Design, synthesis, and biological evaluation of 1,3,6,7-tetrahydroxyxanthone derivatives as phosphoglycerate mutase 1 inhibitors

Article abstract of DOI:10.1016/j.bmcl.2021.127820

Phosphoglycerate mutase 1 (PGAM1) is a promising target for cancer treatment. Herein, we found that α-mangostin and γ-mangostin exhibited moderate PGAM1 inhibitory activities, with IC₅₀ of 7.2 μM and 1.2 μM, respectively. Based on α-mangostin, a series of 1,3,6,7-tetrahydroxyxanthone derivatives were designed, synthesized and evaluated in vitro for PGAM1 inhibition. The significant structure–activity relationships (SAR) and a fresh binding mode of this kind of new compounds were also clearly described. This study provides valuable information for further optimization of PGAM1 inhibitors with 1,3,6,7-tetrahydroxyxanthone backbone or de novo design of novel inhibitor.

Full text of DOI:10.1016/j.bmcl.2021.127820

Bioorg. Med. Chem. Lett. 36 (2021) 127820
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design, synthesis, and biological evaluation of 1,3,6,7-tetrahydroxyxan-
thone derivatives as phosphoglycerate mutase 1 inhibitors
a
,
1
a,
1
a
b
a
a
a
Kaixuan Jiang , Biao Gao , Jing Yu , Lulu Jiang , Ao Niu , Yihe Jia , Tao Meng ,
b,
*
a,*
Lu Zhou , Jinxin Wang
a
Jiangsu Key Laboratory of Drug Design and Optimization, Department of Medicinal Chemistry, School of Pharmacy, China Pharmaceutical University, Nanjing 210009,
China
b
Department of Medicinal Chemistry, School of Pharmacy, Fudan University, No. 826, Zhangheng Rd, Shanghai 201203, China
A R T I C L E I N F O
A B S T R A C T
Keywords:
Phosphoglycerate mutase 1 (PGAM1) is a promising target for cancer treatment. Herein, we found that
PGAM1 inhibitor
Mangostin
α
-mangostin and γ-mangostin exhibited moderate PGAM1 inhibitory activities, with IC50 of 7.2
μM and 1.2 µM,
respectively. Based on -mangostin, a series of 1,3,6,7-tetrahydroxyxanthone derivatives were designed, syn-
α
1
,3,6,7-Tetrahydroxyxanthone derivatives
thesized and evaluated in vitro for PGAM1 inhibition. The significant structure–activity relationships (SAR) and a
fresh binding mode of this kind of new compounds were also clearly described. This study provides valuable
information for further optimization of PGAM1 inhibitors with 1,3,6,7-tetrahydroxyxanthone backbone or de
novo design of novel inhibitor.
Introduction
carcinoma,7 and colorectal cancer.8 Jiang et al. had demonstrated that
PGAM1 simultaneously affects anabolic and glycolysis reactions in cells
Increased glucose metabolism is one of the distinguishing features
between normal cells and highly proliferating cells like cancer, stem and
by modulating the cellular levels of two key intermediates 3-PG and 2-
PG. Elevated 3-PG is able to competitively inhibit 6-phosphogluconate
(6-PGD), resulting in reduced oxidative pentose phosphate pathway
1
immune cells. German scientist Otto Warburg and his coworkers found
9
,10
that tumor tissues metabolize nearly ten-fold more glucose to lactate in a
given time than normal tissues under aerobic conditions. Even in the
non-hypoxia environment, tumor cells prefer the metabolic pathway of
glycolysis to the more efficient oxidative phosphorylation for energy
production.2 This observation is called Otto Warburg effect.3 Cancer
cells tend to competently coordinate bioenergetics, anabolic biosyn-
thesis, and balanced redox status to provide an optimum microenvi-
ronment for cancer cell proliferation and tumor growth.4 Therefore,
aerobic glycolysis has been recognized as a hallmark of cancer, and
targeting the key enzyme involved in aerobic glycolysis may be a
promising therapeutic approach for cancer treatment.
(PPP) flux and ultimately inhibition of nucleotide synthesis.
Mean-
while, 2-PG is a positive regulator of 3-phosphoglycerate dehydrogenase
9
(3-PHGDH) which participates in amino acid synthesis. In conclusion,
inhibition of PGAM1 is not only reducing the rate of glycolysis but also
can affect the biosynthetic pathway of nucleotide and amino acid.
Developing PGAM1 inhibitors is a promising direction toward potential
cancer treatments.
To date, only a few types of PGAM1 inhibitors have been reported.
The reported inhibitors can generally be divided into two types: (i)
substrate competition inhibitor, such as MJE3, which was found to
covalently label PGAM1 on Lys 100 by its unique spiroepoxide phar-
macophore.11 (ii) allosteric site inhibitors, such as PGMI-004A, which is
obtained from the structure modification of Alizarin Red S (ARS) to
improve permeability (Fig. 1).10 Moreover, although (-)-Epi-
gallocatechin-3-gallate (EGCG) was also identified as a non-substrate
Phosphoglycerate mutase 1 (PGAM1), an important mutase in the
glycolytic pathway that catalyzes the conversion of 3-phosphoglycerate
(
3-PG) into 2-phosphoglycerate (2-PG), plays an important role in
coordinating glycolysis and anabolic activity to promote the prolifera-
5
12
tion of cancer cells. In humans, PGAM1 is found to be overexpressed in
competitive inhibitor (Fig. 1), it belongs to pan-assay interference
various types of cancer, such as breast cancer,6 hepatocellular
compounds (PAINS) due to its promiscuous actions.
13
*
Corresponding authors.
E-mail addresses: zhoulu@fudan.edu.cn (L. Zhou), jinxinwang@163.com (J. Wang).
These authors contributed equally to this work.
1
https://doi.org/10.1016/j.bmcl.2021.127820
Received 7 December 2020; Received in revised form 8 January 2021; Accepted 18 January 2021
Available online 26 January 2021
0
960-894X/© 2021 Elsevier Ltd. All rights reserved.

