Semisynthesis and quantitative structure-activity relationship (QSAR) study of some cholesterol-based hydrazone derivatives as insecticidal agents

Article abstract of DOI:10.1016/j.bmcl.2013.06.099

In continuation of our program aimed at the discovery and development of natural-product-based insecticidal agents, four series of novel cholesterol-based hydrazone derivatives were synthesized, and their insecticidal activity was tested against the pre-third-instar larvae of oriental armyworm, Mythimna separata (Walker) in vivo at 1 mg/mL. All the derivatives showed the better insecticidal activity than their precursor cholesterol. Quantitative structure-activity relationship (QSAR) model demonstrated that six descriptors such as RDF085v, Mor06u, Mor11u, Dv, HATS0v and H-046, are likely to influence the insecticidal activity of these compounds. Among them, two important ones are the Mor06u and RDF085v.

Full text of DOI:10.1016/j.bmcl.2013.06.099

Accepted Manuscript
Semisynthesis and quantitative structure-activity relationship (QSAR) study of
some cholesterol-based hydrazone derivatives as insecticidal agents
Chun Yang, Yonghua Shao, Xiaoyan Zhi, Qu Huan, Xiang Yu, Xiaojun Yao,
Hui Xu
PII:
S0960-894X(13)00833-0
DOI:
http://dx.doi.org/10.1016/j.bmcl.2013.06.099
BMCL 20664
Reference:
To appear in:
Bioorganic & Medicinal Chemistry Letters
Received Date:
Revised Date:
Accepted Date:
24 May 2013
24 June 2013
28 June 2013
Please cite this article as: Yang, C., Shao, Y., Zhi, X., Huan, Q., Yu, X., Yao, X., Xu, H., Semisynthesis and
quantitative structure-activity relationship (QSAR) study of some cholesterol-based hydrazone derivatives as
insecticidal agents, Bioorganic & Medicinal Chemistry Letters (2013), doi: http://dx.doi.org/10.1016/j.bmcl.
2013.06.099
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Semisynthesis and quantitative structure-activity relationship (QSAR)
study of some cholesterol-based hydrazone derivatives as insecticidal
agents
Chun Yang,1 Yonghua Shao,2 Xiaoyan Zhi,1 Qu Huan,1 Xiang Yu,1 Xiaojun Yao,2 and
Hui Xu*,1
1Laboratory of Pharmaceutical Design & Synthesis, College of Sciences, Northwest
A&F University, Yangling 712100, Shaanxi Province, People′s Republic of China
2Department of Chemistry, Lanzhou University, Lanzhou 730000, Gansu Province,
People′s Republic of China
* Author to whom correspondence should be addressed [(H.X.) telephone
+86(0)29-87091952; fax: +86(0)29-87091952; e-mail: orgxuhui@nwsuaf.edu.cn].
1

Abstract: In continuation of our program aimed at the discovery and development of
natural-product-based insecticidal agents, four series of novel cholesterol-based
hydrazone derivatives were synthesized, and their insecticidal activity was tested
against the pre-third-instar larvae of oriental armyworm, Mythimna separata (Walker)
in vivo at 1 mg/mL. All the derivatives showed the better insecticidal activity than
their precursor cholesterol. Quantitative structure-activity relationship (QSAR) model
demonstrated that six descriptors such as RDF085v, Mor06u, Mor11u, Dv, HATS0v
and H-046, are likely to influence the insecticidal activity of these compounds.
Among them, two important ones are the Mor06u and RDF085v.
KEYWORDS: Cholesterol; Insecticidal activity; Botanical insecticide; Structural
modification; QSAR; Mythimna separata Walker
2

