Structure-Activity Relationship Study Enables the Discovery of a Novel Berberine Analogue as the RXRα Activator to Inhibit Colon Cancer

Article abstract of DOI:10.1021/acs.jmedchem.0c00088

We reported recently that berberine (Ber), a traditional oriental medicine to treat gastroenteritis, binds and activates retinoid X receptor α (RXRα) for suppressing the growth of colon cancer cells. Here, we extended our studies based on the binding mode of Ber with RXRα by design, synthesis, and biological evaluation of a focused library of 15 novel Ber analogues. Among them, 3,9-dimethoxy-5,6-dihydroisoquinolino[3,2-a]isoquinolin-7-ium chloride (B-12) was identified as the optimal RXRα activator. More efficiently than Ber, B-12 bound and altered the conformation of RXRα/LBD, thereby suppressing the Wnt/β-catenin pathway and colon cancer cell growth via RXRα mediation. In addition, B-12 not only preserved Ber's tumor selectivity but also greatly improved its bioavailability. Remarkably, in mice, B-12 did not show obvious side effects including hypertriglyceridemia as other RXRα agonists or induce hepatorenal toxicity. Together, our study describes an approach for the rational design of Ber-derived RXRα activators as novel effective antineoplastic agents for colon cancer.

Full text of DOI:10.1021/acs.jmedchem.0c00088

Subscriber access provided by Uppsala universitetsbibliotek
Article
Structure-Activity Relationship Study Enables the Discovery of a
Novel Berberine Analogue as RXR# Activator to Inhibit Colon Cancer
Beibei Xu, Xunjin Jiang, Jing Xiong, Jun Lan, Yuan Tian, LINHAI ZHONG, Xinquan Wang, Ning Xu,
Hanwei Cao, Wenqing Zhang, Hao Zhang, Xiaoting Hong, Yan-yan Zhan, Yandong Zhang, and Tianhui Hu
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.0c00088 • Publication Date (Web): 11 May 2020
Downloaded from pubs.acs.org on May 13, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Structure-Activity Relationship Study Enables
the Discovery of a Novel Berberine Analogue as
RXRα Activator to Inhibit Colon Cancer
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
†
,#
‡,#
†
§
†
†
§
Beibei Xu, Xunjin Jiang, Jing Xiong, Jun Lan, Yuan Tian, Linhai Zhong, Xinquan Wang,
§
†
†
‖
†
*,†
Ning Xu, Hanwei Cao, Wenqing Zhang, Hao Zhang, Xiaoting Hong, Yan-yan Zhan, Yandong
*,‡
*,†
Zhang and Tianhui Hu
†Cancer Research Center, School of Medicine, Xiamen University, Xiamen, 361005, China
‡Department of Chemistry and Key Laboratory of Chemical Biology of Fujian Province, iChEM,
College of Chemistry and Chemical Engineering, Xiamen University, Xiamen, Fujian 361005,
China
§Department of Biological Sciences, Tsinghua University, Beijing 100084, China;
‖Department of pathology, Jinan University Medical College, Guangzhou, 510632, China
#These authors contributed equally.
*To whom all correspondence should be addressed. Emails: yyzhan@xmu.edu.cn,
ydzhang@xmu.edu.cn, thu@xmu.edu.cn
KEY WORDS: berberine analogue; RXRα activator; structure-activity analysis; B-12; colon
cancer
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ABSTRACT ： We reported recently that berberine, a traditional oriental medicine to treat
gastroenteritis, binds and activates Retinoid X receptor α (RXRα) to suppress the growth of colon
cancer cells. Here, we extended our studies based on the binding mode of berberine with RXRα
by design, synthesis and biological evaluation of a focused library of 15 novel berberine analogues.
Among them, 3,9-dimethoxy-5,6-dihydroisoquinolino[3,2-a]isoquinolin-7-ium chloride (B-12)
was identified as the optimal RXRα activator. More efficiently than berberine, B-12 bound and
altered the conformation of RXRα/LBD, thereby suppressing Wnt/β-catenin pathway and colon
cancer cell growth via RXRα mediation. In addition, B-12 not only preserved berberine’s tumor
selectivity but also greatly improved its bioavailability. Remarkably, in mice, B-12 did not show
obvious side effects including hypertriglyceridemia as other RXRα agonists, or induce hepatorenal
toxicity. Together, our study describes an approach for the rational design of berberine-derived
RXRα activators as novel effective antineoplastic agents for colon cancer.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
■
INTRODUCTION
Colon cancer is the third most common malignancy and the fourth leading cause of cancer-
related deaths worldwide. It is initiated in the epithelial cells lining the colon or rectum and
frequently results from Wnt/β-catenin signaling pathway overactivation that is driven by mutations
1
in pathway components, such as adenomatous polyposis coli (APC) and β-catenin. In the past two
decades, the survival rate has increased significantly profiting from multiple new therapeutic
methods. However, the 5-year disease-specific overall survival rate is still lower than 60% due to
tumor recurrences and drug resistances.2, 3 New agents targeting important molecules in colon
cancer development, with potent antitumor efficacy and low toxicity, is still urgently needed.
ACS Paragon Plus Environment
2