K. Jiang et al.
Bioorganic & Medicinal Chemistry Letters 36 (2021) 127820
Fig. 1. The chemical structures of mentioned PGAM1 inhibitors.
Fig. 2. xanthone as a novel scaffold.
PGMI-004A contains an anthraquinone core, which belongs to
structural alerts in medicinal chemistry and may lead to potential side
effects due to its high reactivity. Base on this, a scaffold hopping was
conducted to replace the anthraquinone scaffold with xanthone scaffold
to avoid quinone structure, and a series of xanthone derivatives were
disclosed. Many of them showed higher potency against PGAM1 than
PGMI-004A (Fig. 2).1
4,15
In the past few years, our research group has devoted to the related
study of natural product -mangostin extracted from the pericarp of
α
mangosteen and exhibiting numerous pharmacological and biological
activities such as antifungal, anticancer, antibacterial, antioxidant,
antimalarial and anti-inflammatory.1
6,17
In vitro screening assays
α
-mangostin (IC50 = 7.2
μ
M) exhibited better PGAM1 inhibitory activity
M) (Fig. 3). We assumed that it may bind
than PGMI-004A (IC50 = 11.3
μ
to the allosteric site due to its high similarity with reported allosteric
inhibitor mentioned above. Herein, we described our efforts aimed at
the discovery of novel potent PGAM1 inhibitors based on -mangostin.
α
Initially, a series of substituents were introduced on the exposed
Fig. 3. The chemical structure of -mangostin.
α
hydroxyl groups at C3 and C6 of ɑ-mangostin (1a-f), which led to a
significant reduction of enzyme activities against PGAM1, with IC50
>
2
0
μM (Fig.4). The results suggested that the hydrogen bond donor of
2

