Structural optimization of berberine as a synergist to restore antifungal activity of fluconazole against drug-resistant Candida albicans

Article abstract of DOI:10.1002/cmdc.201300332

We have conducted systematic structural modification, deconstruction, and reconstruction of the berberine core with the aim of lowering its cytotoxicity, investigating its pharmacophore, and ultimately, seeking novel synergistic agents to restore the effectiveness of fluconazole against fluconazole-resistant Candida albicans. A structure-activity relationship study of 95 analogues led us to identify the novel scaffold of N-(2-(benzo[d][1,3]dioxol-5-yl)ethyl)-2- (substituted phenyl)acetamides 7 a-l, which exhibited remarkable levels of in vitro synergistic antifungal activity. Compound 7 d (N-(2-(benzo[d][1,3]dioxol- 5-yl)ethyl)-2-(2-fluorophenyl)acetamide) significantly decreased the MIC ₈₀ values of fluconazole from 128.0 μg mL^-1 to 0.5 μg mL^-1 against fluconazole-resistant C. albicans and exhibited much lower levels of cytotoxicity than berberine toward human umbilical vein endothelial cells. Build it better: Structural optimization of berberine led to the identification of the novel scaffold of N-(2-(benzo[d][1,3]dioxol-5-yl) ethyl)-2-(substituted phenyl)acetamides 7 a-l, which exhibited remarkable in vitro synergistic antifungal activity against fluconazole-resistant Candida albicans in combination with fluconazole. Compound 7 d exhibited much lower cytotoxicity than berberine toward human umbilical vein endothelial cells. Copyright

Full text of DOI:10.1002/cmdc.201300332

CHEMMEDCHEM
FULL PAPERS
DOI: 10.1002/cmdc.201300332
Structural Optimization of Berberine as a Synergist to
Restore Antifungal Activity of Fluconazole against Drug-
Resistant Candida albicans
Hong Liu,[b] Liang Wang,[a] Yan Li,[a] Jiang Liu,[d] Maomao An,[c] Shaolong Zhu,[a]
Yongbing Cao,[a] Zhihui Jiang,[a] Mingzhu Zhao,[a] Zhan Cai,[a] Li Dai,[a] Tingjunhong Ni,[a]
Wei Liu,[c] Simin Chen,[c] Changqing Wei,[a] Chengxu Zang,[a] Shujuan Tian,[a] Jingyu Yang,[d]
Chunfu Wu,[d] Dazhi Zhang,*[a] Hua Liu,*[b] and Yuanying Jiang*[a]
We have conducted systematic structural modification, decon-
struction, and reconstruction of the berberine core with the
aim of lowering its cytotoxicity, investigating its pharmaco-
phore, and ultimately, seeking novel synergistic agents to re-
store the effectiveness of fluconazole against fluconazole-re-
sistant Candida albicans. A structure–activity relationship study
of 95 analogues led us to identify the novel scaffold of
N-(2-(benzo[d][1,3]dioxol-5-yl)ethyl)-2-(substituted phenyl)acet-
amides 7a–l, which exhibited remarkable levels of in vitro syn-
ergistic antifungal activity. Compound 7d (N-(2-(benzo[d]-
[1,3]dioxol-5-yl)ethyl)-2-(2-fluorophenyl)acetamide) significantly
decreased the MIC80 values of fluconazole from 128.0 mgmLÀ1
to 0.5 mgmLÀ1 against fluconazole-resistant C. albicans and ex-
hibited much lower levels of cytotoxicity than berberine
toward human umbilical vein endothelial cells.
Introduction
Invasive fungal infections cause significant levels of morbidity
and mortality in immunocompromised patients, including
those with AIDS, undergoing cancer chemotherapy, or having
undergone organ transplantation.[1,2] Candida albicans is a par-
ticularly common pathogen that is capable of causing life-
threatening infections in these patient populations.[2] Flucona-
zole, originally developed in the early 1990s, is the frontline
treatment for systematic C. albicans infections.[3] Unfortunately,
however, the effective use of fluconazole has been limited sig-
nificantly by the development of azole resistance, which has
been attributed to the failure of antifungal treatments in the
clinic.[4–6] One promising approach aimed at overcoming azole
resistance has been reported that involves sensitizing C. albi-
cans toward fluconazole with small molecules, including piper-
azinyl quinolines,[7] amiodarone,[8] tacrolimus,[9] honokiol,[10] and
Allicin.[11]
The recent screening of a natural product library in our re-
search group revealed berberine (1) as the most active syner-
gist in a series of antifungal agents, including fluconazole,[12]
ketoconazole, amphotericin B, and miconazole. As a single
agent, berberine showed weak antifungal activity (MIC80:
32.0 mgmLÀ1) against the fluconazole-resistant isolates tested,
whereas the MIC80 value of fluconazole against the same iso-
lates was 128.0 mgmLÀ1. When the activity of the two agents
in combination was evaluated, the MIC80 value of berberine
was markedly decreased to 1.0 mgmLÀ1, and the MIC80 value of
fluconazole was also decreased to a range of ꢀ0.125–
2.0 mgmLÀ1. Moreover, the occurrence of similar synergistic ef-
fects was also reported against disseminated candidiasis in
vivo for the combination of berberine with amphotericin B.[13]
These synergistic effects indicated the potential for berberine
to be used as an antidrug-resistance agent.
[a] L. Wang,+ Y. Li,+ S. Zhu, Y. Cao, Z. Jiang, M. Zhao, Z. Cai, L. Dai, T. Ni,
C. Wei, C. Zang, S. Tian, Prof. Dr. D. Zhang, Prof. Y. Jiang
School of Pharmacy, Second Military Medical University
325 Guohe Road, Shanghai 200433 (China)
E-mail: zdzhang_yjhx@smmu.edu.cn
jiangyy@smmu.edu.cn
[b] H. Liu,+ Prof. Dr. H. Liu
School of Pharmacy, Jiangxi University of Traditional Chinese Medicine
18 Yunwan Road, Nanchang, Jiangxi 330004 (China)
E-mail: winner@163.com
Berberine itself does not represent a good starting point for
the development of a synergist to be used in combination
with antifungal agents because of its poor properties, includ-
ing its poor solubility,[14] poor rate of absorption,[15] low level of
bioavailability in humans,[16] and high toxicity.[17,18] Furthermore,
the logP value of berberine was determined to be À1.5.[14] We
decided to carry out a series of structural modifications, decon-
struction, and reconstruction of the berberine core, according
to the strategy shown in Figure 1. Pleasingly, the reconstructed
compounds 7a–i restored the effectiveness of fluconazole
[c] M. An, W. Liu, S. Chen
Department of Pharmacology, Tongji University School of Medicine
1239 Siping Road, Shanghai 200092 (China)
[d] J. Liu, Prof. J. Yang, Prof. Dr. C. Wu
School of Life Science and Biopharmaceutics
Shenyang Pharmaceutical University
103 Wenhua Road, Shenyang, Liaoning 110016 (China)
[+] These authors contributed equally to this work.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201300332.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 207 – 216 207

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
10, the designed quaternary am-
monium salts 3a–e were ob-
tained by holding a mixture of
10 at reflux with the correspond-
ing substituted benzyl halides in
acetonitrile (Scheme 2).[19]
The 3,4-dihydroisoquinolin-2-
iums 4a–o were synthesized in
three steps. The initial amidation
of 2-(benzo[d][1,3]dioxol-5-yl)e-
thanamine (15) with a series of
carboxylic acids gave the corre-
sponding intermediates 11 a–d,
which underwent the Bischler–
Napieralski reaction in the pres-
ence of phosphoryl chloride in
toluene at reflux to afford 12a–
d.[20] In the final stage, the nitro-
gen atoms in 12a–d were alky-
lated with benzyl halides to give
4a–o (Scheme 3).[21] Compound
19, shown in Figure 1, was syn-
thesized by holding a solution of
12a and methyl iodide at reflux
in acetonitrile.
