Hairy root mediated functional derivatization of artemisinin and their bioactivity analysis

Article abstract of DOI:10.1016/j.molcatb.2015.01.007

Biotransformation of artemisinin (1) with the selected hairy root clones of three medicinally important plants, i.e., Atropa belladonna, Hyoscyamus muticus and Ocimum basilicum, yielded two biotransformed products, which were identified as 3-α-hydroxy-1-deoxyartemisinin (2) and 4-hydroxy-9,10-dimethyloctahydrofuro-(3,2-i)-isochromen-11(4H)-one (3). Their structures were elucidated through spectroscopic analysis (NMR/MS) and X-ray crystallography. The relative transformation efficiencies of the tested hairy root clones differed concerning individual bioconversion reactions. Consequently, the HR clones of H. muticus and A. belladonna accomplished the highest conversion of (1) to (2) and (3) respectively, while that of O. basilicum imparted an intermediate response. In-silico and in-vitro bioactivity analysis of the derivatives revealed promising anti-plasmodial activity profile in tandem with notable TNF level lowering potential of compound (2), indicating thereby its prospective therapeutic merit in ameliorating the severity of malarial infection.

Full text of DOI:10.1016/j.molcatb.2015.01.007

Journal of Molecular Catalysis B: Enzymatic 113 (2015) 95–103
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Hairy root mediated functional derivatization of artemisinin and their
bioactivity analysis
a
a
b
c
d
Pallavi Pandey , Sailendra Singh , Nimisha Tewari , K.V.N.S. Srinivas , Aparna Shukla ,
e
d
d
b
f
Namita Gupta , Prema G. Vasudev , Feroz Khan , Anirban Pal , Rajendra Singh Bhakuni ,
g
c,∗
a,∗
Sudeep Tandon , Jonnala Kotesh Kumar , Suchitra Banerjee
Plant Biotechnology Division, Central Institute of Medicinal and Aromatic Plants (CSIR-CIMAP), Lucknow 226015, Uttar Pradesh, India
Molecular Bioprospection Division, Central Institute of Medicinal and Aromatic Plants (CSIR-CIMAP), Lucknow 226015, Uttar Pradesh, India
Natural Product Chemistry Division, Central Institute of Medicinal and Aromatic Plants (CSIR-CIMAP), Research Centre, Boduppal, Hyderabad - 500092,
a
b
c
Telangana, India
d
Metabolic and Structural Biology Division, Central Institute of Medicinal and Aromatic Plants (CSIR-CIMAP), Lucknow 226015, Uttar Pradesh, India
Analytical Chemistry Division, Central Institute of Medicinal and Aromatic Plants (CSIR-CIMAP), Lucknow 226015, Uttar Pradesh, India
Medicinal Plant Chemistry Division, Central Institute of Medicinal and Aromatic Plants (CSIR-CIMAP), Lucknow 226015, Uttar Pradesh, India
Process Chemistry and Chemical Engineering Division, Central Institute of Medicinal and Aromatic Plants (CSIR-CIMAP), Lucknow 226015,
e
f
g
Uttar Pradesh, India
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 12 September 2014
Received in revised form
Biotransformation of artemisinin (1) with the selected hairy root clones of three medicinally
important plants, i.e., Atropa belladonna, Hyoscyamus muticus and Ocimum basilicum, yielded two
biotransformed products, which were identified as 3-␣-hydroxy-1-deoxyartemisinin (2) and 4-hydroxy-
2
4 December 2014
9
,10-dimethyloctahydrofuro-(3,2-i)-isochromen-11(4H)-one (3). Their structures were elucidated
Accepted 13 January 2015
through spectroscopic analysis (NMR/MS) and X-ray crystallography. The relative transformation
efficiencies of the tested hairy root clones differed concerning individual bioconversion reactions. Con-
sequently, the HR clones of H. muticus and A. belladonna accomplished the highest conversion of (1) to (2)
and (3) respectively, while that of O. basilicum imparted an intermediate response. In-silico and in-vitro
bioactivity analysis of the derivatives revealed promising anti-plasmodial activity profile in tandem with
notable TNF level lowering potential of compound (2), indicating thereby its prospective therapeutic
merit in ameliorating the severity of malarial infection.
Available online 22 January 2015
Keywords:
Biotransformation
Artemisinin
Hairy root
Bioactivity
Molecular docking
© 2015 Elsevier B.V. All rights reserved.
1
. Introduction
Malaria ranks as one of the most alarming infectious parasitic
increasingly popular as effective candidates of artemisinin-based
combination therapy (ACT) for the treatment of drug-resistant
malaria [3]. Although the ACT treatment regime showed reduc-
tion in the transmissibility of malaria by preventing gametocyte
development [4], but delayed parasite clearance due to rapid emer-
gence of artemisinin resistance [5] necessitated the discovery of
new antimalarials with novel mechanism of actions [6].
Many efforts have been made to prepare simpler antimalarial
molecules based on the trioxane ring of artemisinin or to produce
semisynthetic and synthetic endoperoxides with greater metabolic
and hydrolytic stability than artemisinin itself [7]. However, as
the processes were complicated with low yields and high cost
[8], efforts were diverted towards exploration of new avenues
to produce more efficient artemisinin analogues or even new
sesquiterpenes. Accordingly, the contemporary research findings
corroborated the superior efficacies of artemisinin analogues for
effective cancer chemotherapy (dihydroartemisinin) as well as in
combination therapies for drug resistance malaria [3,9]. Currently
diseases of the world and is imposing severe threat on approx-
imately half of the global populations as per the existing WHO
survey [1]. In spite of the tremendous worldwide research efforts
to combat malaria, the global morbidity and mortality rates have
not been ameliorated significantly over the last 50 years and re-
emergence of malaria in many parts of the world is currently
imposing fresh challenges due to rapid acquirement of drug-
resistance by the parasite [2].
In general, the discovery of artemisinin, a sesquiterpene lactone
containing an endoperoxide bridge and/or its analogues become
∗ Corresponding author. Tel.: +91 522 2718628; fax: +91 522 2342666.
