Structure-activity relationships of simplified resiniferatoxin analogues with potent VR1 agonism elucidates an active conformation of RTX for VR1 binding

Article abstract of DOI:10.1016/j.bmc.2003.12.005

We previously described a series of N-(3-acyloxy-2-benzylpropyl) homovanillate and N′-(4-hydroxy-3-methoxybenzyl) thiourea derivatives that were potent VR1 agonists with high-affinities and excellent analgesic profiles. The design of these simplified RTX analogues was based on our RTX-derived pharmacophore model which incorporates the 4-hydroxy-3- methoxyphenyl (A-region), C₂₀-ester (B-region), orthophenyl (C1-region) and C₃-keto (C2-region) groups of RTX. For the purpose of optimizing the spatial arrangement of the four principal pharmacophores on the lead agonists (1-4), we have modified the distances in the parent C-region, 3-acyloxy-2-benzylpropyl groups, by lengthening or shortening one carbon to vary the distances between the pharmacophores. We find that two of the amides, 4 and 19, possess EC₅₀ values <1 nM for induction of calcium influx in the VR1-CHO cells. As observed previously, the structure-activity relations for inhibition of RTX binding to VR1 and for induction of calcium uptake were distinct, presumably reflecting both intrinsic and methodological factors. In order to find the active conformation of VR1 ligands, the energy-minimized conformations of seven selected agonists were determined and the positions of their four pharmacophores were matched with those of five low energy RTX conformations. The rms values for the overlaps in the pharmacophores were calculated and correlated with the measured binding affinities (K _i) and calcium influx (EC₅₀) values. The binding affinities of the agonists correlated best with the RMS values derived from RTX conformation E (r²=0.92), predicting a model of the active conformation of RTX and related vanilloids for binding to VR1. Poorer correlation was obtained between any of the conformations and the EC ₅₀ values for calcium influx.

Full text of DOI:10.1016/j.bmc.2003.12.005

Bioorganic & Medicinal Chemistry 12 (2004) 1055–1069
Structure–activity relationships of simplified resiniferatoxin
analogues with potent VR1 agonism elucidates an active
conformation of RTX for VR1 binding
Jeewoo Lee,a,* Su Yeon Kim,a Soyoung Park,a Ju-Ok Lim,a Ji-Min Kim,a
Myungshim Kang,a Jiyoun Lee,a Sang-Uk Kang,a Hyun-Kyung Choi,a Mi-Kyung Jin,a
Jacqueline D. Welter,b Tamas Szabo,b Richard Tran,b Larry V. Pearce,b Attila Tothb
and Peter M. Blumbergb
aLaboratory of Medicinal Chemistry, Research Institute of Pharmaceutical Sciences, College of Pharmacy,
Seoul National University, Shinlim-Dong, Kwanak-Ku, Seoul 151-742, South Korea
bLaboratory of Cellular Carcinogenesis and Tumor Promotion, Center for Cancer Research,
National Cancer Institute, NIH, Bethesda, MD 20892-4255, USA
Received 27 September 2003; accepted 9 December 2003
Abstract—We previously described a series of N-(3-acyloxy-2-benzylpropyl) homovanillate and N0-(4-hydroxy-3-methoxybenzyl)
thiourea derivatives that were potent VR1 agonists with high-affinities and excellent analgesic profiles. The design of these simpli-
fied RTX analogues was based on our RTX-derived pharmacophore model which incorporates the 4-hydroxy-3-methoxyphenyl (A-
region), C20-ester (B-region), orthophenyl (C1-region) and C3-keto (C2-region) groups of RTX. For the purpose of optimizing the
spatial arrangement of the four principal pharmacophores on the lead agonists (1–4), we have modified the distances in the parent
C-region, 3-acyloxy-2-benzylpropyl groups, by lengthening or shortening one carbon to vary the distances between the pharmaco-
phores. We find that two of the amides, 4 and 19, possess EC50 values <1 nM for induction of calcium influx in the VR1-CHO
cells. As observed previously, the structure–activity relations for inhibition of RTX binding to VR1 and for induction of calcium
uptake were distinct, presumably reflecting both intrinsic and methodological factors. In order to find the active conformation of
VR1 ligands, the energy-minimized conformations of seven selected agonists were determined and the positions of their four
pharmacophores were matched with those of five low energy RTX conformations. The rms values for the overlaps in the pharma-
cophores were calculated and correlated with the measured binding affinities (Ki) and calcium influx (EC50) values. The binding
affinities of the agonists correlated best with the RMS values derived from RTX conformation E (r2=0.92), predicting a model of
the active conformation of RTX and related vanilloids for binding to VR1. Poorer correlation was obtained between any of the
conformations and the EC50 values for calcium influx.
# 2004 Elsevier Ltd. All rights reserved.
1. Introduction
subfamily of transient receptor potential (TRP) ion
channels.11 It is a cation channel with modest selectivity
for calcium. Consistent with its complicated regulation,
multiple sites of phosphorylation have been identi-
fied12,13 as well as multiple sites mediating acid respon-
siveness,14,15 and a site involved in inhibition by
phosphoinositides.16 Since VR1 is present pre-
dominantly in nociceptors on primary sensory neuron,
it has been an attractive therapeutic target for the
treatment of pain and for other indications in which C-
fiber sensory neurons are involved. Its validation as a
target for analgesia has been confirmed recently by the
observation that VR1 knockout mice exhibit deficits in
the thermal hyperalgesia that accompanies tissue injury
The vanilloid receptor 1 (VR1)1 is a polymodal noci-
ceptor activated by protons,2 heat,3 vanilloids such as
capsaicin (CAP)4 and resiniferatoxin (RTX),5 and lipid
mediators such as anandamide6 and the lipid metabolic
products of lipoxygenases.7 VR1 has been cloned from
rat,3 human,8 chicken9 and guinea pig,10 and has been
shown to be a member of the vanilloid receptor (TRPV)
Keywords: Vanilloid receptor 1; Resiniferatoxin.
* Corresponding author. Tel.:+82-2-880-7846; fax:+82-2-888-0649;
e-mail: jeewoo@snu.ac.kr
0968-0896/$ - see front matter # 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2003.12.005