Page 3 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Retinoid X receptor α (RXRα), a unique member of the nuclear receptor superfamily,
participates in a wide range of cellular processes under physiological and pathophysiological
conditions through modulating gene expression in a ligand-dependent manner. The abnormal
expression or post-translational modification of RXRα is closely related to the development of
several types of cancers, such as colon cancer.4-6 Hyperphosphorylation of RXRα, resulting in its
transactivation inhibition, was detected in a series of human colon cancer cell lines and primary
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
8
human colon cancer tissues. RXRα expression was also downregulated in APC mutant mice.
Proteomics analyses demonstrated that more than 70% of the significantly down-regulated proteins
9
are associated with abnormal RXRα signaling pathway in colon cancer tissues. Additionally,
extensive crosstalk between RXRα and Wnt/β-catenin signaling pathway was also observed in
colon cancer.10, 11 Suppression of RXRα and aberrant β-catenin expression significantly associate
with progression of colon cancer.12, 13 RXRα agonists increased β-catenin and RXRα protein
interaction, leading to the degradation of β-catenin and the inhibition of cell growth in colon
cancer.14, 15 Therefore, RXRα is considered to be an important target for colon cancer treatment.
Identifying and optimizing molecules that bind to its canonical ligand-binding pocket (LBP) has
long been the focus of drug discovery efforts targeting RXR, and a large pool of such RXR
ligands have been reported, such as 9-cis-retinoid acid (9-cis-RA), targretin (LGD1069) and
sulindac.16, 17 However, there are key limitations of treatment with these ligands including
unwanted side effects such as hypertriglyceridemia, mucosal toxicity, hypercalcemia, and
hepatomegaly.4, 16-18 The current challenge is to discover selective RXR modulators with desired
pharmacological activities and minimal undesired side effects.4, 16, 17
Berberine (Figure 1a), an isoquinoline alkaloid isolated from several Chinese herbs such as
Coptis chinensis, is a nonprescription medicine traditionally used to treat gastroenteritis. In recent
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
years, pharmacological and clinical studies have displayed a broad range of benefits from
berberine treatment, including antidiabetic, antihyperlipidemic, and antitumor effects.19-22 In
several cancer models, berberine has been proven to reduce cell proliferation, induce cell apoptosis
and inhibit tumor angiogenesis and metastasis, which makes berberine a promising drug for cancer
therapy.23, 24 Notably, berberine is a radiosensitizer of tumor cells but not normal, untransformed
cells, and also displays high selectivity for colon cancer cells relative to normal colon epithelial
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
25
cells. Our recent study demonstrated that berberine acted as a novel RXRα activator to suppress
β-catenin signaling and cell growth in colon cancer through directly binding to a unique site (Site
X) on the ligand-binding domain (LBD) of RXRα, which is likely the main reason for the tumor
selectivity of berberine and thus provided new strategies for the design of new RXRα-based
antitumor agents.26 However, some limitations of berberine, including poor solubility, low
bioavailability, and moderate tumor-suppressing activity,27, 28 restrict its clinical application to
treat colon cancer. Structural modification of berberine would be an effective way to solve this
dilemma. Some modifications of berberine have been carried out trying to improve its bioactivity
and bioavailability in antimicrobial, antidiabetic, antihyperlipidemic and antitumor effects.
Modification at C8 and C13 greatly increased the antimicrobial activity of berberine, which is
closely related to the substituent chain length.29-31 Most of the berberine derivatives with halogen
and heteroaromatics on C8, C9 position displayed stronger hypoglycemic ability than berberine.32,
3
3
Structure-activity studies on the antitumor effect of berberine mainly focused on the C8, C9, and
C13 positions; for example, 13-alkyl derivatives of berberine showed higher antitumor activities
than berberine, and a moderate alkyl chain length was beneficial for its antitumor activity.34-36
However, almost all these modifications are randomly made and effect-oriented, lacking a
structural basis for the mechanism of berberine’s effects. Most of these derivatives, by inserting
ACS Paragon Plus Environment
4

Page 5 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
either a long aliphatic chain or a heterocyclic ring, may have significant effects but lose the
biosafety of berberine since the structures were changed too much. Moreover, these derivatives
were mostly accessed by modification of natural berberine, thus significantly restricting the
structural diversity necessary for structure-activity relationship (SAR) studies.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Our previous studies identified RXRα as an intracellular target of berberine to suppress colon
2
6
cancer growth and disclosed a unique binding mode of berberine with RXRα, which provides an
ideal theoretical base for SAR design of new anti-tumor berberine analogues. In the present study,
based on our previous findings, we designed and prepared 15 berberine analogues (nine of them
are new compounds) in a total synthesis manner with modifications on the C2/3/9/10/11/12
positions of berberine, in an attempt to find a molecule possessing similar tumor-selectivity to that
of berberine but with enhanced antitumor activity and bioavailability.
■
RESULTS
Design and synthesis of berberine analogues targeting RXRα
We described recently how berberine (Figure 1a) acts as an RXRα activator through direct
binding to the Site X (non-classical ligand-binding pocket) in the ligand-binding domain (LBD)
of RXRα to activate its transactivating function and its regulation on Wnt/β-catenin signaling to
inhibit colon cancer growth; the C2, C3 groups on the head of berberine form H-bonds with the
Arg371 residue of RXRα, and the C9, C10 methoxy groups on the tail region of berberine form
2
6
Van Der Waals contacts with RXRα (Figure 1b). To obtain more efficient RXRα activator and
overcome aforementioned drawbacks residing in berberine as a promising antineoplastic agent, we
designed 15 berberine analogues with modifications on C2, 3, 9, 10, 11 and 12 positions based on
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the mode of berberine binding to RXRα. The structures of these analogues were shown in Table
1
.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1. Chemical structure of berberine (a) and the binding site in RXRα-LBD (b)
(a) Chemical structure of berberine. Red arrows indicated the sites to be remodeled. (b) Detailed
view of the interaction between berberine (yellow) and RXRα/LBD (cyan) (PDB:3FUG) in Site
X. Key residues involved in berberine binding were shown in sticks; potential hydrogen-bonding
interaction between berberine and RXRα-LBD residues Arg371 was shown as dashed red line.
3
7
The analogues of berberine were prepared according to Donohoe’s protocol and a general
synthetic sequence was illustrated in Scheme 1. The acetophenone 3 could be synthesized from
the substituted phenylacetic acids through reduction followed by the protection of the resulting
alcohol as a pivalate and a regioselective Friedel-Crafts acylation reaction with Ac O and ZnCl .
2
2
Aryl bromide 6 was prepared from the corresponding o-bromobenzaldehyde through acetal
protection. Then in the presence of [(Amphos) PdCl ] and Cs CO , the cross-coupling between
2
2
2
3
acetophenone 3 and aryl bromide 6 delivered 7. Finally, the desired berberine analogue was formed
by heating the mixture of compound 7 and NH Cl in EtOH/H O. Through this reaction sequence,
4
2
15 berberine analogues, named B-1 to B-15, were synthesized smoothly (Figure 2b). Among them,
ACS Paragon Plus Environment
6

