Synthesis, extracorporeal nephrotoxicity, and 3D-QSAR of andrographolide derivatives

Article abstract of DOI:10.1111/cbdd.13796

Andrographolide is a traditional Chinese medicine monomer with many pharmacological activities, but has potential nephrotoxicity. Here, we aim to investigate the relationship between modification of andrographolide structure and its nephrotoxicity. Twenty-three andrographolide derivatives were synthesized, and their structures were confirmed by ¹H-NMR and HRMS. Nephrotoxicity of these compounds against human renal tubular epithelial (HK-2) cells was evaluated using the MTT assay. The results indicated that most of them had significantly reduced nephrotoxicity, especially compounds III, V, and IX_c, with IC₅₀ values of 1,985, 1,300, and 806.9?μmol/L, respectively, which were obviously superior to andrographolide (IC₅₀ 30.60?μmol/L). However, compounds I_a–I_f (IC₅₀ values?2?=?0.639, r²?=?0.951; COMSIA: q²?=?0.569, r²?=?0.857). This model allowed proposing five new compounds with lower theoretical nephrotoxicity, which would be worthwhile to synthesize and evaluate. We believe that predicted models will help us to understand the structural modification requirements of andrographolide to reduce the nephrotoxicity, and further investigations will be needed to determine the mechanism involved in the effect of less nephrotoxic compounds.

Full text of DOI:10.1111/cbdd.13796

PROFESSOR XINYUE ZHANG (Orcid ID : 0000-0002-5027-6596)
Article type
: Research Article
Synthesis, Extracorporal Nephrotoxicity and 3D-QSAR of
Andrographolide Derivatives
Lili Gu1, Qingqing Yu2, Qin Li1, Wenhai Huang1, Ningzi Wu1, Jiaqi Lu1, Hong Lu2, Xinyue Zhang1*
1Key Laboratory of Neuropsychiatric Drug Research of Zhejiang Province, Hangzhou medical college, Hangzhou, 310013, China;
17826865042@163.com (G.L.); liqin@zjams.com.cn(L.Q.); cyj@zju.edu.cn(H.W.); wnzyky@126.com (W.N.);
lujiaqi2018@sina.com(L.J.);
2
College of Pharmaceutical science, Zhejiang Chinese Medical University, Hangzhou, 310053, PR China;
m15757196155_2@163.com(Y,Q.); luhong@zcmu.edu.cn (L,H);
* Correspondence: zhangxy@zjams.com.cn; Tel.: +0571-8821-5600
Abstract: Andrographolide is a traditional Chinese medicine monomer with many pharmacological activities, but
has potential nephrotoxicity. Here, we aim to investigate the relationship between modification of andrographolide
structure and its nephrotoxicity. Twenty-three andrographolide derivatives were synthesized and their structures
were confirmed by 1H-NMR and HRMS. Nephrotoxicity of these compounds against human renal tubular epithelial
(HK-2) cells were evaluated using the MTT assay. The results indicated that most of them had significantly reduced
nephrotoxicity, especially compounds III, V, IXc, with IC50 values of 1985, 1300, 806.9 μmol·L-1, respectively,
which were obviously superior to andrographolide (IC50 30.60 μmol·L-1). However, compounds Ia-If(IC50 values＜
30 μmol·L-1) ， the 14-OH derivatives of andrographolide, showed higher nephrotoxicity than that of
andrographolide. Three dimensional quantitative structure-activity relationship (3D-QSAR) models of COMFA and
This article has been accepted for publication and undergone full peer review but has not been
through the copyediting, typesetting, pagination and proofreading process, which may lead to
differences between this version and the Version of Record. Please cite this article as doi:
10.1111/CBDD.13796
This article is protected by copyright. All rights reserved

COMSIA were established (COMFA: q2 = 0.639, r2 = 0.951; COMSIA: q2 = 0.569, r2 = 0.857). This model allowed
proposing five new compounds with lower theoretical nephrotoxicity, which would be worthwhile to synthesize and
evaluate. We believe that predicted models will help us to understand the structural modification requirements of
andrographolide to reduce the nephrotoxicity, and further investigations will be needed to determine the mechanism
involved in the effect of less nephrotoxic compounds.
Keywords: andrographolide derivatives; nephrotoxicity; synthesis; 3D-QSAR
1. INTRODUCTION
Andrographolide ,a lactone diterpenoid compound, which can be isolated from traditional Chinese medicine
(TCM) Andrographis paniculata (Burm. F.) Wall. Ex Nees (Family: Acanthaceae)(Islam, 2017), exhibits various
biological activities such as anti-inflammatory (Li et al., 2017; W. Wang et al., 2019), antivirus (Paemanee, Hitakarun,
Wintachai, Roytrakul, & Smith, 2019; Uttekar et al., 2012), anticancer (Peng et al., 2018; Reabroi et al., 2018; Wei,
Zhang, & He, 2018), liver protection (Abdullah & Ismail, 2018), cardiovascular (Wu et al., 2018), neuroprotective
effect (Tao et al., 2018), and is mainly used in clinic for treating cold, fever, gastroenteritis and upper respiratory tract.
The nephrotoxicity of TCM has been reported since the 1980s (Upreti, Das, Kumar, Singh, & Khanna, 1989), and has
attracted widespread attention (Feng, Fang, Gao, Liu, & Chen, 2018; Song, Y.Y., & Miao, 2018; Vanherweghem et
al., 1993). The two androrapholide derivatives, andrographolide sodium bisulfite (Lianbizhi) and potassium
dehydroandrograpolide succinate (Chuanhuning), were used for the treatment of pneumonia, bacillary dysentery and
other diseases (Xu & Chen, 2014; Zhao & Cai, 2012). However it has been an increasing number of reports on their
adverse reactions, particularly renal adverse reactions in clinic (Xiang et al., 2016; Yang et al., 2013). In the our
previous studies, Lianbizhi and Chuanhuning injection had certain kidney damage effects in model of mice and
rabbits, and the main injury site was the renal tubule (Lu, Zhang, Zhou, & Jin, 2010; Xing, Xu, & Lu, 2013). Further
studies showed that Andro and its water soluble derivatives had a damage effect on human renal tubular epithelial
(HK-2) cells (L.L. Gu, Xing, Wang, Zheng, & Lu, 2016; L. L. Gu, Zhang, Xing, Xu, & Lu, 2016; Lu et al., 2014).
Quantitative structure activity relationship (QSAR) analysis is an important computer simulation method (Neves
et al., 2018; Potemkin, Potemkin, & Grishina, 2018), and is mainly used to study the relationship between compound
structure and physicochemical properties or biological activity (Brahmbhatt, Molnar, Pavi, & Rastijac, 2018;
Flores-Sumoza et al., 2018; Marquina et al., 2019; Toropov & Toropova, 2018). It was reported that QSAR was used
to furnish relationship between molecular descriptors and the nitric oxide synthase inhibition activity, antiproliferative
activity or alpha-glucosidase inhibiting activity in order to the search of more potent drugs in the andrographolide
derivative family of compounds (A, Prasana, Muthu, Abraham, & David, 2019; Hazra et al., 2015; Moorthy, Ramos,
& Fernandes, 2011). QSAR has also been widely used to study the relationship between compound structure and
toxicity (Secretan et al., 2019; Shao, Hollert, Tarcai, Deutschmann, & Seiler, 2019). Among the QASR analysis
This article is protected by copyright. All rights reserved

methods, comparative molecular field analysis (CoMFA) and comparative molecular similarity index analysis
(CoMSIA) are the most commonly used (Chen et al., 2018; M. Wang et al., 2018).
Therefore, in this study, we focused on the relationship between modification of andrographolide structure and
1
its nephrotoxicity. Firstly, a series of 23 andrographolide derivatives were synthesized..Then we used H-NMR and
HRMS and the model of HK-2 cells to characterize all target compounds and their nephrotoxicity , respectively.
QSAR models were constructed with CoMFA and CoMSIA to explore the relationship between structural changes
and nephrotoxicity, which can be applied to predict the nephrotoxicity of andrographolide derivatives and provide a
reference for the development of andrographolide drugs.
2. MATERIALS AND METHODS
2.1. General Information
Andrographolide (98.56%, TC) were provided by Shanxi YuNing Biotechnology Co., Ltd., (Shanxi, China), and
all other chemicals and solvents were of analytical grade and purchased from commercial suppliers. Melting points
1
were determined on the M-565 Melting-Point Apparatus (BUCHI Labortechnik AG, Switzerland). H-NMR spectra
were recorded on ACF400 NMR spectrometer (Bruker Spectroscopic Instruments Co., Rheinstetten, Germany) with
CDCl3 as solvent. Coupling constants (J) and chemical shifts (δ) values were reported in hertz and ppm, respectively.
High resolution mass spectra were carried out with Synapt G2 HDMS (Waters Co., America). Monitoring of the
reactions was performed using silica gel TLC plates (Qingdao Haiyang Chemical Co., Ltd., Qingdao, China) and
visualized under ZF-2 ultraviolet analyzer (Shanghai Anting electronic Instrument Factory, Shanghai, China) under
light 254 nm. Column chromatography was performed on silica gel (200 to 300 mesh).
2.2. Chemistry
2.2.1. Procedure for the synthesis of Andrographolide 14-OH esterification derivatives (Ia-f)
Preparation of 3,19-isopropylidene andrographolide (1)
Following a literature method, andrographolide (7.0 g, 20 mmol), 4-toluenesulfonic acid pyridine salt (PPTS)
(0.5 g, 2 mmol), 2,2-dimethoxypropane(18 mL) were added into dichloromethane (20 mL), and then the reaction
mixture was stirred at 45 °C for 2h (the reaction progress was monitored by TLC with UV detection). The reaction
mixture was evaporated under reduced pressure and washed with petroleum ether to give 3,19-isopropylidene
andrographolide (1) crude product as a white solid. ESI-MS（m/z）:391.2 [M+H]+.
Preparation of 14-OH esterification 3,19-isopropylidene andrographolide (2a-f)
3,19-isopropylidene andrographolide (1, 0.39 g, 1 mmol), triethylamine (0.42 mL) were added into
dichloromethane (15 mL) and chilled to 0 °C. A solution of DL-thioctic acid/ p-chlorobenzoyl chloride / pivaloyl
chloride / acryloyl chloride / carbonochloridic acid / acetyl chloride (1 mmol) dissolved in dichloromethane (5 mL)
was added dropwise, then the reaction mixture was moved to room temperature and stirred for 3-6h. The reaction
This article is protected by copyright. All rights reserved