1056
J. Lee et al. / Bioorg. Med. Chem. 12 (2004) 1055–1069
and inflammation17,18 as well as by emerging studies
characterizing VR1 antagonists.19
to a series of N-(3-acyloxyl-2-benzylpropyl)-N0-[(4-
methylsulfonylamino)benzyl]thiourea analogues, which
have shown high affinity antagonism/partial antagonism
and agonism on VR1.33ꢀ35
VR1 agonists initially activate the receptor to trigger
cation influx resulting in excitation of primary sensory
neurons. Subsequent desensitization leads to the block
of pain perception by the central nervous system. This
desensitization forms the basis for the primary ther-
apeutic use of VR1 agonists as potent analgesics.1,20
Reflecting the complicated pharmacology of VR1 and
the complicated nature of the clinical endpoint of pain,
agonists such as resiniferatoxin and olvanil have shown
that it is possible at least partially to separate the
induction of acute pain resulting from the initial chan-
nel activation and the subsequent desensitization.21
Potential therapeutic applications of VR1 agonists
include not only chronic pain associated with diabetic
neuropathy, post-herpetic neuralgia, arthritis and clus-
ter headache, but also urologic problems, pruritus and
bowel dysfunction.22
As part of our ongoing program to clarify the structural
interactions of VR1 ligands with the capsaicin binding
site of VR1 and eventually to generate simplified, potent
RTX surrogates as clinical drug candidates for analge-
sia, we have modified the 3-acyloxy-2-benzylpropyl
group in the C-region of our lead structures. The strat-
egy was that we varied the spatial arrangement of the
four pharmacophores in the model by lengthening or
shortening one carbon between the C-region pharma-
cophores, thereby producing seven different sets of dis-
tances between the four pharmacophores (Fig. 2). We
probed their activities on VR1 both for inhibition of
ligand binding (measured by competition of [3H]RTX
Several types of VR1 agonists have been reported and
are being developed: capsaicinoids such as capsaicin
(currently marketed) or DA-501823 and SDZ-24966524
(undergoing clinical trial); resiniferatoxin;25 and natural
dialdehydes.26 Resiniferatoxin (RTX), a natural diter-
pene isolated from the cactus-like succulent Euphorbia
resinifera, has been of particular interest both as a
therapeutic candidate per se and as a lead compound
for designing novel VR1 ligands. Not only does RTX
display extraordinarily high potency compared to cap-
saicin (its binding potency was Ki=0.13 nM in CHO/
VR1 versus Ki=1,700 nM for capsaicin),27 but it mark-
edly dissociates desensitization, for which it is ultra-
potent, from pungency, for which it is only slightly more
potent than capsaicin. From the perspective of medic-
inal chemistry, the identification of RTX as an ultra-
potent vanilloid strongly argued that the vanilloid
pharmacophore possessed novel elements not previously
appreciated from the studies with capsaicin and its
derivatives.
Over the past few years, we have demonstrated that N-
(3-acyloxy-2-benzylpropyl)-N0 -(4 -hydroxy-3-methoxy-
benzyl)thiourea (1, 2)28 and N-(3-acyloxy-2-benzyl-
propyl) homovanillic amide (3, 4)29,30 possess potent
VR1 agonism with high affinity in rat DRG and excel-
lent analgesic activity. These agonists were designed
based on an RTX-derived pharmacophore model which
incorporates the 4-hydroxy-3-methoxyphenyl (A-
region), C20-ester (B-region), orthophenyl (C1-region)
and C3-keto (C2-region) groups of RTX, which are
suggested from previously published SAR studies on
RTX31 to be the principal pharmacophores for interac-
tion with the capsaicin binding site of VR1 (Fig. 1) (the
structural regions of the vanilloids are denoted based on
the terminology of Novartis32). The particular struc-
tural feature of these compounds is the replacement of
the diterpene moiety of resiniferatoxin with a 3-acyloxy-
2-benzylpropyl moiety,28,29 which represents a motif
conferring significantly enhanced potency for VR1 ago-
nists compared to capsaicin. Recently, combination of
this template together with an antagonistic A-region led
Figure 1. Resiniferatoxin and its simplified analogues.
Figure 2. SAR of C region in simplified RTX analogues.