Page 7 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
B-2 (pseudocoptisine),38 B-4 (pseudoberberine),39 B-5 (pseudopalmatine),40 and B-11
(
palmatine)41 are natural products and B-142 and B-942 are known berberine analogues. The
structure and purity (above 95% pure) of all berberine analogues were confirmed by qNMR
analysis techniques. Berberine and its 15 analogues were screened by PAINS-remover
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(http://www.cbligand.org/PAINS) to filter out potential false-positive molecules and all these
analogues were proved to be PAIN-less.
Reagents and conditions: (a) BH ·HF (2.0 equiv), THF, 0 °C→rt; (b) PivCl (2.0 equiv), pyridine
3
(
2.0 equiv), CH Cl , 0 °C→rt; (c) ZnCl (5.0 equiv), Ac O, 0 °C→rt; (d) ethylene glycol (5.0
2 2 2 2
equiv), PTSA (0.1 equiv), HC(OEt) (3.0 equiv), toluene, 90 °C; (e) [(Amphos) PdCl ] (5 mol%),
3
2
2
Cs CO (2.5 equiv), THF, 90 °C; (f) NH Cl, EtOH/H O (3:1), 90 °C then 110 °C.
2
3
4
2
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 1. Structures of the berberine analogues
R3
R4
Cl
N
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
R5
R6
R1
R2
Compound
R1
R2
R3
R4
R5
R6
B-1
OCH O
H
H
H
H
H
H
2
B-2
OCH O
OCH O
2
2
B-3
OCH3
H
H
OCH3
OCH3
H
B-4
OCH O
H
OCH3
OCH3
H
2
B-5
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
H
H
OCH3
OCH3
H
B-6
H
H
H
H
H
H
B-7
H
H
B-8
H
H
OCH O
H
2
B-9
OCH O
H
OCH3
OCH3
OCH3
H
H
H
H
H
H
H
H
H
2
B-10
B-11
B-12
B-13
B-14
B-15
OCH3
OCH3
OCH3
OCH3
OCH3
H
OCH3
OCH3
OCH3
OCH3
H
H
OCH3
H
H
H
OCH3
OCH3
H
H
H
OCH3
OCH3
OCH O
H
H
2
ACS Paragon Plus Environment
8

Page 9 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Structure-activity analyses identified B-12 as the most effective RXRα activator among these
berberine analogues
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. Structure-activity analyses identified B-12 as the most effective RXRα activator among
these berberine analogues. (a-d) Close-up view of the berberine analogues binding to RXRα-LBP
(
PDB code 3FUG). Compounds B-1, B-3, B-9, and B-12 were shown in yellow; and compounds
B-4, B-5, B-6 and Berberine (Ber) were indicated in gray. Key residues involved in RXRα
binding were shown in sticks. Potential hydrogen-bonding interaction between berberine
analogues and RXRα/LBD Arg371 was shown as a dashed red line. (e) KM12C cells were co-
transfected with pG5luc plasmid and pBIND-RXRα-LBD or its mutants. Twenty hours after
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
transfection, cells were treated with 50 μM berberine or B-12; then the activities of the reporter
gene were determined by luciferase assays and the basal level was normalized to 1. (f) KM12C
cells co-transfected with pBIND-RXRα/LBD and pG5luc were treated with berberine or its
analogues at 50 μM for 15 h.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
With these compounds in hand, we carried out SAR studies by evaluating binding affinity to
RXRα through molecular docking studies and further by biological tests (Figure 2 and Table 2).
Firstly, the docking calculations suggested that the compounds with 2,3-methylenedioxy group
have much better binding conformation than that with 2,3-dimethoxy group. As shown in Figure
2a, the C3-oxygen of B-1 formed two H-bonds with the backbone of Arg371 in RXRα/LBD, while
B-6 could not form any effective interactions with the surrounding amino acid residues. We next
found that the single C3 methoxy group of B-3 is also beneficial for the compound to bind RXRα-
LBD, while the C2 methoxy group in compound B-5 seemed to hamper the compound to enter the
binding-pocket and form H-bond with Arg371 of RXRα by comparing B3 and B5 (Figure 2b),
which might result from the larger steric hindrance induced by the coexistence of C2 and C3
methoxy groups. Next, we investigated the roles of functional groups present on the C9/10/11/12
positions (the tail) of berberine core when the compounds have the same head structure. By
comparing B-4 with B-9, we found 10,11-dimethoxy groups seemed to impede the binding of the
compounds with RXRα/LBD, because even though B-4 has a methylenedioxy group at C2-C3
positions, which is more favorable for binding to RXR protein, it is still almost outside the pocket
(Figure 2c). Similar docking results mentioned above were obtained when comparing B-14 and B-
1
5, B-12 and B-13, B-2 and B-4, respectively (Figure S1a-c). It is worth noting, when the heads
of compounds bear C3 single methoxy or 2,3-methylenedioxy groups, some analogues with
ACS Paragon Plus Environment
1
0

Page 11 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
substituents on the C9/10/11 positions of berberine, except the analogues with bismethoxy groups
on C10 and C11 (B-3 and B-4), could well match the binding-pocket and form H-bonds with
Arg371, such as B-2, B-8, B-9, B-10 and B-12 (Figure S1d).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
For further verification, we ranked the binding potential of berberine analogues with
RXRα/LBD through molecular docking by four indexes, including shape matching, H-bond
energy, protein solvation, and small molecule solvation. As shown in Table 1, the five top-scoring
compounds are B-12, B-8, B-9, B-10 and B-2, which is consistent with the above conformational
analysis. Next, the physiological consequences of the interactions between these compounds and
RXRα/LBD were validated by their abilities to transactivate RXRα (Figure 2f), as well as to inhibit
colon cancer cell growth, which were illustrated as IC in Table 2. The dose-response curves of
5
0
representative compounds (Ber, B-8, B-9, B-10, B-12), along with 9-cis-RA as positive control,
were shown in Figure S1e. The above results illustrated that compounds with higher binding
potential to RXRα as indicated in molecular docking analysis indeed had a stronger ability to
stimulate the transactivation activities of RXRα/LBD and inhibit the growth of colon cancer cells,
such as B-8, B-9, B-10 and B-12; reversely, those showing poor binding potential exhibited little
effect on the RXRα-LBD transactivation activity and inhibition of colon cancer cell proliferation,
including B-4, B-5, B-6, B-7, B-13, and B-14. Among all these compounds, B-12, as expected,
exhibited the strongest transcriptional activation ability to RXRα and the greatest inhibitory effect
on colon cancer cell growth (Figure 2f and Table 2). B-12 has the lowest IC value to KM12C
5
0
cell growth (31.10 μM), which was much smaller than that of berberine (76.34 μM). Based on
these results, we further explored the binding conformation of compound B-12, and the results
indicated that compound B-12 manifested a better binding conformation with RXRα/LBD (Figure
2
d), and its H-bond lengths with Arg371 of RXRα (3.0 Å, 3.11 Å) was far shorter than that of
ACS Paragon Plus Environment
1
1

