Synthesis, cytotoxic activity and binding model analysis of novel isoxazole-docetaxel analogues with C3′-N modification

Article abstract of DOI:10.1007/s00044-018-2151-7

Structure–activity relationship (SAR) studies confirm that modifications at C-3′ position can lead to the development of highly potent novel taxoids. We designed and synthesized a series of novel isoxazole-docetaxel analogues A1–A5 by introducing isoxazol

Full text of DOI:10.1007/s00044-018-2151-7

MEDICINAL
CHEMISTRY
RESEARCH
Medicinal Chemistry Research
https://doi.org/10.1007/s00044-018-2151-7
ORIGINAL RESEARCH
Synthesis, cytotoxic activity and binding model analysis of novel
isoxazole-docetaxel analogues with C3′-N modification
2
Ming Chen1 Jiyuan Liu2,3 Zhen Tian3 Xueying Liu2 Shengyong Zhang
●
●
●
●
Received: 16 October 2017 / Accepted: 1 February 2018
© Springer Science+Business Media, LLC, part of Springer Nature 2018
Abstract
Structure–activity relationship (SAR) studies confirm that modifications at C-3′ position can lead to the development of
highly potent novel taxoids. We designed and synthesized a series of novel isoxazole-docetaxel analogues A1–A5 by
introducing isoxazolyl groups to C3′-N position. All of the synthesized compounds exhibited similar to better cytotoxic
activities than docetaxel against human cancer cell lines, Hela, A2780, A549, MCF-7, and SK-OV-3. These compounds also
possessed higher inhibition than docetaxel against drug-resistant cancer cell lines, A2780-MDR and MCF-7-MDR. Binding
model analysis of A1–A5 molecule to microtubule (MT) showed that these compounds were anchored to the active site,
explaining their inhibitory effects on MT in vitro. The calculated binding free energy values were in positive correlation with
the IC50 values of A1–A5 compounds against cancer cells. These results strongly support the statement that the introduction
of isoxazolyl groups to C-3′ position indeed improves the cytotoxic activities of taxoids.
●
●
●
●
Keywords Docetaxel Isoxazole Cytotoxic Inhibition Binding model analysis
Introduction
fight against various cancers, such as metastatic breast
cancer, advanced ovarian cancer, non-small cell lung can-
cer, and Kaposi’s sarcoma (Rowinsky 1997; Mekhail and
Markman 2002). However, the treatment with these taxoids
caused intractable obstacles such as multidrug resistance
(MDR). Therefore, it is essential to develop novel taxoid
anticancer drugs with fewer side effects, superior pharma-
cological properties, and improved activity against various
classes of cancers, especially drug-resistant cancers. Related
studies on structure–activity relationship (SAR) of taxoids
have indicated that the C3′ amide-acyl group at the side-
chain is an extremely crucial site for the bioactivity (Kirikae
et al. 1996, 1998; Fu et al. 2009), but many different N-acyl
groups may be tolerated and some groups have improved
activity. Currently, numerous analogues with C3′-N mod-
ification have been designed and prepared to develop novel
taxoids (Georg et al. 1992a, b, 1994; Gunda et al. 1992;
Roh et al. 1999, 2002; Ojima et al. 2003; Lu et al. 2011;
Chang et al. 2013). Roh et al. had synthesized a series of
C3′-N-acyl-N-debenzoyl paclitaxel analogues, some of
which exhibited higher cytotoxicities (up to 20-fold) and
stronger abilities to induce apoptosis than paclitaxel (Roh
et al. 1999, 2002). In 2003, through the replacement of the
C3′-N-benzoyl group with the different aromatic rings,
Ojima et al. pointed out that paclitaxel analogues bearing a
4-methylbenzoyl and a 4-chlorobenzoyl group showed
higher activities than paclitaxel (Ojima et al. 2003).
Cancer has been one of the most deadly diseases in the
world characterized by uncontrolled cell divisions. Che-
motherapy is one of the most common cancer treatments.
Among a variety of chemotherapeutic drugs, taxoids
paclitaxel, and docetaxel (Fig. 1), which act as promoters of
tubulin assembly, as well as inhibitors of the disassembly
process, are two of the most important drugs currently in the
These authors contributed equally: Ming Chen and Jiyuan Liu.
Electronic supplementary material The online version of this article
(https://doi.org/10.1007/s00044-018-2151-7) contains supplementary
material, which is available to authorized users.
* Xueying Liu
xyliu0427@163.com
* Shengyong Zhang
syzhang@fmmu.edu.cn
1
Department of Pharmacy, The 309th hospital of PLA, 17
Heishanhu Road, Haidian District, Beijing 100091, China
2
Department of Medicinal Chemistry, School of Pharmacy, Fourth
Military Medical University, Xi’an 710032 Shanxi, China
3
Key Laboratory of Plant Protection Resources & Pest Management
of the Ministry of Education, College of Plant Protection,
Northwest A&F University, Yangling, Shanxi 712100, China

Medicinal Chemistry Research
Fig. 1 The structure of
paclitaxel, docetaxel, and
intermediate 13
Just recently, Chang et al. synthesized a series of novel C3′-
N-alkoxycarbonyl docetaxel analogues, some of which were
fluorinated. These compounds exhibited similar potency to
docetaxel and greater potency than paclitaxel against the
SK-OV-3 and A549 cell lines (Chang et al. 2013). Mean-
while, Kingston et al. also concluded that design and
synthesis of a highly cytotoxic tubulin-assembly agent
should be based on the effective binding site of the taxane
(Kingston and Snyder 2014). In general, these reported
results encourage us to make further efforts to investigate
the critical site.
docetaxel would improve its affinity to the target protein,
further enhancing the cytotoxic activity. Therefore, in the
present study, novel isoxazole-docetaxel analogues were
designed and synthesized by introducing 3-aryl-isoxazole-
5-carbonyl groups to the C3′-N-amino-docetaxel, and their
cytotoxic activities on different cancer cell lines were
evaluated. Molecular simulations including molecular
docking and binding free energy calculations were per-
formed as well to reveal the binding modes between the
synthesized compounds and microtubule (MT). Our study
probes the optimization of novel taxoids with higher cyto-
toxic activity.