mixture was washed with saturation sodium bicarbonate and water, dried over Na2SO4, and concentrated in vacuo to
get 14-OH esterification 3,19-isopropylidene andrographolide (2a-f) crude product.
Preparation of 14-OH esterification andrographolide (Ia-f)
Dilute acetic acid (1 mmol) were added into a solution of 14-OH esterification 3,19- isopropylidene
andrographolide (2a-f) in THF (10 mL), and the reaction mixture was stirred at room temperature for 2 h. The reaction
mixture was evaporated under reduced pressure and dissolved with ethyl acetate, washed with saturation sodium
bicarbonate, water and saturation salt solution, successively, dried over Na2SO4, and concentrated in vacuo. The
residue was purified by column chromatography using ethyl acetate/petroleum ether 2:1 to yield products Ia-f.
14-DL-thioctic acid ester andrographolide (Ia)
1
Light yellow solid; 67% yield; m.p. 65.1-67.3 °C, H-NMR (CDCl3, δ ppm): 7.04 (t, 1H) , 5.93 (d, J = 5.9 Hz,
1H), 4.88 (s, 1H), 4.88 (s, 1H), 4.50 (s, 1H), 4.23 (d, 1H), 4.16 (d, 1H), 3.55-3.44 (m, 2H), 3.33 (d, J = 10.7 Hz, 1H),
3.25-3.05 (m, 2H), 2.76(s, 3H), 2.41-2.31 (m, 6H), 1.97-1.80 (m, 6H), 1.68-1.60 (m, 5H), 1.46 (m, 2H), 1.40 –1.34
(m,1H), 1.25 (s, 4H), 0.67 (s , 3H). ESI-MS (m/z):561.2[M+Na]+.
14-P-chlorobenzoyl acid ester andrographolide (Ib)
White solid; 59% yield; m.p. 98.5-100.9 °C, 1H-NMR (CDCl3, δ ppm): 7.96 (d, J = 8.6 Hz, 2H), 7.45 (d, J = 8.6
Hz, 2H), 7.09 (t, J = 6.2 Hz, 1H), 6.17 (d, J = 5.8 Hz, 1H), 4.84 (s, 1H), 4.67-4.59 (m, 1H), 4.48 (s, 1H), 4.37 (d, J =
1.7 Hz, 1H), 4.13 (d, 1H), 3.77-3.66 (m, 1H), 3.47 (dd, J = 13.7, 5.2 Hz, 3H), 3.30 (d, J = 11.0 Hz, 1H), 2.62 -2.27 (m,
4H), 2.24-1.90 (m, 6H), 1.89-1.59 (m, 5H), 1.42-1.24 (m, 2H), 1.24 (s, 4H), 0.59 (s, 3H). ESI-MS(m/z):
511.2[M+Na]+.
14-pivaloyl acid ester andrographolide (Ic)
Light yellow solid;71% yield;m.p. 157.8-158.6 °C, 1H-NMR (CDCl3, δ ppm): 6.94 (t, J = 6.6Hz, 1H), 5.81 (d, J
= 6.2 Hz, 1H), 4.82 (s, 1H), 4.54 (m, 1H), 4.42 (s, 1H), 4.14 (m, 1H), 4.07 (d, J = 11.2Hz, 1H), 3.48-3.35 (m, 3H),
3.25 (d, J = 11.2 Hz, 1H), 2.43-2.21 (m, 3H), 2.05-1.55 (m, 8H), 1.46-1.19 (m, 4H), 1.16 (s, 9H), 0.58 (s, 3H).
ESI-MS (m/z): 457.3[M+Na]+.
14-acryloyl acid ester andrographolide (Id)
1
White solid;74% yield; m.p. 127.6-128.3 °C, H-NMR (CDCl3, δ ppm): 7.03 (t, J = 5.9 Hz, 1H), 6.48 (dd, J =
17.3, 3.3 Hz, 1H), 6.15 (dd, J = 17.2, 10.4Hz, 1H), 6.01 (d, J = 5.2 Hz, 1H), 5.96 (d, J = 10.5 Hz, 1H), 4.86 (s, 1H),
4.59 (dd, J = 11.2, 6.2 Hz, 1H), 4.49 (s, 1H), 4.28 (d, J = 11.2 Hz, 1H), 4.17 (d, J = 11.0 Hz, 1H), 3.55–3.41 (m, 1H),
3.31 (d, J=11.0Hz,1H), 2.53–2.31 (m, 3H), 2.03–1.67 (m, 6H), 1.44–1.06 (m, 7H), 0.65 (s,3H). ESI-MS (m/z):
427.2[M+Na]+.
14-carbonochloridic acid ester andrographolide (Ie)
This article is protected by copyright. All rights reserved

1
Yellow solid;47% yield;m.p. 78.8-80.2 °C, H-NMR (CDCl3, δ ppm): 7.07 (t, J = 6.6 Hz, 1H), 5.82 (d, J = 5.7
Hz, 1H), 4.82 (s, 1H), 4.60-4.54 (m, 1H), 4.51 (s, 1H), 4.24 (d, J = 9.4Hz, 2H), 4.20 (d, J = 2.6 Hz, 1H), 4.12 (d, J =
7.1 Hz, 1H), 3.48 (dd, J = 10.9, 4.7 Hz, 2H), 3.35 (d, J = 9.5 Hz, 1H), 3.29 (d, J = 21.5 Hz, 2H), 2.46- 2.27 (m, 6H),
1.64-1.42 (m, 3H), 1.34 (t, 3H), 1.25 (s, 4H), 0.67 (d, J = 8.0 Hz, 3H). ESI-MS (m/z):445.2[M+H]+.
14-acetyl acid ester andrographolide (If)
White solid;57% yield;m.p. 170.3-172.5 °C, 1H-NMR (CDCl3, δ ppm): 6.94 (t, J = 6.5 Hz, 1H), 5.85 (d, J = 5.7
Hz, 1H), 4.81 (s, 1H), 4.48 (m, J = 11.2 Hz, 1H), 4.43 (s, 1H), 4.18 (d, J = 1.2 Hz, 1H), 4.12 (d, J = 10.7 Hz, 1H),
3.49-3.36 (m, 1H), 3.26 (d, J = 10.6 Hz, 1H), 3.11(m，2H),2.55 (s, 2H), 2.06 (s, 3H), 2.35（m,3H）1.80 (m, 6H), 1.18
(s, 4H), 0.60 (s, 3H). ESI-MS (m/z): 415.2[M+Na]+.
2.2.2. Procedure for the synthesis of andrographolide derivatives (II-IV)
Preparation of dehydrated andrographolide (II)
Andrographolide (17.5 g, 0.05 mol) was dissolved in pyridine (25 mL) at 100 °C and then succinic anhydride (5
g, 0.05 mol) was added. The reaction mixture was reflux and stirring at 105 °C for 2h. The reaction mixture was
cooled and poured into 80 mL of water and stirring to crystallize. Placed for 6 h, washed the solid with water, dried
over, purified by column chromatography using ethyl acetate/ petroleum ether 1:1 to yield products (II). White solid;
88% yield; m.p. 200.1-202.3 °C, 1H-NMR (CDCl3, δ ppm): 7.11 (s, 1H), 6.80 (dd, J = 15.8, 10.1 Hz, 1H), 6.05 (d, J =
15.8 Hz, 1H), 4.75 (s, 2H), 4.71 (s, 1H), 4.46 (s, 1H), 4.15 (d, J = 11.1 Hz, 1H), 3.46–3.36 (m, 1H), 3.28 (d, J = 11.1
Hz, 1H), 2.39 (dd, J = 13.6, 1.4 Hz, 1H), 2.28-1.27 (m, 8H), 1.19 (s, 3H), 0.74 (s, 3H). ESI-MS (m/z): 332.2[M+H]+.
Preparation of 19-O-succinate ester dehydrated andrographolide (III)
Dehydrated andrographolide (II, 0.332 g,1 mmol), succinic anhydride (0.202 g,
2
mmol),
4-dimethylarminopyridine (DMAP) were dissolved in dichloromethane (20 mL). Then, the whole solution was stirred
under room temperature for 8 h. The reaction mixture was washed with water for several times firstly and saturation
salt solution for three times seconrly, dried over, purified by column chromatography using ethyl acetate/ petroleum
ether 5:1 to yield products (III). Yellow solid; 90% yield; m.p.132.1-133.7 °C, 1H-NMR (CDCl3, δ ppm):7.17 (s, 1H),
6.88 (dd, J = 15.6, 10.0 Hz, 1H), 6.12 (d, J = 15.8 Hz, 1H), 4.81 (s, 2H), 4.55 (s, 1H), 4.42 (d, J = 11.6 Hz, 1H), 4.09
(d, J = 6.7 Hz, 1H), 3.72 (d, J = 7.0 Hz, 1H), 2.67-2.63(m, 2H), 2.33(d, J=10.0Hz,1H), 2.10–2.04(m, 2H),
1.77-1.39(m, 8H), 0.99(d, J= 6.7Hz, 3H), 0.84(s, 3H). ESI-MS (m/z):455.2[M+Na]+.
Preparation of 3-O- chloroacetyl -19-O-(2-chloroacetyl)dehydrated andrographolide dehydrated andrographolide(3)
Dehydrated andrographolide (II, 0.664 g, 2 mmol), triethylamine(0.6 mL, 4.3 mmol), were dissolved in
dichloromethane (15 mL) and chilled to 0 °C. Chloroacetyl chloride (0.33 mL, 4 mmol) was dissolved in
dichloromethane (5 mL) added dropwise, and the reaction mixture was moved to room temperature stirred 1h. The
reaction mixture was washed with water and saturation salt solution, concentrated in vacuo to get 3-O- chloroacetyl
-19-O-(2-chloroacetyl) dehydrated andrographolide (3) crude product. ESI-MS (m/z):507.2 [M+Na]+.
This article is protected by copyright. All rights reserved