E-mail addresses: koteshkumarj@yahoo.com (J.K. Kumar),
suchitrabanerjee07@yahoo.com (S. Banerjee).
http://dx.doi.org/10.1016/j.molcatb.2015.01.007
1
381-1177/© 2015 Elsevier B.V. All rights reserved.

9
6
P. Pandey et al. / Journal of Molecular Catalysis B: Enzymatic 113 (2015) 95–103
the WHO recommended antimalarial formulations that are being
used as ACT consist of several artemisinin derivatives having
long half-life than artemisinin, such as artemether-lumefantrine,
artesunate-mefloquine, artesunate-amodiaquine, artesunate-
sulfadoxine-pyrimethamine, dihydroartemisinin-piperaquine and
arteether-curcumin [10–12].
compound in acetonitrile. IR spectra were measured by Spectrum
BX Perkin Elmer and Spectronic® Genesys , respectively.
TM
2.2. Hairy root cultures
Two pre-selected hairy root clones of A. belladonna [29] and H.
muticus [30] were used for the present study, which were main-
tained through sub-culturing in 1/2 strength Murashige and Skoog
medium [31] supplemented with 3% sucrose (pH 5.88) and incu-
bated with continuous shaking on a rotary shaker (80 rpm) at
Diversified efforts have been focused towards generating novel
analogues either through combinatorial biosynthesis in microbes
[
8] or through bioconversion of artemisinin or its analogues
(
artemisinic acid, ␣ and ␤ artemether, arteannuin B) using micro-
◦
bial [13–18] and plant cell/tissue culture systems [16,19–22].
Judicious exploitation of the biotransformation proficiency of
hairy root cultures (HR) is gaining world-wide attention due to
the practical merit of this tool in generating novel products with
widened bioactivities [23]. Besides possessing characteristically
distinctive and coherent growth/enzymatic profiles, long-lasting
operational stability and reduced cost involvement [24], HR cul-
tures also enjoy the added benefits of inter-clonal variations in
metabolic framework through the gainful assimilation of Ri T-DNA
mediated insertional mutagenesis. Such attributes not only broad-
ens the range of substrate adaptability but also modulate the regio
and stereo-selective reaction specificity and thereby renders this
system as a potent biotransformation tool in medicinal and ther-
apeutic chemistry [25]. Several reports document the excellent
biotransformation capabilities of hairy roots which consistently
catalyze reactions in a stereospecific manner, resulting in chirally
pure products [23,26]. It is however pertinent to mention that HR
mediated biotransformation of artemisinin is less explored and till
date only two reports of its conversion to deoxy derivative are avail-
able through the use of two different plant systems, i.e., Cyanotis
arachnoidea [27] and Rheum palmatum L. [28], which leaves ample
scope for further exploration.
The necessity for identification of novel targets through
diversity-based generation of molecules has already been unan-
imously acknowledged which can ideally combat the rapid
emergence of parasite resistance [6]. Accordingly, the structural
modifications of the functional groups of artemisinin holds much
promise in fighting against drug resistance malaria. Under such
circumstances, the accredited uniqueness of hairy root cultures in
performing regio-specific modifications [26], which are otherwise
arduous to carry out by microorganisms or synthetic chemi-
cal methods, further reiterates the potential of such exploration
involving hairy roots of diverse plant systems with regard to
artemisinin [23].
25 ± 1 C under dark condition. Additionally, one recently estab-
lished hairy root clone of O. basilicum, having rapid growth potential
and rol positive (both B and C) traits (data not presented), had also
been utilized for the present biotransformation study following its
maintenance under the aforesaid conditions.
2.3. Biotransformation procedures and Isolation of transformed
products
The substrate (1), was dissolved in MeOH (40 mg/mL) and 1.0 mL
of the solution was added to 50 mL of half-strength MS (3% sucrose)
liquid media. These feeded media were dispensed in 2 weeks old
hairy root cultures (∼5.0 g FW) which were subsequently incu-
◦
bated at 25 ± 2 C on rotary shaker (80 rpm) in dark. Two controls
(substrate control and culture control) were also established, using
MeOH instead of DMSO, following our previously reported protocol
[26]. All the experiments were repeated thrice with three replicates
for each category.
After co-incubation with the substrate, the cultures were har-
vested, the hairy root tissues were separated from the media and
each were extracted with ethyl acetate in triplicates as per our
earlier reported protocol [26]. The extracts were subjected to TLC
analysis (silica gel 60 F254 Plate) using the optimized solvent sys-
tems (Diethyl ether: Hexane:: 7:3) followed by the UV detection at
254 nm and after spraying with anisaldehyde solution for visual-
ization of the transformed products.
The time course study was performed in triplicate at weekly
intervals (from 7 to 21 days) and quantification of the biotrans-
formed products in both the media and HR tissues of all the three
plant systems were carried out through HPTLC following our earlier
reported method [29].
Consequently, for the isolation of the biotransformed prod-
ucts, the ethyl acetate extracts (250 mg) were subjected to column
chromatography on silica gel (20 g, 60–120 mesh, 1 × 20 cm glass)
and was eluted with increasing polarity mixture of ethyl acetate-
hexane. The fractions collected in 5% ethyl acetate in hexane yielded
compound (2) (48 mg), while that collected in 8% ethyl acetate in
hexane gave compound (3) (22 mg). The structure of the isolated
biotransformed products were elucidated by 1D/2D NMR, ESI–MS
and further validated through X-ray crystallography.