J. Lee et al. / Bioorg. Med. Chem. 12 (2004) 1055–1069
1057
Scheme 1. Synthesis of 17–20.
binding) and for stimulation of calcium uptake. We
further determined their energy minimized conform-
ations and correlated the goodness of fit of these con-
formations relative to the preferred conformations of
RTX with the biological potency data. Our studies
advance our understanding of vanilloid structure activ-
ity relationships and identify a preferred conformation
of the pharmacophoric elements in the binding to VR1.
ethyl] thiourea and amide analogues (40–44) are shown
in Scheme 2. Alkylation of N-(diphenylmethylene)gly-
cine ethyl ester with the corresponding halides under
potassium hydroxide in DMSO provided 22–24. The
diphenylmethylene group of 22–24 was hydrolyzed
under acidic condition and then reprotected with t-Boc
group to afford 28–30. LiAlH4-reduction followed by
pivaloylation produced 34–36. The ensuing deprotection
of t-Boc furnished the corresponding amines, which were
utilized using the condensation method described in
Scheme 1 to provide thioureas (40–42) and amides (43–44).
2. Chemistry
The syntheses of N-[3-pivaloyloxy-2-(2-phenylethyl)-
propyl] thiourea and amide analogues (17–20) are out-
lined in Scheme 1. Alkylation of diethyl malonate with
2-(3,4-dimethylphenyl)ethyl bromide (5) or 2-(4-t-butyl-
phenyl)ethyl bromide (6), prepared from 3,4-dimethyl-
benzyl chloride and 4-tert-butylstyrene by the
conventional route, produced mono-alkylated malonate
(7, 8). LiAlH4-reduction followed by selective mono-
pivaloylation afforded monoesters (11, 12) and, subse-
quently, intact free hydroxyl was converted into the
corresponding azides (13, 14) using PPh3, DEAD, and
diphenylphosphoryl azide.36 Azides of 13, 14 were
reduced to the amines, which, without further purific-
ation, were directly condensed with 4-methoxymethoxy-
3-methoxybenzyl isothiocyanate29 or pentafluorophenyl
homovanillate30 to produce thiourea (15, 16) and final
amide (19, 20) analogues. The ensuing MOM deprotec-
tion of 15, 16 afforded the final thioureas (17, 18). The
syntheses of N-[2-acyloxy-1-(benzyl or phenylethyl)-
3. Results and discussion
3.1. Biological results
The agonistic activities of the synthesized VR1 agonists
were assessed in vitro by a 45Ca2+ influx assay, which
was carried out using VR1 heterologously expressed in
Chinese hamster ovary (CHO) cells as previously
described.27,33 The potencies of the compounds were
expressed as their EC50 values. Their receptor binding
affinities were determined by competition of binding of
[3H]RTX to VR1 in CHO cells and were expressed as
their Ki values. All values represent the mean of at least
three experiments (Table 1). Since only the agonistic
activities of the parent agonists (1–4) in rat DRG were
reported previously,28,30 their binding affinities and
potencies for induction of calcium uptake were reeval-
uated in our CHO/VR1 system for direct comparison.