Preparation of andrographolide derivatives (IVa-d)
3-O-chloroacetyl-19-O-(2-chloroacetyl) dehydrated andrographolide (3, 0.50 g, 1 mmol), diisopropylethylamine
(0.60 mL, 3 mmol) were dissolved in tetrahydrofuran (10mL) and chilled to 0 °C. Morpholine/piperidine/pyrrolidine/
(4-hydroxypiperidino) acetyl was dissolved in tetrahydrofuran (10 mL) added dropwise, and the reaction mixture was
moved to room temperature stirred 3-8 h. The reaction mixture was evaporated under reduced pressure and dissolved
with ethyl acetate, washed with water and saturation salt solution, successively, dried over Na2SO4, purified by
column chromatography using ethyl acetate / petroleum ether 10:1 to yield products IVa-d.
3-O-morpholinoacetyl-19-O-(2-morpholinoacetyl) dehydrated andrographolide (IVa)
1
Yellow solid; 77% yield ;m.p.121.7-123.5 °C, H-NMR (CDCl3, δ ppm): 7.17 (s, 1H), 6.93 (dd, J = 15.7, 10.1
Hz, 1H), 6.14 (d, J = 15.8 Hz, 1H), 4.82 (s, 2H), 4.81 (s, 1H), 4.69 (dd, J = 10.5, 5.8 Hz, 1H), 4.57 (s, 1H), 4.37 (d, J
= 11.8 Hz, 1H), 4.29 – 4.22 (m, 1H), 3.78 – 3.72 (m, 8H), 3.19 (s, 4H), 2.59 (d, J = 3.5 Hz, 8H), 2.11–1.40 (m, 6H),
1.37–1.12 (m, 4H), 1.01 (s, 3H), 0.90 (s, 3H). ESI-MS (m/z):587.3[M+H]+.
3-O-piperidinoacetyl-19-O-(2-piperidinoacetyl) dehydrated andrographolide (IVb)
Yellow solid; 80% yield; m.p.119.6-121.5 °C, 1H-NMR(CDCl3,δ ppm): 7.10 (s, 1H), 6.86 (dd, J = 15.7, 10.0 Hz,
1H), 6.06 (d, J = 15.8 Hz, 1H), 4.75 (s, 2H), 4.73 (s, 1H), 4.64 – 4.57 (m, 1H), 4.49 (s, 1H), 4.29 (d, J = 11.8 Hz, 1H),
4.17 (d, J = 11.8 Hz, 1H), 3.13 (d, J = 7.9 Hz, 4H), 2.44 (d, J = 26.6 Hz, 8H), 1.80–1.22 (m, 22H), 0.94 (s, 3H), 0.82
(s, 3H).ESI-MS(m/z):583.2[M+H]+.
3-O-pyrrolidinoacetyl -19-O-(2-pyrrolidinoacetyl) dehydrated andrographolide (IVc)
Yellow solid; 67%yield;m.p.127.8-129.1 °C,1H-NMR(CDCl3,δ ppm):7.27(s,1H),6.13 (dd, J = 10.5, 4.4 Hz,
1H), 6.06 (d, J = 6.9 Hz, 1H), 4.83 – 4.74 (m, 2H), 4.68 (dd, J = 11.0, 4.9 Hz, 2H), 4.42 – 4.32 (m, 3H), 4.33 (d, J =
11.6 Hz, 1H), 4.29 – 4.19 (m, 2H), 4.18 – 4.00 (m, 1H), 3.33 (s, 4H), 2.66 (s, 8H), 1.82 (s,8H), 1.74 – 1.47 (m,6H),
1.01 (s, 3H), 0.80 (s, 3H). ESI-MS (m/z):555.3[M+H]+.
3-O-(4-hydroxypiperidino) acetyl-19-O-[2-(4-hydroxypiperidino) acetyl] dehydrated andrographolide (IVd)
1
Yellow solid; 60%yield;m.p.107.2-108.8 °C, H-NMR (CDCl3,δ ppm): 7.17 (s, J = 1.9 Hz, 1H), 6.98 – 6.87
(dd, 1H), 6.16 (d J = 24.5Hz, 1H), 4.82 (d, J = 1.7 Hz, 2H), 4.81 (d, J = 1.3 Hz, 1H), 4.68 (dd, J = 10.7, 5.8 Hz, 1H),
4.56 (d, J = 1.2 Hz, 1H), 4.37 (d, J = 11.8 Hz, 1H), 4.25 (d, J = 11.8 Hz,1H), 3.78 – 3.69 (m, 2H), 3.23 (s, 2H),
3.32(m,4H),2.31 ( m, 8H), 1.72 – 1.66 (m, 4H), 1.62 – 1.53 (m, 4H), 1.35-1.21(m,8H),1.01 (s, 3H), 0.89 (s,
3H).ESI-MS(m/z):615.3[M+H]+.
2.2.3. Procedure for the synthesis of Andrographolide derivatives (V-IX)
Preparation of 19-O- triphenylmethyl dehydrated andrographolide (4)
This article is protected by copyright. All rights reserved

Dehydrated andrographolide (II, 2 g, 6 mmol), 4-nethylmorpholine (0.76 ml, 6.9 mmol) were dissolved in
dichloromethane (20 mL). A solution of trityl chloride (1.7 g, 7.8 mmol) was dissolved in dichloromethane (10 mL)
was added dropwise, ang then the reaction was stirred overnight. The reaction mixture was washed with water and
saturation sodium bicarbonate, dried over Na2SO4, purified by column chromatography using ethyl acetate / petroleum
ether 1:2 to yield products (5). ESI-MS（m/z）:575.2 [M+H]+.
Preparation of 3-O-succinate ester 19-O- triphenylmethyl dehydrated andrographolide (5)
19-O-triphenylmethyl dehydrated andrographolide (4, 0.610 g, 1 mmol), succinic anhydride (0.101 g, 1 mmol),
4-dimethylarminopyridine (DMAP) were dissolved in dichloromethane (20 mL). Then, the whole solution was heated
to reflux for 24h. After cooling the reaction mixture was washed with water and saturation salt solution, dried over
Na2SO4, and concentrated in vacuo to get 3-O-succinate ester 19-O- triphenylmethyl dehydrated andrographolide (5)
crude product.
Preparation of 3-O-succinate ester dehydrated andrographolide (V)
3-O-succinate ester 19-O-triphenylmethyl dehydrated andrographolide (5, 0.35 g, 0.5 mmol) was dissolved in
dichloromethane (10 mL) and then formic acid was added to the reaction. Then the reaction mixture was stirred under
room temperature for 2h, washed with saturation sodium bicarbonate, water and saturation salt solution, successively,
dried over Na2SO4, purified by column chromatography using ethyl acetate / petroleum ether 2:1 to yield products
(V). White solid; 63% yield; m.p. 121.7-123.3 °C, 1H-NMR (CDCl3, δ ppm): 7.66 (s, 1H), 6.77 (dd, J = 15.8, 10.1 Hz,
1H), 6.15 (d, 1H), 4.89 (s, 2H), 4.50 (dd, J = 12.0, 4.1 Hz, 1H), 4.43 (s, 1H), 4.17 – 3.99 (d, 1H), 3.54 (d, J = 11.4 Hz,
1H), 2.46 (t, 2H), 2.36 (d, J = 12.6 Hz, 1H), 2.00 – 1.90 (t,2H),1.60–1.17 (m, 8H), 0.92 (s, 3H),0.86(s,3H). ESI-MS
(m/z):455.2 [M+Na]+.
Preparation of 3-O- acetyl-19-O- triphenylmethyl dehydrated andrographolide (6)
Acetic anhydride (1.98 mL, 2.1), 19-O-triphenylmethyl dehydrated andrographolide (4,0.6 g, 1.05 mmol),
dimethylpyridine (0.112 g,1 mmol) were dissolved in dichloromethane (25 mL) and chilled to 0 °C.
1-ethylene-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC. HCl, 0.4 g, 2.1 mmol) was dissolved in
dichloromethane (15 mL) added dropwise, and the reaction mixture was moved to room temperature stirred overnight.
The reaction mixture was washed with saturation sodium bicarbonate, water and saturation salt solution, successively,
dried over Na2SO4, purified by column chromatography using ethyl acetate / petroleum ether 1:2 to yield products (6).
ESI-MS (m/z): 639.2 [M+Na]+.
Preparation of 3-O- acetyl dehydrated andrographolide (VI)
3-O- acetyl-19-O- triphenylmethyl dehydrated andrographolide (6, 0.62 g, 1 mmol) was dissolved in
dichloromethane (10 mL), formic acid (2 mL) was added to the reaction, and the reaction mixture was stirred under
room temperature for 2h. The reaction mixture was washed with saturation sodium bicarbonate, water and saturation
salt solution, successively, dried over Na2SO4, purified by column chromatography using ethyl acetate / petroleum
This article is protected by copyright. All rights reserved