In recent years, isoxazole scaffold has attracted more and
more attention because of its potent bioactivity. A number
of novel compounds containing isoxazole scaffold were
prepared and evaluated as different inhibitors depending on
protein-ligand binding (Ji et al. 2008; Lilienkampf et al.
2009; Baruchello et al. 2011; Rowbottom et al. 2012).
Michaelides et al. developed a series of benzoisoxazoles
analogues and evaluated their bioactivity as inhibitors of
receptor tyrosine kinases (RTKs) (Ji et al. 2008). SAR
studies confirmed these analogues had potent inhibitory
effects on both the vascular endothelial growth factor
receptor (VEGFR) and platelet-derived growth factor
receptor families of RTKs. Recently, Baruchello et al.
launched some novel 3,4-isoxazolediamides as potent
inhibitors of chaperone heat shock protein 90 through a
structural investigation on the isoxazole scaffold (Bar-
uchello et al. 2011).
Results and discussion
Synthesis
Under the conditions of classic amide formation, five
isoxazole-docetaxel analogues (A1–A5) were prepared
through the acylation of C3′-amino-docetaxel with 3-aryl-
isoxazole-5-carboxylic acids. Thus, the reactions of the key
intermediate 7 with 3-aryl-isoxazole-5-carboxylic acid
chloride 12 in pyridine using 4-dimethylamiopryidine
(DMAP) as catalyst at room temperature gave C3′-N-(3-
aryl-isoxazole-5-acyl)docetaxel (A1–A5) in good yields
(Scheme 1). It is noteworthy that the replacement of con-
ventional triethylamine–dichloromethane reaction system
with current pyridine condition could ensure selective
acylation of amino prior to other active groups (Swindell
and Krauss 2001).
Based on available information, we hypothesized that the
introduction of isoxazolyl group to the C3′-N site of

Medicinal Chemistry Research
Scheme 1 General synthetic
route for compounds A1–A5
Scheme 2 The synthetic route of
7
The synthesis of 7 started from (2R,3S)-3-phenylisoser-
ine methyl ester 1 in five steps. We found that the com-
pound 13 (Fig. 1) which is an intermediate in an alternative
synthetic route to compound 7, is unstable and prones to
remove the protecting group in acidic condition if H atom at
N site is exposed. Based on this, we designed the synthetic
route of 7 outlined in Scheme 2, the advantage being that p-
methoxybenzaldehyde was removed automatically accom-
panied with de-protection of Troc. Compounds 2, 3, and 5
were prepared according to the literature (Hayashi et al.
2003). Then, the coupling of carboxylic acid 5 with 7,10-
diTroc-10-deacetylbaccatin 4 (Fig. 1) in the presence of
dicyclohexylcarbodiimide (DCC) and DMAP furnished the
corresponding esters 6 with a high yield (94.8%). Finally,
the de-protection of all of the protective groups simulta-
neously at C-7, C-10, and C-13 side chain sites of 6 with
zinc powder in AcOH/AcOEt system at room temperature
gave the key intermediate 7, a white powder, with a good
yield after purification (91.2%). Our design ensured a higher
yield and purity of compound 7 comparing with Commer-
con’s (Didier et al. 1994).
The synthesis of 3-aryl-isoxazole-5-carboxylic acid
chloride 12 were carried out using the inexpensive sub-
stituted benzaldehydes 8 as the starting materials in four
steps according to the literatures (Scheme 3) (Shang and
Wang 2002). The reactions of acids 11 with thionyl chloride
allowed for the corresponding acyl chloride 12.
Cytotoxicity
The cytotoxicities of these novel analogues (A1–A5) were
evaluated against five normal human cancer cell lines
(HeLa, A2780, A549, MCF-7, and SK-OV-3), and two
drug-resistant tumor cell lines (A2780-MDR and MCF-7-
MDR) by standard MTT assay (Mosmann 1983; Lamberth
et al. 2007), docetaxel was used as positive control. As
shown in Table 1, all five novel analogues exhibited similar
to better inhibitory activities against the five cancer cell

Medicinal Chemistry Research
Scheme 3 The synthetic route of
12
Table 1 The cytotoxic activities of analogues (A1–A5) against human
cancer cell lines
compounds have negative effects on cytoskeleton in the
present study.
Analogues Cancer cell cytotoxicity IC50 (nM)a
Hela A2780 A2780- A549 MCF-7 MCF-7- SK-
Binding model analysis
MDRb
MDRc OV-3
By using molecular docking, the accurate binding mode of
paclitaxel-tubulin was obtained (reference to supporting
information Fig. S1). The conformations superimposition of
paclitaxel from the crystal structure of paclitaxel-tubulin
complex (PDB ID: 2HXF) (color: gray)and the docking one
(color: cyan) revealed quite minute root-mean-square
deviation (RMSD, 0.075 Å), indicating that the docking
works of the compounds (A1–A5) were accurate in the
binding site. As shown (Fig. 3), hydrogen bond nets were
vital to the binding of compounds A1–A5 to tubulin. The
five compounds all formed hydrogen bonds with the
backbone of Arg275. Specifically, for A1, A4, and A5, the
hydrogen bond interaction was detected between the back-
bone OH (1 position) and the O atom on Arg275, whereas
for A2 and A3, the hydrogen bond occurred between the O
from isoxazole group and the NH from Arg275. Besides,
the hydroxyl (10 position) on the backbone of A3 formed
ahydrogen bond with NH group from the sidechain of
Arg275 as well. Gly359 from tubulin was involved in the
hydrogen bond interaction, the O from its sidechain formed
a hydrogen bond with the backbone OH (7 position)of A1,
A2, A4 and A5, respectively. For compounds A3 and A4, it
is noteworthy that two extra hydrogen bonds were detected
in their interactions with tubulin. By introducing nitro to the
phenyl-isoxazole group in A3, the nitro could form hydro-
gen bonds with backbone NH and side chain O of Thr281,
respectively. While in compound A4, the addition of
methoxyto the phenyl-isoxazole group led to the formation
of hydrogen bond between the oxygen bridge (5 position) of
its backbone and the side chain NH of Arg281. Meanwhile,
the hydrogen bond interaction also occurred between the
backbone NH of Arg281 and the carbonyl (4 position) from
the skeleton of A4. As a whole, the introduction of phenyl-
A1
A2
A3
A4
A5
35.4 9.0
25.1 12.3
38.8 13.4
75.8 13.7
24.0 8.9
40.8
60.5
66.5
65.6
46.4
126
32.7 38.2
23.2 16.6
35.6 39.0
30.7 57.8
39.5 20.0
35.2 38.5
191.0
116.8
112.1
115.2
143.8
180
37.2
26.9
28.8
20.3
15.1
37.1
Docetaxel 50.1 33.1
aIC50: concentration which produces 50% inhibition of proliferation
after 72 h of incubation. HeLa: cervical cancer, A2780: ovarian cancer,
A549: non-small cell lung cancer, MCF-7: breast cancer, SK-OV-3:
ovarian cancer
bA2780-MDR: multidrug-resistant ovarian cancer
cMCF-7-MDR: multidrug-resistant breast cancer
lines comparing with docetaxel. Among these five novel
analogues, A1 and A5 were approximately four times
stronger than docetaxel in the activity against A2780, A5
possessed the highest potencies against Hela, A2780, and
SK-OV-3. Meanwhile, A2 exhibited the best activity
against A549 and MCF-7. What’s more, all of these com-
pounds improved the sensitivity of those drug-resistant
tumor cells, A2780-MDR and MCF-7-MDR. Especially,
A1and A5 exhibited about three times stronger activity
against A2780-MDR.