Although synthetic agrochemicals have played a dominant role in modern
agricultural pest management, the repeated applications over the years have resulted
in emergence of resistance in insect pest populations and environmental problems.1
Hence development of new effective, selective and safe pesticides is still a
challenging task. Nowadays, a component of integrated pest management (IPM)
programs is application of pesticides from natural products, which have received
much research attention worldwide.2-7 Some pesticides from natural products such as
nicotine, avermectins, azadirachtin, pyrethrum and neem extracts are characteristic
examples as defenses against pests.8-10
Cholesterol (1, Figure 1), a waxy steroid, is transported in the blood plasma of all
animals. Compound 1 is the main sterol synthesized by animals, and small amounts
are also found in plants and fungi. Recently, compound 1 has been used as the
readily available lead compound for the preparation of potent anticancer agents,11-15
antifungal agents, 16-19 and antileishmanial agents.20 By contrast, to the best of our
knowledge, little attention has been paid to compound 1 acting as the starting point
for structural modifications as insecticidal agents until now. More recently, we found
that introduction of hydrazone fragments into fraxinellone could lead to more
pronounced derivatives as compared with toosendanin, a commercial botanical
insecticide isolated from Melia azedarach.21 Meanwhile, Brunel et al. described
tosylhydrazone cholesterol derivatives exhibited antifungal activity against Candida
albicans.22 Encouraged by the above-mentioned interesting results, and in
continuation of our program aimed at the discovery and development of novel
3

natural-product-based pesticidal agents,23-26 we herein synthesized four series of
novel cholesterol-based hydrazone derivatives by introduction of hydrazone
fragments on the cholesterol skeleton. Their insecticidal activity was tested against
the pre-third-instar larvae of oriental armyworm, Mythimna separata (Walker) in
vivo. In addition, their quantitative structure-activity relationships (QSAR) were also
reported.
As described in Scheme 1, firstly oxidation of 1 in the presence of chromium
trioxide (CrO3) and pyridine afforded cholest-4-ene-3,6-dione (2), which then
regioselectively reacted with hydrazides or hydrazines to give 3a-j. Among them, the
single-crystal structures of 3f and 3h were illustrated in Figure 2 and 3,27 respectively.
It demonstrated that hydrazone fragments of 3f and 3h were at the C-3 position and
the substituents on the C=N double bond of 3f and 3h adopted E configuration. As
shown in Scheme 2, in the presence of CrO3 and pyridine, 7-keto cholesterol
benzenesulfonate (5) was obtained by regioselective allylic oxidation of cholesterol
benzenesulfonate (4), which was derived from 1 reacting with benzenesulfonyl
chloride. Compounds 6a-c, f and g were then produced by reaction of 5 with
hydrazides or hydrazines. 7-Keto cholesterol acetate (8), as shown in Scheme 3, was
firstly obtained by regioselectively allylic oxidation of cholesterol acetate (7) in the
presence of CrO3 and pyridine. Then removing acetyloxy of 8 in the presence of
p-toluenesulfonic acid (p-TSA) afforded cholest-3,5-diene-7-one (9), which finally
reacted with hydrazides or hydrazines to give 10b, and e-g. As described in Figure
4,27 the configuration of 10f was unambiguously confirmed by single-crystal X-ray
4

diffraction. It suggested that the substituents on the C=N double bond at the C-7
position of 10f adopted configuration. As depicted in Scheme 4,
E
cholest-1,4,6-triene-3-one (11) was obtained by oxidation of 1 in the presence of
2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ). Then reduction of 11 in the
presence of NaBH4 afforded cholest-4,6-diene-3-ol (12), followed by oxidation in the
presence of pyridinium chlorochromate (PCC) to give cholest-4,6-diene-3-one (13).
Compound 13 then reacted with hydrazides or hydrazines to give 14a-g. Similarly, the
configuration of 14g was determined by single-crystal X-ray diffraction (Figure 5),27
and obviously the substituents on the C=N double bond at the C-3 position of 14g
1
adopted E configuration. All the target compounds were characterized by H NMR,
IR, HRMS and mp (see Supplementary data).
The insecticidal activity of compounds 1, 2, 3a-j, 5, 6a-c, f and g, 8, 9, 10b and
e-g, 11-13, and 14a-g against the pre-third-instar larvae of M. separata in vivo was
tested by the leaf-dipping method at the concentration of 1 mg/mL.28 As outlined in
Table 1, the corresponding mortality rates of tested compounds after 35 days were
generally higher than those after 10 and 20 days. Therefore, these compounds
exhibited the delayed insecticidal activity. Moreover, the symptoms of the tested M.
separata were also characterized by the same way as our previous reports.23-26 To our
delight, all the derivatives showed the better insecticidal activity than their precursor
cholesterol. Especially 3g, 3h, 8, 10e and 13 exhibited more promising insecticidal
activity than toosendanin. Among 3a-j, 3g (containing 2,4-dinitro group on the phenyl)
and 3h (containing 2,3,5,6-tetrafluoro group on the phenyl) showed the potent
5