K. Jiang et al.
Bioorganic & Medicinal Chemistry Letters 36 (2021) 127820
Fig. 4. The chemical structures of 1a-g.
Scheme 1. Synthesis of compounds 1a-1f. Reagent and condition:(a) Isocyanate, triethylamine, acetonitrile, reflux, 5 h; (b) (i) Ethyl chloroacetate, Potassium
carbonate, acetone, reflux, 6h; (ii) NaOH ,H O, methanol, reflux, 4 h;(c) dimethylamine solution in water (40%), EDCI, HOBT, TEA, RT, 10 h; (d) 1,2-Dibromoethane,
2
Potassium carbonate, acetone, reflux, 20 h; (e) Secondary amine, Potassium carbonate, acetone, reflux, 10 h; (f) 2-Bromoethanol, Potassium carbonate, acetone,
reflux, 20 h.
hydroxyls at C3 and C6 may play an important role in its activity. The
synthetic pathways for the target compounds 1a-g are provided in
Scheme 1.
higher inhibitory activity than
α-mangostin with IC50 of 1.2 µM. Then,
we retained the fully exposed hydroxyls and transformed the iso-
pentenyl at C2 into hydrogen atoms (2b) to elucidate the role of iso-
pentenyls for enzymatic inhibitory activity. Compound 2b possessed a
similar inhibitory effect against PGAM1 compared with 2a, which sug-
gested that the isopentenyl at the C2 is not necessary for enzyme
To further explore the structure–activity relationship (SAR), we
exposed the last phenolic hydroxyl at C7 in ɑ-mangostin to obtain 2a
(
also called γ-mangstion). As shown in Table 1, compound 2a showed
3

K. Jiang et al.
Bioorganic & Medicinal Chemistry Letters 36 (2021) 127820
Table 1
substituents of different lengths and volumes.
Moreover, the enzyme inhibitory activity of methylation precursor
(
2f) of compound 2e was also determined, with a complete loss of po-
The structures of 2a-f and observed activities.
tency (IC50 > 20 M), suggesting that exposed hydroxyl groups at C3 and
μ
C6 were essential to activities (Table 1). The synthetic pathways for the
target compounds 2a-e are shown in Scheme 2 and Scheme 3.
1
2
a
NO
R
R
IC50(µM)
In addition, we found that introducing lipophilic groups on the hy-
droxyl group at C1, without any substituent at C8, could also lead to
good potencies (Table 2). The inhibitory activities increased with the
extension of the hydrophobic alkane chain substituents on the hydroxyl
a
2
2
2
a (γ-M)
1.9
b
c
H
H
3.4
20
(
3a-b). Unfortunately, the introduction of a piperidine ring at the end of
the alkane chain (3c-d) resulted in a significant drop in activities. By
contrast, a relatively lipophilic indoline substituent (3e) could obtain a
comparable potency with 3b. This might be due to the decrease of hy-
drophobic interactions when putting relatively polar groups (tertiary
amine 3c or amide group 3f) into a completely lipophilic cavity.
Calculated LogP of the terminal structures in the substituents are also
shown in Fig. 5 (by Chemdraw professional 17.0), which is consistent
with our hypothesis. The synthetic pathways for the target compounds
2
2
d
e
H
H
2.8
1.5
2
f
>20
3
a-e are shown in Scheme 4 and Scheme 5.
Cellular level experiments on the selected compounds with good
b
PGMI-004A
13.0
PGAM1 inhibitory activity were also carried out (Table 3). In vitro cell
anti-proliferation assays indicated that 2a and 3b exhibited good
inhibitory activities against Human lung adenocarcinoma cell H1299
a
γ-M: γ-mangostin.
Active control.
b
with EC50 of 2.2
μ
M and 3.4 M, respectively.
μ
To further understand the binding mode of this kind of inhibitors, we
carried out docking studies of compounds 2b and 3b with PGAM1(PDB
code: 5Y2I). A comparison with the co-crystal of PGMI-004A with
PGAM1(PDB code: 5Y2I)18 was also made (Fig.6).
inhibition. Afterwards, replacements of the isopentenyl at C8 with
shortened (2b) and extended (2c) aliphatic chains and bulky phenyl
substituents (2e) were conducted. Compound 2c showed an obvious
reduction of potency, but extended lipophilic chain (2c) and bulky
phenyl group (2e) led to comparable enzyme inhibition with 2b. Based
on the results, we hypothesized that the lipophilic substituents in this
position may extend into a hydrophobic pocket in the binding site. En-
tropy loss caused by long flexible chains may explain the similar activ-
ities of 2d and 2e with 2b, although they owned hydrophobic
The docking result of 2b (Fig.6, B) revealed that the 1,3,6,7-tetrahy-
droxyxanthone scaffold bound into a shallow groove in the binding site,
and the isopentenyl substituent, instead of the anthraquinone scaffold of
PGMI-004A in the co-crystal (Fig.6, A),18 extended into the primary
hydrophobic pocket (constituted by Phe22, Leu95, Val112 and Trp115).
Besides, the hydroxyls at C3 and C6 respectively formed hydrogen bonds
with Asn 20, Glu 19 and Arg 10, which may illustrate the reason for the
loss of potency when methylations were performed on hydroxyls in
◦
Scheme 2. Synthesis of key intermediate 8. Reagent and condition: (a) (i) SOCl2, CH
2
Cl
2
, reflux, 10 h; (ii) AlCl
3
, CH
2
Cl
.
2
, 0 C to reflux, 10 h; (b) Acetic acid, HBr,
2 2
reflux, N , 3 h; (c) Allyl bromide, Potassium carbonate, acetone, reflux, 4 h; (d) N,N-Dimethylaniline, reflux, 3 h, N
4