Effects on cytoskeleton
To further investigate whether the tested drugs have effects
on cytoskeleton, we employed immunofluorescence tech-
niques. As shown in Fig. 2, when exposed to the five tested
compounds, the cytoskeleton of Hela cell lines were
obviously disrupted and displayed the disperse condition
compared to the integrity of control, suggesting that these

Medicinal Chemistry Research
A1
A2
A3
A4
A5
control
Fig. 2 Confocal images of Hela cell lines left untreated (control), or exposed to the tested compounds (A1–A5) for 24 h.The compounds disrupted
the microfilament-associated structure of F-actin (green) and Tublin (red). Nuclear DNA was labelled in blue with Hoechst (color figure online)
isoxazole group to the C3′-N position of docetaxel helps to
form more hydrogen bonds in the active cavity of tubulin,
this change may contribute to the higher cytotoxic activity
of A1–A5. Not only that, the differences between the
hydrogen bond nets of each complex are considered as a
factor in affecting the activity of A1–A5.
individually to the para-position of phenyl-isoxazole groups
significantly raised the S(lipo) and S(hbond) values. Fur-
thermore, the introduction of nitrile to the ortho-position of
phenyl-isoxazole group in A1 also shed positive effects on
S(hbond) contributions. To reveal the reasons involved in
the energy changes mentioned above, the superimposition
of the binding modes of A1–A5 was performed (Fig. S2).
As shown, the binding modes of A2, A4, and A5 were of
high similarity, particularly for A2 and A5, their con-
formations were close to perfect overlap. However, A4 was
characterized by the special H-bond nets. By introducing
the phenyl-isoxazole group, a new H-bond was formed
between the oxygen bridge from A4 (5 position) and the
main chain NH of Arg281 (4 position), hence higher S
(hbond) was detected in A4. The differences between S
(lipo) values of A2 and A5 can be attributed to the fact that
the chlorine on para-position of phenyl (A5) contributed
more to the S(lipo) than the fluorine on meta-position of
phenyl (A2). Comapring with the chlorine group (A5), the
greater exposure of methoxy group (A4) to the solvent
contributed to the lower S(lipo) value. Similar phenomenon
was also detected in A3, the introduction of nitro to the
meta-position of phenyl-isoxazole group led to the exposure
of 10-DAB (A3), causing the loss of S(lipo). Based on
available data above, it is reasonable to speculate that
introducing nitrile group and chlorine group to the ortho-
position and para-position of phenyl-isoxazole group
respectively has potential in increasing S(lipo) and S
Binding free energy
The binding free energy and the energy contribution com-
ponents for compounds A1–A5 and docetaxel were listed in
Table 2. For the five tested compounds, A1, A3, A4, and A5
caused higher binding free energy changes (ΔG) of tubulin
than docetaxel, with A2 being an exception. This fitted well
with their corresponding cytotoxic activities, the one (A5)
with the highest ΔG possessed the highest activity against
cancer as well. By using docetaxel as control, the binding
free energy items of A1–A5 were analyzed and compared in
detail. For A2 which is characterized by the introduction of
fluorine to the meta-position of phenyl-isoxazole group, the
optimization shed negative effects on its lipophilic energy
term[S(lipo)] contributions, with S(lipo) less than the
counterpart of docetaxel. As for A3, the introduction of
nitro caused the significant increase of hydrogen bond
energy [S(hbond)] contribution but decrease of S(lipo),
making its resultant ΔG only slightly higher than the
counterpart caused by docetaxel. In the present study,
introducing methoxy group (A4) and chlorine group (A5)

Medicinal Chemistry Research
Fig. 3 The carton representation
of tubulin 3D model around the
binding site of Docetaxel and its
analogs (A1–A5). The binding
model of tubulin with A1 (a),
with A2 (b), with A3 (c), with
A4 (d), with A5 (e), and with
docetaxel (f). Docetaxel and its
analogs (A1–A5) are presented
as stick and sphere model. Cyan
C; Red O; Blue N; Light blue F;
Green Cl. Key residues are
presented with the stick model.
Color code: Gray C; Red O;
Blue N. the yellow dotted lines
show hydrogen bonds among
the atoms from amino acid
residues and docetaxel and its
analogs (A1–A5) (color figure
online)
(hbond) contributions simultaneously, further enhancing
ΔG values of taxoids. However, future experiments are
required to verify this speculation.
docetaxel analogues to tubulin, consequently improving
cytotoxic activities of taxoids. These facts proved that the t-
Boc group at the C3′-N position, which is the “gold stan-
dard,” can be replaced by isoxazolyl groups without losing
potency. Due to these findings, further SAR studies at C3′-N
position of docetaxel are underway in our lab.