The 1,2,3,4-tetrahydroisoqui-
Figure 1. General structures of the designed analogues of berberine.
nolin-2-iums 5a–e were synthe-
sized according to a three-step
against drug-resistant C. albicans, with some of the synthesized
compounds exhibiting similar levels of activity to berberine.
Furthermore, the cytotoxicities of compounds 7b, 7d, and 7e
were much lower than that of berberine.
procedure. Starting material 15 was treated with a variety of
different aldehydes under Pictet–Spengler reaction conditions
to afford compounds 13a and b,[22,23] the nitrogen atoms of
which were alkylated via substitution with alkyl halides to give
corresponding alkylated derivatives 14a–d.[22] Further alkyla-
tion of compounds 14a–d with benzyl bromides gave com-
pounds 5a–e (Scheme 4). Compound 20, shown in Figure 1,
Results
Chemistry
The structural modification pro-
cess began with reduction of the
iminium ion in berberine. Ber-
berine (1) was treated with po-
tassium carbonate, and the re-
sulting mixture was subsequent-
ly reacted with sodium borohy-
dride in a 5% aqueous sodium
hydroxide solution to give dihy-
droberberine 9. In a separate ex-
periment, berberine was treated
with sodium borohydride in
methanol at reflux to give tetra-
hydroberberine 10 in good yield.
Compounds 2a–j were obtained
by holding a mixture of 9 at
reflux with the appropriate
alkyl halide in acetonitrile
(Scheme 1).[19] In the alkylation
Scheme 1. Synthesis of derivatives 2a–j. Reagents and conditions: a) NaBH4, MeOH, K2CO3, 5% NaOH(aq), RT, 1 h;
b) CH3CN, KI, reflux, 4 h.
Scheme 2. Synthesis of derivatives 3a–e. Reagents and conditions: a) NaBH4, MeOH, reflux, 1 h; b) CH3CN, reflux,
reactions of tetrahydroberberine 6 h.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 207 – 216 208

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
Scheme 3. Synthesis of derivatives 4a–o. Reagents and conditions: a) HOBt, DCC, THF, RT, 12 h; b) POCl3, toluene, reflux, 8–12 h; c) CH3CN, KI, reflux, 6 h.
cal and Laboratory Standards In-
stitute, USA (formerly the Na-
tional Committee for Clinical
Laboratory Standards).[26] When
fluconazole-resistant C. albicans
was treated with fluconazole or
berberine analogues the individ-
ual MIC80 values were deter-
mined and defined as the lowest
Scheme 4. Synthesis of derivatives 5a–e. Reagents and conditions: a) TFA, reflux, 12–24 h, NaOH; b) CH2Cl2, K2CO3,
reflux, 8–12 h; c) CH2Cl2, KI, reflux, 8 h.
concentration of the agents that
inhibited growth by 80%. When
C. albicans was treated with
was synthesized by holding a solution of 14a at reflux with
methyl iodide in acetonitrile.
a combination of fluconazole and berberine analogues, the in-
teraction MIC80 value of each agent was obtained. Further-
more, the fractional inhibitory concentration index (FICI) of
each agent was calculated by determining the ratio of the in-
teraction MIC80/individual MIC80. The interaction modes, either
synergistic or indifferent, were defined according to FICI values
of ꢀ0.5 or >0.5, respectively.[27] In this assay, the final concen-
tration of fluconazole (FLC, in Tables 1 and 2) in each well was
fixed at a single value (16.0 mgmLÀ1), whereas the final concen-
tration of the berberine analogues was set at a range of 0.13–
64.0 mgmLÀ1 in a double dilution series. Four clinical isolates of
fluconazole-resistant C. albicans (with MIC80 values determined
as 128.0 mgmLÀ1) were used in this study. In this paper, we
have chosen to only report the results obtained with C. albi-
cans 103, which were found to be consistent with that of the
other three isolates (C. albicans 100, J28, and 953).
A series of 2-(substituted phenyl)ethylamines, including 15,
16, 17, and 18, were reacted as starting materials with a series
of different carboxylic acids,[24] isocyanates,[25] and isothiocya-
nates[25] to give the corresponding amides 6a–e, 7a–m, and
8a–n, ureas 8o–v, and thioureas 8w–z, respectively, as shown
in Schemes 5 and 6.
As shown in Table 1, dihydroberberine 9 and tetrahydrober-
berine 10 did not show synergistic activity in combination
with fluconazole (16.0 mgmLÀ1). However, compounds 2a-j,
which were derived from 9, showed good synergistic activity.
The interaction MIC80 values of compounds 2a–j were in the
range of 0.25–8.0 mgmLÀ1, and the FICI values of compounds
2a, 2b, 2c, 2d, 2e, 2 f, and 2j in particular were in the range
of 0.129–0.141, indicating that these compounds exhibited
a higher level of synergistic activity than berberine. On the
basis of these results, it was concluded that the introduction of
a benzyl group at the C13 position could improve the synergis-
tic activity of berberine. One question raised by this result,
however, was whether functionality at the C13 position was
critical to the activity of berberine, and if so, whether the intro-
duction of a bulky benzyl group at this position would affect
the ability of the derivatives to bind to the target and conse-
quently lead to a decrease in their activity. To address these
issues, compounds 4a–o were designed and prepared as de-
constructed berberine analogues without the C13 moiety.
Scheme 5. Synthesis of amide derivatives. Reagents and conditions: a) HOBt,
DCC, THF, RT, 6–12 h.
Scheme 6. Synthesis of urea and thiourea derivatives. Reagents and condi-
tions: a) for ureas: THF, RT, 6–12 h; for thioureas: THF, RT, 4–6 h.
Biological evaluation
The in vitro synergistic antifungal activities of the newly syn-
thesized berberine analogues were tested using the micro-
broth dilution method, according to the standards of the Clini-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 207 – 216 209

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
cluded that the E ring in berber-
ine was important to the syner-
gistic activity.
Table 1. Structures and interaction modes of the title compounds and their MIC80 and FICI values.
Compd
R1
R2
R3
MIC80
[mgmLÀ1
MIC80
with FLC[a]
FICI
Mode of
Interaction
]
To investigate the effect of the
iminium ion in berberine, qua-
ternary ammoniums 3a–3e
were prepared from the corre-
sponding inactive tetrahydro-
genberberine 10. These quater-
nary ammonium compounds
3a–e exhibited synergistic activi-
ty. Perhaps more importantly,
the nitrogen atom in com-
pounds 3a–e was in the sp3 hy-
bridization state, and the quater-
nary ammonium was therefore
three-dimensional and bulky,
whereas the nitrogen atom in
berberine and in compounds
2a–j was in the sp2 hybridization
state, and the quaternary imini-
um ion was structurally planar.