ether 1:2 to yield products (VI). White solid; 87% yield; m.p. 103.7-106.3 °C, 1H-NMR (CDCl3, δ ppm): 7.17 (s, 1H),
6.89 (dd, J = 15.8, 9.7 Hz, 1H), 6.12 (d, J = 15.8 Hz, 1H), 4.81 (s, 2H), 4.80 (s, 1H), 4.67 (dd, J = 11.6, 4.6 Hz, 1H),
4.55 (s, 1H), 4.35 (d, J = 11.7 Hz, 1H), 4.17 (d, 1H), 2.06 (s,3H), 1.77 – 1.49 (m, 8H), 0.97 (s, 3H), 0.78 (s,
3H).ESI-MS (m/z): 397.2[M+Na]+.
Preparation of 3-O- acetyl-19-O-(2-chloroacetyl) dehydrated andrographolide(7)
3-O- acetyl dehydrated andrographolide (VI, 0.62 g, 1 mmol), triethylamine (0.6mL, 4.3mmol), were dissolved
in dichloromethane (10 mL) and chilled to 0 °C. Chloroacetyl chloride (0.33 mL, 4 mmol) was dissolved in
dichloromethane (5mL) added dropwise, and the reaction mixture was moved to room temperature stirred 2h. The
reaction mixture was washed with water and saturation salt solution, dried over Na2SO4, and concentrated in vacuo to
get 3-O-acetyl-19-O-(2-chloroacetyl) dehydrated andrographolide (7) crude product.
Preparation of andrographolide derivatives (VIIa-d)
3-O-acetyl-19-O-(2-chloroacetyl) dehydrated andrographolide (7, 0.54 g, 1.23 mmol), diisopropylethylamine
(0.64 mL, 3.69 mmol) were dissolved in tetrahydrofuran (10mL) and chilled to
0
°C.
Morpholine/piperidine/pyrrolidine/ (4-hydroxypiperidino) acetyl was dissolved in tetrahydrofuran (10 mL) added
dropwise, and the reaction mixture was moved to room temperature stirred 3-8 h. The reaction mixture was
evaporated under reduced pressure and dissolved with ethyl acetate, washed with water and saturation salt solution,
successively, dried over Na2SO4, purified by column chromatography using ethyl acetate / petroleum ether 2:1 to yield
productsVIIa-d.
3-O-acetyl-19-O-(2-morpholinoacetyl) dehydrated andrographolide (VIIa)
Light yellow solid; 73% yield; m.p. 113.7-115.3 °C, 1H-NMR (CDCl3, δ ppm): 7.17 (s, 1H), 6.94 (dd, J = 15.7,
10.1 Hz, 1H), 6.14 (d, J = 15.7 Hz, 1H), 4.82 (s, 2H), 4.61 (s,1H), 4.57 (s, 1H), 4.43 (dd, J = 11.8,4.5 Hz, 1H), 4.28 (d,
J = 11.6 Hz, 1H), 4.12 (d, J = 14.3Hz, 1H), 3.83 (t, 4H), 3.28 (s, 2H), 2.66-2.15 (t, 4H), 2.05 (s, 3H),
2.05-1.32(m,8H),1.03 (s, 3H), 0.89 (s, 3H).ESI-MS (m/z): 502.2[M+H]+.
3-O-acetyl-19-O-(2-piperidinoacetyl) dehydrated andrographolide (VIIb)
Yellow solid; 60% yield; m.p. 139.2-141.6 °C, 1H-NMR (CDCl3, δ ppm):. 7.16 (s, 1H), 6.93 (dd, J = 15.8, 10.1
Hz, 1H), 6.13 (d, J = 15.8 Hz, 1H), 4.81 (s, 2H), 4.61 ( s, 1H), 4.56 (s, 1H), 4.42 (dd, J = 11.7,4.4 Hz, 1H), 4.24 (d, J =
17.1Hz, 1H), 3.41 (s, 2H), 2.51 ( m, 4H), 2.04 (s, 3H), 1.72-1.22 (m, 14H), 1.03 (s, 3H), 0.89 (s, 3H).ESI-MS (m/z):
500.3[M+H]+.
3-O-acetyl-19-O-(2-pyrrolidinoacetyl) dehydrated andrographolide (VIIc)
1
Yellow solid; 61% yield;m.p.123.8-125.1 °C, H-NMR (CDCl3, δ ppm): 7.16 (s, 1H), 6.93 (dd, J = 15.8, 10.1
Hz, 1H), 6.13 (d, J = 15.8 Hz, 1H), 4.81 (s, 2H), 4.61 (dd, J = 9.4, 4.8 Hz, 1H), 4.56 (s, 1H), 4.42 (s, 1H), 4.26 (d, J =
11.8 Hz, 1H), 3.41 (s, 2H), 2.77 (t, 4H), 2.46 (d, J = 13.6 Hz, 1H), 2.04 (s, 3H), 1.94-1.79 (m, 4H), 1.72-1.29 (m, 8H),
1.03 (s, 3H), 0.89 (s, 3H).ESI-MS (m/z): 486.2[M+H]+.
This article is protected by copyright. All rights reserved