Conclusion
In summary, a series of novel C3′-N-(3-aryl-isoxazole-5-
acyl)docetaxel analogues (A1–A5) were synthesized and
characterized by H nuclear magnetic resonance (NMR),
Experimental
1
13CNMR, and high resolution mass spectrometry (HRMS)
analysis. The cytotoxicities were evaluated against various
human cancer cell lines. The results showed that each of
these five novel analogues possessed similar to better inhi-
bitory activity than docetaxel. As revealed by binding mode
and binding free energy analysis, the formation of more
hydrogen bonds and differences in hydrogen bond nets were
considered to be related to the cytotoxic activities of com-
pounds A1–A5. The introduction of foreign groups (fluor-
ine, nitro, nitrile,methoxy group and chlorine group) to the
phenyl-isoxazole group did help in increasing the affinity of
General analytical procedures
Reagents and solvents were obtained commercially and
used without further purification. Thin layer chromato-
graphy was carried out on Merck silica gel 60F plates. Flash
column chromatography was performed on silica gel
(200–300 meshes) from Qingdao Marine Chemical Factory
in China. Melting points were determined on a Yanaco MP-
500D melting point apparatus and were uncorrected.
1HNMR and 13CNMR spectra were recorded on a Bruker
NMR AVC500 spectrometer in dimethyl sulfoxide

Medicinal Chemistry Research
Table 2 The predicted binding free energy changes and the individual energy terms for the ligands
Analogues
Structure
S(hbond)
S(lipo)
H(rot)
5.58
ΔG (KJ/
mol)
A1
2.13
318.29
−35.55
A2
A3
1.11
3.19
239.73
230.82
5.50
5.66
−23.15
−28.67
A4
A5
2.30
1.89
323.46
354.13
5.62
5.39
−36.62
−39.42
Docetaxel
0.37
269.76
5.67
−23.78
All values are given in KJ/mol
(DMSO)-d6 or CDCl3 using TMS as an internal standard.
HRMS (FAB) analysis was carried out by the Central Lab
of Nankai University.
isoxazole-5-carboxylic acid chloride 12 (0.154 mmol)
at room temperature. After 1 h, the mixture was diluted
with methylene chloride, washed with 1N HCl, water,
dried over MgSO4 and concentrated to get a crude com-
pound, which was purified by column chromatography
(CH2Cl2:MeOH = 10:1) to furnish the corresponding C3’-
N-(3-aryl-isoxazole-5 -acyl)docetaxel (A1–A5) as a white
solid.
General procedure for the synthesis of C3′-N-(3-aryl-
isoxazole-5-acyl) docetaxel (A1–A5)
To a stirred solution of 7 (0.1 g, 0.14 mmol) and DMAP
(2 mg, 0.014 mmol) in pyridine (3 mL) was added 3-aryl-

Medicinal Chemistry Research
C3′-N-[3-(2-cyanophenyl) isoxazole-5-acyl]docetaxel (A1)
4.97 (d, J = 9.5 Hz, 1H), 4.82 (d, J = 2.4 Hz, 1H),
4.39–4.33 (m, 1H), 4.32–4.20 (m, 3H), 3.94 (d, J = 7.0 Hz,
1H), 2.65–2.54 (m, 1H), 2.45 (s, 3H), 2.38–2.25 (m, 4H),
1.94–1.84 (m, 5H), 1.78 (s, 3H), 1.21 (s, 3H), 1.13 (s, 3H).
13C NMR (125 MHz, CDCl3) δ 211.17, 172.06, 170.61,
166.82, 163.99, 161.48, 155.01, 148.54, 137.98, 137.06,
136.26, 133.83, 132.34, 130.33, 129.33, 129.24, 129.10,
128.83, 128.63, 127.10, 125.17, 121.75, 105.78, 84.20,
81.26, 78.88, 74.82, 74.47, 73.34, 72.25, 71.93, 57.69,
54.80, 46.46, 43.08, 36.88, 35.92, 26.55, 22.64, 20.73,
14.44, 9.97. HRMS (FAB): calcd for C48H49N3O16 (M +
NH4+): 941.3451, found: 941.3437.
Yield: 74.2%. m.p. 202–203.1 °C. 1H NMR (500 MHz,
CDCl3) δ 8.16 (d, J = 7.2 Hz, 2H), 7.89–7.79 (m, 3H), 7.71
(td, J = 7.8, 1.2 Hz, 1H), 7.67–7.59 (m, 2H), 7.57–7.51 (m,
4H), 7.45 (t, J = 7.6 Hz, 2H), 7.39 (d, J = 7.3 Hz, 1H), 7.36
(s, 1H), 6.31 (t, J = 8.5 Hz, 1H), 5.83 (dd, J = 9.2, 2.5 Hz,
1H), 5.70 (d, J = 7.1 Hz, 1H), 5.25 (s, 1H), 4.97 (d, J =
9.4 Hz, 1H), 4.83 (d, J = 2.7 Hz, 1H), 4.33 (d, J = 8.5 Hz,
1H), 4.29–4.22 (m, 2H), 3.93 (d, J = 7.0 Hz, 1H), 3.50 (s,
2H), 2.64–2.53 (m, 1H), 2.42 (s, 3H), 2.37–2.29 (m, 2H),
2.27–2.22 (m, 1H), 2.09 (br, s, 1H), 1.91 (d, J = 13.9 Hz,
1H), 1.85 (s, 3H), 1.79 (s, 3H), 1.23 (s, 3H), 1.14 (s, 3H).
13C NMR (125 MHz, CDCl3) δ 211.15, 171.94, 170.70,
166.85, 163.68, 161.05, 155.07, 137.91, 137.25, 136.29,
134.32, 133.81, 133.39, 130.91, 130.65, 130.21, 129.55,
129.32, 129.05, 128.72, 128.56, 127.11, 117.45, 110.97,
107.24, 84.26, 81.32, 78.66, 74.82, 74.53, 73.45, 72.30,
71.81, 57.72, 54.84, 50.82, 46.53, 43.05, 36.91, 35.91,
26.58, 22.61, 20.57, 14.47, 9.94. HRMS (FAB): calcd for
C49H49N3O14 (M + Na+): 926.3107, found: 926.3103.