Journal of Medicinal Chemistry
Page 12 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
berberine (3.9 Å, 3.2 Å). Subsequent mutation analysis confirmed that Arg371 is essential for the
binding of RXRα/LBD with both B-12 and berberine (Figure 2e). We further confirmed the
superiority of B-12 to berberine on cancer cell growth and RXR-dependent transactivation in
another two colon cancer cell lines HCT116 and SW620. Results showed that B-12 exhibited
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
excellent IC values in the growth inhibition of HCT116 and SW620 cells (27.17 μM and 36.19
50
μM, respectively), which were less than one-third of that of berberine (83.09 μM and 126.20 μM,
respectively) (Figure S2a). B-12 also acted much better than berberine to suppress colon cancer
cell growth indicated by EdU (5-Ethynyl-2’-deoxyuridine) assay and PCNA (proliferating cell
nuclear antigen, a cell proliferation marker) expression level (Figure S2b-c), and to enhance the
transactivation activity of endogenous RXRα on RXR DNA response element (RXRE) in KM12C,
HCT116 and SW620 cells (Figure S2d).
Table 2. Berberine analogues’ molecular docking conformational analysis score sheet and
their inhibitory effects (IC ) on KM12C cell proliferation
50
Total
Score
5.42
5.60
5.38
4.84
Hydrogen
Bond
Ligand
Desolvation
1.03
Protein
Desolvation
6.15
IC50
(μM)
79.70
57.18
83.50
467.00
Compound
Shape
B-1
B-2
B-3
B-4
-0.92
-11.68
-13.55
-12.35
-8.45
-0.94
1.37
7.39
-0.98
0.65
7.30
-0.96
1.28
6.68
B-5
B-6
4.64
4.00
-0.21
-0.49
-10.67
-10.66
0.76
1.18
6.13
5.96
＞600
＞600
B-7
B-8
B-9
5.39
5.90
5.81
-0.83
-0.69
-0.81
-11.47
-13.36
-12.52
0.60
1.17
0.71
6.30
6.97
6.82
128.80
46.37
52.69
ACS Paragon Plus Environment
1
2

Page 13 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
B-10
B-11
B-12
B-13
B-14
B-15
Ber
5.77
3.78
6.01
3.59
3.94
4.88
5.53
-0.98
-0.89
-0.95
-0.11
-0.47
-0.81
-0.90
-13.39
7.83
1.04
1.31
0.90
1.06
0.97
1.35
1.37
7.35
-11.58
7.09
37.88
393.00
31.10
-13.39
-10.90
-9.10
6.35
166.00
270.00
118.40
76.34
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3.97
-11.56
-13.24
6.14
7.39
Evaluation of B-12 binding ability with RXRα
Molecular docking and bio-activity analysis above picked B-12 out as the most distinguished
activator of RXRα among these 15 newly-synthesized berberine analogues (Figure 2). Several
biophysicochemical methods were then employed to examine the binding ability of B-12 with
RXRα. Firstly, we determined the physical interaction between B-12 and RXRα/LBD by
1
saturation transfer difference (STD)-NMR spectroscopy. As shown in Figure 3a, the H NMR
spectra with and without B-12 were obtained, and most signals were observed in STD-NMR
spectra showing that RXRα/LBD protein interacts with B-12. We then used fluorescence
quenching analysis to evaluate the binding affinity of B-12 to RXRα/LBD. His-RXRα/LBD
displayed maximal fluorescence at 332 nm (Figure 3b, left panel). When the concentration of B-
12 or berberine was increased from 0 to 30 μM, the fluorescence intensity of RXRα/LBD gradually
decreased, without a significant shift of the maximum fluorescence emission wavelength (Figure
3
b, left panel). Plotting of the data yielded titration curves (Figure 3b, right panel), from which the
-5
-5
apparent dissociation constants (K ) value was calculated as 1.89×10 M and 2.48×10 M for B-
d
1
2 and berberine binding to RXRα/LBD proteins, respectively. Finally, CD spectroscopic assay
was performed to test whether B-12 could induce a change in the conformation of His-RXRα/LBD.
The far-UV CD spectrum of His-RXRα/LBD exhibited two negative peaks at 208 nm and 222 nm
ACS Paragon Plus Environment
1
3

Journal of Medicinal Chemistry
Page 14 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
and a positive peak at 192 nm, suggesting the presence of α-helical structure (Figure 3c). Upon B-
1
2 binding, this conformation of His-RXRα/LBD was significantly changed, to a greater extent
than that induced by berberine (Figure 3c). These results indicated that B-12 directly interacted
with RXRα/LBD and altered its conformation, and the binding affinity of B-12 with RXRα/LBD
was stronger than that of berberine.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3. B-12 directly bound to RXRα/LBD. (a) STD spectrum of B-12 and RXRα/LBD. The
1
1
green spectra represented the H NMR spectra of B-12 alone. The red spectra represented the H
NMR spectra of B-12 and RXRα incubated together. (b) Fluorescence quenching spectra (left)
ACS Paragon Plus Environment
1
4

Page 15 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
and Stern-Volmer plot (right) for RXRα/LBD-berberine/B-12 interaction. (c) CD spectroscopic
analysis of the binding of berberine and B-12 (5 μM) to RXRα/LBD (1 μM) with consequent
conformational changes.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment
1
5

Journal of Medicinal Chemistry
Page 16 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 4. B-12 inhibited β-catenin signaling and cell proliferation in colon cancer cells. (a, b)
Effects of berberine and B-12 on β-catenin signaling and β-catenin expression were respectively
detected by TOP/FOP luciferase system and western blot. KM12C, HCT116, and SW620 cells
were co-transfected with pRL-TK, TOPflash or FOPflash plasmids (for a). (c, d) Effect of
berberine or B-12 on the binding ability of RXRα with β-catenin or and β-catenin with c-Cbl was
detected by the mammalian two-hybrid assay. KM12C, HCT116, and SW620 cells were co-
transfected with pBIND-β-catenin/S33Y and pACT-RXRα (for c) or pBIND-β-catenin/S33Y and
pACT-c-Cbl (for d) and then treated with berberine or B-12 for 15 h. (e, f) Western blot analyses
and real-time PCR assays were performed to detect the protein/mRNA expression of Cdc2, c-Myc,
and p21WAF1/CIP1 in KM12C, HCT116, and SW620 cells. Significant differences compared with
the vehicle were indicated as **p<0.01 and ***p<0.001. Significant differences of B-12 vs Ber at
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
the same dose were indicated as p<0.05, p<0.01 and ###p<0.001.
#
##
B-12 inhibited β-catenin signaling and colon cancer cell growth via RXRα mediation
We reported recently that berberine suppressed colon cancer cell proliferation through binding
to RXRα and promoting its interaction with nuclear β-catenin, leading to c-Cbl-mediated
2
6
degradation of β-catenin. Therefore, we examined the effect of B-12 on Wnt/β-catenin signaling
pathway in KM12C, HCT116, and SW620 colon cancer cell lines. B-12 inhibited β-catenin-
dependent transcriptional activity and induced degradation of β-catenin protein more obviously
than berberine, as indicated by the TOP/FOP luciferase assay and western blot analysis (Figure 4a,
b). Mammalian two-hybrid assay further demonstrated that the interaction of RXRα with β-catenin
and β-catenin with c-Cbl was enhanced by B-12 more effectively than berberine (Figure 4c, d). In
ACS Paragon Plus Environment
1
6