K. Jiang et al.
Bioorganic & Medicinal Chemistry Letters 36 (2021) 127820
◦
Scheme 3. Synthesis of compounds 2a-2e. Reagent and condition: (a) NaH, DMF, ethyl mercaptan, reflux, N
2
, 5 h; (b) Pd/C, ethanol, 40 C, 10 h, H
2
; (c) Acetic acid,
◦
HBr, reflux, N
2
, 20 h; (d) Dimethyl sulfate, Potassium carbonate, acetone, reflux, 6 h; (e) Olefins, Grubbs II, CH
2
Cl
2
, 40 C, 12 h, N
2
.
Table 2
The structures of 3a-e and observed activities.
NO
R
IC50(µM)
3a
3b
3c
12.8
2.8
>20
Fig. 5. cLogP of terminal structures in substituents of 3c-e.
3
3
d
e
>20
5.7
occupying the same groove (Fig. 6, C). The aliphatic chain on the C1
phenolic hydroxyl bent and then occupied the main hydrophobic cavity.
The flipped binding mode may be due to the high symmetry of the
scaffold and hydroxyl substitution, and interactions observed in the
docking study of 3b were similar to 2b. Overall, the meaningful results
suggested an apparently different binding mode compared to previously
reported allosteric site inhibitors PGMI-004A.1
a
PGMI-004A
13.0
a
Active control.
0,18
In summary, we first reported that
α-Mangostin and γ-mangostin
these positions (2f). A
π
-
π
stacking interaction was also observed be-
(
2a) target PGAM1 in the glycolytic pathway. Subsequently, a series of
tween Phe 22 and 1,3,6,7-tetrahydroxyxanthone scaffold (Fig. 6, D). The
docking result of 3b showed an analogous bind mode compared to 2b,
tetrahydroxyxanthone derivatives were designed, synthesized and
evaluated for their potential as novel PGAM1 inhibitors. In vitro cell
growth inhibition assays indicated that compound 2a, 3b exhibited
with
a horizontally flipped 1,3,6,7-tetrahydroxyxanthone scaffold
5