insecticidal activity. To 6a-c, f and g, introduction of 2-thienylacyl or cyanoacetyl
hydrazone resulted in less active derivatives (e.g., 6b and 6c). Among 10b and e-g,
10e displayed the potent insecticidal activity. To compounds 14a-g, introduction of
cyanoacetyl hydrazone could result in the active derivative 14c with the final
mortality rate of 48.3%. Finally, among those intermediates (2, 5, 8, 9, and 11-13), 8
and 13 exhibited the pronounced insecticidal activity. Introduction of acetyloxy at the
C-3 position of 7-keto cholesterol led to the more potent compound than that
containing benzenesulfonyloxy at the same position (55.2% for 8 versus 41.4% for 5).
It suggested that different kinds of acyloxy substituents could be introduced at the C-3
position of 7-keto cholesterol to prepare its esters as insecticidal agents. Meanwhile,
reduction of double bond at the C-1 position of 11 or oxidation of hydroxyl group at
the C-3 position of 12 to 3-keto could result in the more active compound 13 (55.2%
for 13 vs. 34.5% for 11 and 31% for 12). Retaining the existing functional groups of
13 such as two double bonds at the C-4 and C-6 positions, and 3-keto, therefore, we
could modify other positions of 13 to investigate its more potent derivatives.
The biological activity of 34 compounds (1, 2, 3a-j, 5, 6a-c, f and g, 8, 9, 10b
and e-g, 11-13, and 14a-g) was expressed by final mortality rate values and used as
dependent variable in the following analyses. The structures of the studied compounds
were drawn and pre-optimized with the MM+ molecular mechanics force field by
using the HYPERCHEM software.29 The final lowest energy conformation was
minimized by means of the semi-empirical AM1 method.30 The theoretical molecular
descriptors of the optimal structure were then calculated using the DRAGON
6

software.31 The details about the meaning of the calculated descriptors could be found
in the Handbook of Molecular Descriptors.32 A total of 1664 descriptors could be
obtained to describe the structure characteristics of the compound, including many
different types of 0D-3D molecular descriptors. In the first pre-selection analysis,
constant or near-constant values were removed. In order to minimize the redundant
information, descriptors with pairwise correlation coefficient greater than 0.99 were
excluded. Thus, 667 molecular descriptors were remained. To extract the most
relevant descriptors with the final mortality rate of the compounds, the 667
descriptors were subjected to genetic algorithm (GA)33 for features selection. Based
on the molecular descriptors selected by GA program, multiple linear regression
(MLR) method was used to build the QSAR model. In this study, the genetic
algorithm and multiple linear regression (GA-MLR) procedure were performed by the
MobyDigs software.34 During variable selection and QSAR model building by
GA-MLR, the best model reached until increasing the number of the variables did not
increase the internal prediction performance, evaluated by leave-one-out
cross-validation correlation coefficient Q2LOO. The parameters used in the GA
calculations were as follows: population size 100, maximum allowed descriptors in
final model 6 and reproduction/mutation trade-off 0.5.
The performance of the proposed model was evaluated by several statistic
parameters, such as squared correlation coefficient (R2) for fitness ability,
leave-one-out cross-validation correlation coefficient (Q2LOO) for internal predictive
ability and root mean square error (RMSE). The plot of the standardized residuals
7