3-O-acetyl-19-O-[2-(4-hydroxypiperidino) acetyl] dehydrated andrographolide (VIId)
1
Yellow solid; 65% yield;m.p.87.5-89.2 °C, H-NMR (CDCl3, δ ppm): 7.17 (s, 1H), 6.93 (dd, J = 15.8, 10.1 Hz,
1H), 6.13 (d, J = 15.8 Hz, 1H), 4.81 (d, J = 1.4 Hz, 3H), 4.64-4.58 (m, 1H), 4.56 (s, 1H), 4.40 (d, J = 11.8 Hz, 1H),
4.24 (d, J = 11.8 Hz, 1H), 3.22 (d, J = 2.0 Hz, 2H), 2.88-2.83 (m, 1H), 2.51-2.31 (m, 4H), 2.04 (s, 3H), 2.00-1.90 (m,
2H), 1.88-1.32 (m, 10H), 1.02 (s, 3H), 0.89 (s, 2H).ESI-MS (m/z): 516.2[M+H]+.
Preparation of 3-oxo-19-O- triphenylmethyl dehydrated andrographolide (8)
19-O-triphenylmethyl dehydrated andrographolide (4, 3 g, 5 mmol), was dissolved in dichloromethane (80 mL)
and then pyridinum dichromate (4.3 g, 10 mmol) was added to the solution. The reaction mixture was refluxed for
48h. After cooling, the whole solution was filtered, and the filtrate was washed with water and saturation salt solution,
dried over Na2SO4, and concentrated in vacuo to get 3-oxo-19-O- triphenylmethyl dehydrated andrographolide (8)
crude product. ESI-MS (m/z): 573.2 [M+H]+.
Preparation of andrographolide derivatives (IXa-d)
The process of producing VIII-IXa-d was similar to VI-VIIa-d.
3-oxo-dehydrated andrographolide (VIII)
White solid; 82% yield;m.p. 92.1-93.7 °C, 1H-NMR (CDCl3, δ ppm): 7.19 (s, 1H), 6.94 (dd, J = 15.8, 10.1 Hz,
1H), 6.15 (d, J = 15.8 Hz, 1H), 4.85 (d, J = 1.5 Hz, 1H), 4.83 (s, 2H), 4.61 (s, 1H), 3.92 (d, J = 11.2 Hz, 1H), 3.54 (d, J
= 11.2 Hz, 1H), 2.43 (d, J = 3.1 Hz, 1H), 2.16 – 1.47 (m, 8H), 1.23 (s, 3H), 1.01 (s, 3H).ESI-MS (m/z): 331.2[M+H]+.
3-oxo-19-O-(2-morpholinoacetyl) dehydrated andrographolide (IXa)
Yellow solid; 77% yield;m.p. 80.5-81.8 °C, 1H-NMR (CDCl3, δ ppm): 7.18 (s, 1H), 6.97 (dd, J = 15.8, 10.1 Hz,
1H), 6.14 (d, J = 15.8 Hz, 1H), 4.83 (s, 2H), 4.72 (s, 1H), 4.63 (s, 1H), 4.58（d,J=11.3Hz,1H）3.98 (d, J = 11.3 Hz,
1H), 3.72 (t,J=4.5Hz, 4H), 3.16 (s, 2H), 2.78 (d, J = 14.9, 5.8 Hz, 1H), 2.47 (m, 4H), 2.14-1.36 (m, 8H), 1.18 (s, 3H),
1.14 (s, 3H). ESI-MS (m/z): 458.2[M+H]+.
3-oxo-19-O-(2-piperidinoacetyl) dehydrated andrographolide (IXb)
Yellow solid; 68% yield; m.p. 75.3-76.9 °C, 1H-NMR (CDCl3, δ ppm): 7.12 (s, 1H), 6.90 (dd, J = 15.8, 10.1 Hz,
1H), 6.07 (d, J = 15.8 Hz, 1H), 4.76 (s, 2H), 4.62 (s, 1H), 4.56 (s, 1H), 3.91 (d, J = 11.3 Hz, 1H) , 3.06 (s, 2H), 2.41
(m, J = 5.3 Hz, 4H), 1.81 –1.50 (m, 8H), 1.47–1.20 (m, 4H), 1.11 (s, 3H), 1.07 (s, 3H). ESI-MS (m/z): 456.2[M+H]+.
3-oxo-19-O-(2-pyrrolidinoacetyl) dehydrated andrographolide (IXc)
Yellow solid; 71% yield; m.p. 71.2-73.1 °C, 1H-NMR (CDCl3, δ ppm): 7.12 (s, 1H), 6.90 (dd, J = 15.7, 10.0 Hz,
1H), 6.09 (d, J = 13.0 Hz, 1H), 4.76 (s, 2H), 4.63 (s, 1H), 4.56 (d, J = 0.9 Hz, 1H), 4.35 (s, 1H), 3.93 (d, J = 11.3 Hz,
1H), 3.22 (s, 2H), 2.61 (m , 4H), 2.33 (d, J = 10.1 Hz, 1H), 2.27 – 1.75 (m, 8H), 1.69 – 1.58(m, 4H),
1.42-1.33(t,4H)1.11 (s, 3H), 1.07 (s, 3H). ESI-MS (m/z): 442.2[M+H]+.
This article is protected by copyright. All rights reserved

3-oxo-19-O-[2-(4-hydroxypiperidino) acetyl] dehydrated andrographolide (IXd)
Yellow solid; 65% yield;m.p. 76.4-77.8 °C, 1H-NMR (CDCl3, δ ppm): 7.12 (s, 1H), 6.90 (dd, J = 15.8, 10.1 Hz,
1H), 6.07 (d, J = 15.8 Hz, 1H), 4.90(s,2H),4.76 (s, J = 1.2 Hz, 1H), 4.64 (d, J = 11.3 Hz, 1H), 4.56 (s, 1H), 3.91 (d, J =
11.3 Hz, 1H), 3.70 – 3.60 (m, 1H), 3.12 (s, 2H), 2.81 – 2.63 (m, 4H), 2.44 (d, J = 10.8 Hz, 1H), 2.27 (m, 6.5 Hz, 4H),
1.81 – 1.42 (m, 8H), 1.11 (s, 3H), 1.07 (s, 3H). ESI-MS (m/z): 472.2[M+H]+.
2.3. Extracorporal nephrotoxicity assay
HK-2 cells were purchased from Cell Center of the Chinese academy of medical sciences (Beijing, China), and
the cells were previously derived from the ATCC. HK-2 cells were cultured in Dulbecco’s Modified Eagle’s Medium
nutrient mixture F-12 (DMEM/F12 medium) supplemented with 10% fetal calf serum (FCS) at 37 °C in a humidified
atmosphere of 5% CO2 in air. The in vitro nephrotoxicity of andrographolide and all synthesized compounds was
evaluated against HK-2 cells by MTT method. Different concentrations of drugs acted on cell for 24h. Then the
optical density (OD) value was detected at wavelength of 490 nm using microplate reader iMark (Molecular Devices,
LLC, Sunnyvale, CA, USA).. Cell survival rate (%) = (OD of administration group−OD of blank group)/(OD of
control group−OD of blank group) × 100%. The value of inhibitory concentration 50 (IC50) was calculated by
GraphPad Prism 5 software (San Diego, CA, USA).
2.4. QSAR study
The QSAR model was constructed with SYBYL6.9 (Tripos Associates, St. Louis, MO, USA). The IC50 for each
compound was transformed into negative logarithm of IC50 (pIC50). The pIC50 values were employed as dependent
variable for the QSAR analysis. The molecular structures of synthesized compounds were generated and then divided
into test (4 compounds) and training (20 compounds) datasets using a random selection method(J.T. & K., 2006). The
training set is used to build the QSAR model, and the test set is used to evaluate the predictability of the generated
models.
Molecules were digitized and optimized. Details are as follows: The compounds are transformed into 3D
structures by Chem 3D software, and then imported into SYBYL-X software. Energy minimum of the compound
molecules were set by Tripos force field, Gradient value as a standard termination is reduced to 0.005. If the
calculation step difference is less than 0.005 for two consecutive times, the calculation is terminated, and the
maximum repetition number (Max, Iterations) is increased to 1000. The molecular load type is Gasteiger-Huckel
charge, and the rest parameters are kept as the default values of the software. Calculate and get the optimized
molecules, which are saved as .mol2 file.
In this paper, the 14-acetyl acid ester andrographolide (compound If)(Figure 1), the most active molecule of the
series with the lowest IC50 value, was selected as the common scaffold for molecular alignment. All other synthesized
compounds were aligned with the 14-acetyl acid ester andrographolide (If). The comparative molecular field analysis
(CoMFA) and comparative molecular similarity indices analysis (COMSIA) were used to get 3D-QSAR model. In
CoMFA, the steric and electrostatic fields were calculated and in COMSIA steric, electrostatic, hydrophobic,
This article is protected by copyright. All rights reserved

hydrogen-bond donor, hydrogen-bond acceptor fields were calculated. The relationship was quantified using partial
least squares (PLS) with biological activity (pIC50 value) as the dependent variable and CoMFA/CoMSIA descriptor
as the independent variable. The cross-validation of the model was carried out using the leave-one-out method.
In order to improve the predicted quality of the model, conformational search method is used to select the
compound molecular superposed conformation. According to the superposition of each molecule and template
molecule, the new superposition conformation was selected and imported into the Training table. After refreshing the
Training table, the steps of CoMFA and COMSIA were repeated, and finally get the 3D-QSAR model with reasonable
cross-validation coefficient q2 value.
3. RESULTS AND DISCUSSION
3.1. Chemistry
The synthesis of andrographolide 14-OH esterification derivatives (Ia-f) was outlined in Scheme 1. The first
reaction of synthesizing 1 was followed as literature description (Jada et al., 2007), the second step was altered by
adding the DL-thioctic acid, cathyl chloride and different acyl chloride. The final step of the reaction yielded the end
product by deprotection under acidic conditions.
The synthesis of II-IV was outlined in Scheme 2. Andrographolide was dehydrated under alkaline conditions to
form compound II. Target compound III was obtained from compound II through a one-step reaction with succinic
anhydride. To get compound IVa-d, firstly we used chloroacetyl chloride to connect to the two hydroxyl groups, then
added different amines to the system.
The synthesis of V-IX was outlined in Scheme 3. The intermediate product 4 was synthesized through the
reaction of dehydroandrographolide (II) with triphenylmethyl chloride. Target compound 5 was synthesized like III.
Then removing protection under acidic conditions and getting the target compound V. We oxidized the hydroxyl
group at position 3 to obtain compound 6 and 8, then removed protection under acidic conditions, connected the
hydroxyl group at position 19 with chloroacetyl chloride. Finally, the target compounds VIIa-d and IXa-d were
synthesized with different amines. The physical characteristics of target compounds were determined by 1H NMR and
HRMS.
3.2. Extracorporal Nephrotoxicity
As shown in Table 1, MTT (3-(4, 5-dimethyl thiazolyl-2)-2, 5-diphenyltetrazolium bromide) assay was carried
out in HK-2 cells after treatment for 24h, which revealed the nephrotoxicity of all the compounds. Exhilaratingly,
most of the title compounds emerged less nephrotoxicity than andrographolide. The IC50 of andrographolide 14-OH
esterification derivatives (Ia-f) were less than that of andrographolide, suggesting that 14 hydroxyl esterification
enhanced nephrotoxicity of andrographolide to some extent. Meanwhile our results showed that the nephrotoxicity of
elimination of 14-OH derivatives were significantly reduced. The IC50 of compound II was about 458.0 μmol·L-1 and
obvious higher than andrographolide. On the basis of compound II, the nephrotoxicity of 3-OH acetylation derivatives
(VI) was significantly reduced. Then connected 19-OH with different amine nephrotoxicity was increased. To the
This article is protected by copyright. All rights reserved