Significant variations in the
shape of the quaternary ammo-
nium between compounds 3d
or 3e and berberine did not
lead to much difference in the
levels of activity observed.
Hence, it was entirely possible
that the iminium ion in berber-
ine was not directly involved in
the binding of berberine to its
target.
1
9
–
–
–
o-Br
m-Br
p-Br
o-F
p-OMe
p-F
m-NO2
m-Me
o-NO2
m-F
o-NO2
p-NO2
m-F
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
Me
32.0
>64.0
>64.0
32.0
32.0
64.0
64.0
32.0
64.0
>64.0
32.0
1.0
>64.0
>64.0
0.25
0.5
0.156 synergy
>0.50 indifferent
>0.50 indifferent
0.133 synergy
0.141 synergy
0.141 synergy
0.129 synergy
0.141 synergy
0.141 synergy
0.187 synergy
0.156 synergy
0.156 synergy
0.133 synergy
0.375 synergy
0.375 synergy
0.250 synergy
0.156 synergy
0.156 synergy
0.156 synergy
0.188 synergy
0.188 synergy
0.250 synergy
0.250 synergy
0.250 synergy
0.375 synergy
0.375 synergy
0.375 synergy
0.375 synergy
0.375 synergy
0.375 synergy
>0.50 indifferent
>0.50 indifferent
>0.50 indifferent
>0.50 indifferent
>0.50 indifferent
0.156 synergy
0.250 synergy
0.250 synergy
0.188 synergy
>0.50 indifferent
10
2a
2b
2c
2d
2e
2 f
2g
2h
2i
1.0
0.25
2.0
1.0
8.0
1.0
1.0
1.0
32.0
32.0
16.0
4.0
4.0
4.0
32.0
2j
>64.0
>64.0
>64.0
>64.0
>64.0
>64.0
>64.0
64.0
>64.0
>64.0
>64.0
>64.0
>64.0
>64.0
>64.0
>64.0
>64.0
>64.0
>64.0
>64.0
>64.0
>64.0
>64.0
3a
3b
3c
3d
3e
4a
4b
4c
4d
4e
4 f
4g
4h
4i
o-Cl
o-Me
2-furanyl
2,4-difluorophenyl
–
–
4-tBu
4-tBu
4-tBu
4-tBu
4.0
8.0
H
Me
H
Me
16.0
16.0
16.0
32.0
32.0
32.0
32.0
32.0
32.0
>64.0
64.0
64.0
>64.0
64.0
4.0
2,3-dimethoxy
2,3-dimethoxy
2,3-dimethoxy
4-Br
2,4-difluorophenyl
Me
2-furanyl
2,4-difluorophenyl
2-furanyl
4-Br
4-Br
4j
4k
4l
4-isopropoxycarbonyl
2,4-difluorophenyl 4-isopropoxycarbonyl
4m
4n
4o
19
5a
5b
5c
5d
5e
20
H
Me
–
–
–
–
2,4-difluorophenyl
–
H
H
H
H
Ph
–
2,3-dimethoxy
4-Br
4-tBu
4-Br
4-Br
–
4-bromobenzyl >64.0
Me
>64.0
>64.0
>64.0
>64.0
16.0
16.0
8.0
Compounds 5a–e were de-
signed and synthesized with
a three-dimensional quaternary
ammonium in which the nitro-
gen atom was in the sp3 hybridi-
zation state. Furthermore, the
2-phenylethyl
Me
–
>64.0
[a] MIC80 value [mgmLÀ1] of compound in column 1 in combination with 16.0 mgmLÀ1 fluconazole.
Of compounds 4a–o, compounds 4a, 4b, and 4c exhibited
the highest synergistic activities. Analogues 4g, 4h, 4i, 4j, 4k,
and 4l, which contained either a 4-bromo or a 4-isopropoxy-
carbonyl group as the R2 substituent, showed similar and mod-
erate levels of synergistic activity, whereas compounds 4m,
4n, and 4o, in which the R2 substituent was a hydrogen atom,
showed no activity. This indicated that the opening of the
D ring in berberine allowed retention of its synergistic antifun-
gal activity; thus, we were able to deduce that the motif at the
C13 position of berberine was not essential to its activity. At
this stage in our investigative program, we began to believe
that one of the key pharmacophores in berberine resided in its
E ring, together with its associated substituents. With this in
mind, we proceeded to evaluate the synergistic activities of
compound 19 and intermediates 12a–d, which do not have
the berberine E ring. As shown in the Supporting Information,
compounds 19 and 12a–d did not demonstrate any synergis-
tic activity with fluconazole. Based on these results, we con-
D ring in these compounds was effectively opened by deleting
the C13 substituents in compounds 3a–e. As shown in Table 1,
compounds 5b–e exhibited synergistic activity, though lower
than that of berberine. Compound 20 and intermediates 13a,
13b, 14a, 14b, 14c, and 14d, as shown in the Supporting In-
formation, possessed no synergistic activity with fluconazole.
Comparison of compounds 5a–e with compounds 20, 13a,
13b, 14a–d, 3a–e and berberine led us to believe that the
E ring of berberine was essential to its activity.
The data collected from preliminary structure–activity rela-
tionship (SAR) studies of the analogues outlined above effec-
tively guided us toward the design of compounds 6a–e and
7a–i. Compounds 6a–e, which did not contain an iminium
ion, represented berberine analogues with the A, B, and
E rings linked together through an amide chain instead of the
C and D rings. Unfortunately, however, these compounds pos-
sessed no activity in terms of their MIC80 and FICI values, as
shown in Table 2. Despite this failure, we decided to extend
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 207 – 216 210

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
our work with the amide linkers
by adding different carboxylic
acids to give compounds 7a–m.
Compounds 7a–i showed excel-
lent levels of activity. Although
they did not show antifungal ac-
Table 2. Structures and interaction modes of the title compounds and their MIC80 and FICI values.
Compd
6a
R1
R2
R3
MIC80
MIC80
FICI
Mode of
Interaction
[mgmLÀ1
]
with FLC[a]
–OCH2O–
>64.0
>64.0
>64.0
>64.0
>64.0
>64.0
>0.50
>0.50
>0.50
indifferent
indifferent
indifferent
tivity
(MIC80 >64.0 mgmLÀ1
)
when used alone, their interac-
tion MIC80 values decreased to
6b
–OCH2O–
–OCH2O–
a
range of 0.13–16.0 mgmLÀ1
when they were used in combi-
nation with fluconazole at
16.0 mgmLÀ1. Their FICI values
were found to be in the range
of 0.126–0.250, indicating that
they possessed significant syner-
gistic activity with fluconazole
6c
6d
6e
–OCH2O–
–OCH2O–
>64.0
>64.0
>64.0
>64.0
>0.50
>0.50
indifferent
indifferent
against
fluconazole-resistant
C. albicans. Compounds 7b and
7e showed the highest levels of
synergistic activity, with interac-
tion
MIC80
values
of
7a
7b
7c
7d
–OCH2O–
–OCH2O–
–OCH2O–
–OCH2O–
>64.0
>64.0
>64.0
>64.0
1.0
0.125
8
0.133
0.126
0.188
0.133
synergy
synergy
synergy
synergy
0.125 mgmLÀ1 and 0.5 mgmLÀ1
and FICI values of 0.126 and
0.129, respectively. We therefore
hypothesized that the A, B, and
E rings might be the pharmaco-
phore of berberine, and then
proceeded to focus the SAR
study on compounds 7a–i by
changing or replacing the
phenyl ring and the linker.