C3′-N-[3-(4-methoxyphenyl) isoxazole-5-acyl]docetaxel (A4)
1
Yield: 78.6%. m.p. 191.8–193.4 °C. H NMR (500 MHz,
CDCl3) δ 8.26 (d, J = 7.5 Hz, 2H), 7.77 (d, J = 9.1 Hz, 1H),
7.67 (t, J = 7.4 Hz, 1H), 7.58 (t, J = 7.6 Hz, 2H), 7.52–7.44
(m, 5H), 7.41 (d, J = 7.1 Hz, 3H), 6.89 (s, 1H), 6.78
(d, J = 7.8 Hz, 2H), 6.45 (s, 1H), 5.88 (d, J = 7.4 Hz, 1H),
5.74 (d, J = 7.2 Hz, 1H), 5.26 (s, 1H), 4.99 (d, J = 9.2 Hz,
1H), 4.82 (s, 1H), 4.38–4.27 (m, 4H), 4.15 (br, s, 1H), 3.97
(d, J = 7.1 Hz, 1H), 3.85 (s, 3H), 2.66–2.58 (m, 1H),
2.49 (s, 3H), 2.43–2.33 (m, 3H), 1.96–1.88 (m, 5H), 1.81
(s, 3H), 1.24 (s, 3H), 1.14 (s, 3H). 13C NMR (125 MHz,
C3′-N-[3-(3-fluorophenyl) isoxazole-5-acyl]docetaxel (A2)
Yield: 72.1%. m.p. 193.5–195 °C. 1H NMR (500 MHz,
CDCl3) δ 8.23 (d, J = 7.3 Hz, 2H), 7.86 (d, J = 9.3 Hz, 1H),
7.67 (t, J = 7.4 Hz, 1H), 7.57 (t, J = 7.7 Hz, 2H), 7.49 (d, J
= 7.4 Hz, 2H), 7.45 (t, J = 7.5 Hz, 2H), 7.40 (d, J = 7.2 Hz,
1H), 7.38–7.35 (m, 1H), 7.34–7.31 (m, 1H), 7.21 (d, J =
9.0 Hz, 1H), 7.12 (t, J = 8.1 Hz, 1H), 6.96 (s, 1H), 6.39 (t, J
= 8.7 Hz, 1H), 5.85 (dd, J = 9.3, 2.6 Hz, 1H), 5.72 (d, J =
7.1 Hz, 1H), 5.25 (s, 1H), 4.98 (d, J = 9.1 Hz, 1H), 4.80 (s,
1H), 4.36 (d, J = 8.3 Hz, 2H), 4.32–4.23 (m, 3H), 3.95 (d, J
= 7.1 Hz, 1H), 2.67–2.55 (m, 1H), 2.46 (s, 3H), 2.41–2.24
(m, 4H), 1.90 (s, 4H), 1.79 (s, 3H), 1.22 (s, 3H), 1.13 (s,
3H). 13C NMR (125 MHz, CDCl3) δ 211.17, 172.11,
170.64, 166.74, 163.38, 162.35, 161.85, 155.22, 137.98,
137.07, 136.26, 133.70, 130.76, 130.37, 129.44, 129.08,
128.78, 128.59, 127.10, 122.44, 117.74, 113.92, 106.03,
84.22, 81.27, 78.81, 74.81, 74.45, 73.42, 72.21, 71.90,
57.67, 54.79, 46.41, 43.12, 36.87, 36.03, 26.53, 22.63,
20.82, 14.42, 9.98. HRMS (FAB): calcd for C48H49N2O14F
(M + NH4+): 914.3506, found: 914.3490.
CDCl3)
δ 211.32, 172.21, 170.56, 166.72, 162.91,
162.74, 161.38, 155.38, 137.95, 137.10, 136.31, 133.65,
130.41, 129.09, 128.76, 128.55, 128.16, 127.03, 114.35,
105.90, 84.18, 81.23, 78.81, 74.81, 74.48, 73.53, 72.24,
71.92, 57.65, 55.35, 54.65, 46.35, 43.16, 36.96, 36.10,
29.71, 26.52, 22.67, 20.90, 14.46, 10.02. HRMS (FAB):
calcd for C49H52N2O15 (M + NH4+): 926.3706, found:
926.3693.
C3′-N-[3-(4-chlorophenyl) isoxazole-5-acyl]docetaxel (A5)
1
Yield: 76.7%. m.p. 190.7–192.3 °C. H NMR (500 MHz,
CDCl3) δ 8.27 (d, J = 7.3 Hz, 2H), 7.92 (d, J = 9.4 Hz, 1H),
7.67 (t, J = 7.4 Hz, 1H), 7.59 (t, J = 7.6 Hz, 2H), 7.50–7.38
(m, 6H), 7.32 (s, 1H), 7.21 (d, J = 8.2 Hz, 2H), 6.85 (s, 1H),
6.47 (t, J = 8.5 Hz, 1H), 5.87 (dd, J = 9.5, 2.6 Hz, 1H), 5.73
(d, J = 7.1 Hz, 1H), 5.25 (s, 1H), 4.99 (d, J = 9.5 Hz, 1H),
4.78 (s, 1H), 4.59–4.25 (m, 5H), 3.96 (d, J = 7.1 Hz, 1H),
2.66–2.59 (m, 1H), 2.50 (s, 3H), 2.44–2.27 (m, 3H),
1.96–1.88 (m, 5H), 1.80 (s, 3H), 1.22 (s, 3H), 1.12 (s, 3H).
13C NMR (125 MHz, CDCl3) δ 211.19, 172.17, 170.63,
166.61, 163.33, 162.25, 155.17, 137.95, 137.00, 136.85,
136.28, 133.66, 130.45, 129.61, 129.25, 129.08, 128.80,
128.58, 127.84, 127.07, 125.48, 105.98, 84.23, 81.25,
78.85, 74.83, 74.40, 73.50, 72.11, 71.87, 57.64, 54.78,
46.33, 43.18, 36.86, 36.16, 26.49, 22.68, 21.00, 14.41,
10.04. HRMS (FAB): calcd for C48H49N2O14Cl (M + NH4
+): 930.3211, found: 930.3207.
