High-throughput screening of ecdysone agonists using a reporter gene assay followed by 3-D QSAR analysis of the molting hormonal activity

Article abstract of DOI:10.1016/j.bmc.2005.09.034

In this study, 172 diacylhydrazine analogs were examined for their ability to activate an ecdysone (molting hormone)-dependent reporter gene in a silkworm (Bombyx mori) cell-based high-throughput screening assay. The measured EC ₅₀ values (concentration required to cause an effect in 50% of the cells) were used to construct a 3-D QSAR model that describes the ecdysone agonist activities of the diacylhydrazine analogs. Of these compounds, 14 exhibited no activity and were excluded from the 3-D QSAR analysis. The resulting equation described ～74% of the activity for 158 compounds. The final equation consisted of 42% electrostatic and 58% steric effects (r ² = 0.74 and q² = 0.45). Comparative molecular field analysis (CoMFA) was used to visualize the steric and electrostatic potential fields that were favorable and unfavorable for biological activity. Of particular interest was the observation that the hydrophobic parameter (log P) was not necessary for describing the observed activities, although previous studies have cited the importance of hydrophobic parameters in both classical and 3-D QSAR analyses of these compounds. Modeling studies of the B. mori ecdysone receptor supported the observed physicochemical parameters required for activity reported by the CoMFA models. Comparison of the present analysis with those performed using other lepidopteran assay systems evidenced a high degree of correlation (r² = 0.81 for a Sf-9 cell-based assay and r ² = 0.89 for a Chilo suppressalis integument-based assay), indicating that it is valid to compare the results generated with the B. mori cell-based system to those generated with previous lepidopteran assays. This novel assay system is amendable to a high-throughput screening format and should greatly increase our ability to discover novel agonists of molting hormone (ecdysone) activity.

Full text of DOI:10.1016/j.bmc.2005.09.034

Bioorganic & Medicinal Chemistry 14 (2006) 1143–1159
High-throughput screening of ecdysone agonists using a
reporter gene assay followed by 3-D QSAR analysis
of the molting hormonal activity
a,b,
a, ,
*
a
Toshiyuki Harada, Nobuhiro Oikawa,
a
Craig E. Wheelock,
Yoshiaki Nakagawa,
c
Miki Akamatsu, Guy Smagghe, Dimitra Stefanou, Kostas Iatrou and Luc Swevers
d
e
e
e
a
Division of Applied Life Sciences, Graduate School of Agriculture, Kyoto University, Kyoto 606-8502, Japan
b
Bioinformatics Center, Institute for Chemical Research, Kyoto University, Uji 611-0011, Japan
Division of Environmental Science and Technology, Graduate School of Agriculture, Kyoto University, Kyoto 606-8502, Japan
c
d
Laboratory of Agrozoology, Department of Crop Protection, Faculty of Bioscience Engineering,
Ghent University, Coupure Links 653, B-9000 Ghent, Belgium
Insect Molecular Genetics and Biotechnology Group, Institute of Biology, National Center for Scientific Research ÔDemokritosÕ,
Aghia Paraskevi, 153 10 Athens, Greece
e
Received 21 July 2005; revised 13 September 2005; accepted 14 September 2005
Available online 24 October 2005
Abstract—In this study, 172 diacylhydrazine analogs were examined for their ability to activate an ecdysone (molting hormone)-de-
pendent reporter gene in a silkworm (Bombyx mori) cell-based high-throughput screening assay. The measured EC50 values (con-
centration required to cause an effect in 50% of the cells) were used to construct a 3-D QSAR model that describes the ecdysone
agonist activities of the diacylhydrazine analogs. Of these compounds, 14 exhibited no activity and were excluded from the 3-D
QSAR analysis. The resulting equation described ꢀ74% of the activity for 158 compounds. The final equation consisted of 42%
2
2
electrostatic and 58% steric effects (r = 0.74 and q = 0.45). Comparative molecular field analysis (CoMFA) was used to visualize
the steric and electrostatic potential fields that were favorable and unfavorable for biological activity. Of particular interest was the
observation that the hydrophobic parameter (logP) was not necessary for describing the observed activities, although previous stud-
ies have cited the importance of hydrophobic parameters in both classical and 3-D QSAR analyses of these compounds. Modeling
studies of the B. mori ecdysone receptor supported the observed physicochemical parameters required for activity reported by the
CoMFA models. Comparison of the present analysis with those performed using other lepidopteran assay systems evidenced a high
2
2
degree of correlation (r = 0.81 for a Sf-9 cell-based assay and r = 0.89 for a Chilo suppressalis integument-based assay), indicating
that it is valid to compare the results generated with the B. mori cell-based system to those generated with previous lepidopteran
assays. This novel assay system is amendable to a high-throughput screening format and should greatly increase our ability to dis-
cover novel agonists of molting hormone (ecdysone) activity.
ꢀ
2005 Elsevier Ltd. All rights reserved.
1
. Introduction
the ligand, the 20E-EcR–USP complex activates tran-
scription through binding to DNA target sites known
as ecdysone response elements (EcREs), which are
located upstream of multiple genes that are involved
Insect molting and metamorphosis are regulated by
the steroidal molting hormone 20-hydroxyecdysone
(20E). The hormone receptor consists of the ecdysone
receptor (EcR) and the ultraspiracle (USP), which
1
4
in molting and metamorphosis. External administra-
tion of excessive doses of molting hormone-active
compounds induces premature molting in larvae lead-
2
,3
together form a heterodimer. Following binding of
5
ing to death. This observation has led to the develop-
ment of a number of compounds designed to target
Keywords: Bombyx mori; Ecdysone receptor; Ecdysone agonists; High-
throughput screening; 3-D QSAR.
6–9
this receptor and exert insecticidal properties.
This
*
Corresponding author. Tel.: +81 75 753 6117; fax: +81 75 753 6123;
e-mail: naka@kais.kyoto-u.ac.jp
These authors contributed equally to this work.
nuclear hormone receptor is not present in mammals
and therefore represents a logical target for insecticide
development.
1
0,11
0
968-0896/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2005.09.034

1144
C. E. Wheelock et al. / Bioorg. Med. Chem. 14 (2006) 1143–1159
0
N-tert-Butyl-N,N -dibenzoylhydrazine and its analogs
are non-steroidal ecdysone agonists that, via binding
to the EcR–USP complex, cause incomplete molting in
insects leading to death. The insecticidal properties of
GFP reporter gene activity, which can be easily quanti-
fied in individual microplate wells using a fluorescence
plate reader. In comparison with other systems that de-
tect ecdysteroid activities, the cell-based system has the
advantage of being extremely rapid and robust, and by
being cell-based it encompasses the necessity of crossing
cellular membranes to exert biological activity. In this
project, we present the first extended 3-D QSAR analy-
sis of diacylhydrazines using this novel assay system. We
measured the activity of 172 diacylhydrazine derivatives
and developed a 3-D QSAR model using comparative
molecular field analysis (CoMFA) to analyze the param-
eters that were beneficial for insecticidal activity. These
effects were then further examined through a series
of modeling studies performed with the B. mori EcR
binding pocket. Results were compared to the CoMFA
fields to provide a comprehensive description of agonist
binding and suggest potential new agonists.
0
N-tert-butyl-N,N -dibenzoylhydrazine were first discov-
ered by Wing et al., who reported that the compound
1
2,13
RH-5849 (1 in Table 1) was an ecdysone agonist.
Since then, a number of studies have examined the range
of activities surrounding this class of compounds and
structure–activity relationships have been extensively
This work has led to the release of a
number of commercial products including tebufenoz-
1
4–21
investigated.
2
2–24
25–27
ide,
halofenozide.
methoxyfenozide,
6
chromafenozide, and
Even though all insects use 20E as
,28,29
a natural molting hormone, their susceptibility to non-
steroidal ecdysteroid agonists exhibits species-specific
variation. Insect order-specific potency was observed
with halofenozide and RH-5849, which have greater
activity towards coleopteran pests such as the Colorado
potato beetle Leptinotarsa decemlineata, and chroma-
fenozide, tebufenozide, and methoxyfenozide which
have extremely low potency toward coleopterans.
However, chromafenozide, tebufenozide and meth-
oxyfenozide are highly toxic to lepidopteran pests such
3
0
2. Results
2
2
2.1. Molting hormonal activity
1
7
as the rice stem borer Chilio suppressalis and the beet
The hormonal activities of 133 compounds are listed in
Table 1. These compounds all have the basic
dibenzoylhydrazine backbone containing an N-tert-bu-
tyl moiety, varying only by their substitution on the
A- and B-rings (see Fig. 1 for A-ring and B-ring nomen-
clature). Among the 133 different dibenzoyl-type com-
pounds tested, the commercially available compound
tebufenozide (107) was the most potent. Another Lepi-
doptera-selective commercial insecticide, methoxyfenoz-
ide (117), was five times less potent than tebufenozide.
Compound 109 containing an i-propyl group at the
para-position of the B-ring was nearly equipotent to
tebufenozide, which contains a para-ethyl group on the
B-ring. RH-5849 (1) and halofenozide (133) were 350-
fold and 70-fold less potent than tebufenozide, respec-
tively. A total of seven compounds, 10, 12, 13, 31, 32,
82, and 100, were inactive.
3
1
armyworm Spodoptera exigua. In addition, these com-
pounds are all ineffective against hemipterans (which in-
cludes beneficial predatory insects such as Geocoris
3
2
punctipes and Orius insidiosus).
The newer generation analogs discussed above have de-
creased mammalian toxicity relative to the initial ana-
3
3
logs reported. Our understanding of the mechanism
of action of these compounds was further complicated
by structural analysis of the EcR–USP heterodimer,
which showed that the region of the complex that binds
a strong agonist with ecdysteroidal structure (ponaster-
one A, Pon A) did not overlap with any of the regions
that bind the non-ecdysteroidal dibenzoylhydrazines
3
4
(
BYI06830, a chromafenozide analog). These findings
suggest that even though both compounds target the
same receptor, they exhibit their biological activity
through different mechanisms. These results have initiat-
ed renewed interest in structure–activity relationships in
an attempt to understand the mechanism by which these
compounds exert their insecticidal effects. The utility of
these compounds and their insecticidal properties as well
as the structural selectivity exhibited by the different
commercial analogs currently available has led to a
number of structure and function studies as described
above. However, a disadvantage of many of these stud-
ies is that they employ in vitro assay systems or time-
consuming and costly radiometric assays. Swevers
et al. recently described a Bombyx mori (silkmoth;
Lepidoptera) cell-based high-throughput screening
system for detecting ecdysteroid agonists and antago-
nists using synthetic compound libraries as well as plant
The activities of 21 compounds (134–154) in which one
of the benzene rings was replaced by a variety of steric
groups (i.e., long-chain alkyl and naphthyl) are listed
in Table 2. Compound 144, which contains an a-naph-
thyl group on the A-ring, was the most potent among
these derivatives. However, 144 was still 10-fold less po-
tent than the corresponding 3,5-dimethyl benzoyl ana-
log, tebufenozide (107). Interestingly, compound 138,
which contained an i-hexyl instead of a naphthyl group,
was inactive.
The bridge moiety of the compound backbone was
modified as shown in Table 3 for compounds 155–
172. Replacement of the tert-butyl group with a CH3
(155) or ethyl (156) resulted in a loss of biological
activity. Substitution with a tert-amyl group (162)
was slightly favored, with this compound exhibiting
the greatest potency of this set of analogs. Compound
162 was also more potent than its tert-butyl analog
compound 1 (pEC50 = 6.81 vs 6.37, respectively). How-
ever, 162 was still 2 orders of magnitude less potent
3
5
extracts. The high-throughput system is based on
transformed Bombyx-derived Bm5 tissue culture cells
that have an ecdysone-inducible green fluorescent
protein (GFP) reporter cassette incorporated into their
genomes. The addition of 20E at 100 nM to 1 lM
quantities results in 500- to 2000-fold stimulation of