versus the leverage value35 (Williams plot), was employed to describe the
applicability domain (AD) of the QSAR model. The Williams plot was used to verify
the presence of response outliers (i.e., compounds with cross validate standardized
residuals greater than three standard deviation units, >3σ) and X outliers (i.e.,
compounds with leverage values greater than warning leverage h*). The warning
leverage used in this work is defined as h* = 3(k+1)/n, where k is the number of model
variables and n is the number of the compounds used to develop the model.
The regression equation and the statistical parameters of the 6-variable model are
as follows:
Y = 0.852RDF085v + 0.882Mor06u – 0.568Mor11u + 0.322Dv – 0.470HATS0v –
0.491H-046 + 1.358
(1)
Where Y is the final mortality rate.
ndataset = 34, R2 = 0.814, RMSE = 0.035, Q2LOO = 0.742, RMSELOO = 0.042
The predicted final mortality rate values of 34 compounds were shown in Table 2.
The plot of the experimental vs. predicted values was shown in Figure 6. In the six
selected descriptors, there is one RDF descriptor, two 3D-MoRSE descriptors, one
WHIM descriptor,36 one GETAWAY descriptor37 and one atom-centred fragment
descriptor. In the Eq. (1), the relative importance of the descriptors was determined by
standardized regression coefficient values. The most relevant descriptor is the Mor06u
(3D-MoRSE-signal 06 / unweighted), which belongs to 3D-MoRSE descriptor.
3D-MoRSE (3D-Molecule Representation of Structure based on Electron diffraction)
descriptors provide 3D information from the 3D coordinates by using the same
8

transform as in electron diffraction. The second important descriptor is RDF085v,
which is one of the radial distribution function (RDF) descriptors. RDF descriptors
are calculated from the molecular geometry, representing the molecular conformation
in 3D with a series of weighting schema, such as weighted by atomic masses, atomic
van der Waals volumes, atomic Sanderson electronegativities and atomic
polarizabilities. The descriptor RDF085v was proposed as Radial Distribution
Function – 8.5 / weighted by atomic van der Waals volumes which relates to van der
Waals volumes of molecule. Dv is a WHIM descriptor, which represents D total
accessibility index / weighted by atomic van der Waals volumes. WHIM (Weighted
Holistic Invariant Molecular) descriptors are built in such a way to capture relevant
molecular 3D information regarding molecular size, shape, symmetry and atom
distribution with respect to invariant reference frames. All above three descriptors
have the positive coefficient values, which indicate that these descriptors had a
positive effect on final mortality rate of compounds. Mor11u is another 3D-MoRSE
descriptor similar to Mor06u, and it is represents the 3D-MoRSE signal 23/
unweighted. HATS0v is a GETAWAY descriptor, which represents leverage –
weighted autocorrelation of lag 0 / weighted by atomic van der Waals volumes.
GETAWAY (Geometry, Topology, and Atom-Weights-Assembly) are derived from the
leverage matrix which is deduced by the centering all of the atomic coordinates.
H-046 is an atom-centred fragment descriptor, which indicates the H attached to CO
(sp3) no X attached to next C. The atom-centred fragment descriptors are simple
molecular descriptors defined as the number of specific atom types in a molecule.
9

They are calculated by knowing the molecular composition and atom connectivity.
The applicability domain of the proposed model was displayed by the Williams plot
(Figure 7). This means that there is no structure outlier in this model and all
compounds with leverage values less than warning value (h* = 0.618). Only 5 (No.13)
was not well predicted with the largest residue of -0.125.
In conclusion, four series of novel cholesterol-based hydrazone derivatives were
synthesized, and evaluated for their insecticidal activity against the pre-third-instar
larvae of M. separata in vivo. All the derivatives displayed the better insecticidal
activity than their precursor cholesterol. Especially 3g, 3h, 8, 10e and 13, whose final
mortality rates were all increased greater than 30% compared to that of cholesterol.
QSAR model demonstrated that six descriptors such as one RDF descriptor, two
3D-MoRSE descriptors, one WHIM descriptor, one GETAWAY descriptor and one
atom-centred fragment descriptor are likely to influence the final mortality rate of
these compounds. Among them, two important ones are the Mor06u and RDF085v. It
will lay the foundation for further structural modification and development of
cholesterol as insecticidal agents.
Acknowledgments
The present research was supported by National Natural Science Foundation of
China (No.31171896), and Special Funds of Central Colleges Basic Scientific
Research Operating Expenses (YQ2013008) to Dr. Xu.
Supplementary data
10