C3′-N-[3-(3-nitrophenyl) isoxazole-5-acyl]docetaxel (A3)
1
Yield: 70.8%. m.p. 195.2–196.7 °C. H NMR (500 MHz,
CDCl3) δ 8.38 (s, 1H), 8.27 (d, J = 8.2 Hz, 1H), 8.22 (d, J
= 7.3 Hz, 2H), 8.03 (d, J = 7.8 Hz, 1H), 7.86 (d, J = 9.3 Hz,
1H), 7.69 (t, J = 7.4 Hz, 1H), 7.59 (t, J = 7.8 Hz, 3H), 7.51
(d, J = 7.4 Hz, 2H), 7.46 (t, J = 7.5 Hz, 2H), 7.40 (d, J =
7.3 Hz, 1H), 7.11 (s, 1H), 6.35 (t, J = 8.6 Hz, 1H), 5.84 (dd,
J = 9.3, 2.5 Hz, 1H), 5.70 (d, J = 7.1 Hz, 1H), 5.26 (s, 1H),

Medicinal Chemistry Research
The synthesis of C3′-N-amino-docetaxel (7)
plate mixer, the absorbance of each well was measured by
amicroplate reader using a test wavelength of 570 nm. The
results were expressed as the IC50, which is the concentra-
tion of the drugs inducing a 50% inhibition of cell growth of
treated cells when compared to the growth of control cells.
Each experiment was performed at least three times. There
was a good reproducibility between replicate wells with
standard errors below 10%.
To a stirred solution of 6 (1.35 g, 1 mmol) in AcOEt
(20 mL) was added zinc powder (1.76 g, 27 mmol) and
AcOH (3.5 mL) at room temperature. After 4 h, zinc powder
was filtered off, water (20 mL) and NaHCO3 (5.2 g) were
added to the solution. The organic phase was washed with
water and brine, dried over MgSO4 and concentrated to get
a crude compound, which was purified by column chro-
matography (CH2Cl2:MeOH = 7:1) to furnish a pure pro-
dect 7 of 0.65 g as a white powder. Yield: 91.2%. m.p.
Immunocytochemistry assay the cell structures
1
166.2–167.6 °C. H NMR (400 MHz, DMSO) δ 7.95 (d, J
Cells were fixed in suspension with 2% paraformaldehyde
in phosphate buffered saline (PBS) on ice for 10 min. The
cells were washed once with wash buffer (PBS containing
0.1% sodium azide), and resuspended in blocking buffer
(PBS containing 5% serum), immobilized on round 12 mm
coverslips by cytocentrifugation, and post-fixed with
methanol at −20 °C for 10 min. The slides were washed
twice with wash buffer and then blocked with blocking
buffer, with shaking at room temperature for 30 min. The
cells were stained with Actin-Tracker Green and Tubulin-
Tracker-Red (Shanghai Yope Biotechnology, Shanghai,
China). The cells were counterstained with Hoechst nuclear
stain (Beyotime, 1:100).
= 7.1 Hz, 2H), 7.73 (t, J = 7.4 Hz, 1H), 7.65 (t, J = 7.5 Hz,
2H), 7.42 (d, J = 4.3 Hz, 4H), 7.28–7.19 (m, 1H), 5.85 (t, J
= 8.7 Hz, 1H), 5.40 (d, J = 7.2 Hz, 1H), 5.09 (s, 1H), 5.04
(d, J = 7.1 Hz, 1H), 4.97 (s, 1H), 4.89 (d, J = 10.0 Hz, 1H),
4.53 (s, 1H), 4.22 (d, J = 7.9 Hz, 1H), 4.16 (d, J = 7.9 Hz,
1H), 4.10–3.96 (m, 3H), 3.63 (d, J = 7.2 Hz, 1H),
2.33–2.20 (m, 1H), 2.11 (s, 3H), 1.74 (s, 3H), 1.69–1.62 (m,
1H), 1.51 (s, 3H), 1.01 (s, 3H), 0.97 (s, 3H). 13C NMR
(100 MHz, DMSO) δ 214.55, 177.74, 174.90, 170.41,
141.99, 141.07, 138.65, 135.24, 134.72, 133.89, 133.53,
133.09, 132.81, 88.93, 85.48, 82.06, 80.63, 80.37, 79.95,
78.97, 75.98, 75.04, 63.33, 62.18, 51.14, 48.09, 41.66,
40.19, 31.73, 27.64, 26.02, 18.94, 15.01. HRMS (FAB):
calcd for C38H45NO12 (M + H+) 708.3015, found:
708.3005.
Molecular docking simulations
In order to obtain the accurate binding poses of tubulin with
the compounds A1–A5, all the docking simulations were
investigated by the program GOLD 5.3 (Jones et al. 1995,
1997).The 3D structure of the compounds A1-A5 were
sketched using Maestro version (Schrodinger Inc.) and
2000 steps of optimization were performed in Amber12
with the GAFF force field (Wang et al. 2004; Case et al.
2012). Base on the high crystallographic resolution and R-
factor, the crystal structure of tubulin from Sus scrofa (PDB
ID: 2HXF, Chain A) was finally selected as the receptor for
docking simulations. The receptor tubulin was performed
5000 steps minimization in Amber12 with the ff99SB force
field (Case et al. 2012; Hummer et al. 2001). To obtain the
best set of the docking parameters and the reliability of the
docking results, the paclitaxel derived from the crystal
structure of the complex tubulin was first docked into the
binding site. The CO2 atom coordinate of the paclitaxel was
defined as the centroid of the binding site with 10 Å radius
sphere. Recent studies demonstrated that ChemPLP was
superior to the other scoring functions in GOLD for pose
prediction (Chen et al. 2014; Yang et al. 2014; Tian et al.
2016a, b). ChemPLP score was employed to obtain the
most accurate binding modes for A1–A5. The 2D interac-
tion diagram was presented by the programMaestro version
10.1. Visualization of the structures was performed by
PyMol1.3r1 edu (DeLano 2002).
Cytotoxicity assay
Cytotoxic activities were evaluated by using standard MTT
assay after exposure of cells to the tested compounds for
72 h. Each experiment was performed at least three times.