The present study explores the competence of the pre-
selected elite HR clones of three medicinally important plants,
i.e., Atropa belladonna, Hyoscyamus muticus and Ocimum basilicum,
for the biotransformation of artemisinin. This communication
reports the first successful hairy-root mediated biotransforma-
tion of artemisinin (1) to 3-␣-hydroxy-1-deoxyartemisinin (2)
and 4-hydroxy-9,10-dimethyloctahydrofuro-(3,2-i)-isochromen-
1
1(4H)-one (3), deduced through spectroscopic analysis (NMR/MS)
2.3.1. 3-˛-Hydroxy-1-deoxyartemisinin (2)
White crystalline solid; H NMR (CDCl , 300 MHz) and 13C NMR
1
and X-ray crystallography. During the present course of study, the
in-silico and in-vitro bioactivity analysis of the derivatives revealed
encouraging activity profile of compound (2) with respect to anti-
plasmodial activity coupled with notable TNF lowering potency.
3
(CDCl , 75 MHz) see Table 1. ESI–MS 282.3.
3
2.3.2. 4-Hydroxy-9,10-dimethyloctahydrofuro-(3,2-i)-
isochromen-11(4H)-one
(
3)
2
. Materials and methods
White crystalline solid; H NMR (CDCl , 300 MHz) and 13C NMR
1
3
(
CDCl , 75 MHz) see Table 1. ESI–MS 240.3.
3
2.1. General experimental procedures
2.4. X-ray crystallographic data
1H, 13C and 2D NMR spectra were recorded using Bruker Avance
3
00 MHz spectrometer and the chemical shifts (ı) were expressed
Single crystals of (2) were obtained by slow evaporation
in ppm with reference to TMS as internal standard. ESI–MS
data were obtained on Shimadzu LC–MS system after dissolving
from chloroform/methanol mixture. Diffraction data were col-
lected on Bruker AXS SMART APEX diffractometer using Mo

P. Pandey et al. / Journal of Molecular Catalysis B: Enzymatic 113 (2015) 95–103
97
Table 1
1
H (300 MHz) and 13C NMR (75 MHz) chemical shift data for compounds (2) and (3) in CDCl3.
Assignment
Compound 2
Compound 3
1H
13C
Nature
1H
13C
Nature
1
2
3
4
5
6
7
8
9
1.66 (m, 1H)
1.65, 2.11 (m, 2H)
3.74 (bs, 1H)
–
5.76 (s, 1H)
–
2.194 (m, 1H)
2.06 (m, 2H)
1.22, 1.95 (m, 2H)
1.434 (m, 1H)
3.325 (m, 1H)
–
1.322 (d, 3H)
1.06 (d, 3H)
1.70 (s, 3H)
41.00
30.73
69.48
109.29
99.34
83.34
42.48
23.92
33.80
35.53
33.10
171.72
12.96
18.77
20.89
CH
CH2
CH
Q
CH
Q
CH
CH2
CH2
CH
CH
Q
CH3
CH3
CH3
1.54 (m, H)
1.43 and 1.88 (m, 2H)
3.96, 4.15 (dd, 2H)
5.88 (br,1H)
–
1.95 (br, H)
1.45 and 2.03 (m, 2H)
1.96 (m, H)
1.98 (m, H)
2.68 (m, H)
–
56.52
27.28
68.26
110.66
86.88
48.33
26.84
35.37
30.97
40.76
179.25
16.21
20.75
–
CH
CH2
CH2
CH
Q
CH
CH2
CH2
CH
CH
Q
10
11
12
13
14
15
1.51 (d, 3H)
1.12 (d, 3H)
–
CH3
CH3
–
–
–
–
K␣ radiation (ꢀ = 0.7173 A˚ ) in ␻-scan mode at room tempera-
ture. Crystal data: C15H22O5, M = 282.3, space group monoclinic,
tautomeric states of amino acid residues such as His, Asp, Ser and
Glu. Binding poses with the lowest docked energy belongs to the
top-ranked cluster was selected as the final model for post-docking
analysis with Discovery Studio v3.5, UCSF Chimera v1.5.3 (Univer-
sity of California, San Francisco) and PyMOL (The PyMOL Molecular
Graphics System, Version 1.5.0.4 Schrodinger, LLC, USA).
P2 ; unit cell dimensions were determined to be a = 5.464 (1) A˚ ,
1
◦
3
b = 13.752 (2) A˚ , c = 9.845 (2) A˚ , ˇ = 101.701 (7) , V = 724.4 (2) A˚
Z = 2, Dcalc = 1.294 mg/m , F(0 0 0) = 304, ␮(Mo K␣) = 0.096 mm
889 unique reflections were collected, of which 1510 reflections
,
3
−1
;
2
2
2
were observed on the basis of F > 4ꢁ(F ). The structure was solved
2
by direct methods using SHELXS [32] and was refined against F
2.6. In vitro antimalarial activity
with full-matrix least squares method by using SHELXL [33]. All the
non hydrogen atoms were refined anisotropically. Hydrogen atoms
were fixed geometrically in idealized positions and were refined
as riding over the atoms to which they were bonded. The final
refinement gave R = 0.0521, Rw = 0.1032, goodness-of-fit, S = 0.943.
Crystals of (3) were obtained by evaporation from methanol.
Diffraction data were collected at room temperature (293 K) on
a Bruker Apex-II CCD diffractometer, using Mo K␣ radiation
2.6.1. Parasite cultures and treatment
Plasmodium falciparum (NF-54 strain) cultures were maintained
as reported earlier [35]. Compounds (1) and (2) were solubilized
in DMSO and further diluted with culture medium to achieve the
required concentrations (final concentration of DMSO <1%). Par-
asite growth was determined spectrophotometrically in control
and drug-treated cultures using a parasite lactate dehydrogenase
assay (PfLDH). Chloroquine diphosphate salt was used as positive
control. After incubation, plates were subjected to three 20 min
freeze–thaw cycles to release cell content. Parasite culture was
carefully mixed and aliquots of 20 ␮L were taken and added to
another flat bottom 96-well plate containing 100 ␮L of Malstat
reagent (0.125% Triton X-100, 130 mM l-lactic acid, 30 mM Tris
buffer and 0.62 ␮M APAD) and 25 ␮L of NBT–PES (1.9 ␮M NBT and
0.24 ␮M PES) solution/well. The plate was incubated in dark for
30 min and absorbance was recorded at 650 nm using a microplate
reader (FLUOStar Omega, BMG Labtech). All the experiments were
done in triplicates. The antiplasmodial activity of the compound
(2) was expressed as IC50 (mean ± SEM), calculated from dose-
response curve data by nonlinear regression analysis.