Supplementary data associated with this article can be found, in the online version.
References and notes
1. Heckel, D. G. Science 2012, 337, 1613.
2. Chen, J. T.; Su, H. J.; Huang, J. W. J. Agric. Food Chem. 2012, 60, 7341.
3. Zhou, Y. Y. ; Luo, S. H.; Yi, T. S.; Li, C. H.; Luo, Q.; Hua, J.; Liu, Y.; Li, S. H. J.
Agric. Food Chem. 2011, 59, 6004.
4. Goel, M.; Dureja, P.; Rani, A.; Uniyal, P. L.; Laatsch, H. J. Agric. Food Chem. 2011,
59, 2299.
5. Caballero-Gallardo, K.; Olivero-Verbel, J.; Stashenko, E. E. J. Agric. Food Chem.
2011, 59, 1690.
6. Isman, M. B. Pest Manag. Sci. 2008, 64, 8.
7. Lin, L.; Mulholland, N.; Wu, Q. Y.; Beattie, D.; Huang, S. W.; Irwin, D.; Clough, J.;
Gu, Y. C.; Yang, G. F. 2012, 60, 4480.
8. Dayan, F. E.; Cantrell, C. L.; Duke, S. O. Bioorg. Med. Chem. 2009, 17, 4022.
9. Morgan E. D. Bioorg. Med. Chem. 2009, 17, 4096.
10. Pitterna, T.; Cassayre, J.; Hü ter, O. F.; Jung, P. M. J.; Maienfisch, P.; Kessabi, F.
M.; Quaranta, L.; Tobler, H. Bioorg. Med. Chem. 2009, 17, 4085.
11. Cui, J.; Wang, H.; Huang, Y.; Xin, Y.; Zhou, A. Steroids 2009, 74, 1057.
12. Cui, J.; Huang, L.; Fan L.; Zhou, A. Steroids 2008, 73, 252.
13. Poza, J.; Rega, M.; Paz, V.; Alonso, B.; Rodriguez, J.; Salvador, N.; Fernandez, A.;
Jimenez, C. Bioorg. Med. Chem. 2007, 15, 4722.
14. Cui, J.; Fan, L.; Huang, L.; Liu, H.; Zhou, A. Steroids 2009, 74, 62.
11

15. Deive, N.; Rodriguez, J.; Jimenez, C. J. Med. Chem. 2001, 44, 2612.
16. Brunel, J. M.; Loncle, C.; Vidal, N.; Dherbomez, M.; Letourneux, Y. Steroids
2005, 70, 907.
17. Banday, M. R.; Farshori, N. N.; Ahmad, A.; Khan, A. U.; Rauf, A. Eur. J. Med.
Chem. 2010, 45, 1459.
18. Khan, S. A.; Kumar, P.; Joshi, R.; Iqbal, P. F.; Saleem, K. Eur. J. Med. Chem.
2008, 43, 2029.
19. El kihel, L.; Choucair, B.; Dherbomez, M.; Letourneux, Y. Eur. J. Org. Chem.
2002, 4075.
20. Bazin, M.; Loiseau, P. M.; Bories, C.; Letourneux, Y.; Rault, S.; El Kihel, L. Eur. J.
Med. Chem. 2006, 41, 1109.
21. Guo, Y.; Yan, Y.; Yang, C.; Yu, X.; Zhi, X.; Xu, H. Bioorg. Med. Chem. Lett. 2012,
22, 5384.
22. Brunel, J. M.; Loncle, C.; Vidal, N.; Dherbomez, M.; Letourneux, Y. Eur. J. Med.
Chem. 2004, 39, 1067.
23. Xu, H.; Wang, Q.; Guo, Y. Chem. Eur. J. 2011, 17, 8299.
24. Guo, Y.; Yan, Y.; Yu, X.; Wang, Y.; Xu, H. J. Agric. Food Chem. 2012, 60, 7016.
25. Wang, Y.; Shao, Y.; Wang, Y.; Fan, L.; Yu, X.; Zhi, X.; Yang, C.; Qu, H.; Yao, X.;
Xu, H. J. Agric. Food Chem. 2012, 60, 8435.
26. Xu, H.; Xiao, X.; Zhao, X. F.; Guo, Y.; Yao, X. J. Bioorg. Med. Chem. Lett. 2011,
21, 4008.
27. Crystallographic data (excluding structure factors) for the structures of 3f, 3h, 10f
12