1.0
Replacement of the R3 sub-
stituent with thiophen-2-yl, 1H-
indol-3-yl, and naphthalen-1-yl
led to compounds 7j, 7k, and
7l, respectively, with retention
of activity. Replacement of the
CH2CH2NHCOCH2 motif with
CH2CH2NHCONH (8o, 8p, and
8q) or with CH2CH2NHCSNH
(8w) resulted in lower levels of
activity than 7a, 7b, and 7 f, re-
spectively. Introduction of a hy-
droxy (8a) or phenyl (8b) group
on the methylene group in the
linker of 7a led to a decrease in
activity.
7e
7 f
–OCH2O–
–OCH2O–
–OCH2O–
–OCH2O–
>64.0
>64.0
>64.0
>64.0
0.5
1.0
8.0
2.0
0.129
0.133
0.188
0.141
synergy
synergy
synergy
synergy
7g
7h
7i
7j
–OCH2O–
–OCH2O–
>64.0
>64.0
16.0
2.0
0.250
0.141
synergy
synergy
Extended analogues were de-
signed to investigate the effect
of the methylenedioxy group in
the A ring. For example, the
methylenedioxy groups in com-
pounds 7b, 7d, 7a, and 7k
were replaced with dimethoxy
groups to give compounds 8c–
f, whereas the methylenedioxy
7k
7l
–OCH2O–
–OCH2O–
>64.0
>64.0
2.0
1.0
0.141
0.133
synergy
synergy
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 207 – 216 211

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
values of each of the com-
pounds as shown in Table 3,
compounds 2g (0.023), 2j
(0.012), 4c (0.023), 5b (0.020),
7a (0.016), 7b (0.012), 7d
(0.012), 7e (0.023), 7j (0.023), 7 f
(0.012), 7l (0.012), and 8b
(0.012) showed higher levels of
synergistic activity than berber-
ine (1) (0.035). The lowest FICI
value was found for compound
7k and was similar to that of
berberine. It was clear from the
data that compounds from
series 7 were superior to berber-
ine in terms of their synergistic
activity.
Table 2. (Continued)
Compd
7m
R1
R2
R3
MIC80
MIC80
FICI
>0.50
0.156
Mode of
Interaction
[mgmLÀ1
]
with FLC[a]
–OCH2O–
>64.0
>64.0
>64.0
indifferent
synergy
8a
–OCH2O–
4.0
8b
8o
8p
–OCH2O–
–OCH2O–
–OCH2O–
>64.0
>64.0
>64.0
2.0
8.0
0.141
0.188
0.250
synergy
synergy
synergy
16.0
Cytotoxicity assay
The synergistically active com-
pounds 1, 2a, 2i, 7b, 7d, and
7e were subjected to a cytotoxic-
ity assay against human umbili-
8q
–OCH2O–
–OCH2O–
>64.0
>64.0
16.0
16.0
0.250
0.250
synergy
synergy
8w
cal
vein
endothelial
cells
(HUVEC) in the presence of flu-
conazole at a concentration of
16.0 mgmLÀ1. The results of the
assay are reported in Table 4.
[a] MIC80 value [mgmLÀ1] of compound in column 1 in combination with 16 mgmLÀ1 fluconazole.
groups in compounds 7b, 7a, 8a, and 7j were replaced with
two hydrogen atoms to give compounds 8g–j. In one last ex-
ample, the methylenedioxy groups in compounds 7b, 7c, 8a,
and 7j were replaced by a single hydrogen and a methoxy
group to give compounds 8k–n. All of these analogues
showed no activity. Furthermore, in compounds 8r–v and 8x–
z, replacement of the methylenedioxy group with two hydro-
gen atoms, or a hydrogen atom and a methoxy group, result-
ed in no activity at all. Taken together, these results clearly
demonstrated that the methylenedioxy group was essential to
their activity. The data associated with the inactive compounds
8c–n, 8r–v, and 8x–z is provided in the Supporting Informa-
tion.
Berberine (1) showed a moderate cytotoxic effect at a concen-
tration of 32.0 mgmLÀ1, with a viability of 64.0% determined
after 24 h and 45.0% after 48 h. Compounds 2a and 2i had vi-
Table 3. Checkerboard microdilution assay of fluconazole with the title
compounds.
[a]
Compd
MIC80 [mgmLÀ1
]
MIC80
with FLC[b]
Lowest
FICI
Mode of
interaction
1
32.0
>64.0
32.0
1.0(0.5)
1.0(2.0)
1.0(1.0)
1.0(1.0)
1.0(0.5)
1.0(8.0)
1.0(8.0)
1.0(2.0)
2.0(0.5)
1.0(1.0)
1.0(0.5)
2.0(32.0)
1.0(0.5)
1.0(2.0)
2.0(1.0)
1.0(0.5)
4.0(1.0)
4.0(0.5)
1.0(0.5)
1.0(0.5)
0.035
0.023
0.039
0.039
0.012
0.070
0.078
0.023
0.020
0.016
0.012
0.266
0.012
0.023
0.023
0.012
0.039
0.035
0.012
0.012
synergy
synergy
synergy
synergy
synergy
synergy
synergy
synergy
synergy
synergy
synergy
synergy
synergy
synergy
synergy
synergy
synergy
synergy
synergy
synergy
2g
2h
2i
32.0
2j
>64.0
>64.0
64.0
4a
4b
4c
5b
7a
7b
7c
7d
7e
7j
Checkerboard microdilution assay
>64.0
>64.0
>64.0
>64.0
>64.0
>64.0
>64.0
>64.0
>64.0
>64.0
>64.0
>64.0
>64.0
Some of the compounds displaying good activity in the pre-
liminary screening assay were then subjected to a checkerboard
microdilution assay. As a result of this analysis, a series of FICI
and interaction MIC80 values were determined for the com-
pounds, and the lowest FICI value of each of the tested com-
pounds is shown in Table 3. All compounds tested showed
high levels of synergistic activity with fluconazole against flu-
conazole-resistant C. albicans. At concentrations of 1.0, 2.0, 4.0,
and 8.0 mgmLÀ1, all tested compounds provided a significant
decrease in the MIC80 values of fluconazole from 128.0 mgmLÀ1
to within a range of 0.25–8.0 mgmLÀ1, as shown in the Sup-
porting Information. Through a comparison of the lowest FICI
7 f
7h
7k
7l
8b
[a] Fluconazole-resistant clinical isolates of C. albicans 103. [b] Interaction
MIC80 value of fluconazole given in parentheses.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 207 – 216 212

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
Table 4. Cytotoxic effects of berberine and its analogues against HUVEC.