(
0.71073 A˚ ) in ␻-scan mode. Crystal data: C13H20O4, M = 240.3,
space group orthorhombic P212121; unit cell dimensions were
determined to be a = 6.347 (7) A˚ , b = 12.002 (12) A˚ , c = 16.532 (16) A˚ ,
3
V = 1259 (2) A˚ 3, Z = 4, Dcalc = 1.267 mg/m , F(0 0 0) = 520, ␮(Mo
−
1
K␣ = 0.093 mm 1; 2736 unique reflections were collected, of
2
2
which 1308 reflections were observed with F > 4ꢁ(F ). The struc-
ture solution and refinement methods were similar as described for
compound (2). The final R and Rw values were 0.0725 and 0.1637,
respectively, with goodness-of-fit S = 1.159. Crystallographic data
(
excluding structure factors) of (2) and (3) have been deposited
with the Cambridge Crystallographic Data Centre as supplementary
publication numbers 1015578 and 1015579, respectively. Copies of
the data can be obtained, free of charge, on application to CCDC, 12
Union Road, Cambridge CB2 1EZ, UK, (fax: +44 0 1223 336033 or
deposit@ccdc.cam.ac.uk).
2.6.2. Effect on suppression of TNF in murine peritoneal
macrophages
2
.5. Molecular docking studies
Macrophages were collected from the peritoneum of mice
injected with sterile 1% peptone and cultured as previ-
ously described [36]. The treatments with lipopolysaccharide
(Escherichia coli 050: B5, Sigma, USA) at 1 ␮g/mL alone and in
combination with compound (2) (1, 10 and 100 ␮g/mL) were
accomplished for 18 h in triplicate. Tumor necrosis factor-␣ (TNF-
␣) activity in culture supernatants were estimated by Enzyme
Immune Assay (EIA) using commercial kits for mouse TNF (BD Bio-
sciences, USA).
Drawing and geometry cleaning of the studied compounds
was performed through ChemBioDraw-Ultra-v12.0 software
http://www.cambridgesoft.com). The 2D structures has trans-
(
formed into 3D structures using converter module of ChemBio-
Draw. The 3D structures then subjected to energy minimization.
The docking study of selected target and ligand was done by using
AutoDock Vina software (http://vina.scripps.edu/). The crystallo-
graphic structure of antimalarial target Plasmodium falciparum lac-
tate dehydrogenase (pfLDH) (PDB ID: 1LDG) was retrieved through
Brookhaven Protein Data Bank (PDB) (http://www.pdb.org). The
valency and hydrogen bonding of the ligand as well as target pro-
tein was subsequently satisfied [34]. Polar Hydrogen atoms were
added to the protein target to achieve the correct ionization and
3. Results and discussion
The HR clones of all the three tested plant systems revealed
artemisinin (1) biotransformation competency and two deriva-
tives, i.e., (2) and (3), could be isolated, which were identified by

9
8
P. Pandey et al. / Journal of Molecular Catalysis B: Enzymatic 113 (2015) 95–103
Fig. 1. Biotransformation of artemisinin (1) to 3-␣-hydroxy-1-deoxyartemisinin (2) and 4-hydroxy-9,10-dimethyloctahydrofuro-(3,2-i)-isochromen-11(4H)-one (3) by the
HR clones of A. belladonna, H. muticus and O. basilicum.
spectroscopy. None of the two controls demonstrated any such con-
versions of the substrate. The structures of the isolated compounds
are represented in Fig. 1.
Crystallization, followed by single crystal X-ray diffraction stud-
ies further confirmed the structures of (2) and (3). The crystal
structure of (2) presented here is the same as that previously
reported [38] (Supplementary Fig. S1a and b).
The molecular conformation of (3), obtained through single
crystal X-ray diffraction, is depicted in Fig. 2a. The crystal struc-
ture confirmed the absence of a peroxide linkage and the presence
of a tetrahydrofuran ring, as suggested by NMR analysis. An enve-
lope conformation for the five-membered tetrahydrofuran ring and
a stable chair conformation for the all-carbon 6-membered ring
are observed in the crystal structure of (3). The other 6-membered
ring containing lactone is flattened at O atom. An intermolecular
hydrogen bond interaction between the O atom of the tetrahy-
3
.1. Identification of the biotransformed products
The structure elucidation of compounds (2) and (3), by various
D and 2D NMR spectroscopy, revealed to be 3-␣-hydroxy-1-
1
deoxyartemisinin (2) and 4-hydroxy-9,10-dimethyloctahydrofuro-
3,2-i)-isochromen-11(4H)-one (3). Although, compound (2) was
(
reported earlier in the Artemisia annua plant itself [37] as well as
in various biotransformation studies [38], but compound (3) has
drofuran ring and the hydroxyl group stabilize the molecular
resulted for the first time from the hairy root mediated biotrans-
formation of artemisinin. The 13C NMR (75 MHz) spectrum coupled
assembly in crystals (Fig. 2b; O3. . .O1 = 2.79 A˚ , H. . .O1 = 1.98 A˚ ,
◦
∠
O3–H. . .O1 = 169.9 ). The carbonyl O atom is involved in a weak
with DEPT-135 showed that there are total 13 carbons in compound
C–H. . .O hydrogen bond interaction with the C6 (H) (Fig. 2b;
C6. . .O2 = 3.25 A˚ , H6. . .O2 = 2.46 A˚ , ∠C6–H. . .O2 = 137 ). Notably,
(
3) of which two were quaternary, five tertiary, four secondary and
◦
two primary carbons (Table 1).
the ring conformations observed for (3) are very similar to that
observed for the O–acetyl derivative reported earlier [38]. How-
ever, the absence of a strong hydrogen bond donor like hydroxyl
group has resulted in a different packing arrangement of the O-
acetyl derivative crystals, in which a weak C–H. . .O hydrogen bond
between the O atom of the five-member ring and the methyl hydro-
gen of acetyl group is observed.