C. E. Wheelock et al. / Bioorg. Med. Chem. 14 (2006) 1143–1159
1145
Table 1. Activity of tert-butyl-containing dibenzoylhydrazine congeners with various substituents on the A and B-rings
t-Bu
A
C
O
N
N
H
C
O
B
Xn
Yn
b
Compound
A-ring
B-ring
Activity (pEC50, M)
Observed
LogP
a
Calculated
pEC50
95% CI
pEC50
c
d
d
d
d
d
d
d
d
d
e
1
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
2-F
2-Cl
2-Br
2-I
6.37
6.51
7.12
6.77
6.98
6.20
6.65
6.65
6.23
<4.00
6.28
<4.00
<4.00
6.07
6.68
6.94
6.29
6.15
5.74
5.13
5.82
7.07
6.65
6.78
6.10
4.88
4.83
4.69
4.67
5.45
<4.00
<4.00
4.72
4.36
5.04
5.84
5.89
6.93
6.24
5.48
6.00
5.36
5.59
6.14
4.33
4.64
7.16
5.40
5.34
6.67
4.34
7.29
6.48
6.50
5.89
5.76
6.21
7.14
4.58
5.79–6.81
5.65–7.33
6.94–7.31
6.47–7.19
6.67–7.30
5.91–6.50
6.28–7.01
6.46–6.83
6.06–6.40
—
6.60
5.96
6.35
6.40
6.71
6.06
6.62
5.94
6.15
5.12
6.05
6.16
6.62
6.19
6.30
6.59
6.34
6.58
5.70
6.07
6.03
6.81
6.02
6.17
5.91
5.82
4.73
5.40
4.98
6.00
5.25
5.23
5.99
4.02
5.61
5.58
6.18
5.71
5.77
6.44
6.42
5.90
5.97
6.07
4.85
5.37
6.81
5.27
6.25
5.68
5.50
6.63
6.62
6.03
6.46
5.93
5.81
6.25
5.18
2.45
2.38
2.59
2.69
2.83
2.85
2.27
2.75
2.96
3.89
2.04
3.18
3.66
2.60
2.78
3.28
3.49
3.72
3.61
2.73
2.34
2.79
2.85
3.42
3.66
3.78
3.77
2.63
2.50
2.99
4.11
4.24
2.56
4.50
3.41
3.57
3.10
3.55
3.18
3.25
3.32
2.16
2.34
3.34
2.09
4.39
3.39
4.68
2.91
3.44
3.25
2.63
2.75
2.91
3.11
3.02
1.99
2.91
3.77
2
3
4
5
6
7
8
9
2-F
2-Cl
2-Br
2-I
2-CF
2-NO
3
2
2-CH
2-Et
3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
2-Ph
d
e
2-OCH
2-O-secBu
2-OCH Ph
3
6.07–6.49
—
—
e
2
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
e
2-SCH
3-F
3
5.88–6.25
6.43–6.84
6.69–7.18
6.02–6.56
5.85–6.44
5.53–5.96
4.93–5.33
5.61–6.02
6.84–7.30
6.41–6.89
6.62–6.95
5.81–6.40
4.46–5.30
4.48–5.18
4.49–4.89
4.34–5.01
5.17–5.72
—
3-Cl
3-Br
3-I
3-CF
3
3-NO
3-CN
2
3-CH
4-F
3
4-Cl
4-Br
4-I
4-CF
3
4-NO
4-CN
2
4-CH
3
4-t-Bu
4-Ph
e
—
d
e
4-OCH
3
4.55–4.90
4.07–4.65
4.64–5.44
5.62–6.07
5.71–6.07
6.78–7.08
6.05–6.43
5.24–5.72
5.76–6.24
5.17–5.55
5.36–5.83
5.99–6.29
4.10–4.67
4.47–4.82
7.00–7.31
5.07–6.03
5.15–5.54
6.46–6.88
4.18–4.50
7.25–7.50
6.26–6.70
6.28–6.73
5.72–6.07
5.48–6.03
5.97–6.45
6.89–7.40
4.25–4.91
4-O(CH
2,3-Cl
2-CH
2,3-(CH
2,4-Cl
2 3
) Ph
d
d
d
d
d
d
e
2
3
-3-Cl
3
)
2
2
2,4-(CH
2,5-(CH
3
)
)
2
2
3
2-OCH
2,6-F
2-F-6-Cl
3,4-(CH
3,4-(OCH
3
-5-n-Pr
d
d
d
d
e
2
3 2
)
3
)
2
2,3,4-Cl
3,5-(CH
2,5-Cl
2-OCH
2,3,4,5-F
3
d
e
3
)
2
2
-3-CF
-3,5-(CH
3
e
3
3
)
2
d
d
d
d
f
4
2,3,4,5,6-F
2-Cl
5
2-Cl
2-Cl
f
2-Cl
f
2-Cl
2-Cl
2-CF
2-NO
3
f
2
d
f
2-Cl
2-Cl
2-CH
2-Ph
3
(continued on next page)

1146
C. E. Wheelock et al. / Bioorg. Med. Chem. 14 (2006) 1143–1159
Table 1 (continued)
b
Compound
A-ring
B-ring
Activity (pEC50, M)
Observed
LogP
a
Calculated
pEC50
95% CI
pEC50
f
6
6
6
6
6
6
6
6
6
6
7
7
7
7
7
7
7
7
7
7
8
8
8
8
8
8
8
8
8
8
9
9
9
9
9
9
9
9
9
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
2-Cl
2-Cl
2-OCH
3
5.21
6.17
7.16
6.81
6.72
6.20
6.03
5.70
6.02
6.22
7.28
6.41
6.94
7.47
8.12
8.00
5.46
6.22
7.73
8.24
8.04
7.61
<4.00
7.29
5.89
6.98
8.03
7.70
7.81
6.65
6.79
5.69
5.79
7.53
6.77
5.17
6.34
7.06
6.42
5.07
<4.00
6.99
7.55
6.47
7.18
6.32
7.92
8.91
7.88
8.87
6.75
8.61
7.29
7.72
8.43
8.62
6.89
8.22
7.28
7.26
7.31
8.01
7.16
6.88
4.61–5.80
5.99–6.36
6.99–7.33
6.62–7.00
6.42–7.02
6.03–6.35
5.60–6.45
5.43–5.96
5.88–6.17
5.98–6.46
7.10–7.47
6.25–6.56
6.76–7.13
7.34–7.61
7.93–8.31
7.79–8.21
5.31–5.61
6.13–6.33
7.52–7.83
8.09–8.38
7.77–8.31
7.40–7.82
—
6.02
6.01
5.87
6.53
6.45
6.73
6.30
6.25
5.82
6.88
6.17
6.15
6.85
7.11
7.68
6.91
5.81
6.61
7.26
7.76
7.73
8.16
7.24
7.46
5.82
6.93
6.70
6.88
7.57
6.67
7.35
6.07
6.22
6.41
5.96
5.78
6.32
7.59
6.11
6.19
6.04
6.86
7.29
6.83
6.86
6.60
8.09
8.40
8.38
8.92
8.05
8.93
8.12
7.53
7.58
7.83
7.10
8.03
8.24
7.25
7.01
7.79
7.10
6.14
2.37
f
2-SCH
3-F
3
2.84
d
2-Cl
2-Cl
2.88
d
3-Cl
3-Br
3-I
3.49
f
2-Cl
2-Cl
3.62
f
3.87
f
2-Cl
3-CF
3
3.70
f
2-Cl
2-Cl
3-NO
3-CN
2
2.73
f
2.48
d
2-Cl
3-CH
3
3.11
f
2-Cl
2-Cl
3-OCH
4-F
3
2.81
d
2.87
d
2-Cl
2-Cl
4-Cl
4-Br
4-I
3.51
f
3.73
f
2-Cl
3.96
f
2-Cl
4-CF
3
3.68
f
2-Cl
2-Cl
4-NO
4-CN
2
2.78
f
2.44
d
2-Cl
2-Cl
4-CH
4-Et
3
3.15
f
3.59
f
2-Cl
2-Cl
4-n-Pr
4-i-Pr
4-Ph
4.06
f
4.11
f
2-Cl
4.49
f
2-Cl
2-Cl
4-OCH
3
7.08–7.50
5.75–6.02
6.84–7.11
7.91–8.15
7.55–7.86
7.63–7.98
6.44–6.85
6.60–6.97
5.39–5.99
5.52–6.05
7.41–7.66
6.55–6.99
5.00–5.35
6.12–6.57
6.82–7.31
6.09–6.75
4.82–5.32
—
2.82
g
4-SO
4-COCH
2,3-Cl
2,3-(CH
2-CH -3-OCH
2,4-Cl
2,4-(CH
2,5-Cl
2,5-(CH
2,6-F
2-F-6-Cl
2
CH
3
1.46
g
2-Cl
3
2.42
f
2-Cl
2
3.75
f
2-Cl
3
)
2
3.40
h
2-Cl
3
3
3.46
f
2-Cl
2
3.71
f
2-Cl
2-Cl
3
3
)
)
2
2
3.38
f
2
3.75
f
2-Cl
3.40
f
2-Cl
2
2.35
f
2-Cl
2.67
f
2-Cl
2-Cl
2,6-Cl
3,4-Cl
2
3.03
f
2
4.25
f
2-Cl
2-Cl
3,4-(CH
3,5-Cl
3,5-(CH
3
)
)
2
2
3.65
f
2
4.29
f
2-Cl
3
3.67
f
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
2-Cl
3,5-(O-n-Bu)
2
5.63
h
3,5-(CH
3,5-(CH
3,5-(CH
3,5-(CH
3,5-(CH
3,5-(CH
3,5-(CH
3,5-(CH
3,5-(CH
3,5-(CH
3,5-(CH
3,5-(CH
3,5-(CH
3,5-(CH
3,5-(CH
3,5-(CH
3,5-(CH
3,5-(CH
3,5-(CH
3,5-(CH
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
)
2
)
2
)
2
)
2
)
2
)
2
)
2
)
2
)
2
)
2
)
2
)
2
)
2
)
2
)
2
)
2
)
2
)
2
)
2
)
2
2-CH
3-CH
3
6.72–7.26
7.35–7.75
6.29–6.65
6.99–7.38
6.00–6.64
7.75–8.10
8.70–9.12
8.04–8.73
8.66–9.08
6.56–6.95
8.43–8.78
7.20–7.39
7.42–8.01
8.21–8.65
8.45–8.79
6.70–7.09
8.05–8.39
7.11–7.45
7.06–7.47
7.16–7.47
7.80–8.22
6.81–7.51
6.57–7.18
3.98
h
3
3.98
h
3-OH
3.34
h
3-OCH
3-OEt
3
3.75
h
4.28
f
4-CH
4-Et
3
3.95
i
f
4.39
f
4-n-Pr
4-i-Pr
4.86
f
4.91
j
4-n-Bu
4-t-Bu
4-n-Pentyl
5.39
5.31
j
5.92
d
4-Cl
4.31
4-CF
2,3-(CH
3
4.66
h
3
)
2
4.43
h
2-CH
2-CH
2-CH
3
3
3
-3-OH
3.84
k
g
-3-OCH
-3-OEt
3
3.93
h
4.78
2,3,4-F
2,4,5-F
3
3.64
3.71
3
f
3,5-Cl
3,5-Cl
2
4-CH
4-Cl
3
4.81
f
2
5.17
3,5-Br
2
2-CH
3
4.93
(
continued on next page)