and 14g have been deposited with the Cambridge Crystallographic Data Centre
as supplementary publication number CCDC 888312, 888770, 912936, and
917434, 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 e-mail: deposit@ccdc.cam.ac.uk].
28. Xu, H.; Wang, J. J.; Sun, H. J.; Lv, M.; Tian, X.; Yao, X. J.; Zhang, X. J. Agric.
Food Chem. 2009, 57, 7919.
29. Hypercube, Inc. HyperChem 7.0; Gainesville, FL, 2002.
30. Dewar, M. J. S.; Zoebisch, E. G.; Healy, E. F.; Stewart, J. J. P. J. Am. Chem. Soc.
1985, 107, 3902.
31. Talete srl, DRAGON for windows (software for molecular descriptor calculations).
Version 5.4, 2006; http://www.talete.mi.it.
32. Todeschini, R.; Consonni, V. Handbook of Molecular Descriptors: Methods and
Principles in Medicinal Chemistry; Wiley-VCH: Weinheim, Germany, 2000.
33. Leardi, R.; Boggia, R.; Terrile, M. J. Chemom. 1992, 6, 267.
34. Todeschini, R.; Consonni, V.; Pavan, M. MOBY DIGS, version 1.2 for Windows,
software for multilinear regression analysis and variable subset selection by
genetic algorithm; Talete Srl, Milan, Italy, 2002.
35. Tropsha, A.; Gramatica, P.; Gombar, V. K. QSAR Comb. Sci. 2003, 22, 69.
36. Todeschini, R.; Gramatica, P. Perspect. Drug Discov. Des. 1998, 9-11, 355.
37. Consonni, V.; Todeschini, R.; Pavan, M. J. Chem. Inform. Comput. Sci. 2002, 42,
682.
13

Table 1
Insecticidal activity of 3a-j, 6a-c, f and g, 10b and e-g, and 14a-g against M. separata
on leaves treated with a concentration of 1 mg/mL.a
Corrected mortality rate (%)
Compound
10 days
20 days
35 days
1
6.7 3.3
13.8 3.3
31.0 3.3
27.6 5.8
34.5 3.3
31.0 3.3
20.7 3.3
44.8 3.3
37.9 5.8
48.3 5.8
2
0
0
3a
3.3 3.3
3.3 3.3
3b
3c
10.0
0
37.9
0
3d
3.3 3.3
3.3 3.3
3.3 3.3
10.0 5.8
17.2
34.5 3.3
48.3
31.0 3.3
31.0 3.3
31.0 6.7
0
27.6 5.8
44.8 3.3
3e
3f
0
48.3
0
3g
55.2 3.3
51.7 6.7
44.8 3.3
3h
10.0
0
3i
3.3 3.3
10.0 5.8
6.7 3.3
10.0 5.8
3.3 3.3
3.3 3.3
3.3 3.3
6.7 3.3
23.3 3.3
3j
37.9
0
48.3
0
5
31.0 3.3
31.0 3.3
41.4 3.3
44.8 3.3
34.5 3.3
34.5 6.7
41.4 3.3
6a
6b
17.2
0
6c
27.6 5.8
34.5 3.3
20.7 3.3
44.8 3.3
20.7 3.3
24.1 3.3
34.5 3.3
17.2 5.8
20.7 3.3
13.8 6.7
13.8 3.3
6f
6g
37.9
0
8
55.2 3.3
41.4 6.7
31.0 6.7
51.7 3.3
34.5 3.3
41.4 6.7
34.5 3.3
31.0 3.3
55.2 3.3
31.0 3.3
37.9 5.8
48.3 5.8
34.5 3.3
37.9 5.8
44.8 3.3
41.4 3.3
48.3 5.8
9
0
0
10b
3.3 3.3
3.3 3.3
3.3 3.3
10e
10f
10g
10.0
0
11
6.7 3.3
12
0
0
13
13.3 3.3
13.3 6.7
6.7 3.3
6.7 3.3
3.3 3.3
3.3 3.3
37.9
0
14a
17.2 5.8
27.6 5.8
34.5 6.7
13.8 3.3
31.0 3.3
37.9 5.8
31.0 3.3
31.0 3.3
14b
14c
14d
14e
14f
0
0
14g
toosendaninb
3.3 3.3
23.3 3.3
a Values are means S.D. of three replicate.
b Toosendanin was used as a positive control at 1 mg/mL.
14