The two quaternary carbons were resonated at ı 179.25 (car-
bonyl carbon) and 86.88 ppm, respectively. The presence of a
carbon resonance at ı 86.88 ppm was due to C-5 and this also sug-
gests the absence of a peroxy bond in the structure. Also, presence of
a carbon resonance at 68.26 ppm is attributed to C-3 and its germi-
nal protons appearing at ı 3.96 and 4.15 ppm showed a long range
correlation in HMBC spectrum with C-5 carbon confirms the ether
linkage of the furan moiety in the structure. Presence of a down
field resonance at ı 110.66 ppm is due to C-4 with a free hydroxyl
attachment and this further confirms the absence of a seven mem-
bered lactone structure. Rest of the carbons were resonated as per
the presently reported observation (Table 1).
3.2. Time course studies
The HPTLC analysis revealed that all the three tested HR clones
displayed the maximum bioconversion of (1) to (2) on the 14th

P. Pandey et al. / Journal of Molecular Catalysis B: Enzymatic 113 (2015) 95–103
99
Fig. 2. (a) ORTEP drawing of compound (3) at 50% ellipsoid probability. (b) Intermolecular hydrogen bonds in the crystals of compound (3).
day of substrate feeding, although their relative efficiencies differed
bioconversion potential (∼3.2 and 5.7 times lower respectively).
(
(
Fig. 3a). The H. muticus HR clone performed best amongst the three
Fig. 3a) with regard to the conversion of (1) to (2) and displayed
The HR clone of the former revealed a total of 18% conversion while
that of the later displayed only 10% conversion of (1) to (2) after the
same period of incubation. Microbial bioconversion of artemisinin
to 3-␣-hydroxy-1-deoxyartemisinin (2) by Aspergillus niger and
two Cunninghamella species (i.e., C. echinulata, C. elegans) had earlier
a maximum of >57% total conversion (i.e., 49% in the medium and
almost 8% in the hairy root tissue). On the other hand, the HR clones
of O. basilicum and A. belladonna demonstrated comparatively lesser
Fig. 3. Time course biotransformation proficiency analysis of the HR clones of different medicinal plants with regard to the conversion of artemisinin (1) to: (a) compound
2) and (b) compound (3) respectively, in the media and root tissues. (M-Media; R-Hairy root).
(

1
00
P. Pandey et al. / Journal of Molecular Catalysis B: Enzymatic 113 (2015) 95–103
Table 2
Comparison of binding affinity of artemisinin and its derivatives (2) and (3) in terms of docking energy and binding site residues of anti-malarial target PfLDH (PDB: 1LDG).
−
˚
Compound name
NADPH
Binding energy (kcal mol 1
)
Binding pocket residues (4 A radius)
Key interacting residues and H-bonds
−11.5
VAL-26, GLY-27, GLY-29, MET-30, ILE-31, PHE-52,
ASP-53, ILE-54, VAL-55, MET-58, TYR-85, THR-97,
ALA-98, GLY-99, PHE-100, THR-101, LEU-112,
ASN-116, ILE-119, GLU-122, VAL-138, THR-139,
ASN-140, VAL-142, LEU-163LEU-167, HIS-195,
PRO-246, PRO-250
MET-30, ILE-31, ASP-53, ILE-54, TYR-85,
ALA-98,GLY-99, PHE-100, VAL-138, ASN-140,
LEU-163
Artemisinin
Compound 2
Compound 3
−7.7
−6.3
−5.6
PHE-52, ASP-53, ILE-54, TYR-85, ALA-98, GLY-99,
PHE-100, ILE-119, ILE-123
GLY-27, SER-28, GLY-29, ASP-53, ILE-54, VAL-55,
THR-97, GLY-99
GLY-29, ASP-53, ILE-54, VAL-55, MET-58, ALA-98,
GLY-99, PHE-100
ILE-54 (2 H-bonds 3.0, 3.5), GLY-99 (2 H-bonds
2.8, 2.9)
ASP-53 (1.9), GLY-29 (3.1), THR-97 (2.9)
ASP-53 (2.1)
Note: The underlined residues of NADPH pocket participate in artemisinin and its derivatives (2) and (3) docking.
been reported with comparatively reduced conversion frequencies
than that of the present findings [13,14]. On the contrary, none of
the formerly reported plant cell/organ culture mediated biotrans-
formation attempts of the substrate (1) had ever displayed similar
bioconverted product formation [16,18–22,27,28].
Then again, the A. belladonna HR demonstrated the highest
conversion efficiency with regard to the bioconversion of (1)
to (3) after the same incubation period (14th day) and a max-
imum of >45% total conversion (i.e., 38% in the medium and
about 7% in the hairy root tissues) could be obtained (Fig. 3b).
Likewise, the HR clones of O. basilicum and H. muticus also exhib-
ited similar biotransformation response with decreased propensity
as a total of 12% and 6% bioconversion of (1) to (3) were
recorded respectively, which were ∼3.75 and 7.5 times lower
in that order than that with the A. belladonna HR mediated
derivatization (Fig. 3b). To the best of our knowledge, similar con-
version of (1) to (3) through the use of any biological systems
has not been documented earlier and accordingly, the present
observations corroborate special significance because of its exclu-
sivity.
Apart from these two derivatives, (1) had also been converted
by all the three tested HR cultures into a most prevalent derivative,
i.e., deoxyartemisinin at a lower proportion (∼3 to 6%) as deter-
mined through HPTLC (Supplementary Fig. S2). The bioconversion
Fig. 4. Docking poses of artemisinin (1) and its derivatives (2) and (3) on PfLDH (PDB: 1LDG). (a) Docking procedure standardization by redocking of co-crystallized
−
1
−1
ligand NADPH with docking energy −11.5 kcal mol , (b) Artemisinin (1) docked on PfLDH with binding energy −7.7 kcal mol , (c) Compound (2) docked on PfLDH with
−1
−1
binding energy −6.3 kcal mol and (d) Compound (3) docked on PfLDH with binding energy −5.6 kcal mol . Note: “H-bond is represented with magenta dashed line”. (For
interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

P. Pandey et al. / Journal of Molecular Catalysis B: Enzymatic 113 (2015) 95–103
101
Table 3
Comparison of binding affinity of artemisinin and its derivatives compound (2) and (3) in terms of docking energy and binding site residues of anti-inflammatory target
TNF-␣ (PDB: 2AZ5).