C. E. Wheelock et al. / Bioorg. Med. Chem. 14 (2006) 1143–1159
1147
Table 1 (continued)
b
Compound
A-ring
B-ring
Activity (pEC50, M)
Observed
LogP
a
Calculated
pEC50
95% CI
pEC50
f
f
124
125
126
127
128
129
130
131
132
133
3,5-Br
3,5-Br
3,5-Br
2
4-CH
4-Et
3
7.68
8.04
6.72
6.55
6.39
7.38
6.64
7.08
8.31
7.06
7.42–7.94
7.87–8.21
6.54–6.91
6.26–6.85
6.17–6.62
7.14–7.62
6.46–6.82
6.75–7.42
8.12–8.49
6.88–7.23
7.99
8.25
6.62
6.42
6.42
7.19
6.95
6.41
7.79
6.89
5.07
5.51
4.38
3.48
4.13
3.86
4.54
3.00
3.51
3.37
2
2
4-NO
2
2-CH
4-Cl
4-Cl
4-Et
3
2-CH
4-Cl
4-CH
4-Et
4-Cl
4-Et
4-Cl
3
3
2,6-F
H
2
j
l
g
H
a
b
c
d
e
f
Calculated by Eq. 2b.
Unless noted, values were calculated for this work using CLOGP (BioByte Corp., Claremont, CA, USA).
RH-5849.
42,43
Experimentally measured by Oikawa et al.
43
Oikawa et al.
42
Oikawa et al.
g
h
i
54
Nakagawa et al.
Nakagawa et al.
21
Tebufenozide (RH-5992).
18
Nakagawa et al.
j
k
Methoxyfenozide.
Halofenozide.
l
2
pEC50 ¼ 0:980 log P ꢁ 0:116ðlog PÞ
þ ½CoMFA termsꢂ þ 4:755;
n ¼ 126; s ¼ 0:457; r ¼ 0:815;
ðq ¼ 0:506; SCV ¼ 0:747; m ¼ 6Þ;
Electrostatic 32:5%; Steric 47.0%; logP=10.9%;
2
F
6;119 ¼ 87:445
2
Figure 1. General template of dibenzoylhydrazines examined in this
study. The A-ring and B-ring nomenclature are used throughout the
work to describe the position of the relative CoMFA fields. The
numbers on the aromatic rings refer to the positions of the different
2
ðlog PÞ ¼ 9:5% log Popt ¼ 4:22.
ð1aÞ
1 2
substituents (X and Y) listed in Tables 1–3. R and R refer to
The resulting Eq. 1a provided an acceptable description
of the biological observations; however, we found that
removal of the hydrophobic parameter logP slightly im-
proved the conventional correlation.
substitutions on the nitrogen atoms as listed in Tables 1–3. The atoms
used for the CoMFA alignment are shown in the bridging backbone
between the A- and B-ring, and consist of C–C–N–N–C–C.
pEC50 ¼ ½CoMFA termsꢂ þ 6:437;
2
than 107. Replacement of the tert-butyl group with a
cyclohexyl (165) as well as a CH -cyclohexyl (166)
was unfavored. Substitution of both N atoms (167–
n ¼ 126; s ¼ 0:436; r ¼ 0:831; F 6;119 ¼ 97:870
2
2
ðq ¼ 0:479; SCV ¼ 0:768; m ¼ 6Þ;
Electrostatic 44:0%; Steric 56:0%
ð1bÞ
1
70) resulted in complete loss of activity, most likely
due to blockage of hydrogen bond formation with
We then attempted to formulate a CoMFA equation for
all the active compounds (n = 158) examined in this
study (1–172), shown with their structure and biological
potency in Tables 1–3. The initial equation was again
generated using logP terms and the resulting Eq. 2a is
shown below.
3
4,36
amino acid residues in the active site.
Replacement
of the ketone alpha to the A-ring with a methylene
moiety resulted in a reduction in the compoundÕs
potency as shown by 171 and 172.
2
.2. Three-dimensional QSAR with CoMFA
2
pEC50 ¼ 1:312 log P ꢁ 0:155ðlog PÞ
We performed CoMFA analyses for all the dib-
enzoylhydrazines in Table 1 (1–133), but excluded com-
pounds that did not exhibit any biological activity (10,
þ ½CoMFA termsꢂ þ 1:188;
2
n ¼ 158; s ¼ 0:591; r ¼ 0:701;
F
4;153 ¼ 89:473
2
ðq ¼ 0:490; SCV ¼ 0:772; m ¼ 4Þ;
Electrostatic 27:6%; Steric 38:7%;
1
2, 13, 31, 32, 82, and 100). Earlier studies had shown
that the hydrophobic parameter (logP) was important
for describing the inhibitory activity of these compounds
and it was therefore used in the development of the ini-
2
log P ¼ 17:8%; ðlog PÞ ¼ 15:9%; log Popt ¼ 4:23.
3
7
tial equations
ð2aÞ

1148
C. E. Wheelock et al. / Bioorg. Med. Chem. 14 (2006) 1143–1159
Table 2. Activity of tert-butyl-containing diacylhydrazines
t-Bu
N
R
C
O
N
H
C
O
R
A
B
b
Compound
R
A
R
B
Activity (pEC50, M)
Observed
95% CI
LogP
a
Calculated
pEC50
pEC50
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
Cyclohexyl
Ph
Ph
4.33
5.14
5.06
4.84
<4.00
5.19
4.26
5.55
6.08
4.93
7.88
6.00
5.03
6.13
5.73
7.10
7.04
7.28
7.36
7.01
5.81
4.06–4.60
4.85–5.43
4.64–5.48
4.38–5.29
—
4.45
4.54
4.71
4.77
4.41
4.54
5.16
6.00
6.17
5.43
7.51
6.80
5.46
6.30
6.45
6.93
6.88
7.22
7.22
6.73
6.45
2.61
3.22
3.69
3.56
4.09
4.75
2.44
3.72
3.59
4.25
4.68
3.15
3.63
3.50
3.16
3.69
3.56
4.22
4.09
4.75
5.80
n-Hexyl (C6)
n-Pentyl (C5)
i-Pentyl (C5)
i-Hexyl (C6)
n-Heptyl (C7)
i-Pr (C3)
Ph(3,5-(CH
Ph(3,5-(CH
Ph(3,5-(CH
Ph(3,5-(CH
Ph(4-Et)
3
3
3
3
)
2
)
2
)
2
)
2
)
)
)
)
4.86–5.51
3.92–4.60
5.13–5.96
5.79–6.38
4.18–5.67
7.68–8.07
5.83–6.16
4.75–5.30
5.81–6.45
5.47–5.99
6.94–7.26
6.90–7.17
7.04–7.52
7.14–7.57
6.80–7.22
5.43–6.18
n-Pentyl (C5)
i-Pentyl (C5)
n-Hexyl (C6)
a-Naphthyl
3-Cyclohexene
Cyclohexyl
Ph(4-Et)
Ph(4-Et)
Ph(4-Et)
Ph(4-Et)
Ph(4-Et)
Ph(4-Et)
Ph(2-Cl)
n-Hexyl (C6)
n-Butyl (C4)
Ph(3,5-(CH
Ph(3,5-(CH
Ph(3,5-(CH
Ph(3,5-(CH
Ph(3,5-(CH
Ph(3,5-(CH
Ph(3,5-(CH
3
3
3
3
3
3
3
)
)
)
)
)
)
)
2
)
2
)
2
)
2
)
2
)
2
)
2
)
n-Pentyl (C5)
i-Pentyl (C5)
n-Hexyl (C6)
i-Hexyl (C6)
n-Heptyl (C7)
n-Nonyl (C9)
a
Values were calculated using Eq. 2b.
Calculated by CLOGP (MacLogP ver. 4.0).
b
However, we again found that exclusion of the hydro-
phobic parameter slightly increased the quality of the
correlation as shown in Eq. 2b.
Additional negative fields were observed around the 2-,
3-, and 6-positions of the B-ring, with positive fields
around the 5-position. A large positive field surrounds
the 2-, 3-, and 4-positions of the B-ring.
pEC50 ¼ ½CoMFA termsꢂ þ 4:041;
2
A few small differences were observed in the CoMFA
fields following the inclusion of logP terms (Eqs. 2a vs
2b; data not shown). The electrostatic potential dis-
played key differences surrounding the 4-position of
the A-ring, with a much larger positive field observed
with the inclusion of logP values. Significant reduc-
tions in both negative and positive electrostatic fields
were observed surrounding the 4-position of the B-
ring. The greatest effect was the lack of large negative
and positive fields at the end of the ethyl substituent
on the 4-position of the B-ring in the absence of logP.
The steric fields, on the other hand, evidenced very
few differences with a slight reduction in favorable ste-
ric activity surrounding the ethyl substituent in the 4-
position of the B-ring in the absence of logP. The
hydrophobic effects were further examined with a
comparative molecular similarity index analysis (CoM-
n ¼ 158; s ¼ 0:554; r ¼ 0:737; F 4;153 ¼ 107:199
2
ðq ¼ 0:447; SCV ¼ 0:781; m ¼ 4Þ;
Electrostatic 42:1%; Steric 57:9%.
ð2bÞ
The superposition of the compounds used in Eq. 2b is
displayed in Figure 2 and the atoms chosen for the align-
ment are shown in Figure 1. In general, the steric and
electronic contours were nearly identical for the inclu-
sion or exclusion of logP parameters and the resulting
CoMFA fields were relatively uncomplicated. Therefore,
the displayed results are only for the exclusion of logP
(
Eq. 2b). Figures 3 and 4 show the respective contour
maps for the steric and electrostatic fields with tebufe-
nozide as the template molecule. Important observa-
tions include a large region of unfavorable steric
activity surrounding the edge of the 4-position of the
A-ring and a sterically favorable region surrounding
the ethyl substituent in the 4-position of the B-ring.
The 3- and 5-positions of the B-ring have sterically unfa-
vorable fields surrounding them. The electrostatic fields
show a positive field around the 2- and 3-positions of the
A-ring, with negative fields over the oxygen atoms of
the bridge and a positive field over the NH moiety.
3
8
SIA); however, the resulting equation was inferior to
the CoMFA analysis (data not shown). Eq. 2b was
subsequently chosen as the final equation to calculate
the pEC50 values shown in Tables 1–3. This equation
was able to describe approximately 74% of the biolog-
ical activity for 158 active compounds as shown in
Figure 5.