There was a good reproducibility between replicate wells
with standard errors below 10%. The in vitro cytotoxicity of
the synthesized compounds against different cancer cell
lines was performed with the MTT assay according to the
Mosmann’s method. The MTT assay is based on the
reduction of the soluble 3-(4,5-dimethyl-2-thiazolyl)-2,5-
diphenyl-2H-tetrazolium bromide (MTT) into a blueepurple
formazan product, mainly by mitochondrial reductase
activity inside living cells. The cells used in cytotoxicity
assay were cultured in improved RPMI 1640 medium
supplemented with 10% fetal calf serum. Cells suspended in
the medium were plated in 96-well culture plates and
incubated at 37 °C in a 5% CO2 incubator. After 24 h, the
test sample was added to the cells in 96-well plates and
cultured at 37 °C for 3days. The cultured cells were mixed
with 50 μL 5mg/mL of MTT solution and incubated for 2 h
at 37 °C. The supernatant was carefully removed from each
well and 200 μL of DMSO was added to each well to dis-
solve the formazan crystals which were formed by the
cellular reduction of MTT. After mixing with amechanical

Medicinal Chemistry Research
Binding energy calculations
Chang J, Hao XD, Hao YP, Lu HF, Yu JM, Sun X (2013) Design,
synthesis and cytotoxicity of novel 3′-n-alkoxycarbonyl doc-
etaxel analogs. Bioorg Med Chem Lett 23:6834–6837
Chen X, Liu J, Zhang Y (2014) Cantharidin impedes the activity of
protein serine/threonine phosphatase in plutella xylostella. Mol
Biosyst 10:240–250
To compare the binding affinity of the analogs A1–A5 with
the receptor tubulin, the top docking pose for each analog
that corresponded to the ChemPLP score was recorded by
the Chemscore (Eldridge et al. 1997; Baxter et al. 1998).
The Chemscore function was applied to measure affinity
data by ranking the Chemscore delta value (Chen et al.
2014; Liu et al. 2014, 2015, 2016). Chemscore estimates the
binding free energy ΔG according to Eq. (1):
DeLano WL (2002) PyMOL molecular graphics system,version
̈
1.3.0.4. LLC, Schrodinger
Didier E, Fouque E, Commerçon A (1994) Expeditious semisynthesis
of docetaxel using 2-trichloromethyl-1,3-oxazolidine as side-
chain protection. Tetrahedron Lett 35:3063–3064
Eldridge MD, Murray CW, Auton TR, Paolini GV, Mee RP (1997)
Empirical scoring functions: I. The development of a fast
empirical scoring function to estimate the binding affinity of
ligands in receptor complexes. J Comput Aided Mol Des
11:425–445
Fu Y, Li S, Zu Y, Yang G, Yang Z, Luo M, Jiang S, Wink M, Efferth
T (2009) Medicinal chemistry of paclitaxel and its analogues.
Curr Med Chem 16:3966–3985
ΔGbinding ¼ ΔGO þ ΔGhbond þ ΔGlipo þ ΔGrot
ð1Þ
The Chemscore function in our work can be written in
the form:
ΔG ¼ À5:4800 þ À3:3400 Ã S ðhbondÞ þ À0:1170
ÃS ðlipoÞ þ 2:5600 Ã HðrotÞ
Georg GI, Cheruvallath ZS, Himes RH, Mejillano MR, Burke CT
(1992a) Synthesis of biologically active taxol analogues with
modified phenylisoserine side chains.
35:4230–4237
J
Med Chem
ð2Þ
Georg GI, Cheruvallath ZS, Himes RH, Mejillano MR (1992b) Novel
biologically active taxol analogues: Baccatin iii 13-(n-(p-chlor-
obenzoyl)-(2′r,3′s)-3′-phenylisoserinate) and baccatin iii 13-(n-
benzoyl-(2′r,3′s)-3′-(p-chlorophenyl)isoserinate). Bioorg Med
Chem Lett 2:295–298
Georg GI, Boge TC, Cheruvallath ZS, Harriman GCB, Hepperle M,
Park H, Himes RH (1994) Cheminform abstract: Schotten-
baumann acylation of n-debenzoyltaxol; an efficient route to n-
acyl taxol analogues and their biological evaluation. ChemInform
25:no–no
Gunda IG, Zacharia SC, Himes RH, Mejillano MR (1992) Semi-
synthesis and biological activity of taxol analogues: Baccatin iii
13-(n-benzoyl-(2′r,3′s)-3′-(p-tolyl)isoserinate), baccatin iii 13-(n-
(p-toluoyl)-2′r,3′s)-3′-phenylisoserinate), baccatin iii 13-(n-ben-
zoyl-(2′r,3′s) -3′-(p- trifluoromethylphenyl)isoserinate), and bac-
catin iii 13-(n-(p-trifluoromethylbenzoyl)- (2′r,3′s)-3′-phenyl-
isoserinate). Bioorg Med Chem Lett 2:1751–1754
Hayashi Y, Skwarczynski M, Hamada Y, Sohma Y, Kimura T, Kiso Y
(2003) A novel approach of water-soluble paclitaxel prodrug with
no auxiliary and no byproduct: design and synthesis of isotaxel. J
Med Chem 46:3782–3784
Each component of this equation is the product of a term
depending on the magnitude of a particular physical con-
tribution to free energy (e.g., hydrogen bonding). The
hydrogen bond term (S(hbond)) is calculated for all com-
plementary possibilities of hydrogen bonds between ligand
atoms and receptor atoms. The lipophilic term (S(lipo)) is
calculated for all lipophilic ligand atoms and all lipophilic
receptor atoms. H(rot) is used to estimate the flexibility
penalty for molecules possessing frozen rotatable bonds
(Eldridge et al. 1997).
Acknowledgements We gratefully acknowledge financial support of
this work by the National Natural Science Foundation of China
(21172262), the Chinese National Science & Technology Major Pro-
ject (grants 2010ZXJ0900X-007), the Foundation of the 309th hospital
(2015MS-011), the National Natural Science Foundation of China
(grant number 21503272) and the General Financial Grant from the
Hummer G, Rasaiah JC, Noworyta JP (2001) Water conduction
through the hydrophobic channel of a carbon nanotube. Nature
414:188–190
China
Postdoctoral
Science
Foundation
(grant
number
2015M572753).
Ji Z, Ahmed AA, Albert DH, Bouska JJ, Bousquet PF, Cunha GA,
Diaz G, Glaser KB, Guo J, Harris CM et al. (2008) 3-amino-
benzo[d]isoxazoles as novel multitargeted inhibitors of receptor
tyrosine kinases. J Med Chem 51:1231–1241
Jones G, Willett P, Glen RC, Leach AR, Taylor R (1997) Develop-
ment and validation of a genetic algorithm for flexible docking. J
Mol Biol 267:727–748
Jones G, Willett P, Glen RC (1995) Molecular recognition of receptor
sites using a genetic algorithm with a description of desolvation. J
Mol Biol 245:43–53
Kingston DG, Snyder JP (2014) The quest for a simple bioactive
analog of paclitaxel as a potential anticancer agent. Acc Chem
Res 47:2682–2691
Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of
interest.