Binding energy (kcal mol 1
−
)
Chain ID (dimer form) Binding pocket residues (4 A radius)
˚
Interacting residues
and length (Å)
Compound name
Re-docking of bound
inhibitor-307* in TNF-␣
dimer complex form
−9.6
Chain A
Chain B
LEU-57, TYR-59, SER-60, TYR-119, LEU-120, GLY-121,
GLY-122, TYR-151
LEU-57, TYR-59, SER-60, TYR-119, LEU-102,
GLY-121TYR-151
–
Artemisinin
Compound 2
−8.2
−7.9
Chain A
Chain B
Chain A
TYR-59, TYR-119, LEU-120, GLY-121, TYR-151
TYR-59, TYR-119
SER-60, TYR-119, LEU-120, GLY-121
–
GLY-121 (2.7), TYR-151
(
3.0)
Chain B
Chain A
Chain B
TYR-59, TYR-119, TYR-151
TYR-119, LEU-120, GLY-121, GLY-122
LEU-57, TYR-59, SER-60, TYR-119, TYR-151
–
–
Compound 3
−7.5
Note: “–” represent no H-bond and
yl}methyl)amino]ethyl}amino)methyl]-4H-chromen-4-one).
*
refer TNF-␣ dimer inhibitor-307 named as (6,7-dimethyl-3-[(methyl{2-[methyl({1-[3-(trifluoromethyl)phenyl]-1H-indol-3-
of artemisinin to deoxyartemisinin have been well documented
through the utilization of diverse biological systems, such as cell
suspension [20,21] HR [27,28] and microbial cultures [15]. In all
the three tested plant systems, the maximum recovery of the bio-
transformed products could be made from the media component
3.3. Molecular docking analysis
The docking of artemisinin (1) and its derivatives (2 and 3) was
performed on PfLDH target enzyme (PDB: 1LDG), which demon-
strated efficient binding to the NADPH binding site pocket (Table 2).
The molecular docking prediction was evaluated by redocking of
co-crystallized ligand NADPH on PfLDH, which showed high bind-
(
∼5–15 times higher than that in the HR tissues) and ethyl acetate
was found to be the best solvent for the maximum recovery of the
bioconverted products.
−
1
ing affinity docking energy i.e., −11.5 kcal mol . Results showed
Fig. 5. Docking pose of artemisinin (1) and compound (2) and (3) on target TNF-␣ (PDB: 2AZ5). (a) Docking software standardization by redocking study of co-crystallized
−
1
−1
ligand 307* on 2AZ5 with docking binding energy −9.6 kcal mol , (b) artemisinin (1) docked on TNF-␣ with binding energy −8.2 kcal mol , (c) compound (2) docked on
−1
−1
TNF-␣ with binding energy −7.9 kcal mol , (d) compound (3) docked on TNF-␣ with binding energy −7.5 kcal mol . Note: Ligand is highlighted in pink color, target chain A
is highlighted in cyan color, target chain B is highlighted in yellow color and H-bond highlighted in magenta color. (For interpretation of the references to color in this figure
legend, the reader is referred to the web version of this article.)

1
02
P. Pandey et al. / Journal of Molecular Catalysis B: Enzymatic 113 (2015) 95–103
Fig. 6. Tumor necrosis factor-␣ (TNF-␣) from the supernatant of macrophages cultured in the presence of compound (2). Peritoneal macrophages were stimulated with
lipopolysaccharide (LPS) to produce TNF and treated with compound (2), solubilized in DMSO in a dose dependent manner, which at 10 and 100 ␮g/mL could significantly
***
lower the level of TNF. MØ–Macrophages. = p < 0.001.
that the (1) docked well on pfLDH with high binding affinity docking
anti-plasmodial activity of 31.88 ± 2.59 ␮M, corroborating the
aforesaid in-silico predictions, wherein the docking energy is com-
parable to that of the parent compound (1). Moreover, the IC50 value
of this compound (2) is comparable to that of our earlier reported
natural molecule—glabridin [35], which has demonstrated anti-
plasmodial activity having the IC50 value of 23.9 ± 0.43 ␮M with
respect to the positive control—artemisinin (0.007 ± 0.004 ␮M).
Additionally, in agreement to the in-silico forecasting, the com-
pound (2) has also been observed to significantly lower the levels
of tumor necrosis factor, as determined through the in-vitro bioac-
tivity analysis (Fig. 6). It has already been established that TNF
and related cytokines initiate events that cause pathology, as
well as parasite death within red cells in such infectious diseases
[41]. Furthermore, TNF level has been clinically proven to rise in
patients infected with Plasmodium falciparum [42]. TNF also acti-
vates a broad array of intracellular signal mechanisms triggering
the inflammatory machinery within the host, through the NFK␤
pathway [43]. Artemisinin derivatives, like artesunate have been
reported to decrease the TNF levels, which in turn do not allow
triggering the release of other pro-inflammatory cytokines, thus
retarding the deterioration of body conditions [44]. The increase in
TNF levels in malaria infected patients might also lead to secondary
complications, like hypoglycaemia, which eventually deteriorates
the prognosis of the disease [44]. In the background of such infor-
mation, the presently observed credentials of the biotransformed
compound (2) in terms of its anti-plasmodial activity coupled with
significant TNF lowering competence, highlight the therapeutic
potential of this biotransformed product in ameliorating the sever-
ity of malarial infection.