C. E. Wheelock et al. / Bioorg. Med. Chem. 14 (2006) 1143–1159
Table 3. Activity of various N-substituted dibenzoylhydrazines and benzyl analogs
1149
R1 R2
C
O
N
N
C
O
b
Compound
R
1
R
2
Activity (pEC50, M)
Observed
LogP
a
Calculated
pEC50
95% CI
pEC50
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
CH
Et
3
H
<4.00
<4.00
5.24
—
—
6.11
6.11
5.86
5.60
5.66
6.15
5.96
5.89
4.57
4.58
6.20
5.43
5.84
6.35
6.18
1.25
1.77
2.08
2.83
2.70
2.61
3.23
3.01
3.15
3.30
3.28
3.90
2.74
3.97
5.56
H
i-Pr
n-Bu
H
5.07–5.42
4.78–5.53
5.37–5.71
5.54–5.85
5.88–6.48
6.65–6.97
4.35–4.83
4.72–6.05
5.90–6.19
5.58–5.95
—
H
5.15
5.54
i-Bu
H
sec-Bu
2-Methylbutyl
tert-Amyl
Ph
H
5.70
6.18
H
H
6.81
4.59
H
Ph(4-F)
Cyclohexyl
H
4.89
6.05
H
CH
CH
2
-Cyclohexyl
H
5.77
3
CH
CH
3
<4.00
<4.00
<4.00
tert-Bu
tert-Bu
3
—
—
n-Bu
H3C
H3C
t-Bu CH3
<4.00
—
8.52
6.00
170
C
O
N
N
C
O
Et
H C
3
t-Bu
5.90
5.69–6.11
4.73–5.34
5.95
5.87
6.09
4.06
1
71
72
CH2
N
N
H
C
O
Et
H3C
1
t-Bu
CH2
5
.04
N
N
H
C
O
a
Values were calculated using Eq. 2b.
Calculated by CLOGP (MacLogP ver. 4.0).
b
Figure 2. Stereoview of the superposition of all 172 compounds analyzed in this study.

1150
C. E. Wheelock et al. / Bioorg. Med. Chem. 14 (2006) 1143–1159
Figure 3. Stereoviews of the CoMFA fields generated with Eq. 2b (absence of logP) using tebufenozide (compound 107) as the template. Contours
are shown to surround regions where increased steric bulk increase (green) or decreases (yellow) the biological activity.
Figure 4. Stereoviews of the CoMFA fields generated with Eq. 3 (absence of logP) using tebufenozide (compound 107) as the template. Contours are
shown to surround regions where a positive (blue) or negative (red) electrostatic potential increases the biological activity.
important for ligand activity. A model of the B. mori
EcR receptor binding pocket was generated based upon
the Heliothis virescens structure previously published
3
4
(
Fig. 6). The results of the CoMFA steric map were
then superimposed over the surface of the EcR binding
pocket to show the steric interactions (Fig. 7). The re-
sults show that the ligand (tebufenozide, 107) fits tightly
into the receptor, but that there is potential room for
expansion (groups with increased steric bulk) in the
upper left-hand portion of the receptor. The limits on
steric bulk (yellow areas) correspond to the boundaries
of the receptor, except for the tert-butyl group, where
no field is visualized. This observation is probably due
to the fact that the structural variety of the N-alkyl
groups is limited and that no biological activity was ob-
served for compounds containing small N-substituted
Figure 5. Linear correlation of predicted versus experimental pEC50
values for the 158 compounds used in Eq. 2b.
groups such as CH (155) or ethyl (156).
3
2
.3. Ecdysone receptor modeling
The modeling studies also showed that the orientation
of tebufenozide in the binding pocket enables it to
undergo both hydrophobic and hydrophilic interactions
Modeling studies were performed in order to further
understand the physicochemical parameters that are

C. E. Wheelock et al. / Bioorg. Med. Chem. 14 (2006) 1143–1159
1151
Figure 6. Sequence alignment of ligand-binding domains between the Bombyx mori, Heliothis virescens, and Bemisia tabaci ecdysone receptors (EcR).
51
The numbering of the first amino acid (P331) of BmEcR is the same as that reported by Kamimura et al. and corresponds to P394 reported by
3
5
34
Swevers et al. The numbering of HvEcR and BtEcR is the same as that reported by Billas et al. and Carmichael et al., respectively. Regions in
39
˚
yellow are highlighted to show areas of sequence conservation. Residues in red are located within 3 A from the ligand.
Figure 7. Surface of the EcR binding pocket overlaid with the CoMFA steric fields. Tebufenozide (compound 107) is used as the template and
contours are shown to surround regions where increased steric bulk increases (green) or decreases (yellow) the activity.
that most likely support its binding geometry. These
interactions were expected based upon previously pub-
3. Discussion
3.1. Molting hormonal activity
3
4,39
lished work with the EcR receptor.
However, it
was interesting to compare our model with published
structures. Results agree with the current theory of bind-
ing of diacylhydrazine, which involves a series of hydro-
phobic and hydrophilic interactions. The tert-butyl
group is surrounded by a number of hydrophobic amino
acids including M409, M503, L507, and L514 as shown
in Figure 8. These groups surround the pocket contain-
ing the tert-butyl and are most likely involved in the
need for a large steric moiety in this portion of the mol-
ecule. The amino acids T339, Y404, and N500 are in a
position to form hydrogen bonds with the bridge of
the diacylhydrazine moiety (Fig. 9). The hydrogen-
Non-steroidal ecdysone agonists such as the dib-
enzoylhydrazines represent an important class of insect
control agents. Their target selectivity makes them a
useful tool in controlling crop pests. It is therefore
important that the mechanism of agonist binding be well
understood to increase our ability to construct new po-
tent and selective agonists. This project was designed to
examine the compound substructures that are essential
for biological activity and to evaluate the utility of a
new high-throughput screening system for agonist activ-
ity. These results are described by Eq. 2b, which shows
that we were able to obtain a reasonable model for the
array of compounds examined in this study without
excluding any active compounds.
bonding groups of all of these amino acids (the NH
2
˚
and OH moieties) are within 2.0–2.5 A of the corre-
sponding groups of the diacylhydrazine moiety.

1152
C. E. Wheelock et al. / Bioorg. Med. Chem. 14 (2006) 1143–1159
Figure 8. Hydrophobic surface gradient. The solvent accessible surface of the optimized structure was expressed using the MOLCAD module of
SYBYL using the Fast Conolley algorithm. There are no physical units for the value (unlike experimental logP); however, the color gradient of the
hydrophobicity of the receptor surface ranges from blue (highest hydrophilic area) to brown (highest lipophilic area). The color scheme is best
interpreted by associating blue with water and brown with oil or fat. M = methionine and L = leucine.
Figure 9. Hydrophilic amino acids that are in a position to form hydrogen bonds with the agonist (tebufenozide, compound 107). The hydrogen-
˚
or OH moiety) are in a position to form hydrogen bonds (2.0–2.5 A) with the diacylhydrazine moiety. The
bonding groups of the amino acids (NH
2
green and yellow maps indicate the solvent accessible surface areas of the amino acids of the receptor and the ligand, respectively. T = threonine,
Y = tyrosine, and N = asparagine.
A necessary exercise was to determine the extent to
which data collected with this new B. mori assay could
be compared to existing data from other assay systems.
We therefore examined the relationship between pEC50
values measured with the B. mori system and the binding
activity of various dibenzoylhydrazines using Spodoptera
pEC50 ðB. moriÞ ¼ 0:931ðꢃ0:122ÞpIC ðChiloÞ
50
þ 0:325ðꢃ0:821Þ;
ð4Þ
2
n ¼ 61; s ¼ 0:497; r ¼ 0:893.
4
0
frugiperda-derived cells (Sf-9) and molting hormonal
activity in an in vitro system using integument parts of
Forty-seven compounds were used to derive Eq. 3: 1,
3
1
6
, 11, 16, 47, 57, 58, 69–81, 83–85, 87, 88, 93, 101–
05, 107, 110, 115–118, 132, 133, and 147–154, and
1 compounds were used in Eq. 4: 1, 3, 7, 8, 11, 15,
1
6,17
C. suppressalis
as shown in Eqs. 3 and 4:
pEC50 ðB. moriÞ ¼ 0:886ðꢃ0:193ÞpIC ðSf9Þ
16, 19–22, 24, 26, 28–30, 33, 52–54, 56, 58, 60, 63,
50
6
1
5–67, 69, 72–78, 83, 106–110, 112, 117, 132–134,
41–146, 148–154, and 171, 172. The observed activity
þ 0:809ðꢃ1:383Þ;
ð3Þ
2
n ¼ 47; s ¼ 0:468; r ¼ 0:809;
using the B. mori reporter gene assay correlated fairly