Table 5. Calculated physicochemical properties of compounds 7b, 7d,
and 7e.[a]
Compd
Viability [%][a]
32.0 mgmLÀ1
24 h 48 h
16.0 mgmLÀ1
24 h
Property
Optimal Range
7b
7d
7e
48 h
Mr [Da]
logP
<500
<5
328
2.8
301
3.0
297
3.3
1
2a
2i
7b
7d
7e
64.0Æ1.7
0
2.0Æ0.5
90.0Æ0.1
96.0Æ1.5
91.0Æ0.1
45.0Æ1.1
0
0
84.0Æ1.0
96.0Æ3.1
91.0Æ3.0
78.0Æ1.8
0
2.0Æ0.3
89.0Æ2.1
96.0Æ2.5
91.0Æ0.7
52.0Æ1.5
0
0
86.0Æ3.0
96.0Æ2.7
95.0Æ1.0
H-bond donors
H-bond acceptors
Rotatable bonds
TPSA
<5
<10
<5
1
7
6
1
4
5
1
4
5
<140
93.4
47.6
47.6
[a] Calculated with Molinspiration property engine ver.2013.09 (http://
www.molinspiration.com).
[a] Data represent arithmetic means ÆSD of at least three independent
experiments.
abilities of 0% and 2.0% after 24 h, and 0% and 0% after 48 h,
respectively, showing much higher cytotoxic effects than ber-
berine. In contrast, compounds 7b, 7d, and 7e showed much
lower cytotoxic effects with viabilities greater than 84.0% after
24 and 48 h. Berberine has been demonstrated to be more cy-
totoxic toward a variety of different cell lines than its deriva-
tives containing tertiary amines,[18] indicating that the quater-
nary iminium ion in berberine is related to its toxicity. Our re-
sults are not only consistent with this conclusion, but have
also provided compounds with higher levels of activity and
lower levels of cytotoxicity by replacing the iminium ion with
an amide.
compounds could serve as promising lead compounds for fur-
ther research.
Experimental Section
Chemistry
All evaporations were conducted in vacuo on a rotary evaporator.
Analytical samples were dried in vacuo (1–5 mmHg) at room tem-
perature. Thin-layer chromatography (TLC) was conducted on silica
gel 60 F254 plates (Yantai Zhifu Chemical Co. Ltd., Yantai, China).
The purities of new compounds were determined using microanal-
ysis (C, H, N) and HPLC and agreed with the theoretical values
1
within Æ0.4% and ꢁ95.0%, respectively. H NMR spectra were re-
corded on a Bruker Avance 300 (300 MHz) spectrometer (Bruker,
Fꢁllanden, Switzerland), using CDCl3 or [D6]DMSO as solvents and
tetramethylsilane (TMS) as the internal standard. The purity and
electrospray ionization mass spectroscopy (ESIMS) data for the title
compounds were determined on an Agilent LC–MS 6120 (Agilent,
Santa Clara, CA, USA) coupled to a Max mass spectrometer (Agi-
lent). The following methods were used for liquid chromatography:
Method 1=Ultimate XB-C18 column (2.1 mmꢂ50 mmꢂ3.5 mm)
with a 12.0 min gradient (buffer: TFA (0.01%)+H2O/CH3CN), from
90:10 to 70:30 over 0.5 min, and a subsequent gradient from 70:30
to 10:90 over 7.5 min, followed by a gradient from 10:90 to 90:10
over 4.0 min. Method 2=Ultimate XB-C18 column (2.1 mmꢂ
50 mmꢂ3.5 mm) with a 12 min gradient (buffer: TFA (0.01%)+
H2O/CH3CN), from 95:5 to 50:50 over 3.0 min, then a subsequent
hold at 50:50 for 5.0 min, followed by a gradient from 50:50 to
95:5 over 4.0 min. Compounds 19 and 20 were analyzed using
method 2, whereas the other compounds were analyzed using
method 1. Compound 15 was purchased from Hengye Zhongyuan
Chemical Ltd., (Beijing, China). Compound 17 was purchased from
Dengguan Chemical Ltd., (Jintan, China). Compound 18 was pur-
chased from Suzhou Yacoo Chemical Reagent Co. Ltd., (Suzhou,
China). The other chemicals were purchased from Aladdin Reagent
Co. Ltd., (Shanghai, China) and Sinopharm Chemical Reagent Co.
Ltd. (Shanghai, China).
Conclusions
We have described the process of structural optimization of
berberine as a synergist to restore antifungal activity of fluco-
nazole against fluconazole-resistant C. albicans. Modification of
berberine by introducing substituted phenyl groups led to an-
alogues of series 2 and 3, with synergistic activities. Compound
2a exhibited higher activity than berberine, but its cytotoxicity
(viability: 0%) is much higher than berberine (64%) at a con-
centration of 32.0 mgmLÀ1. The SAR study of series 2 and 3
guided us to deconstruct berberine and design series 4 and 5,
with similar or lower activity in general than berberine. The
data collected from the preliminary SAR study of the ana-
logues outlined above effectively guided us toward the design
of compounds 6a–e and 7a–i with novel scaffolds. Series 6 ex-
hibited no activity; however, series 7 was found to possess
active synergistic antifungal activity, low levels of cytotoxicity,
and good structural characteristics. Subsequent SAR studies of
series 7 indicated that the pharmacophore of berberine includ-
ed the A, B, and E rings, which should be linked using appro-
priate linkers. Compounds 7b, 7d, 7 f, and 7l (1.0 mgmLÀ1), in
combination with fluconazole, showed remarkable levels of in
vitro synergistic antifungal activity against fluconazole-resistant
C. albicans, providing a decrease in the MIC80 values of flucona-
zole from 128.0 to 0.5 mgmLÀ1. Compounds 7b, 7d, and 7e
demonstrated much lower levels of cytotoxicity toward HUVEC
than did parent berberine.
Synthesis of dihydroberberine (9): Berberine (1) (3.7 g, 10 mmol)
and K2CO3 (3.6 g, 30 mmol) were suspended in a mixture of MeOH
(60 mL), and a solution of NaBH4 (0.3 g, 7.5 mmol) in 5% aqueous
NaOH (5 mL) was added to the mixture in a dropwise manner.[19]
The resulting mixture was stirred at room temperature for 1 h. The
precipitated product was then filtered, and the filter cake was
washed sequentially with a 20% aqueous EtOH mixture (20 mL)
and an 80% aqueous EtOH (20 mL) mixture, before being collected
and purified by flash chromatography over silica gel, to give title
compound 9 (2.6 g, 77%) as a yellow solid.
The physicochemical parameters of 7b, 7d, and 7e were
calculated using Molinspiration Cheminformatics software
(2013), and none were found to violate the optimal require-
ments for druggability (Table 5). This suggested that these
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 207 – 216 213

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
Synthesis of tetrahydroberberine (10): NaBH4 (1.2 g, 30 mmol)
was added in a portionwise manner to a stirred solution of
1 (3.7 g, 10 mmol) and K2CO3 (3.6 g, 30 mmol) in MeOH (60 mL),
and the resulting mixture was heated at reflux for 1 h, leading to
the formation of a yellow solid.[19] The solid was collected by filtra-
tion and purified by flash chromatography over silica gel to give
title compound 10 (2.5 g, 74%) as a yellow solid.
a dropwise manner to a solution of corresponding intermediate
12a–d (2.0 mmol), respectively, in CH3CN (50 mL), and the resulting
mixture was heated at reflux for 6 h.[21] The mixture was then
cooled to ambient temperature and the solvent removed in vacuo
to give the crude residue, which was purified by flash chromatog-
raphy over silica gel to give the final compounds 4a–o, respective-
ly. Compound 4a was isolated as a yellow solid (543 mg, 58%):
1H NMR (300 MHz, CDCl3): d=7.89 (s, 1H), 7.50–7.49 (m, 1H), 7.43–
7.36 (m, 4H), 6.87–6.82 (m, 3H), 6.17–6.15 (d, 2H, J=6.9 Hz), 5.64
(s, 2H), 4.21–4.16 (t, 2H, J=7.4 Hz), 3.50–3.33 (t, 2H, J=7.4 Hz),
1.30 ppm (s, 9H); ESIMS m/z: 388.2 [MÀBr]+; Purity: 99.6%.