−1
−1
energy of −7.7 kcal mol , than the compound (2) (−6.3 kcal mol
)
−1
and compound (3) (−5.6 kcal mol ). The docked binding site
residues of PfLDH for (1) within 4 A˚ region were PHE-52, ASP-
5
3, ILE-54, TYR-85, ALA-98, GLY-99, PHE-100, ILE-119, ILE-123
and the H-bonds forming residues were ILE-54 (H-bond of 3.0 A˚
and 3.5 A˚ ), GLY-99 (H-bond of 2.8 A˚ and 2.9 A˚ ). Similarly, for com-
pound (2), the docked binding site residues of PfLDH were GLY-27,
SER-28, GLY-29, ASP-53, ILE-54, VAL-55, THR-97, GLY-99 and the
H-bonds forming residues were ASP-53 (1.9 A˚ ), GLY-29 (3.1 A˚ ),
THR-97 (2.9 A˚ ) (Table 2). Likewise for compound (3), the docked
binding site residues of PfLDH were GLY-29, ASP-53, ILE-54, VAL-
5
5, MET-58, ALA-98, GLY-99, PHE-100 and the H-bond forming
residue was ASP-53 (2.1 A˚ ) (Fig. 4). The conserved residue was ASP-
5
3, forming H-bonds in both binding sites.
Similarly, the docking of compounds (1), (2) and (3) was also per-
formed on anti-inflammatory target-TNF-␣ (PDB: 2AZ5) (Table 3).
Severe malaria has been associated with high TNF-␣ plasma lev-
els, increased production of other cytokines (IFN-␥ and IL-1␤)
and decreased production of anti-inflammatory cytokines. Pro-
inflammatory Th1-type cytokines (e.g., TNF-␣, IFN-␥, interleukin
IL-1␤, and IL-6) are found critical for the control of exoerythro-
cytic and erythrocytic Plasmodium infection, but their increased
production may also contribute to organ damage, particularly in
the brain [39]. The clinical studies also demonstrated TNF to cause
toxic side effects such as headache, nausea, vomiting, fever, chills
and myalgia [40], which justifies for the selection of this cytokine
as the target of the present docking studies. The docking proto-
col standardization was done by redocking study of co-crystallized
ligand 307 (PDB: 2AZ5) on TNF-␣ dimer, which showed high bind-
−
1
ing affinity to the receptor with binding energy of −9.6 kcal mol
.
4. Conclusions
Results showed that (1) docked well on TNF-␣ with docking binding
−1
energy −8.2 kcal mol , consecutively followed by its derivatives:
The present findings evidently elucidated that the unanimously
recognized goal of generating novel target- based functional deriva-
tives of artemisinin can be effectively achieved through such
HR mediated derivatization of the parent molecule to cope with
the impending emergence of drug resistance malaria. As per
our knowledge, the present investigation records the first suc-
cessful utilization of the HR clones of three different medicinal
plants (A. belladonna, H. muticus and O. basilicum) to biocatalyze
artemisinin into two atypical derivatives other than the most
recurrently observed one (deoxyartemisinin). Rational exploita-
tion of the biochemical machinery of hairy root clones to carry
out such biotransformation reactions leading to the production of
−1
−1
compound 2 (−7.9 kcal mol ) and compound 3 (−7.5 kcal mol ).
The chemical nature of binding site residues of TNF-␣ dimer struc-
tural unit (Chain A and B) were aliphatic due to LEU-57, LEU-120,
GLY-121, GLY-122 and aromatic due to TYR-59 and TYR-119. The
hydroxyl group containing residue within the binding site was SER-
6
0 (Fig. 5).
3.4. Biological activity evaluation
The in-vitro bioactivity analysis against P. falciparum revealed
that the biotransformed compound (2) holds prospective

P. Pandey et al. / Journal of Molecular Catalysis B: Enzymatic 113 (2015) 95–103
103
artemisinin derivatives with interesting anti-malarial activity and
TNF lowering competence can undoubtedly broaden the prospect
of such research for future drug delivery.
[14] I.A. Parshikov, B. Miriyala, K.M. Muraleedharan, A. Illendula, M.A. Avery, J.S.
Williamson, Pharm. Biol. 43 (2005) 579–582.
[
[
[
15] S. Srivastava, S. Luqman, A. Fatima, M.P. Darokar, A.S. Negi, J.K. Kumar, K.
Shanker, Sci. Pharm. 77 (2009) 87–95.
16] S.G. Musharraf, J. Uddin, M. Akhter, M. Parvez, S.S. Khan, S. Yousuf, S. Khan, M.I.
Choudhary, J. Mol. Catal. B: Enzym. 82 (2012) 80–85.
Acknowledgement
17] M.N. Omar, N.T. Khan, N.H.M. Hasali, S.F. Moin, H.Y. Alfarra, Adv. Biores. 3 (2012)
2
7–31.
[18] R. Gaur, S. Patel, R.K. Verma, A. Mathur, R.S. Bhakuni, Med. Chem. Res. 23 (2014)
202–1206.
19] S. Patel, R. Gaur, P. Verma, R.S. Bhakuni, A. Mathur, Biotechnol. Lett. 32 (2010)
167–1171.
The authors wish to express their sincere thanks to the Direc-
tor, CSIR-CIMAP, for providing the facilities to carry out this
research. Financial supports from Council of Scientific and Indus-
trial Research and DST-SERB in the form of fellowships to PP, SS,
NT and AS are also gratefully acknowledged. The work has been
carried out under a CSIR-CIMAP in-house project (OLP-16).
1
[
1
[20] S. Patel, R. Gaur, M. Upadhyaya, A. Mathur, A.K. Mathur, R.S. Bhakuni, J. Nat.
Med. 65 (2011) 646–650.