C. E. Wheelock et al. / Bioorg. Med. Chem. 14 (2006) 1143–1159
1153
well with the activity measured using other assay sys-
tems. As these assays all employ Lepidoptera insects, a
high degree of correlation was expected. However, the
correlation data are useful in that they enable us to
directly compare results measured with the different
assay systems.
increases with the lipophilicity of a molecule. Tables
1–3 list the predicted pEC50 values calculated using
Eq. 2b (exclusion of logP). We also performed the cal-
culation with Eq. 2a (CoMFA views not shown) and
found that it was generally inferior to Eq. 2b at pre-
dicting the observed biological activity. For example,
the three commercial compounds examined in this
study (tebufenozide 107, methoxyfenozide 117, and
halofenozide 133) resulted in the following: for 107
the observed pEC50 was 8.91 and predicted with Eq.
2a was 8.14 versus 8.40 with Eq. 2b, for 117, observed
pEC50 was 8.22 and predicted with Eq. 2a was 7.94
versus 8.03 with Eq. 2b, and for 133 the observed
pEC50 was 7.06 and predicted with Eq. 2a was 6.75
versus 6.98 with Eq. 2b. These results were typical in
that Eq. 2a consistently underpredicted the pEC50
value.
Although the three assay systems are specific to lepidop-
teran insects (S. frugiperda, C. suppressalis, and B. mori),
clear differences exist with respect to the assessment of
molting hormone activity. In the Sf-9 cell-based assay,
binding activity is measured against protein extracts de-
rived from the cellular nuclei. Therefore, no inhibition is
expected that is related to the necessity of the compound
to transverse cellular membranes. Moreover, simple
binding does not necessarily implicate that the bound
compound is capable of activating transcription. The
B. mori-based cell assay and C. suppressalis-based integ-
ument assay therefore more accurately (although not
perfectly) reflect in vivo conditions. Differences between
the B. mori- and C. suppressalis-based systems probably
relate to penetration differences between the two systems
Earlier work with these compounds in a different assay
system had shown that logP was important for describ-
ing activity. We performed similar 3-D QSAR CoMFA
studies using data obtained from the examination of the
ability of dibenzoylhydrazine derivatives to displace
(
e.g., the existence of a basal membrane barrier in the
3
37
integument assay as opposed to cells in tissue culture
that do not have a basal membrane). The duration of
the required response to detect an effect can also differ
between the two assays. In the B. mori-based system, a
direct response to ecdysone is detected, while the re-
sponse in the integument system is achieved only after
the completion of the ecdysone regulatory cascade that
involves different layers of primary and secondary re-
sponse genes as well as feedback mechanisms that occur
over a period of several days. Differences between the
assay construction and physical parameters could there-
fore explain the species-specific variations observed for
compound activity. Alternatively, the differences may
be due to variation in the structure of the individual
ecdysone receptors. To distinguish between the possibil-
ities, differently constructed assays should be performed
for the same species or assays of the same set-up should
be carried out for different species. For instance, the
molting hormonal activity in B. mori can be compared
[ H]PonA binding to nuclear extracts of Sf-9 cells.
This equation consisted of 50 compounds including 46
compounds, 1, 3, 11, 16, 47, 57, 58, 69–81, 83–85, 87,
88, 93, 101–105, 107, 110, 115–118, 132, 133, and 148–
154, from the current study. Since four compounds did
not overlap between the present and previous study,
those compounds were omitted and the remaining com-
pounds were reanalyzed to derive Eq. 5a as the best
3
7
descriptor of biological activity
2
pIC50 ¼ 1:772logP ꢁ 0:228ðlogPÞ þ ½CoMFA termsꢂ
þ 2:913;
ð5aÞ
2
n ¼ 46; s ¼ 0:269; r ¼ 0:868;
F
5;40 ¼ 52:636;
2
ðq ¼ 0:486; SCV ¼ 0:530; m ¼ 5Þ;
Electronic 16:7%; Steric 29:8%; logP ¼ 27:8%;
2
ðlogPÞ ¼ 25:8%; logPopt ¼ 3:89:
However, a reanalysis of the same data shows that
3
by (1) competition with [ H]PonA for binding to nuclear
removal of the logP terms provides a significant equa-
tion, with increased r and F values, but using an addi-
extracts; (2) reporter gene activation in cell lines; and (3)
induction of molting in larval integument in vitro. A
more attractive possibility would be the generation of
high-throughput systems based on the cell lines of other
lepidopteran species, in a similar fashion as carried out
for B. mori.
2
tional component (m = 6 for Eq. 5b vs m = 5 for Eq. 5a)
and giving poor q and S values in the cross-valida-
tion analysis
2
CV
pIC50 ¼ ½CoMFA termsꢂ þ 5:713;
ð5bÞ
6;39 ¼ 63:660
2
n ¼ 46; s ¼ 0:228; r ¼ 0:907;
F
3
.2. Three-dimensional QSAR with CoMFA
2
ðq ¼ 0:334; SCV ¼ 0:611; m ¼ 6Þ;
Electronic 37:3% Steric 62:7%.
The CoMFA analyses provided a number of insights
into the mechanism of agonist binding. Of particular
interest was the observation that logP effects were
not important for describing biological activity. LogP
is the octanol–water partition coefficient and is defined
as the ratio of the concentration of a chemical in oct-
anol and in water at equilibrium (at a specified temper-
ature). This partition coefficient has been adopted as
the standard measure of lipophilicity, which governs
the mass flux of a molecule in a physical or biological
These observations are interesting in that the CoMFA
fields clearly show that there are distinct areas of favor-
able and unfavorable steric fields. However, these effects
do not appear to be correlated explicitly with hydropho-
bicity. Furthermore, given the similarities between Eqs.
5a and 5b (and Eqs. 2a and 2b), it would seem that in
general hydrophobicity parameters are not necessary
for describing the activity of these dibenzoylhydrazine
analogs. Another possible explanation lies in the inher-
system. Solubility in
a lipophilic organic phase

1154
C. E. Wheelock et al. / Bioorg. Med. Chem. 14 (2006) 1143–1159
ent nature of the CoMFA analysis. It is difficult to sep-
arate logP effects from the electrostatic potential and
steric fields generated by the CoMFA model. The nature
of the analysis for especially the steric fields contains a
hydrophobic component, essentially inherently account-
ing for logP effects. It is therefore not surprising that di-
rect inclusion of the logP term does not have a
significant impact upon the generated CoMFA equa-
tions. However, other CoMFA studies have reported
that the inclusion of logP terms can dramatically affect
without the use of logP parameters. Further work could
examine this premise by performing a more thorough
classical QSAR analysis that included multiple substitu-
tions on the A- and B-rings.
None of the new dibenzoylhydrazines tested exhibited
biological potency greater than that of tebufenozide
(107). This observation is displayed in the CoMFA fields
in that any substitution on the A-ring was sterically
unfavorable. The electronic fields did not show any sig-
nificant areas of increased activity around the A-ring in
the absence of logP. However, with the inclusion of
logP a field of positive electrostatic potential was ob-
served around the A-ring (data not shown). Although
this type of observation has been shown before to be
associated with positive interactions of hydrogen atoms,
it nevertheless does suggest one potential novel area of
chemistry to be explored in future studies. For instance,
the incorporation of particular electron deficient moie-
ties could lead to improvements in activity. However,
given the requirement for the compound to cross-cellu-
lar membranes, the addition of charged moieties is not
advised.
4
1
the outcome of analyses.
Interestingly, we previously reported that molecular
hydrophobicity was an important parameter in classical
1
7,42,43
QSAR for the activity against Lepidoptera.
In
light of the above discussion, these results are not neces-
sarily at odds as it first appears. LogP is undoubtedly
important in the biological activity of these compounds,
as evidenced by numerous structure–activity relation-
ships. However, in the creation of CoMFA models,
hydrophobic effects are not separated from the rest of
the model generation. It is possible that if the hydropho-
bic contribution is small compared with other steric and
electrostatic effects, then the hydrophobic terms will be
reflected in the steric and electrostatic terms. However,
attempts to further break down the fields into their indi-
vidual components with the CoMSIA analysis were not
useful. Limited improvement in CoMFA model perfor-
mance upon the addition of hydrophobicity parameters
The electrostatic fields surrounding the bridging back-
bone (area between the A- and B-ring) displayed fields
3
7
similar to those previously reported. The large positive
electrostatic fields display the requirement for at least
one hydrogen-substituted nitrogen atom on the hydra-
zine bridge for biological activity as evidenced by the
lack of activity for compounds 167–170. The negative
field over the oxygen atom of the ketone displays the
importance of this atom in activity. Substitution of this
ketone with a methylene group resulted in a 3 orders of
magnitude reduction in potency (171 vs 107). Also the 3-
position of the B-ring displayed a region where negative
electrostatic potential is favorable for activity. This
activity was shown by compounds containing electron-
rich groups in this position such as 70, 88, and 103–
105, 116–118, which all contain an oxygen-substituted
group in the meta position. It therefore may be useful
to prepare a molecule that incorporates an electron-rich
group in the 3-position in addition to the key ethyl moi-
ety in the 4-position.
4
4
(
such as logP) has been previously reported. To fur-
ther explore these observations, we conducted a classical
QSAR analysis of a similar set of compounds. To sim-
plify the analysis, we focused on only mono-substituted
compounds. There were 29 compounds containing
mono-substitution on the A-ring, which gave the follow-
ing equation:
para
pEC50 ¼ ꢁ0:483ðꢃ0:127ÞB5 þ 0:939ðꢃ0:229Þr
þ 6:018ðꢃ0:157Þ;
ð6aÞ
2
n ¼ 29; s ¼ 0:321; r ¼ 0:861; F 2;26 ¼ 81:166.
para
In Eq. 6a, B5 is the steric parameter defined by Ver-
4
5
loop and r is HammettÕs electronic parameter. The
values in parentheses are 95% confidence intervals of
the regression coefficients. Similar studies performed
with the 34 mono-substituted compounds on the B-ring
provided a poorer correlation.
Other areas for potential improvements in agonist activ-
ity include the substituent on the hydrazine moiety. It is
vital to have an N-substituted sterically bulky group
(e.g., tert-butyl) as well as an unsubstituted (free hydro-
gen) group for compound activity. However, the opti-
mal size of the bulky group is still in question.
Comparison of compound 162 with 1 shows that the
tert-amyl substitution slightly increases agonist potency.
However, replacement with a cyclohexyl (165) reduces
potency. The differences in the Connolly-solvent exclud-
ed volumes between a tert-amyl group and cyclohexyl
ortho
5
pEC50 ¼ ꢁ1:160ðꢃ0:716Þr ꢁ 0851ðꢃ0:326ÞB
meta
ꢁ
0:424ðꢃ0:417ÞB5 þ 7:409ðꢃ0:340Þ
ð6bÞ
2
n ¼ 34; s ¼ 0:622; r ¼ 0:551;
F
3;30 ¼ 12:284.
Eq. 6b shows that electron-donating groups are favor-
able (as opposed to the A-ring) and that the steric effect
of para-substituents is insignificant, although ortho- and
meta-substituents are sterically unfavorable. These re-
sults are consistent with the CoMFA fields shown in
Figures 3 and 4. While the size and statistical strength
of Eqs. 6a and 6b are limited, the equations do suggest
that logP terms are not necessary for describing activity.
These results combined with the CoMFA studies in
this work suggest that the biological activity of these
dibenzoylhydrazines can be adequately described
3
˚
group are small (94.7 and 100.8 A , respectively) relative
3
to that of a tert-butyl group (77.0 A ). An examination
˚
of Fig. 7 shows that there appears to be more room in
the binding pocket surrounding the tert-butyl group.
In addition, Figure 8 shows that this portion of the
binding pocket is very lipophilic. It is very possible that
a tert-amyl group is the optimal molecular volume;