General procedure for the synthesis of 13-benzylberberine de-
rivatives 2a–j: o-Bromo benzyl bromide (1.0 mmol) was added in
a dropwise manner to a stirred solution of KI (310 mg, 2.06 mmol)
and 9 (337 mg, 1.0 mmol) in CH3CN (40 mL), and the resulting mix-
ture was held at reflux for 4 h.[19] The reaction mixture was then fil-
tered, and the filtrate was collected and distilled to dryness in
vacuo to give the crude residue, which was purified by flash chro-
matography over neutral alumina to give the final compound 2a
General procedure for the synthesis of 5,6-disubstituted-5,6,7,8-
tetrahydro[1,3]dioxolo[4,5-g]isoquinolin-6-ium bromide deriva-
tives 5a–e: A solution of 2-benzo[1,3]dioxol-5-ylethylamine 15
(16.5 g, 100 mmol) and the corresponding aldehyde (100 mmol) in
formic acid (50 mL) was stirred at reflux for 24 h. The mixture was
then cooled to room temperature and diluted with H2O
(100 mL).[22,23] The solution was then adjusted to pH 8–9 with a 1n
aqueous NaOH solution and extracted with CH2Cl2 (3ꢂ100 mL).
The combined organics were then dried for 1 h over anhydrous
MgSO4 before being distilled to dryness in vacuo to give the crude
residue, which was purified by flash chromatography over silica gel
to give intermediate products 13a–b, respectively, as white solids.
Iodomethane or the substituted benzyl bromide was added to
a stirred mixture of intermediate 13a or 13b (10 mmol) and anhy-
drous K2CO3 (1.38 g, 10 mmol) in EtOH (50 mL), and the resulting
mixture was heated at reflux for 12 h. The mixture was then
cooled and the resulting precipitation removed by filtration. The
filtrates were then collected and distilled to dryness in vacuo to
give the crude residue, which was purified by flash chromatogra-
phy over silica gel to give the intermediate products 14a–d, re-
spectively, as white solids. 2,3-Dimethoxybenzyl bromide (264 mg,
2.0 mmol) was added in a dropwise manner to a stirred solution of
intermediate 14a–d (2.0 mmol) in CH3CN (50 mL), and the resulting
mixture was heated at reflux for 8 h. The mixture was then cooled
and the solvent was removed in vacuo to give the crude residue,
which was purified by flash chromatography over silica gel to give
products 5a–e, respectively. Compound 5a was isolated as a white
solid (700 mg, 83%): 1H NMR (300 MHz, CDCl3): d=7.44–7.41 (d,
1H, J=8.1 Hz), 7.13–7.08 (m, 1H), 7.01–6.98 (d, 1H, J=8.1 Hz), 6.59
(s, 1H), 6.50 (s, 1H), 5.91 (s, 2H), 5.17–5.13 (d, 1H, J=12.3 Hz),
5.04–4.00 (d, 1H, J=12.3 Hz), 4.73–4.68 (d, 1H, J=15.0 Hz), 4.50–
4.45 (d, 1H, J=15.0 Hz), 4.12–3.92 (m, 2H), 3.87 (s, 3H), 3.84 (s,
3H), 3.19 (s, 3H), 3.15–3.03 ppm (m, 2H); ESIMS m/z: 342.1
[MÀBr]+; Purity: 95.2%.
General procedure for the synthesis of amide derivatives 6a–e,
7a–m, and 8a–n: A mixture of the starting material 15 (1.65 g,
0.01 mol), HOBt (1.35 g, 0.01 mol), the corresponding carboxylic
acid (0.01 mol), and DCC (2.06 g, 0.01 mol) in THF (20 mL) was
stirred at room temperature for 12 h. The mixture was then filtered
to remove the white precipitate, and the filtrates were collected
and distilled to dryness in vacuo to give the crude residue, which
was dissolved in CH2Cl2 and washed three times with a saturated
solution of Na2CO3. The organic layers were then dried over anhy-
drous MgSO4 for 1 h and the solvent removed in vacuo to give the
crude residue, which was purified by flash column chromatography
over silica gel to give final compounds 6a–d and 7a–m, respec-
tively. Compounds 8a–n were also synthesized according to the
same procedure from 16, 17, or 18, respectively. Compound 7a
was isolated as a white solid (2.01 g, 71%): 1H NMR (300 MHz,
CDCl3): d=7.36–7.25 (m, 3H), 7.20–7.17 (m, 2H), 6.67–6.64 (d, 1H,
J=8.1 Hz), 6.53–6.52 (d, 1H, J=1.8 Hz), 6.45–6.42 (m, 1H), 5.92 (s,
1
as a yellow solid (350 mg, 60%): H NMR (300 MHz, [D6]DMSO): d=
10.08 (s, 1H), 8.13–8.10 (d, 1H, J=9.3 Hz), 7.85–7.83 (d, 1H, J=
9.3 Hz), 7.67–7.64 (d, 1H, J=9.3 Hz), 7.31–7.26 (m, 2H), 7.17 (s, 1H),
6.84–6.81 (m, 1H), 6.71 (s, 1H), 6.08 (s, 2H), 4.88 (s, 2H), 4.61 (2H,
s), 4.12 (s, 3H), 4.03 (s, 3H), 3.15 ppm (s, 2H); ESIMS m/z: 504.1
[MÀBr]+; Anal. (C27H23Br2NO4) calcd: C 55.41, H 3.96, N 2.39, found:
C 55.62, H 3.90, N 2.59. Compounds 2b–j were also synthesized ac-
cording to the same general procedure.
General procedure for the synthesis of 7-benyzltetrahydrober-
berine bromide derivatives 3a–e: A solution of o-nitrobenzyl bro-
mide (1.0 mmol) in CH3CN (40 mL) was added to a stirred solution
of 10 (339 mg, 1.0 mmol) in CH3CN (60 mL) in a dropwise manner,
and the resulting reaction mixture was held at reflux for 6 h.[19] The
mixture was then cooled and the solvent removed in vacuo to
give the crude residue, which was purified by flash chromatogra-
phy over neutral alumina to give title compound 3a as a yellow
solid (361 mg, 65%): 1H NMR (300 MHz, CDCl3): d=8.85–8.83 (d,
1H, J=7.2 Hz), 7.98–7.95 (d, 1H, J=8.1 Hz), 7.89–7.85 (t, 1H, J=
6.6 Hz), 7.73–7.68 (t, 1H, J=6.6 Hz), 6.91–6.84 (m, 3H), 6.73 (s, 1H),
6.02 (dd, 2H, J1 =1.2 Hz, J2 =3.6 Hz), 5.88–5.83 (d, 1H, J=13.2 Hz),
5.74–5.71 (m, 1H), 5.68–5.64 (d, 1H, J=13.2 Hz), 5.19–5.13 (d, 1H,
J=12.6 Hz), 4.45–4.40 (d, 1H, J=12.6 Hz), 3.99–3.94 (m, 1H), 3.86
(s, 3H), 3.83 (s, 3H), 3.56–3.40 (m, 2H), 3.37–3.29 (m, 1H), 3.22–
3.08 ppm (m, 2H); ESIMS m/z: 475.2 [MÀBr]+; Purity: 96.7%.
