-epoxy-14β-dihydroxy-1-oxowitha-2,16,24-trienolide-4β-yl
-((3-fluoro-4-morpholinophenyl)amino)-5-oxopentanoate (7). Pale yellow amorphous
1
powder. Yield: 17.1 mg (54.9%). H NMR (500 MHz, CDCl ), δ 7.78 (1H, s, -CONH- ), The
3
hydrogen signals of the phenyl ring [7.59 (1H, d, J = 14.4 Hz), 7.15 (1H, d, J = 7.6 Hz), 6.96
(1H, dd, J = 9.6, 5.9 Hz)], 7.07 (1H, dd, J = 9.8, 6.0 Hz, H-3), 6.27 (1H, d, J = 9.8 Hz, H-2),
5
4
=
.70 (1H, d, J = 2.6 Hz, H-16), 5.25 (1H, d, J = 2.7 Hz, H-15), 4.79 (1H , d, J = 6.0 Hz, H-4),
.22 (1H, ddd, J = 12.2, 7.4, 3.6 Hz, H-22), 3.87 (4H, t, J = 4.2 Hz, O(CH -) ), 3.06 (4H, t, J
2
2
3.9 Hz, -N(CH -) ), 1.95 (3H, s, -COCH ), 1.94 (3H, s, Me-28), 1.86 (3H, s, Me-27), 1.35
2
2
3
13
(
3H, s, Me-19), 1.12 (3H, d, J = 7.0 Hz, Me-21), 1.07 (3H, s, Me-18). C NMR (126 MHz,
CDCl ), δ 202.04 (C-1), 172.3 (-CONH-), 170.3 (-COO-), 169.7 (-COCH ), 166.1 (C-26),
3
3
1
62.8 (C-17), 148.5 (C-24), 140.6 (C-3), 133.3 (C-2), 122.4 (C-25), 120.9 (C-16), The carbon
signals of the phenyl ring [156.6, 154.6, 115.6, 115.6, 109.0, 108.8], 83.7 (C-15), 81.6 (C-14),
9.7 (C-22), 72.0 (C-4), 67.0 (C-5), 62.0 (C-6), 60.8 (-OCHCHN-), 60.5 (-OCHCHN-), 52.1
7
(
C-13), 51.4 (-NCHCHO-), 51.4 (-NCHCHO-), 48.2 (C-10), 39.6 (C-9), 37.5 (C-12), 36.3
C-20), 35.6 (C-8), 34.5 (C-23), 33.6 (-NHCOCH -), 33.4 (-OCOCH -), 24.6 (C-7), 21.5
(
2
2
2
8