C. E. Wheelock et al. / Bioorg. Med. Chem. 14 (2006) 1143–1159
1155
however, the remaining structure of 162 has not been
optimized and lacks the 3,5-dimethyl substitution on
ring A as well as the para-ethyl on ring B. This com-
pound would be a logical choice for further synthesis ef-
forts. The other area of agonist structure that should be
examined includes the permissible size of the A- and B-
rings. Figure 3 shows that there is a favorable area of
steric bulk surrounding the para position (site of the
para-ethyl group) on the B-ring. This area could be ex-
plored by substitution with a bulky group such as an
adamantyl moiety. The CoMFA fields suggest that the
area surrounding the A-ring is at a steric maximum
and further steric bulk in this area will not be favorable.
These different areas can be targeted in an attempt to in-
crease the potency and/or specificity of ecdysone
agonists.
include them in CoMFA analyses performed with
dibenzoylhydrazines. Of particular importance is the
high degree of correlation between QSAR analyses
(both classical and 3-D) performed with this new B. mori
system and those conducted with other screening
systems (Chilo integument or Sf-9 cells). These results
suggest that it is appropriate to compare the results, at
least rank order potency, between different studies. This
new assay system should greatly increase the speed and
ease with which screening for novel ecdysone agonists
can be performed.
5. Experimental
5.1. Bioassays
3
.3. Ecdysone receptor modeling
GFP fluorescence reporter gene assays were performed
using the cell-based screening method of Swevers et al.
3
5
The modeling studies supported the results of the CoM-
FA analyses. Of particular interest was the observation
that the three key hydrophilic amino acid residues ap-
peared to be serving the same role in the B. mori EcR
as those observed in the lepidopteran H. virescens and
hemipteran Bemisia tabaci EcRs.
T339, Y404, and N500 triad appears to be in an equiv-
˚
alent position to form hydrogen bonds (2.0–2.5 A) with
the diacylhydrazine moiety (Fig. 9). The importance of
these hydrogen-bonding interactions can be observed
by examining compounds in which the key hydrogen-
bonding moiety has been removed. In B. mori, the OH
group of Y404 forms a hydrogen bond with the N–H
moiety of the hydrazine and replacement of the hydro-
gen atom (107) with a methyl group (170) results in a
complete loss of compound activity (pEC50 < 4.00).
The B. mori T339 forms a hydrogen bond to the carbon-
yl moiety alpha to the A-ring. Replacement of the car-
bonyl with a methylene group results in a 3 orders of
magnitude loss in activity (107, pEC50 = 8.91 to 171,
pEC50 = 5.90). Therefore, any further attempts at refin-
ing the structure–activity relationships of these com-
pounds must ensure that these hydrogen-bonding
moieties are not replaced.
and are briefly described here. Fifty thousand cells of
one of the Bm5/ERE.gfp clonal cell lines (permanently
transformed with an ecdysone-responsive GFP reporter
cassette L5Q or L6Q) were distributed in individual
wells of a 96-well plate and treated with different con-
centrations of dibenzoylhydrazine compounds dissolved
in 1 ll of dimethylsulfoxide (DMSO). Twenty-four
hours after challenge, GFP fluorescence was quantified
using a FLUOstar Galaxy Unit microplate fluorometer
(BMG Lab Technologies, Offenburg, Germany) using
a filter set with excitation filter 485 ± 10 nm and emis-
sion filter 530 ± 20 nm. In all measurements, back-
ground fluorescence was subtracted and values were
expressed as percent of maximum fluorescence. Initially,
3
4,39
In B. mori, the
ꢁ
5
ꢁ11
a wide range of concentrations (10 to 10 with 10-
fold differences between each concentration) was used
to determine the approximate potency of the com-
pounds. To determine more exact EC50 values, interme-
diary concentrations (2- to 3-fold differences) were also
tested. EC50 values were calculated using the GraphPad
Prism Program software (GraphPad Software, San Die-
go, CA) and were derived from at least three different
series of measurements at concentrations that span the
activation curve of each compound.
5.2. QSAR analyses
4. Conclusion
The Hansch–Fujita QSAR analysis was performed using
4
6
We have shown that a cell-based reporter assay system
using B. mori cells is useful for measuring dib-
enzoylhydrazine molting hormone activity in a high-
throughput screening assay. 3-D QSAR analyses with
CoMFA in conjunction with modeling of the B. mori
EcR receptor show several key areas for agonist interac-
tion with the binding pocket. These interactions can be
possibly further optimized as discussed above. Analyses
of the observed biological activity using CoMFA dem-
onstrated that logP is not explicitly important for gener-
ating a significant equation. However, earlier results
using classical QSAR had demonstrated that logP is
an important parameter for describing biological
activity. These divergent findings suggest that hydro-
phobic parameters are implicitly included within the
CoMFA analysis and it is not necessary to explicitly
QREG system ver.2.05. LogP (where P is the partition
coefficient in a 1-octanol–water system) values were tak-
1
6,42,43
en from the literature
CLOGP program MacLogP Ver.4.0 (BioByte Corp.,
or calculated using the
4
7
Claremont, CA, USA). In all equations, n is the num-
ber of compounds used for the regression analyses, s is
the standard deviation, r is the correlation coefficient,
and the values in parentheses represent the 95% confi-
dence intervals of the regression coefficient and the inter-
cept. All CoMFA computations were performed with
the molecular modeling software package Sybyl ver
6.91 (Tripos Co., St. Louis, MO, USA). The structures
were generated by modifying the structure of RH-5849
(1), whose X-ray structure has been previously report-
1
5
ed. The geometry of all structures was fully optimized
using PM3 and charges were calculated using AM1.

1156
C. E. Wheelock et al. / Bioorg. Med. Chem. 14 (2006) 1143–1159
Compounds were automatically aligned using the Sybyl
module Align Database and the common skeletal chain
C–C–N–N–C–C shown in Figure 1 was used as a tem-
plate. The molecules were superposed in a lattice space
to the correlation equations is also shown. Results were
visualized using contour maps that have connected lat-
tice points that contain an equivalent coefficient level
for each molecular field.
˚
of 25.04 · 28.65 · 21.61 A (X = ꢁ9.84 to 15.20,
Y = ꢁ14.83 to 13.82, and Z = ꢁ9.93 to 11.68). The
5.3. Modeling
superposition of all 172 compounds is shown in Figure
. The CoMFA analyses were performed using the Sybyl
2
The homology modeling software PDFAMS (Protein
Discovery Full Automatic Modeling System; In-Silico
Sciences, Inc.; Tokyo, Japan), whose proto-type soft-
˚
QSAR module with a lattice spacing of 2 A. The electro-
static and steric fields were estimated using a +1 charge
and a sp carbon, respectively. The electrostatic and ste-
3
49
ware FAMS was developed by Ogata and Umeyama,
ric potential energies at each lattice point were calculat-
ed using Coulombic and Lennard–Jones potential
functions, respectively. Hydrophobic effects were evalu-
was used to construct the ligand-binding region of B.
mori EcR. In addition to the basic module PDFAMS
pro, the optional module, PDFAMS-ligand is useful
for proteins that are combined with ligands. The mod-
ule was used to predict the protein structures in this
study. These programs construct a tertiary structure
of the target protein using the information from
homologous proteins. The method consists of a data-
base search, with the resulting structure minimized
by the simulated annealing procedure. Processes such
as searches for homologous proteins, alignment, con-
struction of Ca-atoms, main-chain construction, and
side-chain construction with the optimization of
main-chain were performed automatically in an itera-
tive fashion.
2
ated using both logP and (logP) as lattice independent
external descriptors. Correlation analyses were analyzed
4
8
using partial least squares (PLS). The number of latent
variables in the set was initially selected using the leave-
one-out method with the column filtering set at 2 kcal/
mol. Conventional analyses were then performed using
the optimum number of components, m. The CoMFA
results are displayed using several statistical values: the
2
leave-one-out cross-validated correlation coefficient, q ,
the cross-validated standard error, S , the conventional
cv
correlation coefficient, r, and the standard deviation, s.
The relative percentage contribution of each descriptor
5
0,51
In the first step, primary sequences of BmEcR
ob-
tained from the NCBI web site were aligned with homol-
ogous sequences in the FAMS database of PDFAMS.
In this alignment, two homologous proteins were listed.
The most homologous protein was 1R1K, a crystal
structure of the ligand-binding domain of the heterodi-
mer EcR–USP of the the lepidopteran H. virescens
3
4
bound to steroidal ligand, PonA. However, in this
study, the alignment file was manually constructed using
1
R20, HvEcR–USP with a DBH-type ligand BYI06830,
as a reference protein. Although 1R20 was absent in the
FAMS alignment database because of its sequence sim-
ilarity to 1R1K (98%), the crystal structures showed that
the ligand-binding pockets of 1R20 and 1R1K were only
partially overlapped. Therefore, 1R20 was considered to
be the best protein for the modeling of the BmEcR–
DBH complex. In the second step, the three-dimensional
structure of the ligand-binding domain of BmEcR was
constructed from 1R20 and optimized using the simulat-
0
Scheme 1. Synthesis of N-tert-butyl-N,N -dibenzoylhydrazines using
tert-butylhydrazine and substituted benzoyl chloride as the starting
materials.
Scheme 2. Alternative synthesis methods for dibenzoylhydrazine derivatives using benzoylhydrazine and the corresponding alkyl halide. The
experimental procedures for pathways A–D are described in the text.