General procedure for the synthesis of 5,6-disubstituted-7,8-
dihydro[1,3]dioxolo[4,5-g]isoquinolin-6-ium bromide derivatives
4a–o: A solution of 15 (16.5 g, 0.1 mol), HOBt (13.5 g, 0.1 mol), car-
boxylic acid (0.1 mol), and DCC (20.6 g, 0.1 mol) in THF (200 mL)
was stirred at room temperature for 12 h. The mixture was then fil-
tered to remove the white precipitate, and the filtrates were col-
lected and distilled to dryness in vacuo to give the crude residue,
which was dissolved in CH2Cl2 and washed three times with a satu-
rated solution of Na2CO3 before being dried over anhydrous
MgSO4 for 1 h. The MgSO4 was then removed by filtration, then
evaporated in vacuo to give the crude residue, which was purified
by flash chromatography over silica gel to give intermediates 11 a–
d, respectively. POCl3 (7.67 g, 50 mmol) was added in a dropwise
manner over a period of 0.5 h to a cooled solution (ice bath) of
11 a–d (10 mmol) in toluene (50 mL), and the resulting mixture was
stirred at reflux for 12 h. The mixture was then cooled to ambient
temperature and quenched by the addition of H2O (100 mL). The
resulting mixture was then adjusted with ammonia to pH 8–9
under stirring, and the resulting aqueous mixture was extracted
with CH2Cl2 (3ꢂ100 mL). The combined organic extracts were then
distilled to dryness in vacuo to give the crude residue, which was
purified by flash chromatography over silica gel to give intermedi-
ates 12a–d, respectively. Benzyl bromide (2 mmol) was added in
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 207 – 216 214

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
2H), 5.39 (s, 1H), 3.53 (s, 2H), 3.41–3.38 (m, 2H), 2.65–2.60 ppm (t,
of fluconazole was set at 16.0 mgmLÀ1 for the preliminary screen-
ing assay, and in the range of 0.125–64.0 mgmLÀ1 for the checker-
board microdilution assay. The assay plates were incubated at
358C for 24 h. Optical density was measured at 630 nm, and the
background optical densities were then subtracted from the value
provided for each well. Each isolate was tested in triplicate. The
MIC80 values were determined as the lowest concentrations of the
drugs (alone or in combination) that inhibited fungal growth by
80% compared with that of the drug-free wells.
2H, J=6.6 Hz); ESIMS m/z: 284.1 (M+H)+; Purity: 96.8%.
General procedure for the synthesis of urea derivatives 8o–v:
The corresponding isocyanate (0.01 mmol) was added to a cooled
solution (ice bath) of starting material 15, 16, 17, or 18 (0.01 mol)
in THF (50 mL), respectively, and the resulting mixture was stirred
at room temperature for 6–12 h.[25] The solvent was then removed
in vacuo to give the crude residue, which was purified by recrystal-
lization to give final products 8o–v, respectively. Compound 8o
was isolated as a white solid (2.61 g, 82%): 1H NMR (300 MHz,
CDCl3): d=8.02–7.99 (d, 1H, J=8.1 Hz), 7.36–7.33 (d, 1H, J=
1.5 Hz), 7.24–7.21 (m, 1H), 7.02–6.97 (m, 1H), 6.77–6.71 (m, 1H),
6.67–6.64(m, 2H), 5.94 (s, 2H), 3.52–3.48 (t, 2H, J=6.6 Hz), 2.82–
2.77 ppm (t, 2H, J=6.6 Hz); ESIMS m/z: 319.1 [M+H]+; Purity:
99.4%.
Cytotoxicity assays: Human umbilical vein endothelial cells (HUVEC)
were maintained in RPMI1640 medium (HyClone, Beijing, China)
supplemented with 10% fetal calf serum (Gibco, CA, USA), and
100 mgmLÀ1 streptomycin. The cells were grown in a humidified in-
cubator in 5% CO2 at 378C and were used for assays during their
exponential growth phase. Cell viability was determined following
24 h and 48 h of incubation with berberine (16.0 or 32.0 mgmLÀ1
or its derivatives in the presence of fluconazole (16.0 mgmLÀ1
)
)
General procedure for the synthesis of thiourea derivatives 8w–
z: The corresponding isothiocyanate (0.01 mmol) was added to
a cooled (ice bath) solution of starting material 15, 16, 17, or 18
(0.01 mol) in THF (50 mL), respectively, and the resulting mixture
was stirred at room temperature for 6–12 h.[25] The solvent was
then removed in vacuo to give the crude residue, which was puri-
fied by recrystallization to give final products 8w–z, respectively.
using a Roche Cell Proliferation Kit II (XTT). Briefly, the XTT mixing
solution (XTT labeling reagent/electron coupling reagent, 5:1) was
added to each well containing 5ꢂ103 HUVEC. After a 6 h reaction,
optical density was measured by a microplate reader (Thermo, MA,
USA) at 450 nm. The cytotoxic activity was measured by the follow-
ing formula: viability (%)=100ꢂOD450 of test/OD450 of control.
1
Compound 8w was isolated as a white solid (1.38 g, 46%): H NMR
(300 MHz, CDCl3): d=8.29 (s, 1H), 7.34–7.31 (t, 2H, J=3.9 Hz),
7.24–7.22 (t, 1H, J=3.9 Hz), 7.05–7.04 (d, 2H, J=3.9 Hz), 6.67–6.65
(d, 2H, J=3.9 Hz), 6.61 (s, 1H), 6.54–6.53 (d, 1H, J=3.9 Hz), 6.01 (s,
1H), 5.89 (s, 2H), 3.82–3.78 (q, 2H, J=3.0 Hz), 2.80–2.78 ppm (t,
2H, J=3.0 Hz); ESIMS m/z: 301.1 (M+H)+; Purity: 99.7%.
Acknowledgements
This research was supported by the National Natural Science
Fund of China (Nos. 21072226, 30672533, 30772644, and
21272270) and the Shanghai Pujiang Program (No. 05PJ14003).
Synthesis of compounds 19 and 20: CH3I (0.6 mL, 5.0 mmol) was
added to a solution of intermediate 12a or 14a (2.0 mmol) in
CH3CN (50 mL), respectively, and the resulting mixture was heated
at reflux for 8 h.[19] The mixture was then cooled, and the precipi-
tate was filtered and washed sequentially with CH2Cl2 and EtOAc
to give desired product 19 or 20, respectively. Compound 19 was
Keywords: antifungal agents · berberine · Candida albicans ·
drug resistance · structure–activity relationships
1
isolated as a yellow solid (285 mg, 90%): H NMR (300 MHz, D2O):
d=8.59 (s, 1H), 7.12 (s, 1H), 6.90 (s, 1H), 6.08 (s, 2H), 3.92–3.86 (t,
2H, J=8.1 Hz), 3.63 (s, 3H). 3.15–3.10 ppm (t, 2H, J=8.1 Hz);
ESIMS m/z: 190.1 [MÀI]+; Purity: 97.2%. Compound 20 was isolat-
[1] O. Gudlaugsson, S. Gillespie, K. Lee, J. V. Berg, J. F. Hu, S. Messer, L. Her-
waldt, M. Pfaller, D. Diekema, Clin. Infect. Dis. 2003, 37, 1172–1177.