Page 17 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
addition, the expression levels of Cdc2 and c-Myc, two target genes of the β-catenin signaling
pathway involved in regulating the G2/M checkpoint, and p21WAF1/CIP1, a downstream effector
4
3
4
4
of c-Myc for G2/M checkpoint regulation, were regulated by B-12 more significantly than
berberine (Figure 4e, f). These results suggested that similar to berberine, B-12 also promotes
RXRα–β-catenin interaction and inhibits β-catenin signaling and colon cancer cell growth, but its
regulatory effect is much stronger than that of berberine.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
We further investigated whether RXRα is essential for B-12 to suppress the hyperactivated β-
catenin signaling and cell growth in KM12C cells. We stably knocked down the expression of
RXRα in KM12C cells with a specific shRNA against RXRα (Figure 5a). Silencing RXRα
expression greatly reduced the ability of B-12 to inhibit cell growth, as detected by the MTT assay
(Figure 5b). Our results also showed that the expression of β-catenin, Cdc2, c-Myc, and
p21WAF1/CIP1 were regulated by B-12 in an RXRα-dependent manner (Figure 5c and Figure S3).
These data suggested that RXRα mediated the inhibitory effect of B-12 on β-catenin signaling and
cell growth in colon cancer cells.
ACS Paragon Plus Environment
1
7

Journal of Medicinal Chemistry
Page 18 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. B-12 suppressed β-catenin signaling and cell proliferation in an RXRα-dependent
manner. (a) KM12C cells were transfected with control shRNA or shRNA targeting RXRα, then
the expression levels of endogenous RXRα were examined by western blotting analyses. (b, c)
KM12C cells expressing control shRNA or shRNA targeting RXRα were treated with indicated
concentrations of berberine or B-12 for 15 h, then the cell viability was measured by MTT assays
(
b) and the protein levels of Cdc2, c-Myc and p21WAF1/CIP1 were examined by western blot analysis
(c).
B-12 inhibited tumor growth in vivo
We assessed the therapeutic effects of B-12 in a xenograft tumorigenesis model by
subcutaneously injecting KM12C cells into nude mice. When the volume of xenografts had
3
reached 80 to 200 mm , the mice with similar tumor load were divided into three groups, and then
respectively treated with vehicle, B-12 (100 mg/kg) and berberine (100 mg/kg), each day through
ACS Paragon Plus Environment
1
8

Page 19 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
intragastric administration. Two weeks later, the tumors were dissected and weighed. Tumor
growth was significantly reduced in B-12-treated and berberine-treated groups as compared to the
vehicle group, and the inhibition rate was 69.1% (P < 0.01) and 33.6% (P < 0.05) for B-12 and
berberine respectively (Figure 6a-c). B-12 also regulated the protein and mRNA expression of
PCNA, Ki67, Cdc2, c-Myc, p21WAF1/CIP1 and β-catenin in the xenograft tumors, more effectively
than berberine, as detected by western blot, real-time PCR and immunohistochemistry (Figure 6d-f
and Figure S4, 5). These results indicated that B-12 inhibited colon cancer cell growth more
efficiently than berberine in vivo.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Unwanted side effects, such as hypertriglyceridemia, mucosal toxicity, hypercalcemia and
hepatomegaly, are the major obstacles of the clinical applications of current RXRα ligands. So we
also tested the possible toxicity or side effects of B-12. Berberine and B-12 had no obvious
hepatotoxicity and nephrotoxicity as indicated by the biochemical indexes of liver and renal
function measured from the serum of mice after drug treatment (Table 3). The conclusion was
further supported by the results from HE staining as no damaged hepatocellular morphology and
augmented necrotic hepatocytes were observed (Figure 6g). The liver to body weight ratio also
proved that berberine and B-12 did not cause hepatomegaly (Figure S6). Moreover, unlike the
classical RXRα agonists, B-12 did not cause unwanted side effects such as hypertriglyceridemia,
hypercalcemia and hypercholesterolemia (Table 3).
ACS Paragon Plus Environment
1
9

Journal of Medicinal Chemistry
Page 20 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6. B-12 inhibited tumor growth in mice. Mice xenografted with KM12C cells were treated
with or without berberine or B-12 as described in the method. (a) Tumor volume was measured
every 2 days. (b-c) Two weeks later, mice were sacrificed and the xenograft tumors were dissected
and weighed. Representative images of tumors were shown in b. (d) Expression levels of β-catenin,
Cdc2, c-Myc, and p21WAF1/CIP1 in xenografts were analyzed by western blot. (e, f) Representative
images of the expression of β-catenin, Ki-67, Cdc2, c-Myc, PCNA and p21WAF1/CIP1 in xenograft
tumors analyzed by immunohistochemistry. (g) Liver and kidney histopathology by HE staining.
ACS Paragon Plus Environment
2
0