C. E. Wheelock et al. / Bioorg. Med. Chem. 14 (2006) 1143–1159
1157
Table 4 (continued)
Table 4. Melting points and elemental analyses of newly synthesized
a
compounds
Compound
Mp (ꢁC)
Elemental analyses
Compound
Mp (ꢁC)
Elemental analyses
Calcd: C, 74.97; H, 8.01; N, 7.95
Found: C, 75.32; H, 8.26; N, 7.46
Calcd: C, 75.38; H, 8.25; N, 7.64
1
70
107–109
111
114
119
120
123
126
127
128
129
130
131
134
135
136
137
138
139
140
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
228–230
Found: C, 75.75; H, 8.47; N, 7.40
Calcd: C, 75.75; H, 8.48; N, 7.36
Found: C, 64.15; H, 6.02; N, 7.12
Calcd: C, 64.28; H, 5.91; N, 7.14
Found: C, 63.38; H, 5.55; N, 7.54
Calcd: C, 63.48; H, 5.59; N, 7.40
Found: C, 63.40; H, 5.61; N, 7.37
Calcd: C, 63.48; H, 5.59; N, 7.40
Found: C, 48.70; H, 4.31; N, 6.02
Calcd: C, 48.74; H, 4.31; N, 5.98
Found: C, 43.49; H, 3.40; N, 8.30
Calcd: C, 43.31; H, 3.43; N, 8.42
Found: C, 73.80; H, 7.39; N, 8.63
Calcd: C, 74.04; H, 7.46; N, 8.64
Found: C, 58.97; H, 5.02; N, 7.54
Calcd: C, 59.19; H, 4.97; N, 7.67
Found: C, 66.10; H, 6.27; N, 8.13
Calcd: C, 66.18; H, 6.14; N, 8.12
Found: C, 74.70; H, 8.12; N, 8.18
Calcd: C, 74.97; H, 8.01; N, 7.95
Found: C, 58.68; H, 4.51; N, 7.63
Calcd: C, 58.94; H, 4.67; N, 7.64
Found: C, 71.25; H, 8.55; N, 9.31
Calcd: C, 71.49; H, 8.67; N, 9.26
Found: C, 71.12; H, 9.27; N, 9.19
Calcd: C, 71.02; H, 9.27; N, 9.20
Found: C, 71.66; H, 9.57; N, 8.81
Calcd: C, 71.66; H, 9.50; N, 8.80
Found: C, 71.56; H, 9.24; N, 8.83
Calcd: C, 71.66; H, 9.50; N, 8.80
Found: C, 71.98; H, 9.73; N, 8.45
Calcd: C, 72.25; H, 9.70; N, 8.43
Found: C, 72.65; H, 9.92; N, 8.06
Calcd: C, 72.79; H, 9.89; N, 8.08
Found: C, 70.25; H, 9.21; N, 9.86
Calcd: C, 70.31; H, 9.02; N, 9.66
Found: C, 70.84; H, 5.47; N, 11.06
Calcd: C, 70.85; H, 5.55; N, 11.02
Found: C, 71.58; H, 5.96; N, 10.38
Calcd: C, 71.62; H, 6.01; N, 10.44
Found: C, 72.20; H, 6.46; N, 9.96
Calcd: C, 72.32; H, 6.43; N, 9.92
Found: C, 72.88; H, 6.87; N, 9.44
Calcd: C, 72.95; H, 6.80; N, 9.45
Found: C, 72.98; H, 6.78; N, 9.52
Calcd: C, 72.95; H, 6.80; N, 9.45
Found: C, 72.91; H, 6.80; N, 9.53
Calcd: C, 72.95; H, 6.80; N, 9.45
Found: C, 73.67; H, 7.14; N, 9.01
Calcd: C, 73.52; H, 7.14; N, 9.03
Found: C, 73.62; H, 7.14; N, 9.12
Calcd: C, 73.52; H, 7.14; N, 9.03
Found: C, 75.77; H, 5.23; N, 8.84
Calcd: C, 75.93; H, 5.10; N, 8.86
Found: C, 71.87; H, 4.59; N, 8.40
Calcd: C, 71.85; H, 4.52; N, 8.38
Found: C, 74.52; H, 6.93; N, 8.70
Calcd: C, 74.51; H, 6.88; N, 8.69
Found: C, 74.98; H, 7.26; N, 8.33
Calcd: C, 74.97; H, 7.19; N, 8.33
Found: C, 71.59; H, 5.93; N, 10.44
Calcd: C, 71.62; H, 6.01; N, 10.44
Found: C, 73.27; H, 7.15; N, 9.00
Calcd: C, 73.52; H, 7.14; N, 9.03
Found: C, 74.95; H, 7.92; N, 7.98
214
a
Analyses were performed at the Center for Organic Elemental
Microanalysis of Kyoto University. Mp = melting point.
242–243
239
ed annealing method. To optimize the protein structure,
the original DBH-type ligand (BYI06830) in 1R20 was
used as a ligand. The coordinates of BYI06830 were
fixed during optimization.
239–242
214–215
201–203
249–250
246–247
180–182
249
The cavity surface at the ligand-binding site of the mod-
eled BmEcR was expressed by the MOLCAD module of
SYBYL using the Fast Conolly algorithm. The gradient
of the hydrophobic property of the cavity surface was
colored from blue (less hydrophobic) to brown (more
hydrophobic) in Figure 8 based on the approach by Hei-
5
2
den et al.
5.4. Synthesis of compounds
213–214
87–88
0
N-tert-Butyl-N,N -dibenzoylhydrazines were synthe-
sized with a range of substituents on the benzene rings
using tert-butylhydrazine and benzoyl chloride accord-
ing to previously published methods (Schemes 1 and 2;
Table 4). Substituted benzoyl chloride was prepared
from the corresponding benzoic acid and arylhydrazines
were used to introduce aryl groups on the N-atom.
4-Fluorophenylhydrazine was derived from 4-fluoroani-
line according to the conventional method. As hydra-
zine compounds are explosive, sufficient caution must
be used in their handling, and particular attention must
be paid to avoid condensation. Other dibenzoylhydr-
azine derivatives containing various alkyl groups on
the N-atoms were synthesized according to the proce-
dures shown in Scheme 2. Introduction of primary and
secondary alkyl groups was executed by reacting ben-
zoylhydrazine (benzoic hydrazide) and the correspond-
ing alkyl halide. A tert-amyl group was introduced by
reaction of the corresponding tert-amyl carbocation
with 2,5-diphenyl-1,3,4-oxazole derived from two
151–153
161–162
159–160
112–113
209–210
145–146
127–128
164–165
103–104
171–173
167–170
167–168
161–162
178–179
173
5
3
1
benzoylhydrazine molecules. The H NMR spectra were
recorded on a JEOL JNM-PMX60 (Tokyo, Japan)
spectrometer with the use of CDCl and/or deuteriodim-
3
ethylsulfoxide (DMSO-d ) with tetramethylsilane (TMS)
6
as the internal standard. Experimental procedures for
steps A–D in Scheme 2 are described below.
0
5.5. N-Benzoyl-N -n-butylhydrazine (step A)
N-Benzoylhydrazine (1.36 g, 10.0 mmol) and n-butyral-
dehyde (0.72 g, 10.0 mmol) in methanol (30 ml) with a
catalytic amount of para-toluenesulfonic acid were stir-
red overnight to obtain the corresponding hydrazone.
To the reaction mixture was added NaBH CN (0.75 g,
3
1
in dioxane to keep the mixture acidic and stirred for
3 h at room temperature. After stripping of the organic
solvents under reduced pressure, water was added to the
199–201
200–201
71–72
2 mmol) with the simultaneous addition of 4 M HCl
119–121
53–56
(
continued on next page)