[2] S. K. Fridkin, W. R. Jarvis, Clin. Microbiol. Rev. 1996, 9, 499–511.
[3] C. Charlier, E. Hart, A. Lefort, P. Ribaud, F. Dromer, D. W. Denning, O. Lor-
tholary, J. Antimicrob. Chemother. 2006, 57, 384–410.
1
ed as a light-yellow solid (546 mg, 82%): H NMR (300 MHz, D2O):
d=6.75 (s, 1H,), 6.62 (s, 1H), 5.91 (s, 2H), 4.40 (s, 2H), 3.62–3.57 (t,
2H, J=6.6 Hz), 3.13–3.06 ppm (m, 8H); ESIMS m/z: 206.1 [MÀI]+;
Purity: 95.3%.
[4] A. Goffeau, Nature 2008, 452, 541–542.
[5] B. C. Monk, A. Goffeau, Science 2008, 321, 367–369.
[6] M. Nucci, J. R. Perfect, Clin. Infect. Dis. 2008, 46, 1426–1433.
[7] W. Youngsaye, B. Vincent, C. L. Hartland, B. J. Morgan, S. J. Bahrlage, S.
Johnston, J. A. Bittker, L. Macpherson, S. Dandapani, M. Palmer, L.
Whitesell, S. Lindquist, S. L. Schreiber, B. Munoz, Bioorg. Med. Chem. Lett.
2011, 21, 5502–5505.
Biological activity assays
Strains and agents stock solution preparation: Four clinical isolates
of fluconazole resistant C. albicans (MIC80 =128.0 mgmLÀ1) were
used in this study, and C. albicans ATCC 90028 was used as a quality
control. The strains were cultured at 308C under constant shaking
(200 rpm) in a liquid complete medium (YPD) consisting of 1% (w/
v) yeast extract, 2% (w/v) peptone, and 2% (w/v) dextrose. The
stock solution of fluconazole (Pfizer, Dalian, China) was prepared in
sterile water, whereas the other compounds, including berberine
(Sigma–Aldrich, MO, USA), were prepared in DMSO.
[8] S. Gamarra, E. M. F. Rocha, Y. Q. Zhang, S. Park, R. Rao, D. S. Perlin, Anti-
microb. Agents Chemother. 2010, 54, 1753–1761.
[9] P. Uppuluri, J. Nett, J. Heitman, D. Andes, Antimicrob. Agents Chemother.
2008, 52, 1127–1132.
[10] J. Jin, N. Guo, J. Zhang, Y. Ding, X. Tang, J. Liang, L. Li, X. Deng, L. Yu,
Lett. Appl. Microbiol. 2010, 51, 351–357.
[11] N. Guo, X. Wu, L. Yu, J. Liu, R. Meng, J. Jin, H. Lu, X. Wang, S. Yan, X.
Deng, FEMS Immunol. Med. Microbiol. 2010, 58, 193–201.
[12] H. Quan, Y. Y. Cao, Z. Xu, J. X. Zhao, P. H. Gao, X. F. Qin, Y. Y. Jiang, Anti-
microb. Agents Chemother. 2006, 50, 1096–1099.
[13] Y. Han, J. H. Lee, Biol. Pharm. Bull. 2005, 28, 541–544.
[14] S. K. Battu, M. A. Repka, S. Maddineni, A. G. Chittiboyina, M. A. Avery, S.
Majumdar, AAPS PharmSciTech 2010, 11, 1466–1475.
[15] X. Zhang, F. Qiu, J. Jiang, C. Gao, Y. Tan, Xenobiotica 2011, 41, 290–296.
[16] W. Hua, L. Ding, Y. Chen, B. Gong, J. He, G. Xu, J. Pharm. Biomed. Anal.
2007, 44, 931–937.
Antifungal susceptibility testing: The in vitro MIC80 values of the
compounds against the clinical isolates of C. albicans were deter-
mined using the microbroth dilution method, according to the
Clinical and Laboratory Standards Institute.[12,26] The initial concen-
tration of the fungal suspension in RPMI1640 medium was
103 colony-forming units (CFU)/mL, and final compound concentra-
tions were in the range of 1.0–8.0 mgmLÀ1. The final concentration
[17] M. M. Kheir, Y. Wang, L. Hua, J. Hu, L. Li, L. Du, Food Chem. Toxicol. 2010,
48, 1105–1110.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 207 – 216 215

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
[18] L. Orfila, M. Rodrı’guez, T. Colman, M. Hasegawa, E. Merentes, F. Arvelo,
J. Ethnopharmacol. 2000, 71, 449–456.
[24] X. Zhao, M. X. Jia, X. K. Jiang, L. Z. Wu, Z. T. Li, G. J. Chen, J. Org. Chem.
2004, 69, 270–279.
[19] Y. H. Li, P. Yang, W. J. Kong, Y. X. Wang, C. Q. Hu, Z. Y. Zuo, Y. M. Wang, H.
Gao, L. M. Gao, Y. C. Feng, N. N. Du, Y. Liu, D. Q. Song, J. D. Jiang, J. Med.
Chem. 2009, 52, 492–501.
[25] J. M. Vega-Pꢄrez, I. PeriÇꢅn, M. ArgandoÇa, M. Vega-Holm, C. Palo-Nieto,
E. Burgos-Morꢃn, M. Lꢃpez-Lꢅzaro, C. Vargas, J. J. Nieto, F. Iglesias-
Guerra, Eur. J. Med. Chem. 2012, 58, 591–612.
[20] A. S. Capilla, M. Romero, M. D. Pujol, D. H. Caignard, P. Renard, Tetrahe-
dron 2001, 57, 8297–8303.
[26] P. A. Wayne, Approved standard M27-A3: Clinical and Laboratory Stand-
ards Institute (CLSI): USA, 2008.
˝
[21] J. Tꢃth, A. Dancsꢃ, G. Blaskꢃ, L. Toke, P. W. Groundwater, M. Nyerges,
[27] F. C. Odds, J. Antimicrob. Chemother. 2003, 52, 1–1.
Tetrahedron 2006, 62, 5725–5735.
[22] F. I. Zubkov, J. D. Ershova, A. A. Orlova, V. P. Zaytsev, E. V. Nikitina, A. S.
Peregudov, A. V. Gurbanov, R. S. Borisov, V. N. Khrustalev, A. M. Maharra-
mov, A. V. A. Varlamov, Tetrahedron 2009, 65, 3789–3803.
Received: August 9, 2013
Revised: October 10, 2013
[23] A. Yokoyama, T. Ohwada, K. Shudo, J. Org. Chem. 1999, 64, 611–617.
Published online on November 4, 2013
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 207 – 216 216