contrary, oxidized 3-OH to the keto group, could reduce the nephrotoxicity, restructured 19-OH had no significant
change in nephrotoxicity. 3-OH or 19-OH (V, III) connected with succinic anhydride nephrotoxicity had a large
degree of change and IC50 were up to 1300 and 1985 μmol·L-1, respectively.
3.3. Statistical Analyses of CoMFA and CoMSIA Models
The COMFA and COMSIA models of 3D-QSAR were generation based on biological activity. The relevant
results of the models were listed in Table 2. The cross-validation coefficient q2 and the non-cross-validation
coefficient R2 represents the robustness and fitness of the model to reproduce the experimental value, and the value of
criteria are q2>0.5 and R2>0.8.
The partial least square (PLS) results of the COMFA and COMSIA models were showed in Table 2. The
optimal COMFA model yielded a q2=0.639 with an optimal number of principal components (ONC) of 2, R2=0.951,
SEE=0.158 and F value of 163.747. The contribution of steric and electrostatic fields was 33.6% and 66.4%,
respectively. The best COMSIA model yielded a q2=0.569 with an ONC of 2, R2 of 0.857, SEE=0.268 and F value of
51.096. The contribution of steric, electrostatic, hydrophobic, hydrogen-bond-donor and hydrogen-bond acceptor were
6.4%, 24.4%, 14.8%, 20.3% and 34.2%, respectively. Based on these field contributions, the electrostatic field was the
most important field in the CoMFA model, whereas the hydrogen-bond acceptor field was the most important field in
the CoMSIA model. All the parameters in the Table 2 indicated that the CoMFA and CoMSIA models were robust
and stable.
Further, based CoMFA and CoMSIA models were used to predict the activities of the training and test dataset
compounds. The value of experimental versus predicted extracorporal nephrotoxicity for COMFA and COMSIA
models were shown in Table 3 and the scatter plots were illuminated in Figure 2. As shown in Figure 2, all points are
situated around the diagonal lines, and there is no obvious deviated point present on them. The correlation coefficients
R2 of the COMFA and COMSIA models were 0.9506 and 0.8574, respectively. The R2 of COMFA was higher than its
COMSIA, which indicated COMFA model had higher predictive ability. In general, the models had a good correlation
and further confirmed that COMFA and COMSIA models were predictive.
3.4 CoMFA Contour Maps
The COMFA and COMSIA contour with the template molecule If were generated. Those contours depicted
default contribution levels. The COMFA steric and electrostatic contour maps were shown in Figure 3. The green
contours indicated that introduction of bulky groups would increase activities, in contrast, the yellow patches
illustrated where bulky groups addition would decrease activities（Figure 3A）. The blue contour in the COMFA
electrostatic field indicates the electron-donating group, and the red contour represents that the electron-withdrawing
group will be favorable to improve the activity.（Figure 3B）. It could be clearly seen in Figure 3A that there was a
green patch at the position of 14-OH, indicating that introduce a large substitution group at this position would
increase nephrotoxicity. Meanwhile there was a big yellow patch near 19-OH, that meaning that introduce bulky
groups in this position would decrease nephrotoxicity. So the nephrotoxicity of the compounds Ia-If, which were
This article is protected by copyright. All rights reserved

introduced a large substitution group at the position of 14-OH, were significantly higher than other compounds. In
Figure 3B, there were two big blue patches near the position of 3-OH, which revealed that the regions with
electron-withdrawing groups would increase the nephrotoxicity. Therefore the nephrotoxicity of the compounds
IXa-IXd, which were formed by oxidation of 3-hydroxyl groups to acetyl groups and nitrogen-containing heterocyclic
ligation with 19-OH were obviously reduced.
3.5 CoMSIA Contour Maps
The steric, electrostatic, hydrophobic and hydrogen bond acceptor contour maps of the COMSIA model were
shown in Figure 4. The steric and electrostatic contour of COMSIA (Figure 4A and 4B) was in agreement well with
that of COMFA.As shown in the COMSIA hydrophobic field (Figure 4C), yellow polyhedras meant that the regions
with addition of hydrophobic groups would increase activity, white polyhedras meant that where introduction of
hydrophilic groups would increase activities. There were large yellow polyhedra near the position of 3,19-OH,
indicating that the nephrotoxicity of compound with hydrophobic group was higher, with hydrophilic groups was
lower. So the nephrotoxicity of compound III and V, that hydrophilic carboxyl groups were introduced at 3 and
19-OH, respectively, were observably decreased. In the CoMSIA hydrogen bond acceptor field (Figure 4D), the
magenta polyhedras showed that the regions with hydrogen-bond acceptor groups were beneficial for activity, while
the red contour specifies the regions where hydrogen bond-accepting moiety may deteriorate the biological activity
and hydrogen-bond donor groups was favorable for activity. Further, the hydrogen bond acceptor contour in Figure
4D corresponds to the electron-donating group in the electrostatic field contour map of Figure 3B and 4B.
3.6 Biological Activities Prediction of Newly Designed Compounds
The biological activities of newly designed compounds listed in Table 4 were predicted using the best CoMFA
and CoMSIA models. Before the activities’ prediction, molecular alignment of the newly designed molecules was
achieved using energy optimization and conformation search method. The predicted activities are reported in Table 4.
The predicted nephrotoxicity of these newly designed compounds are lower than most of training dataset compounds.
These results prove that generated 3D-QSAR models with significant predictive ability could be used for structural
optimization of the newly designed compounds.
4. CONCLUSIONS
In this present study, twenty three of andrographolide derivatives were synthesized and the structures were
1
confirmed by H-NMR and HRMS, then their nephrotoxicity against human proximal tubular HK-2 cell were
evaluated. Most compounds exhibited lower nephrotoxicity than that of andrographolide, with compound III being
the least active of the series and compound If being the most active. The conformation search-based alignment method
with Gasteiger-Huckel charges yielded the best CoMFA and CoMSIA models. The contour maps suggest that the
bulky hydrophilic group near the 19-OH position, the hydrophilic electron-donating groups near the 3-OH position
and the lighter group near the 14-OH position will help to enhance the biological activities of this series of
compounds. Based on contour maps information of the models, five new compounds were designed, and their
biological activities were predicted. Hence, this work offered an important structural insight for design of
This article is protected by copyright. All rights reserved

andrographolide derivatives with a lower nephrotoxicity, and further investigations will be needed to determine the
mechanism involved in the effect of less nephrotoxic compounds.
ACKNOWLEDGMENTS
This study was supported by a grant from the Natural Science Foundation of Zhejiang province (LY18H280013), Chinese
Medicine Reasearch Program of Zhejiang Province (2019ZQ004) and Key Laboratory of Neuropsychiatric Drug Research of
Zhejiang Province (2019E10021).
CONFILCT OF INTEREST
The authors declare no conflict of interest.
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.
REFERENCES
A, S. B., Prasana, J. C., Muthu, S., Abraham, C. S., & David, H. A. (2019). Spectroscopic and quantum/classical
mechanics based computational studies to compare the ability of Andrographolide and its derivative to
inhibit Nitric Oxide Synthase. Spectrochim Acta
A
Mol Biomol Spectrosc, 218, 374-387.
doi:10.1016/j.saa.2019.04.040
Abdullah, N. H., & Ismail, S. (2018). Inhibition of UGT2B7 Enzyme Activity in Human and Rat Liver Microsomes by
Herbal Constituents. Molecules : a journal of synthetic chemistry and natural product chemistry.(10).
Brahmbhatt, H., Molnar, M., Pavi, V., & Rastijac, V. (2018). Synthesis, Characterization, Antibacterial and Antioxidant
Potency of N-Substituted-2-Sulfanylidene-1,3-Thiazolidin-4-one Derivatives and QSAR Study. Med Chem.
doi:10.2174/1573406415666181205163052
Chen, J., Chen, Y., Gan, X., Song, B., Hu, D., & Song, B. (2018). Synthesis, Nematicidal Evaluation, and 3D-QSAR
Analysis of Novel 1,3,4-Oxadiazole-Cinnamic Acid Hybrids. J Agric Food Chem, 66(37), 9616-9623.
doi:10.1021/acs.jafc.8b03020
Feng, X., Fang, S. N., Gao, Y. X., Liu, J. P., & Chen, W. (2018). Current research situation of nephrotoxicity of Chinese
herbal medicin. Chin J Chin Mater Med, 43(03), 417-424.
Flores-Sumoza, M., Alcázar, J., Márquez, E., Mora, J., Lezama, J., & Puello, E. (2018). Classical QSAR and Docking
Simulation of 4-Pyridone Derivatives for Their Antimalarial Activity. Molecules : a journal of synthetic
This article is protected by copyright. All rights reserved