1158
C. E. Wheelock et al. / Bioorg. Med. Chem. 14 (2006) 1143–1159
residue followed by extraction with ether. The ether
layer was dried over anhydrous Na SO and concentrat-
ed in vacuo to obtain a yellow oil containing N-benzoyl-
Science and Technology, Greek Ministry of Develop-
ment. This study was supported, in a part, by the 21st
century COE program for Innovative Food and Envi-
ronmental Studies Pioneered by Entomomimetic Scienc-
es, from the Ministry of Education, Culture, Sports,
Science and Technology of Japan. The authors thank
2
4
0
N -n-butylhydrazine (0.80 g), which was purified by
silica gel column (Walogel C-300) chromatography (hex-
1
ane/ethyl acetate = 2/1). H NMR (CDCl ) d 0.90 (3H, t,
3
˚
J = 7 Hz), 1.48 (6H, m), 2.00 (1H, s), 6.00–7.00 (1H, br),
7
Asa Wheelock for assistance with the CoMFA figures.
.20–7.90 (5H, m).
5
.6. 2,5-Diphenyl-1,3,4-oxadiazole (step B)
References and notes
Benzoic
3
hydrazide
6.8 mmol) was dissolved in polyphospholic acid
(benzoylhydrazine,
5.00 g,
1. Nijhout, H. F. Insect Hormones; Princeton University
Press: New Jersey, 1994.
(
100 g) and stirred for 40 min at 170 ꢁC. After cooling,
2. Yao, T.-P.; Segraves, W. A.; Oro, A. E.; McKeown, M.;
Evans, R. M. Cell 1992, 71, 63.
the reaction mixture was poured into water. The solid
material was filtered and washed with water, and then
crystallized from ethanol to obtain 2,5-diphenyl-1,3,4-
3
. Thomas, H. E.; Stunnenberg, H. G.; Stewart, A. F. Nature
993, 362, 471.
4. Yao, T.-P.; Forman, B. M.; Jiang, Z.; Cherbas, L.; Chen,
1
1
oxadiazole (3.20 g, 78.0%). H NMR (CDCl ) d 7.00–
3
J.-D.; McKeown, M.; Cherbas, P.; Evans, R. M. Nature
993, 366, 476.
7
5
2
.60 (10H, mj).
1
5
. Hsu, A. C. -T. In Synthesis and Chemistry of Agrochem-
icals II; Baker, D. R., Fenyes, J. G., Moberg, W. K., Eds.;
American Chemical Society: Washington, DC, 1991; Vol.
443, p.478.
0
0
.7. N,N -Dibenzoyl-N -tert-amylhydrazine (step C)
,5-Diphenyl-1,3,4-oxadiazole (1.11 g, 5.0 mmol) was
dissolved in 2-methyl-2-butene (1.40 g, 10.0 mmol) and
concentrated H SO (0.98 g, 10.0 mmol) and stirred for
6. Dhadialla, T. S.; Carlson, G. R.; Le, D. P. Annu. Rev.
Entomol. 1998, 43, 545.
2
4
7
. Sawada, Y.; Yanai, T.; Nakagawa, H.; Tsukamoto, Y.;
Tamagawa, Y.; Yokoi, S.; Yanagi, M.; Toya, T.;
Sugizaki, H.; Kato, Y.; Shirakura, H.; Watanabe, T.;
Yajima, Y.; Kodama, S.; Masui, A. Pest Manag. Sci.
three days. After the addition of water, the mixture
was extracted with ether. The ether layer was washed
with saturated NaHCO , followed by brine and then
3
dried over anhydrous Na SO . The solvent was stripped
2
4
2
003, 59, 49.
under reduced pressure and the residue was purified by
silica gel chromatography (hexane/ethyl acetate = 2/1)
8
. Sawada, Y.; Yanai, T.; Nakagawa, H.; Tsukamoto, Y.;
Yokoi, S.; Yanagi, M.; Toya, T.; Sugizaki, H.; Kato, Y.;
Shirakura, H.; Watanabe, T.; Yajima, Y.; Kodama, S.;
Masui, A. Pest Manag. Sci. 2003, 59, 25.
0
0
to afford N,N -dibenzoyl-N -tert-amylhydrazine as a col-
1
orless oil. H NMR (CDCl ) d 0.92 (3H, t, J = 7 Hz),
3
1
.30–2.00 (6H, m), 7.00–7.60 (10H, m), 9.80 (1H, s).
9. Sawada, Y.; Yanai, T.; Nakagawa, H.; Tsukamoto, Y.;
Yokoi, S.; Yanagi, M.; Toya, T.; Sugizaki, H.; Kato, Y.;
Shirakura, H.; Watanabe, T.; Yajima, Y.; Kodama, S.;
Masui, A. Pest Manag. Sci. 2003, 59, 36.
0. Henrich, V. C. In Comprehensive Molecular Insect Science;
Gilbert, L. I., Iatrou, K., Gill, S. S., Eds.; Elsevier: Oxford,
0 0
.8. N,N -Dibenzoyl-N-n-butyl-N -tert-butylhydrazine
step D)
5
(
1
1
0
0
N,N -Dibenzoyl-N -tert-butylhydrazine was prepared
according to the conventional method by the gradual
addition of NaH (0.20 g) to 20 ml of N,N-dimethylform-
amide (DMF) followed by stirring for 30 min. Next, n-
butyl iodide (0.74 g, 4.0 mmol) was added dropwise to
the mixture and stirred for 2 h at room temperature.
The reaction mixture was diluted with 50 ml of water,
extracted with ether and the organic layer was washed
with 1 M HCl and brine, and then dried over anhydrous
Na SO . After stripping the solvent, the residue was
2
005; p 243.
1. Laudet, V.; Bonneton, F. In Comprehensive Molecular
Insect Science; Gilbert, L. I., Iatrou, K., Gill, S. S., Eds.;
Elsevier: Oxford, 2005; p 287.
12. Wing, K. D. Science 1988, 241, 467.
13. Wing, K. D.; Slawecki, R. A.; Carlson, G. R. Science
1
988, 241, 470.
1
1
4. Nakagawa, Y.; Akagi, T.; Iwamura, H.; Fujita, T. Pestic.
Biochem. Physiol. 1989, 33, 144.
5. Nakagawa, Y.; Shimizu, B.; Oikawa, N.; Akamatsu, M.;
Nishimura, K.; Kurihara, N.; Ueno, T.; Fujita, T. In
Classical and Three-Dimensional QSAR in Agrochemistry;
Hansch, C., Fujita, T., Eds.; American Chemical Society:
Washington, DC, 1995; Vol. 606; p 288.
2
4
purified by silica gel column chromatography to give
0
0
N,N -dibenzoyl-N-n-butyl-N -tert-butylhydrazine as a
1
yellowish oil. H NMR (CDCl ) d 0.90 (3H, t,
3
J = 7 Hz), 1.00–1.50 (4H, m), 1.66 (9H, s), 3.05–3.40
(
16. Nakagawa, Y.; Hattori, K.; Shimizu, B.; Akamatsu, M.;
Miyagawa, H.; Ueno, T. Pestic. Sci. 1998, 53, 267.
2H, J = 8 Hz, t), 7.20–7.70 (10H, br s).
1
1
1
2
2
7. Nakagawa, Y.; Hattori, K.; Minakuchi, C.; Kugimiya, S.;
Ueno, T. Steroids 2000, 65, 117.
8. Nakagawa, Y.; Minakuchi, C.; Ueno, T. Steroids 2000, 65,
Acknowledgments
5
37.
9. Nakagawa, Y.; Minakuchi, C.; Takahashi, K.; Ueno, T.
Insect Biochem. Mol. Biol. 2002, 32, 175.
0. Nakagawa, Y.; Smagghe, G.; Tirry, L.; Fujita, T. Pest
Manag. Sci. 2002, 58, 131.
1. Nakagawa, Y.; Takahashi, K.; Kishikawa, H.; Ogura, T.;
Minakuchi, C.; Miyagawa, H. Bioorg. Med. Chem. 2005,
13, 1333.
Craig E. Wheelock was funded by a Japanese Society for
the Promotion of Science (JSPS) postdoctoral fellow-
ship. Guy Smagghe acknowledges the Fund for Scientif-
ic Research (FWO-Vlaanderen, Brussels, Belgium).
Dimitra Stefanou, Kostas Iatrou and Luc Swevers
acknowledge the support of the General Secretariat of

C. E. Wheelock et al. / Bioorg. Med. Chem. 14 (2006) 1143–1159
1159
2
2
2. Smagghe, G.; Degheele, D. Pestic. Biochem. Physiol. 1994,
9, 224.
3. Hsu, A. C.-T; Fujimoto, T. T.; Dhadialla, T. S. In
Phytochemicals for Pest Control; Hedin, P. A., Holling-
worth, R. M., Masler, E. P., Miyamoto, J., Thompson, D.
G., Eds.; American Chemical Society: Washington, DC,
37. Nakagawa, Y.; Takahashi, K.; Kishikawa, H.; Ogura, T.;
Minakuchi, C.; Miyagawa, H. Bioorg. Med. Chem. 2005,
13, 1333.
38. Klebe, G.; Abraham, U.; Mietzner, T. J. Med. Chem.
1994, 37, 4130.
39. Carmichael, J. A.; Lawrence, M. C.; Graham, L. D.;
Pilling, P. A.; Epa, V. C.; Noyce, L.; Lovrecz, G.; Winkler,
D. A.; Pawlak-Skrzecz, A.; Eaton, R. E.; Hannan, G. N.;
Hill, R. J. J. Biol. Chem. 2005, 280, 22258.
40. Ogura, T.; Nakagawa, Y.; Minakuchi, C.; Miyagawa, H.
J. Pestic. Sci. 2005, 30, 1.
41. Wheelock, C. E.; Nakagawa, Y.; Akamatsu, M.; Ham-
mock, B. D. Bioorg. Med. Chem. 2003, 11, 5101.
42. Oikawa, N.; Nakagawa, Y.; Nishimura, K.; Ueno, T.;
Fujita, T. Pestic. Biochem. Physiol. 1994, 48, 135.
43. Oikawa, N.; Nakagawa, Y.; Nishimura, K.; Ueno, T.;
Fujita, T. Pestic. Sci. 1994, 41, 139.
44. Fleischer, R.; Wiese, M. J. Med. Chem. 2003, 46, 4988.
45. Hansch, C.; Leo, A. J.; Hoekman, D. Exploring QSAR:
Hydrophobic, Electronic, and Steric Constants; American
Chemical Society: Washington, DC, 1995.
4
1
997; p 206.
2
4. Carlson, G. R. In Green Chemical Syntheses and Processes;
Anastas, P. T., Heine, L. G., Williamson, T. C., Eds.;
American Chemical Society: Washington, DC, 2000; Vol.
7
67, p 8.
5. Trisyono, A.; Chippendale, M. J. Econ. Entomol. 1997, 90,
486.
6. Trisyono, A.; Chippendale, M. Pestic. Sci. 1998, 53,
77.
7. Carlson, G. R.; Dhadialla, T. S.; Hunter, R.; Jansson, R.
K.; Jany, C. S.; Lidert, Z.; Slawecki, R. A. Pest Manag.
Sci. 2001, 57, 115.
2
2
2
1
1
2
2
3
3
8. Cowles, R. S.; Alm, S. R.; Villani, M. G. J. Econ. Entomol.
999, 92, 427.
9. Farinos, G. P.; Smagghe, G.; Tirry, L.; Castanera, P.
Arch. Insect Biochem. Physiol. 1999, 41, 201.
0. Smagghe, G.; Degheele, D. Pestic. Biochem. Physiol. 1993,
46, 149.
1. Smagghe, G.; Nakagawa, Y.; Carton, B.; Mourad, A. K.;
Fujita, T.; Tirry, L. Arch. Insect Biochem. Physiol. 1999,
1
46. Asao, M.; Shimizu, R.; Nakao, K.; Fujita, T. Japan
Chemistry Program Exchange; Society of Computer
Chemistry: Japan, 1997.
47. Leo, A. J. Chem. Rev. 1993, 93, 1281.
48. Cramer, R. D., III Perspect. Drug Discov. Des. 1993, 1, 269.
49. Ogata, K.; Umeyama, H. J. Mol. Graph. Model. 2000, 18,
258.
50. Swevers, L.; Drevet, J. R.; Lunke, M. D.; Iatrou, K. Insect
Biochem. Mol. Biol. 1995, 25, 857.
4
2. Elzen, G. W. J. Econ. Entomol. 2001, 94, 55.
3. Le, D. P.; Thirugnanam, M.; Lidert, Z.; Carlson, G. R.;
Ryan, J. B. Proc. Brighton Crop Prot. Conf.—Pests Dis.
1, 42.
3
3
1
996, 481.
51. Kamimura, M.; Tomita, S.; Fujiwara, H. Comp. Biochem.
Physiol. B 1996, 113, 341.
52. Heiden, W.; Moeckel, G.; Brickmann, J. J. Comput. Aided
Mol. Des. 1993, 7, 503.
53. Coleman, G. H. In Organic Synthesis Collection; Gilman,
H., Eds.; Wiley: New York, 1967; Vol. 1, p 442.
54. Nakagawa, Y.; Smagghe, G.; Kugimiya, S.; Hattori, K.;
Ueno, T.; Tirry, L.; Fujita, T. Pestic. Sci. 1999, 55, 909.
3
3
4. Billas, I. M. L.; Iwema, T.; Garnier, J. M.; Mitschler, A.;
Rochel, N.; Moras, D. Nature 2003, 426, 91.
5. Swevers, L.; Kravariti, L.; Ciolfi, S.; Xenou-Kokoletsi,
M.; Ragoussis, N.; Smagghe, G.; Nakagawa, Y.;
Mazomenos, B.; Iatrou, K. FASEB J. 2004, 18, 134.
6. Toya, T.; Yamaguchi, K.; Endo, Y. Bioorg. Med. Chem.
3
2
002, 10, 953.