K. Jiang et al.
Bioorganic & Medicinal Chemistry Letters 36 (2021) 127820
◦
Scheme 4. Synthesis of key intermediate 15. Reagent and condition: (a) Acetic acid, HBr, N
2
, reflux, 24 h; (b) Benzyl bromide, potassium carbonate, DMF, 40 C, 4 h.
◦
Scheme 5. Synthesis of compounds 3a-3e. Reagent and condition:(a) Brominated alkanes, Caesium carbonate, acetonitrile, reflux, 5 h; (b) Pd/C, ethanol, 40 C, 10
2 2
h, H ; (c) (i) Ethyl 4-bromobutyrate, Caesium carbonate, acetonitrile, reflux, 5 h; (ii) NaOH, H O, methanol, reflux, 4 h; (d) Piperidine, EDCI, HOBT, TEA, RT, 10 h;
(
e) Dibromoalkane, Caesium carbonate, acetonitrile, reflux, 5 h; (f) amine substrates, Potassium carbonate, acetone, reflux, 6 h.
Table 3
better inhibitory activities against Human lung adenocarcinoma cell
H1299 with EC50 of 2.2 M and 3.4 M, respectively. The SAR studies
Antiproliferative activities of selected compounds against H1299 cells.
μ
μ
a
a
suggested that exposed hydroxyls at C3 and C6 of 1,3,6,7-tetrahydroxyx-
anthone scaffold and suitable hydrophobic substituents at C1 or C8 are
beneficial for the potency. Notably, the docking studies revealed a novel
binding mode that hydroxyls at C3 and C6 formed hydrogen bond
interaction with the protein and lipophilic substituents at C1 or C8
extended into the unique hydrophobic pocket in the allosteric site. This
research provides valuable information and guideline for further
exploration of novel PGAM1 inhibitors.
Compound
EC50
(μM)
compound
EC50( M)
μ
2
2
2
2
a
b
d
e
2.2
47.7
13.5
41.6
3a
60
3b
3.4
3e
28.1
26.2
a
PGMI-004A
a
Active control.
6

K. Jiang et al.
Bioorganic & Medicinal Chemistry Letters 36 (2021) 127820
Fig. 6. The binding modes of PGAM1- inhibitors, with carbon atoms colored in gray, oxygen atoms in red, nitrogen atoms in blue and hydrogen atoms in off-white
A: the crystal structure of PGMI-004A-PGAM1 complex{PDB 5Y2I}; B: the docking result of 2b; C: the docking result of 3b; D: interactions observed in 2b).
(
Declaration of Competing Interest
References
1
2
3
Lunt SY, Vander Heiden MG. Aerobic Glycolysis: Meeting the Metabolic
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Requirements of Cell Proliferation. Annu Rev Cell Dev Biol. 2011;27(1):441–464.
Warburg O, Wind FNE. The metabolism of tumors in the body. J Gen Physiol. 1927;8:
519–530.
Racker E. Bioenergetics and the Problem of Tumor Growth: An understanding of the
mechanism of the generation and control of biological energy may shed light on the
problem of tumor growth. Am Sci. 1972;60(1):56–63.
Acknowledgements
4
5
6
Cairns RA, Harris IS, Mak TW. Regulation of cancer cell metabolism. Nat Rev Cancer.
2
011;11(2):85–95.
This work was supported by the Natural Science Foundation of
Jiangsu Province (BK20191318), the National Natural Science Foun-
dation of China (No 82073710, 21877014), the Double First-Class
initiative new drug development and construction project of China
Pharmaceutical University (CPU2018PZQ07).
Fothergill-Gilmore LA, Watson HC. The phosphoglycerate mutases. John Wiley & Sons,
Ltd; 2006:227–313. doi:10.1002/9780470123089.ch6.
Durany N, Joseph J, Jimenez OM, et al. Phosphoglycerate mutase, 2,3-bisphospho-
glycerate phosphatase, creatine kinase and enolase activity and isoenzymes in breast
carcinoma. Br J Cancer. 2000;82(1):20–27.
7
Ren F, Wu H, Lei Y, Zhang H, Liu R, Zhao Y, Chen X, Zeng D, et al. Quantitative
proteomics identification of phosphoglycerate mutase 1 as a novel therapeutic target
in hepatocellular carcinoma. Mol Cancer. 2010;9(1):81. https://doi.org/10.1186/
Appendix A. Supplementary data
1
476-4598-9-81.
8
9
Usuba T, Ishibashi Y, Okawa Y, Hirakawa T, Takada K, Ohkawa K. Purification and
identification of monoubiquitin-phosphoglycerate mutase B complex from human
colorectal cancer tissues. Int J Cancer. 2001;94(5):662–668.
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.bmcl.2021.127820.
Jiang X, Sun Q, Li H, Li K, Ren X. The role of phosphoglycerate mutase 1 in tumor
aerobic glycolysis and its potential therapeutic implications: Phosphoglycerate
mutase 1 in tumor aerobic glycolysis. Int J Cancer. 2014;135(9):1991–1996.
7