21] F. Sabir, A. Kumar, P. Tiwari, N. Pathak, R.S. Sangwan, R.S. Bhakuni, N.S. Sangwan,
Z. Naturforsch. C (2010) 607–612.
[
[
22] R. Gaur, S. Tiwari, A. Jakhmola, J.P. Thakur, R.K. Verma, R. Pandey, R.S. Bhakuni,
J. Mol. Catal. B: Enzym. 106 (2014) 46–55.
Appendix A. Supplementary data
[
[
23] S. Banerjee, S. Singh, L.U. Rahman, Biotechnol. Adv. 30 (2012) 461–468.
24] P. Pandey, R. Kaur, S. Singh, S.K. Chattopadhyay, S.K. Srivastava, S. Banerjee,
Biotechnol. Lett. 36 (2014) 1523–1528.
25] M. Kawauchi, T. Arima, O. Shirota, S. Sekita, T. Nakane, Y. Takase, M. Kuroyanagi,
Chem. Pharm. Bull. 58 (2010) 934–938.
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.molcatb.
[
[
[
[
2
015.01.007.
26] V. Srivastava, R. Kaur, S.K. Chattopadhyay, S. Banerjee, Ind. Crops Prod. 44 (2013)
1
71–175.
27] L.G. Zhou, D.C. Ruan, Z.D. He, H.T. Zhu, C.R. Yang, J.J. Wang, Acta Bot. Yunnanica
0 (1998) 229–232.
28] L. Lixin, S. Yanfang, L. Xiaofeng, G. Dean, J. Chin. Pharm. Sci. 4 (2002) 122–124.
References
2
[
[
[
[
1] World Health Organization, World Malaria Report 2011, World Health Organi-
zation, Geneva, 2011.
2] S.I. Hay, C.A. Guerra, A.J. Tatem, A.M. Noor, R.W. Snow, Lancet Infect. Dis. 4
[29] V. Srivastava, A.S. Negi, P.V. Ajayakumar, S.A. Khan, S. Banerjee, Appl. Biochem.
Biotechnol. 166 (2012) 1401–1408.
[30] M. Zehra, S. Banerjee, A.A. Naqvi, S. Kumar, Plant Sci. 136 (1998) 93–99.
[31] T. Murashige, F. Skoog, Physiol. Plant. 15 (1962) 473–497.
[32] G.M. Sheldrick, SHELXS-97, University of Gottingen, Gottingen, Germany, 1997.
[33] G.M. Sheldrick, SHELXL-97, University of Gottingen, Gottingen, Germany, 1997.
[34] G.M. Morris, D.S. Goodsell, R.S. Halliday, R. Huey, W.E. Hart, J. Comput. Chem.
19 (1998) 1639–1662.
(
2004) 327–336.
3] A. Martensson, J. Stromberg, C. Sisowath, M.I. Msellem, J.P. Gil, S.M. Mont-
gomery, P. Olliaro, A.S. Ali, A. Bjorkman, Clin. Infect. Dis. 41 (2005) 1079–1086.
4] P.J. Guerin, P. Olliaro, F. Nosten, P. Druilhe, R. Laxminarayan, F. Binka, W.L.
Kilama, N. Ford, N.J. White, Lancet Infect. Dis. 2 (2002) 564–573.
[
[
5] K. Na-Bangchang, K. Congpuong, Tohuku J. Exp. Med. 211 (2007) 99–113.
6] S. Dandapani, E. Comer, J.R. Duvall, B. Munoz, Future Med. Chem. 4 (2012)
[35] H.S. Cheema, O. Prakash, A. Pal, F. Khan, D.U. Bawankule, M.P. Darokar, Parasitol.
Int. 63 (2014) 349–358.
[36] J. Agrawal, K. Shanker, D. Chanda, A. Pal, Parasitol. Res. 112 (2013) 2601–2609.
[37] C. Jing, Z. Yu-Bo, Z. Xin, H. Li, S. Wei, W. Jin-Hui, J. Shenyang Pharm. Univ. 25
(2008) 866–870.
[38] A.J. Lin, D.L. Klayman, J.M. Hoch, J. Org. Chem. 50 (1985) 4504–4508.
[39] P. Kinra, V. Dutta, Trop. Biomed. 30 (2013) 645–653.
[40] B. Beutler, D. Greenwald, J. Hulmes, M. Chang, Y. Pan, J. Mathison, R. Ulevitch,
A. Cerami, Nature 316 (1985) 552–554.
2
279–2294.
[
[
[
7] D.D. Begue, Bonnet, Drugs Future 30 (2005) 509–518.
8] C. Liu, Y. Zhao, Y. Wang, Appl. Microbiol. Biotechnol. 72 (2006) 11–20.
9] N.P. Singh, H.C. Lai, Anticancer Res. 25 (2005) 4325–4331.
[
[
10] D. Sinclair, B. Zani, S. Donegan, P. Olliaro, P. Garner, The Cochrane collabora-
tion, in: The Cochrane Library, 4, John Wiley & Sons, Ltd., UK, 2009, pp. 1–127,
http://dx.doi.org/10.1002/14651858.
11] P.G. Vathsala, C. Dende, V.A. Nagaraj, D. Bhattacharya, G. Das, P.N. Rangara-
jan, G. Padmanaban, PLoS ONE 7 (1) (2012) e29442, http://dx.doi.org/10.1371/
journal.pone.0029442.
[41] I. Clark, Cytokine Growth Factor Rev. 18 (2007) 335–343.
[42] P. Tangteerawatana, S. Krudsood, N. Kanchanakhan, M. Troye-Blomberg, S. Khu-
smith, J. Trop. Med. Public Health 45 (2014) 517–530.
[43] H. Xu, Y. He, X. Yang, L. Liang, Z. Zhan, Y. Ye, X. Yang, F. Lian, L. Sun, Rheumatology
46 (2007) 920–926.
[12] R.N. Price, J. Infect. Dis. (2013) 1–3.
[13] J.-X. Zhan, Y.-X. Zhang, H.-Z. Guo, J. Han, L.-L. Ning, D.-A. Guo, J. Nat. Prod. 65
(
2002) 1693–1695.
[44] K. Elased, J. Taverne, J. Playfair, Clin. Exp. Immunol. 105 (1996) 443–449.