Page 21 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
Table 3. Biochemical analyses of the serums from berberine or B-12 treated mice
Biochemical
Vehicle
(n=6)
Ber
B-12
P values
Ber
P values
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
indexes
(n=6)
(n=6)
VS B12 VS Vehicle
Vehicle
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
TBIL (μmol/L)
IBIL (μmol/L)
ALT (U/L)
0.28±0.024
0.55±0.096
24.18±0.77
157.15±8.41
76.53±4.31
27.02±1.22
8.21±0.26
0.26±0.032
0.40±0.057
27.90±1.21
177.86±7.88
71.28±3.16
24.36±1.68
8.08±0.62
0.21±0.029
0.52±0.083
27.20±1.57
158.36±6.68
78.35±5.99
28.95±1.40
8.39±0.41
0.4713
0.9617
0.2423
0.2958
0.9143
0.8739
0.6713
0.7600
0.8605
0.1635
0.2239
0.8728
0.4813
0.8449
AST (U/L)
ALP (U/L)
CREA (μmol/L)
UREA
(mmol/L)
Ca (mmol/L)
TG (mmol/L)
TC(mmol/L)
2.27±0.059
0.96±0.056
2.32±0.11
2.13±0.034
0.77±0.069
2.35±0.12
2.23±0.021
0.74±0.031
2.49±0.087
0.1413
0.1241
0.9908
0.9079
0.0574
0.9991
Data were presented as the mean ± SEM. Statistical significance is analyzed by one-way analysis
of variance (ANOVA). P <0.05 was considered statistically significant.
Detection of tumor selectivity and bioavailability of B-12
Berberine has a significant advantage of high tumor selectivity (killing tumor cells but not
normal cells), which makes it a promising anti-tumor drug. Therefore, we tested whether B-12
could still retain this advantage of berberine, by detecting the cell viability in response to a series
of concentrations of B-12 and berberine in human normal intestinal epithelial cell line HIEC-6 and
ACS Paragon Plus Environment
2
1

Journal of Medicinal Chemistry
Page 22 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
mouse fibroblast epithelial cell line MEF. Treatment with B-12 or berberine caused only a slight
decrease in viability of HIEC-6 and MEF cells, suggesting that B-12 still preserved the high tumor
selectivity of berberine (Figure 7a-b). Poor solubility and low bioavailability are two major
weaknesses of berberine. Excitingly, the solubility of B-12 was increased to 10.66 times compared
with berberine (Supplementary Table 1). The increasing solubility of B-12 is probably due to the
decreasing intermolecular π, π-interaction caused by the more bulky C-3 methoxy group. We thus
test whether the bioavailability of B-12 is also better than berberine. SD rats were treated with
berberine (100 mg/kg) or B-12 (100 mg/kg) by gavage, then the blood was taken from the tail vein
at different periods, and the drug concentration in the blood was detected by LC-MS. The results
showed that both berberine and B-12 reached the maximum drug concentration at approximately
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
hours, but the maximum drug concentration of B-12 was 167 times higher than that of berberine
in the blood of rats (Figure 7c and Table 4). The results of the pharmacokinetic parameters of
berberine and B-12 were shown in Table 4. Compared with berberine, the peak concentration
(Cmax) was increased to 167.5 times; the area under the curve (AUC) was increased to 21.25
times; the bioavailability (F%) of B-12 was raised to 26.10 times (Table 4), which might also
contribute to B-12’s stronger in vivo inhibitory effect on tumor growth. These results indicate that
B-12 not only reserves the high tumor selectivity of berberine but also greatly improves its
bioavailability.
ACS Paragon Plus Environment
2
2

Page 23 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Figure 7. Detection of tumor selectivity and bioavailability of B-12. (a, b) MEF and HIEC-6 cells
were treated with berberine or B-12 for 15 h. The viability of cells was determined by MTT assay.
(c) Berberine (100 mg/kg) or B-12 (100 mg/kg) were given to SD rats by intragastric
administration, and plasma samples were collected at 0.038, 0.25, 0.5, 1, 2, 4, 6, 8, 12, 24 and 36
h for analyses of berberine and B-12 by LC-MS.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 4. The pharmacokinetic parameters of berberine
and B-12 (PO 100mg/kg)
Compound
T (h)
Ber
B-12
11.59
2.000
486.0
2104
15.67
14.20
28.98
2.000
2.893
1
/2
Tmax (h)
Cmax (ng/mL)
AUC (ng·h/mL) 99.00
0
-t
MRTINF (h)
F%
31.51
0.544
■ DISCUSSION AND CONCLUSIONS
Many studies have identified RXRα as an important target for cancer therapy, especially in colon
cancer.5, 45-48 We recently reported that berberine inhibits colon cancer cell growth through direct
binding to a unique site (non-classical ligand-binding pocket) in RXRα/LBD to activate its
transactivating function and its regulation on β-catenin pathway.26 To obtain more efficient
berberine analogues as RXRα activators, which are very likely to retain the advantage of berberine,
we designed and synthesized a series of novel berberine analogues with modifications on the C2,
ACS Paragon Plus Environment
2
3

Journal of Medicinal Chemistry
Page 24 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
2
, 9, 10, 11 and 12 positions of berberine core based on its binding mode with RXRα (Figure 1-
). By analyzing the structure-activity relationship of berberine analogues and RXRα, we
identified B-12 as an effective analogue with stronger anticancer activity, comparative tumor
selectivity, and significantly improved bioavailability, notably with no adverse side effects caused
by traditional RXRα agonists, and no obvious toxicity to mice, which made it a promising
therapeutic agent for colon cancer treatment. (Figure 2-7).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Through structure-bioactivity analysis, we have summarized the important relationships
between the structural components of berberine analogues and their binding and activation of
RXRα. First, the presence of particular groups on the C2, C3 positions (the head moiety) of
berberine to form H-bonds with the Arg371 residue of RXRα may be required. Single C3 methoxy
group or C2-C3 methylene-dioxy group can easily and closely interact with RXRα through
formation of H-bonds with RXRα Arg371, and consequently exhibiting superior bio-activities in
the activation of RXRα and the inhibition of colon cancer cell growth; moreover, single C3
methoxy group is a little more favorable than the C2-C3 methylenedioxy group probably due to
its less steric hindrance; however, the coexistence of C2 methoxy and C3 methoxy groups with
larger steric hindrance seems to prevent the compound from approaching and forming H-bonds
with RXRα Arg371, thereby failing to activate RXRα and suppress colon cancer. Second, the
presence of single C9 or C10 methoxy group, C10-C11 methylenedioxy group or C9/C10
bismethoxy groups on the tail region of berberine may be necessary to make maximal interaction
with the hydrophobic groups in the Site X of RXRα/LBD through Van Der Waals contacts; while
the presence of 10,11-dimethoxy groups or single C12 methoxy group led to poor RXRα binding
potential and biological activity, which may be attributed to the unfavorable steric effect. These
features may help guide the further improvement of current compounds.
ACS Paragon Plus Environment
2
4