K. Jiang et al.
Bioorganic & Medicinal Chemistry Letters 36 (2021) 127820
1
1
1
0
1
2
Hitosugi T, Zhou Lu, Elf S, et al. Phosphoglycerate mutase 1 coordinates glycolysis
and biosynthesis to promote tumor growth. Cancer Cell. 2012;22(5):585–600.
https://doi.org/10.1016/j.ccr.2012.09.020.
Evans MJ, Morris GM, Wu J, Olson AJ, Sorensen EJ, Cravatt BF. Mechanistic and
structural requirements for active site labeling of phosphoglycerate mutase by
spiroepoxides. Mol BioSyst. 2007;3(7):495. https://doi.org/10.1039/b705113a.
Li X, Tang S, Wang Q-Q, et al., Identification of epigallocatechin-3- gallate as an
inhibitor of phosphoglycerate mutase 1. Front. Pharmacol 8:325. doi: 10.3389/
fphar.2017.00325.
15 Wang P, Jiang L, Cao Y, Ye D, Zhou L. The Design and Synthesis of N-Xanthone
Benzenesulfonamides as Novel Phosphoglycerate Mutase 1 (PGAM1) Inhibitors.
Molecules. 2018;23(6). https://doi.org/10.3390/molecules23061396.
16 Shan T, Ma Q, Guo K, et al. Xanthones from mangosteen extracts as natural
chemopreventive agents: potential anticancer drugs. Curr Mol Med. 2011;11(8):
666–677. https://doi.org/10.2174/156652411797536679.
17 Ibrahim MY, Hashim NM, Mariod AA, et al. -Mangostin from Garcinia mangostana
α
Linn: An updated review of its pharmacological properties. Arabian J Chem. 2016;9
(3):317–329. https://doi.org/10.1016/j.arabjc.2014.02.011.
1
1
3
4
Baell J, Walters MA. Chemistry: Chemical con artists foil drug discovery. Nature.
18 Huang K, Liang Q, Zhou Y, Jiang LL, Gu WM, Luo MY, et al. A novel allosteric
inhibitor of phosphoglycerate mutase 1 suppresses growth and metastasis of non-
small-cell lung cancer. Cell Metab. 2019;30(6):1107–1119.e8. https://doi.org/
10.1016/j.cmet.2019.09.014.
2
014;513(7519):481–483. https://doi.org/10.1038/513481a.
Wang P, Jiang L, Cao Y, Zhang X, Chen B, et al. Xanthone derivatives as
phosphoglycerate mutase 1 inhibitors: Design, synthesis, and biological evaluation.
Bioorg Med Chem. 2018;26(8):1961–1970. https://doi.org/10.1016/j.
bmc.2018.02.044.
8