Table 2
Experimental and predicted activity by developed QSAR model.
Number
1
Compound
1
Experimental activity Predicted activity
0.207
0.448
0.379
0.483
0.379
0.276
0.448
0.483
0.552
0.517
0.448
0.483
0.414
0.448
0.345
0.345
0.414
0.379
0.552
0.414
0.31
0.228
0.467
0.364
0.456
0.434
0.3
2
2
3
4
5
6
7
8
9
3a
3b
3c
3d
3e
0.405
0.469
0.537
0.522
0.504
0.434
0.47
3f
3g
3h
3i
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
3j
5
6a
0.455
0.352
0.347
0.362
0.375
0.534
0.435
0.328
0.499
0.344
0.368
0.347
0.339
0.443
0.332
0.435
0.482
0.316
0.371
0.446
0.465
6b
6c
6f
6g
8
9
10b
10e
10f
10g
11
0.517
0.345
0.414
0.345
0.31
12
13
0.552
0.31
14a
14b
14c
14d
14e
14f
14g
0.379
0.483
0.345
0.379
0.448
0.414
15

12
11
9
17
16
13
14
1
15
2
10
8
7
3
5
4
6
Figure 1. The chemical structure of cholesterol (1).
Figure 2. The X-ray crystal structure of 3f.
16

Figure 3. The X-ray crystal structure of 3h.
Figure 4. The X-ray crystal structure of 10f.
17

Figure 5. The X-ray crystal structure of 14g.
18

0.6
0.5
0.4
0.3
0.2
0.1
Dataset
0.1
0.2
0.3
0.4
0.5
0.6
The experimental values
Figure 6. Plot of experimental and predicted biological activity values of 34
compounds by GA-MLR model.
19

3
2
Dataset
1
0
-1
-2
-3
13
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Leverages
Figure 7. Williams plot for the GA-MLR model with six descriptors.
20

Scheme 1. The synthetic route for the preparation of 3a-j.
H
H
H
BsCl/Py
rt/12 h
H
CrO3/Py
rt/12 h
H
H
O
1
H
H
BsO
1
44% from to
5
5
BsO
4
H
hydrazides
or hydrazines
O
O
H
S
reflux/5-8 h
33-64%
Bs =
H
H
H
N
6a-c,f,g
BsO
N
R
R =
O
O
O2N
O
S
g
a
c
b
f
NC
NO2
NO2
Scheme 2. The synthetic route for the preparation of 6a-c, f and g.
21

H
H
CrO3/Py
t-BuOOH(70%)
H
CH3COCl/Et3N
H
1
H
H
0 oC-rt,17 h
90%
rt/6 h
60%
H
H
7
O
AcO
H
8
AcO
H
H
hydrazides
or hydrazines
p-TSA
H
reflux,4 h
H
H
H
H
H
reflux/5-48 h
19-60%
N
10b,e-g
94%
9
N
R
O
R =
O
O
O2N
S
g
f
e
b
H3CO
NO2
NO2
Scheme 3. The synthetic route for the preparation of 10b and e-g.
Scheme 4. The synthetic route for the preparation of 14a-g.
22

Graphic Abstract
Semisynthesis and quantitative structure-activity relationship (QSAR) study of
some cholesterol-based hydrazone derivatives as insecticidal agents
Chun Yang, Yonghua Shao, Xiaoyan Zhi, Huan Qu, Xiang Yu, Xiaojun Yao, and Hui
Xu*
24