References
Baruchello R, Simoni D, Grisolia G, Barbato G, Marchetti P, Ron-
danin R, Mangiola S, Giannini G, Brunetti T, Alloatti D et al.
(2011) Novel 3,4-isoxazolediamides as potent inhibitors of cha-
perone heat shock protein 90. J Med Chem 54:8592–8604
Baxter CA, Murray CW, Clark DE, Westhead DR, Eldridge MD
(1998) Flexible docking using tabu search and an empirical
estimate of binding affinity. Protein: Struct, Funct, Bioinforma
33:367–382
Kirikae T, Ojima I, Kirikae F, Ma Z, Kuduk SD, Slater JC, Takeuchi
CS, Bounaud PY, Nakano M (1996) Structural requirements of
taxoids for nitric oxide and tumor necrosis factor production by
murine macrophages. Biochem Biophys Res Commun
227:227–235
Case TADDA, Cheatham ITE, Simmerling JWCL (2012) AMBER12.
University of California, San Francisco

Medicinal Chemistry Research
Kirikae T, Ojima I, Ma Z, Kirikae F, Hirai Y, Nakano M (1998)
Structural significance of the benzoyl group at the c-3’-n position
of paclitaxel for nitric oxide and tumor necrosis factor production
by murine macrophages. Biochem Biophys Res Commun
245:698–704
Lamberth C, Kempf HJ, Kriz M (2007) Synthesis and fungicidal
activity of n-2-(3-methoxy-4- propargyloxy) phenethyl amides.
Part 3: stretched and heterocyclic mandelamide oomyceticides.
Pest Manage Sci 63:57–62
Roh EJ, Kim D, Lee CO, Choi SU, Song CE (2002) Structure–activity
relationship study at the 3′-n-position of paclitaxel: Synthesis and
biological evaluation of 3′-n-acyl-paclitaxel analogues. Bioorg
Med Chem 10:3145–3151
Roh EJ, Song CE, Kim D, Pae HO, Chung HT, Lee KS, Chai KB, Lee
CO, Un CS (1999) Synthesis and biology of 3′-n-acyl-n-deben-
zoylpaclitaxel analogues. Bioorg Med Chem 7:2115–2119
Rowbottom MW, Faraoni R, Chao Q, Campbell BT, Lai AG, Setti E,
Ezawa M, Sprankle KG, Abraham S, Tran L et al. (2012) Iden-
tification of 1-(3-(6,7-dimethoxyquinazolin- 4-yloxy)phenyl)-3-
Lilienkampf A, Mao J, Wan B, Wang Y, Franzblau SG, Kozikowski
AP (2009) Structure-activity relationships for
a
series of
(5-(1,1,1-trifluoro-2-methylpropa
n-2-yl)isoxazol-3-yl)urea
quinoline-based compounds active against replicating and non-
hydrochloride (cep-32496), a highly potent and orally efficacious
inhibitor of v-raf murine sarcoma viral oncogene homologueb1
(braf) v600e. J Med Chem 55:1082–1105
replicating mycobacterium tuberculosis.
J
Med Chem
52:2109–2118
Liu JY, Chen XE, Zhang YL (2015) Insights into the key interactions
between human protein phosphatase 5 and cantharidin using
molecular dynamics and site-directed mutagenesis bioassays. Sci
Rep 5:12359
Liu J, Tian Z, Zhang Y (2016) Structure-based discovery of potentially
active semiochemicals for cydia pomonella (l.). Sci Rep 6:34600
Liu J, Yang X, Zhang Y (2014) Characterization of a lambda-
cyhalothrin metabolizing glutathione s-transferase cpgstd1 from
cydia pomonella (l.). Appl Microbiol Biotechnol 98:8947–8962
Lu HF, Xie C, Chang J, Lin GQ, Sun X (2011) Synthesis, cytotoxicity,
metabolic stability and pharmacokinetic evaluation of fluorinated
docetaxel analogs. Eur J Med Chem 46:1743–1748
Rowinsky EK (1997) The development and clinical utility of the
taxane class of antimicrotubule chemotherapy agents. Annu Rev
Med 48:353–374
Shang YJ, Wang YG (2002) Synthesis of isoxazolines and isoxazoles
using poly (ethylene glycol) as Suppor. Synthesis 12:1663–1668
Swindell CS, Krauss N (2001) Syntheses of taxol, taxol analogs and
their intermediates with variable A-ring side chain structures and
compositions thereof. US Patent 6262281B1
Tian Z, Liu J, Zhang Y (2016a) Key residues involved in the inter-
action between cydia pomonella pheromone binding protein 1
(cpompbp1) and codlemone. J Agric Food Chem 64:7994–8001
Tian Z, Liu J, Zhang Y (2016b) Structural insights into cydia pomo-
nella pheromone binding protein 2 mediated prediction of
potentially active semiochemicals. Sci Rep 6:22336
Mekhail TM, Markman M (2002) Paclitaxel in cancer therapy. Expert
Opin Pharmacother 3:755–766
Mosmann T (1983) Rapid colorimetric assay for cellular growth and
survival: application to proliferation and cytotoxicity assays. J
Immunol Methods 65:55–63
Wang J, Wolf RM, Caldwell JW, Kollman PA, Case DA (2004)
Development and testing of a general amber force field. J Comput
Chem 25:1157–1174
Ojima I, Fumero-Oderda CL, Kuduk SD, Ma Z, Kirikae F, Kirikae T
(2003) Structure–activity relationship study of taxoids for their
ability to activate murine macrophages as well as inhibit the
growth of macrophage-like cells. Bioorg Med Chem
11:2867–2888
Yang XQ, Liu JY, Li XC, Chen MH, Zhang YL (2014) Key amino
acid associated with acephate detoxification by cydia pomonella
carboxylesterase based on molecular dynamics with alanine
scanning and site-directed mutagenesis. J Chem Inf Model
54:1356–1370