Page 25 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
As we mentioned above, berberine binds RXRα in a unique mode. Unlike other RXRα ligands
such as 9-cis-RA, berberine lacks a classic carboxyl group which is crucial for 9-cis-RA to interact
with RXRα through forming a salt bridge with Arg316 in the classical ligand-binding mode of
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
49
RXRα . These distinct structures and binding modes of berberine and 9-cis-RA lead to different
conformational changes of RXRα and consequently some diverse biological functions.
Furthermore, the concomitant binding of berberine and 9-cis-RA to RXRα/LBD is possibly the
main reason for their synergistic effect on RXRα-dependent transcriptional activity and colon
cancer cell growth, suggesting a great potential for the combination of berberine and 9-cis-RA to
2
6
treat colon cancer . Indeed, combination therapy with different drugs for cancer has long been a
successful way to increase drug efficacy, reduce toxicity and reverse drug resistance. For instance,
clinical trials have proven that the combination of all-trans retinoic acid and arsenic trioxide is a
5
0
feasible treatment in acute promyelocytic leukemia with a higher cure rate and less relapse . In
the present study, B-12 combined with 9-cis-RA had a more significant synergistic effect than
berberine with 9-cis-RA (Figure S7), suggesting that B-12 might be a better drug than berberine
for combined drug therapy. Besides, in mouse model we also found that, unlike other RXRα
ligands, B-12 did not show side effects such as hypertriglyceridemia, hypercalcemia, and
hypercholesterolemia (Table 3), which are the main reasons for the limited use of the existing
RXRα agonists in clinical practice; moreover, B-12 did not induce obvious hepatorenal toxicity or
hepatomegaly as most anti-tumor drugs did (Table 3, Figure 6 and Figure S6). Further studies,
including in vitro and in vivo experiments, are still required to elucidate its potential.
In conclusion, this study describes a novel SAR design of berberine analogues based on the
structural features of berberine analogues as RXRα activators, which might serve as a theoretical
basis for designing and evaluating the next generation of berberine-based drugs against colon
ACS Paragon Plus Environment
2
5

Journal of Medicinal Chemistry
Page 26 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
cancer. Especially, we developed B-12 as a better RXRα activator than berberine, with stronger
anticancer activity, higher bioavailability, and comparative tumor selectivity.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
■
EXPERIMENTAL SECTION
Chemistry
All reactions were carried out under an argon atmosphere with dry solvents under anhydrous
conditions unless otherwise stated. All the chemicals were purchased commercially and used
without further purification unless otherwise stated. The boiling point of petroleum ether (PE) is
between 60-90 ºC. Tetrahydrofuran (THF) was distilled from sodium-benzophenone;
Dichloromethane (CH Cl ) was distilled from calcium hydride. Toluene was distilled from sodium
2
2
under argon atmosphere. Reactions were monitored by thin-layer chromatography (TLC) carried
out on 0.25 mm Qingdao silica gel plates (F254) using UV lights as the visualizing agent and
KMnO . Flash column chromatography was performed over Qingdao silica gel (200-300 mesh).
4
Infrared spectra were recorded on a Nicolet AVATER FTIR330 spectrometer as thin film and are
-1
reported in reciprocal centimeter (cm ). High-resolution mass spectra (HRMS) were recorded on
a Micromass QTOF2 Quadrupole/Time-of-Flight Tandem mass spectrometer using electron spray
ionization. NMR spectra were recorded on Bruker AV-400 and Bruker AV-500 instruments and
were calibrated using residual undeuterated solvents (CHCl , δ = 7.26; DMSO-d , δ = 2.50;
3
H
6
H
CD OD, δ = 3.31) and deuterated solvents (CDCl , δ = 77.0; CD OD, δ = 49.0; DMSO-d , δ
3
H
3
C
3
C
6
C
1
=
39.5) as internal references. The data for H NMR are reported as follows: chemical shift,
multiplicity (s = singlet, d = doublet, t = triplet, m = multiplet or unresolved, brs = broad singlet),
coupling constants (Hz) and integration. The structure and purity of B1-B15 was determined by
qNMR with triphenylmethane as an internal standard. The purity of key compounds (B-2, B-8, B-
ACS Paragon Plus Environment
2
6

Page 27 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
9
, B-10, B-12) were also analyzed by HPLC. The purity of all the compounds were determined to
be > 95%.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
General Synthetic Procedures for the Berberine Analogues B1-B15 (B-12 as an Example)
2
-(2-Bromo-6-methoxyphenyl)-1,3-dioxolane (5a). With a modified Donohoe’s procedure: To a
stirred mixture of 2-bromo-6-methoxybenzaldehyde (2.15 g, 10.0 mmol) and ethylene glycol (2.78
mL, 50 mmol) in toluene (200 mL) was added HC(OEt) (5.0 mL, 30.0 mmol) and p-TSA (190
3
mg, 1.0 mmol) in sequence. The resultant mixture was heated at 90 C for 18 h and then cooled to
room temperature and extracted with Et O. The combined layers were washed with saturated brine,
2
dried over MgSO and concentrated in vacuo. Purification of the residue by flash chromatography
4
1
(PE/EtOAc = 10:1) afforded 5a as a colorless oil (2.56 g, 99%). H NMR (500 MHz, CDCl ) δ
3
7
4
1
2
.17 – 7.14 (m, 1H), 7.10 (dd, J = 10.6, 5.6 Hz, 1H), 6.83 (dd, J = 8.2, 1.0 Hz, 1H), 6.44 (s, 1H),
13
.27 – 4.19 (m, 2H), 4.01 – 3.96 (m, 2H), 3.78 (s, 3H). C NMR (125 MHz, CDCl ) δ 159.7,
3
+
30.8, 126.1, 123.7, 123.5, 110.7, 101.1, 65.7, 55.9. HRMS (ESI, m/z) calcd for C H BrO , [M] :
1
0
11
3
57.9892, found: 257.9898.
3-Methoxyphenethyl pivalate (2a). To a solution of 3-methyoxyphenylacetic acid (1.66 g, 10.0
mmol) in THF at 0 C was added borane-THF complex solution (20.0 mL, 20.0 mmol). The
mixture was allowed to warm to room temperature and stirred for 2 h before being quenched with
1
M aqueous solution of NaOH. The resulting mixture was extracted with Et O. The combined
2
organic layers were washed with saturated brine, dried over MgSO and concentrated in vacuo.
4
The residue was redissolved in CH Cl (50 mL) and cooled to 0 C. To the mixture was added
2
2
pyridine (0.8 mL, 20.0 mmol) and pivaloyl chloride (2.78 mL, 20.0 mmol) sequentially. The
resulting mixture was allowed to warm to room temperature and stirred for 18 h. The reaction was
ACS Paragon Plus Environment
2
7