chemistry and natural product chemistry., 23(12).
Gu, L. L., Xing, W. M., Wang, Y. Q., Zheng, X. L., & Lu, H. (2016). Toxicity screening of andrographolide and its
watersoluble derivatives on HK-2 cells and relative toxicity mechanism induced by andrographolide sodium
bisulfite. Chin J Pharmacol Toxicol, 30(03), 229-235.
Gu, L. L., Zhang, X. Y., Xing, W. M., Xu, J. D., & Lu, H. (2016). Andrographolide-induced apoptosis in human renal
tubular epithelial cells: Roles of endoplasmic reticulum stress and inflammatory response. Environ Toxicol
Pharmacol, 45, 257-264. doi:10.1016/j.etap.2016.02.004
Hazra, A., Mondal, C., Chakraborty, D., Halder, A. K., Bharitkar, Y. P., Mondal, S. K., . . . Mondal, N. B. (2015). Towards
the development of anticancer drugs from andrographolide: semisynthesis, bioevaluation, QSAR analysis
and pharmacokinetic studies. Curr Top Med Chem, 15(11), 1013-1026. Retrieved from
https://www.eurekaselect.com/129592/article
Islam, M. T. (2017). Andrographolide, a New Hope in the Prevention and Treatment of Metabolic Syndrome. Front
Pharmacol, 8(23), 571-579. doi:10.3389/fphar.2017.00571
J.T., L., & K., R. (2006). On selection of training and test sets for the development of predictive QSAR models. QSAR
and Combinatorial Science, 25(3).
Jada, S. R., Subur, G. S., Matthews, C., Hamzah, A. S., Lajis, N. H., Saad, M. S., . . . Stanslas, J. (2007). Semisynthesis
and in vitro anticancer activities of andrographolide analogues. Phytochemistry, 68(6), 904-912.
doi:10.1016/j.phytochem.2006.11.031
Li, Y., He, S., Tang, J., Ding, N., Chu, X., Cheng, L., . . . Wu, J. (2017). Andrographolide Inhibits Inflammatory Cytokines
Secretion in LPS-Stimulated RAW264.7 Cells through Suppression of NF-kappaB/MAPK Signaling Pathway.
Evid Based Compl Alt, 2017, 8248142. doi:10.1155/2017/8248142
Lu, H., Zhang, X. Y., Wang, Y. Q., Zheng, X. L., Zhao, Y., Xing, W. M., & Zhang, Q. (2014). Andrographolide sodium
bisulfate-induced apoptosis and autophagy in human proximal tubular endothelial cells is a ROS-mediated
pathway.
Environ
Toxicol
Pharmacol,
37(2),
718-728.
Retrieved
from
https://www.sciencedirect.com/science/article/pii/S1382668914000210?via%3Dihub
Lu, H., Zhang, X. Y., Zhou, Y. Q., & Jin, S. S. (2010). Toxic actions of andrographolide sodium bisulfite on kidneys of
mice and rabbits. Chin J Pharmacol Toxicol, 24(03), 223-227.
This article is protected by copyright. All rights reserved

Marquina, S., Maldonado-Santiago, M., Sánchez-Carranza, J. N., Antúnez-Mojica, M., González-Maya, L.,
Razo-Hernández, R. S., & Alvarez, L. (2019). Design, synthesis and QSAR study of 2'-hydroxy-4'-alkoxy
chalcone derivatives that exert cytotoxic activity by the mitochondrial apoptotic pathway. Bioorgan med
chem(1), 43-54.
Moorthy, N. S., Ramos, M. J., & Fernandes, P. A. (2011). Prediction of the relationship between the structural
features of andrographolide derivatives and alpha-glucosidase inhibitory activity: a quantitative
structure-activity relationship (QSAR) study.
doi:10.3109/14756361003724760
J
Enzyme Inhib Med Chem, 26(1), 78-87.
Neves, B. J., Braga, R. C., Melo-Filho, C. C., Moreira-Filho, J. T., Muratov, E. N., & Andrade, C. H. (2018). QSAR-Based
Virtual Screening: Advances and Applications in Drug Discovery. Front Pharmacol, 9, 1275.
doi:10.3389/fphar.2018.01275
Paemanee, A., Hitakarun, A., Wintachai, P., Roytrakul, S., & Smith, D. R. (2019). A proteomic analysis of the
anti-dengue virus activity of andrographolide. Biomed Pharmacother, 109, 322-332.
doi:10.1016/j.biopha.2018.10.054
Peng, Y., Wang, Y., Tang, N., Sun, D. D., Lan, Y. L., Yu, Z. L., . . . C.Z., L. (2018). Andrographolide inhibits breast cancer
through suppressing COX-2 expression and angiogenesis via inactivation of p300 signaling and VEGF
pathway. J exp clin canc res(1), 248.
Potemkin, V., Potemkin, A., & Grishina, M. (2018). Internet Resources for Drug Discovery and Design. Curr Top Med
Chem, 18(22), 1955-1975. doi:10.2174/1568026619666181129142127
Reabroi, S., Saeeng, R., Boonmuen, N., Kasemsuk, T., Saengsawang, W., Suksen, K., . . . Chairoungdua, A. (2018). The
anti-cancer activity of an andrographolide analogue functions through a GSK-3beta-independent
Wnt/beta-catenin signaling pathway in colorectal cancer cells. Sci Rep, 8(1), 7924.
doi:10.1038/s41598-018-26278-8
Secretan, P. H., Karoui, M., Sadou-Yaye, H., Levi, Y., Tortolano, L., Solgadi, A., . . . Do, B. (2019). Imatinib: Major
photocatalytic degradation pathways in aqueous media and the relative toxicity of its transformation
products. Sci Total Environ, 655, 547-556. doi:10.1016/j.scitotenv.2018.11.270
Shao, Y., Hollert, H., Tarcai, Z., Deutschmann, B., & Seiler, T. B. (2019). Integrating bioassays, chemical analysis and in
This article is protected by copyright. All rights reserved

silico techniques to identify genotoxicants in surface water. Sci Total Environ, 650(Pt 2), 3084-3092.
doi:10.1016/j.scitotenv.2018.09.288
Song, Y. G., Y.Y., M., & Miao, M. S. (2018). Analysis on toxicity of traditional Chinese medicine containing aristolochic
acid. Chin J Tradit Chin Med Pharm, 33(05), 1950-1954.
Tao, L., Zhang, L., Gao, R., Jiang, F., Cao, J., & Liu, H. (2018). Andrographolide Alleviates Acute Brain Injury in a Rat
Model of Traumatic Brain Injury: Possible Involvement of Inflammatory Signaling. Front Neurosci, 12, 657.
doi:10.3389/fnins.2018.00657
Toropov, A. A., & Toropova, A. P. (2018). Predicting Cytotoxicity of 2-Phenylindole Derivatives Against Breast Cancer
Cells Using Index of Ideality of Correlation. Anticancer res, 38(11), 6189-6194. Retrieved from
http://ar.iiarjournals.org/content/38/11/6189.full.pdf
Upreti, K. K., Das, M., Kumar, A., Singh, G. B., & Khanna, S. K. (1989). Biochemical toxicology of argemone oil. IV.
Short-term oral feeding response in rats. Toxicology, 58(3), 285-298. doi:10.1016/0300-483x(89)90142-x
Uttekar, M. M., Das, T., Pawar, R. S., Bhandari, B., Menon, V., Nutan, . . . Bhat, S. V. (2012). Anti-HIV activity of
semisynthetic derivatives of andrographolide and computational study of HIV-1 gp120 protein binding. Eur J
Med Chem, 56, 368-374. doi:10.1016/j.ejmech.2012.07.030
Vanherweghem, J. L., Depierreux, M., Tielemans, C., Abramowicz, D., Dratwa, M., Jadoul, M., . . . et al. (1993).
Rapidly progressive interstitial renal fibrosis in young women: association with slimming regimen including
Chinese herbs. Lancet, 341(8842), 387-391. doi:10.1016/0140-6736(93)92984-2
Wang, M., Wang, Y., Kong, D., Jiang, H., Wang, J., & Cheng, M. (2018). In silico exploration of aryl sulfonamide
analogs as voltage-gated sodium channel 1.7 inhibitors by using 3D-QSAR, molecular docking study, and
molecular dynamics simulations. Comput Biol Chem, 214-225.
Wang, W., Wu, Y., Chen, X., Zhang, P., Li, H., & Chen, L. (2019). Synthesis of new ent-labdane diterpene derivatives
from andrographolide and evaluation of their anti-inflammatory activities. Eur J Med Chem, 162, 70-79.
doi:10.1016/j.ejmech.2018.11.002
Wei, R. J., Zhang, X. S., & He, D. L. (2018). Andrographolide sensitizes prostate cancer cells to TRAIL-induced
apoptosis. Asian J Androl, 20(2), 200-204. doi:10.4103/aja.aja_30_17
Wu, T., Chen, X., Wang, Y., Xiao, H., Peng, Y., Lin, L., . . . Shuai, X. (2018). Aortic plaque-targeted andrographolide
This article is protected by copyright. All rights reserved

delivery with oxidation-sensitive micelle effectively treats atherosclerosis via simultaneous ROS capture and
anti-inflammation. Nanomedicine, 14(7), 2215-2226. doi:10.1016/j.nano.2018.06.010
Xiang, D., Wang, M. D., Wang, W. Q., Shi, C. Y., Xie, X. J., & Fang, J. G. (2016). Analysis and exploration on adverse
reactions of four kinds of andrographolide injections. Chin J Chin Mater Med, 41(12), 2350-2355.
Xing, W. M., Xu, Z. L., & Lu, H. (2013). Preliminary study on the nephrotoxicity and its mechanism of Chuanhuning
injection. Chin J Exp Tradit Med, 19(20), 263-267.
Xu, H., & Chen, F. G. (2014). Clinical observation of potassium dehydrographolide succinate injection in the
treatment of viral pneumonia. Chin Comm Doc, 30(12), 92-93.
Yang, W. B., Wang, X. W., Yu, J. Y., Jing, Z. W., Wang, Z., Cao, J. L., & Jiang, F. (2013). Analysis of 253 Cases of ADR/AE
Induced by Lianbizhi Injection. Chin J Pharmacov, 10(01), 46-50.
Zhao, X. Y., & Cai, X. H. (2012). Clinical application of Lianbizhi injection in pediatrics. Mil Med J Southeast Chin,
14(04), 359+378.
O
O
O
O
O
O
HO
O
HO
O
R
O
O
R
O
O
a
b
c
H
H
H
H
HO
HO
O
O
O
O
HO
HO
Andro
1
2a-f
Ia-f
Ia R=
Cl
CH3
OC2H5
If R=
Ie R=
Ib R=
Id R=
C(CH3)3
CH=CH2
Ic R=
S
S
Scheme 1. Synthesis of andrographolide 14-OH esterification derivatives (Ia-f). Reagents
and conditions: (a) H3CC(OCH3)2CH3, PPTS, CH2Cl2, reflux, 2h; (b) NEt3, CH2Cl2, rt, RCOOH/ RCOCl
/ClCOOEt, 2h; (c)AcOH:H2O(7:3), 3h.
This article is protected by copyright. All rights reserved

O
O
b
H
O
O
O
HO
HO
O
O
COOH
O
a
III
O
O
H
H
O
O
HO
HO
HO
HO
c
d
II
Andro
O
O
H
H
Cl
R
O
O
3
O
O
IVa-d
Cl
R
O
O
N
N
IVa R=
N
O
IVc R=
IVb R=
N
OH
R=
IVd
Scheme 2. Synthesis of target compounds II-IV. Reagents and conditions: (a) Succinic anhydride pyridine,
refiux, 2h; (b) succinic anhydride, DMAP, CH2Cl2, 5h; (c) Chloroacetyl chloride, Et3N, CH2Cl2, rt ,2h; (d)
Amine, DIPEA, THF, rt, 3h.
O
O
O
O
O
O
O
O
e
g
f
O
H
O
O
O
H
O
H
H
O
O
O
O
O
TrO
6
HO
VI
O
O
R
Cl
7
VIIa-dO
d
O
O
O
O
O
O
O
c
a
b
H
O
O
H
O
H
H
HO
HO
HO
TrO
HOOC
O
HOOC
O
II
4
HO
TrO
5
V
h
O
O
O
O
O
O
O
i
j
k
H
O
H
O
H
H
O
O
O
O
O
O
Cl
R
OTr
VIII
OH
IXa-d
8
9
N
O
N
VIIa,IXa R=
VIIc,IXc R=
N
VIIb,IXb R=
N
OH
R=
VIId,IXd
Scheme 3. Synthesis of target compounds V-IX. Reagents and conditions: (a) TrCl, N-methymorpholine,
24h; (b) succinic anhydride, DMAP, CH2Cl2, refiux, 24h; (c) HCOOH, CH2Cl2, rt, 2h; (d) Ac2O, EDC, DMAP,
CH2Cl2, rt, 12h; (e) HCOOH, CH2Cl2, rt, 2h; (f) Chloroacetyl chloride, Et3N, CH2Cl2, rt, 2h; (g) Amine,
This article is protected by copyright. All rights reserved

DIPEA, THF, rt, 2h; (h) PDC, CH2Cl2, reflux, 48h; (i) HCOOH, CH2Cl2, rt, 2h; (j) Chloroacetyl chloride, Et3N,
CH2Cl2, rt, 2h; (k) Amine, DIPEA, THF, rt, 2h.
Table 1. The IC50 value of andrographolide and its derivatives (I-IX) in HK-2 cells.
Compounds
IC50/μmol·L-1
30.60
20.63
17.06
28.66
17.84
25.60
17.05
458.0
1985
Compounds
IVd
IC50/μmol·L-1
338.7
Andrographolide
Ia
Ib
V
1300
VI
685.0
Ic
VIIa
VIIb
VIIc
VIId
VIII
IXa
201.8
Id
114.0
Ie
307.7
If
162.3
II
757.1
III
IVa
IVb
IVc
786.6
416.8
190
IXb
460.0
IXc
806.9
75.84
IXd
622.0
Table 2. Summary of the results obtained from COMFA and COMSIA analyses.
Parameter
CoMFA
0.951
COMSIA
0.857
R2 (correlation coefficient squared)
ONC (the optimal number of components)
2.000
2.000
q2 LOO (leave-one-out cross validation
correlation coefficient squared)
0.639
0.569
F value
163.747
51.096
This article is protected by copyright. All rights reserved

SEE
Field distribution (%)
Steric
0.158
0.268
33.6
6.4
Electrostatic
Hydrophobic
Donor
66.4
24.4
14.8
20.3
34.2
-
-
-
Acceptor
This article is protected by copyright. All rights reserved

Table 3. .Experimental and predicted extracorporal nephrotoxicity (pIC50) from 3D-QSAR of andrographolide
and its derivatives.
CoMFA
Predicted pIC50
CoMSIA
Predicted pIC50
Compounds
Experimental pIC50
Residual
-0.03
Residual
0.016
Andrographolide
Ia
4.51
4.69
4.76
4.477
4.794
4.143
4.526
4.763
3.979
0.104
0.073
*
-0.617
-0.781
Ib
*
4.54
4.75
4.59
4.77
3.34
2.70
3.38
3.72
4.12
3.47
2.89
3.16
3.70
3.94
3.51
3.79
4.471
4.490
4.780
4.535
3.186
2.851
3.521
3.387
3.833
3.519
2.797
3.257
3.840
4.011
3.718
4.015
-0.069
-0.26
4.144
4.449
4.662
4.694
3.591
2.997
3.534
3.573
3.200
3.552
2.800
3.189
3.676
3.889
3.149
3.889
-0.396
-0.301
0.072
-0.076
0.251
0.297
0.154
-0.207
-0.920
0.082
-0.09
Ic
Id
Ie
0.190
-0.235
-0.154
0.151
0.141
-0.333
-0.287
0.049
-0.093
0.097
0.140
0.071
0.208
0.225
If
II
III
IVa
*
IVb
IVc
IVd
V
VI
0.029
-0.024
-0.051
-0.361
0.099
VIIa
VIIb
*
VIIc
VIId
This article is protected by copyright. All rights reserved

VIII
IXa
IXb
IXc
3.12
3.10
3.34
3.09
3.21
3.042
3.104
3.350
3.041
3.217
-0.078
0.004
0.010
-0.049
0.007
3.092
3.274
3.467
3.068
3.347
-0.028
0.174
0.127
-0.022
0.137
IXd
*Test set compounds: pIC50 = -log(IC50);
Table 4. Newly designed compounds structures with predicted biological activities.
Predicted pIC50
No.
Compound
CoMFA Model
3.710
CoMSIA Model
O
1
2
3.615
O
O
N
N
O
H
H
HO
O
3.412
3.359
O
HO
H
O
O
N
HOOC
This article is protected by copyright. All rights reserved

O
3
4
3.339
3.414
3.356
3.405
O
O
N
N
O
H
O
H
O
N
COOH
O
O
O
HOOC
N
O
H
H
O
O
N
C
O
O
H
O
5
3.652
3.300
O
O
H
O
O
N
HOOC
B
A
O
15
14
O
O
16
O
HO
13
12
O
O
11
9
20
1
17
2
10
H
8
5
3
7
4
HO
OH
OH
6
19
HO
18
If
Andro
Figure 1. The structure of andrographolide (A) and compound If. (B).
This article is protected by copyright. All rights reserved

COMSIA
COMFA
5
4.5
4
5
4.5
4
training
testing
3.5
3
3.5
3
training
testing
2.5
2.5
Experimental activity
pIC50(mol·L-1)
Experimental activity
pIC50(mol·L-1)
2
4
2
4
Figure 2. Plot of experimental and predicted pIC50 for the COMFA and COMSIA models.
A
B
Figure 3. Space field(A) and static electric field(B) three-dimensional equipotential diagram of COMFA. The
molecular structure is the template molecule compound If, the most active compound of the series.
A
B
This article is protected by copyright. All rights reserved

C
D
Figure 4. COMSIA contour maps of steric (A), electrostatic (B), hydrophobic(C), and H-bond acceptor fields(D). The
molecular structure is the template molecule compound If, the most active compound of the series.
This article is protected by copyright. All rights reserved