Journal of Medicinal Chemistry
Page 28 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
quenched with H O (5 mL) and extracted with CH Cl . The combined organic layers were washed
2
2
2
with saturated brine, dried over MgSO and concentrated in vacuo. Purification of the residue by
4
1
flash chromatography (PE/EtOAc = 20:1) afforded 2a as a colorless oil (2.12 g, 90%). H NMR
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
500 MHz, CDCl ) δ 7.17 (t, J = 7.75 Hz, 1H), 6.78 (d, J = 7.6 Hz, 1H), 6.77 – 6.72 (m, 2H), 4.25
3
1
3
(
t, J = 6.8 Hz, 2H), 3.75 (s, 3H), 2.88 (t, J = 6.8 Hz, 2H), 1.15 (s, 9H). C NMR (125 MHz, CDCl )
3
δ 178.1, 159.5, 139.3, 129.2, 121.1, 114.4, 111.7, 64.6, 54.8, 38.6, 35.0, 26.9. HRMS (ESI, m/z)
+
calcd for C H O , [M+H] : 237.1491, found: 237.1487.
1
4
20
3
2
-Acetyl-5-methoxyphenethyl pivalate (3a). To a solution of 2a (1.89 g, 8.0 mmol) in acetic
anhydride (20 mL) at 0 C was added ZnCl (5.45 g, 40.0 mmol). The mixture was allowed to
2
room temperature and stirred for 18 h. The solution was concentrated in vacuo and then the residue
was redissolved in EtOAc. The mixture was washed with saturated aqueous solution of NaHCO3,
water and brine. The organic phase was dried over MgSO and concentrated in vacuo. Purification
4
of the residue by flash chromatography (PE/EtOAc = 10:1) afforded 3a as a colorless oil (1.66g,
1
7
5%). H NMR (500 MHz, CDCl ) δ 7.78 (d, J = 8.5 Hz, 1H), 6.82 – 6.76 (m, 2H), 4.30 (t, J = 6.6
3
1
3
Hz, 2H), 3.84 (s, 3H), 3.26 (t, J = 6.6 Hz, 2H), 2.56 (s, 3H), 1.14 (s, 9H). C NMR (125 MHz,
CDCl ) δ 199.1, 178.3, 151.4, 146.8, 133.4, 129.3, 114.9, 113.5, 64.8, 55.8, 38.6, 33.6, 29.1, 27.1.
3
+
HRMS (ESI, m/z) calcd for C H O , [M+Na] : 301.1416, found: 301.1410.
1
6
22
4
2
-(2-(2-(1,3-Dioxolan-2-yl)-3-methoxyphenyl)acetyl)-5-methoxyphenethyl pivalate (6a). To a
stirred solution of 5a (596 mg, 2.3 mmol) and 3a (1.28 g, 4.6 mmol) in THF (10 mL) in a 30 mL
vial was added Cs CO (1.87 g, 5.75 mmol) and (Amphos) PdCl (81 mg, 0.115 mmol). The
2
3
2
2
resulting solution was bubbled with argon for a minimum of 10 min and then quickly equipped
with a Teflon septum. The mixture was heated at 90 °C for 18 h. The mixture was then cooled to
room temperature and quenched with water and extracted with Et O. The combined organic layers
2
ACS Paragon Plus Environment
2
8

Page 29 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
were washed with saturated brine, dried over MgSO and concentrated in vacuo. Purification of
4
the residue by flash chromatography (PE/EtOAc = 4:1) afforded 6a as a colorless oil (756 mg,
1
72%). H NMR (500 MHz, CDCl ) δ 7.90 – 7.85 (m, 1H), 7.34 – 7.28 (m, 1H), 6.87 – 6.77 (m,
3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
H), 6.24 (s, 1H), 4.40 (s, 2H), 4.33 (t, J = 6.6 Hz, 2H), 3.84 (s, 3H), 3.81 (s, 3H), 3.80 – 3.75 (m,
1
3
2
H), 3.74 – 3.67 (m, 2H), 3.24 (t, J = 6.6 Hz, 2H), 1.15 (s, 9H). C NMR (125 MHz, CDCl ) δ
3
198.7, 178.5, 161.5, 159.0, 142.2, 137.1, 131.7, 130.3, 129.9, 125.4, 121.2, 117.9, 111.5, 109.9,
+
9
8.7, 64.9, 64.5, 55.8, 55.3, 44.9, 38.7, 34.1, 27.3. HRMS (ESI, m/z) calcd for C H O , [M+H] :
2
6
32
7
4
57.2226, found: 457.2221.
,9-Dimethoxy-5,6-dihydroisoquinolino[3,2-a]isoquinolin-7-ium chloride (B-12). To 6a (723
mg, 1.58 mmol) in a 15 mL vial was added 1 M solution of NH Cl in 3:1 of EtOH/H O (4 mL).
3
4
2
The solution was bubbled with argon for a minimum of 10 min and quickly equipped with a Teflon
septum. The stirred mixture was heated at 90 °C for 18 h. Then the solution was heated at 110 C
for another 48 h. The mixture was cooled to room temperature, and concentrated with silica gel in
vacuo. Purification of the residue by flash chromatography (CH Cl /MeOH = 10:1) afforded a
2
2
crude product. The crude product was recrystallized in methanol/Et O to give B-12 as a pale yellow
2
1
solid (368 mg, 71%). The purity of the title compound was determined to be > 95% by qNMR. H
NMR (500 MHz, DMSO-d ) δ 9.92 (s, 1H), 8.96 (s, 1H), 8.21 (d, J = 8.8 Hz, 1H), 8.11 (t, J = 8.1
6
Hz, 1H), 7.78 (d, J = 8.3 Hz, 1H), 7.38 (d, J = 7.9 Hz, 1H), 7.12 (dd, J = 8.7, 2.6 Hz, 1H), 7.08 (d,
1
3
J = 2.5 Hz, 1H), 4.94 (t, J = 6.3 Hz, 2H), 4.13 (s, 3H), 3.88 (s, 3H), 3.31 – 3.26 (m, 2H). C NMR
125 MHz, DMSO-d ) δ 171.5, 166.7, 155.0, 149.8, 148.8, 147.9, 147.1, 137.7, 128.9, 128.8,
(
6
1
28.2, 127.2, 123.9, 122.7, 118.2, 66.2, 65.2, 64.3, 36.1. HRMS (ESI, m/z) calcd for C H NO ,
19 18 2
+
[M] : 292.1332, found: 292.1329.
ACS Paragon Plus Environment
2
9