1058
J. Lee et al. / Bioorg. Med. Chem. 12 (2004) 1055–1069
Scheme 2. Synthesis of 40–44.
In the CHO/VR1 system, the parent thioureas, 1 and 2,
showed high binding affinities with Ki values of 17.4 and
6.35 nM, respectively, and potent agonism for calcium
uptake with EC50 values of 1.97 and 2.83 nM, respec-
tively, as expected. Their potencies thus represented,
respectively, 100- and 300-fold enhancements in binding
affinity and 20- and 15-fold enhancements in agonism
compared to capsaicin. The parent amides, 3 and 4, also
showed high binding affinities, Ki=157 and 15 nM, and
potent agonism, EC50=12.7 and 0.29 nM. These values
represented 10- and 150-fold increases in binding affi-
nity and 3.5- and 150-fold increases in agonism com-
pared to capsaicin. In particular, the high agonism of 4
(EC50=0.29 nM) in the CHO/VR1 cells approaches
that of RTX (EC50=0.27 nM in this system).
EC50=6.3 and 18.7 nM (3-fold in 17, 6.5-fold in 18) as
compared to the parent thioureas, 1 and 2, respectively.
One-carbon shortening (m=0, n=1 in thiourea) pro-
duced 2-pivaloyloxy-1-benzylethyl analogues, 40 and
41, in which their binding affinities and potencies as
agonists were also reduced to a similar extent as that of
lengthening: Ki=45 and 34.6 nM (2.5-fold in 40, 5.5-
fold in 41), EC50=19.6 and 5.7 nM (10-fold in 40, 2-
fold in 41). Interestingly, 42 (m=0 and n=2 in
thiourea), modified both in m and n, showed a compli-
cated result. Whereas its binding affinity was reduced by
53-fold compared to the parent 1, its potency as an
agonist was unaltered. This finding is yet another
example that receptor binding affinity is not fully corre-
lated with agonistic activity. Such divergence between
binding affinity and agonism has been described by us
previously.27,37 Since we have shown that RTX binding,
like calcium uptake, is mediated through VR1, we
assume that the divergence reflects both the influence of
factors differentially mediating the intrinsic SAR for the
two assays, as well as some contribution from metho-
dological differences in the two assays.17,27
One-carbon lengthening of the thiourea lead com-
pounds, 1 and 2, providing 3-pivaloyloxy-2-phenethyl-
propyl analogues, 17 and 18 (m=1, n=2 in thiourea),
led to modestly reduced potencies both in binding affi-
nities with values of Ki=25.0 and 18.3 nM (1.5-fold in
17, 3-fold in 18) and in agonism with values of

J. Lee et al. / Bioorg. Med. Chem. 12 (2004) 1055–1069
1059
In the similar manner, one-carbon elongation of parent
amides 3 and 4 produced compounds 19 and 20 (m=1,
n=2 in amide). As found with 42, compound 19 dis-
played high agonistic potency for calcium uptake with a
value of EC50=0.78 nM, which is 16- and 60-fold more
potent, respectively, than parent compound 3 and cap-
saicin; its binding affinity, with a value of Ki=153 nM,
was virtually the same as parent compound 3 (Ki=157
nM). Once again, this result highlights the distinct
structure activity relations of VR1 ligands for binding to
VR1 and for stimulation of calcium uptake. Compound
20 displayed a 2-fold reduction in binding affinity with a
value of Ki=36 nM and a 30-fold reduction in potency
as an agonist with a value of EC50=8.4 nM as com-
pared to the parent 4. Analogues shortened by one-car-
bon, 43 and 44 (m=0, n=1 in amide), showed
significantly reduced binding affinities (45-fold in 43,
60-fold in 44) and potencies as agonists (40-fold in 43,
550-fold in 44) compared to parent compounds. We
conclude that the in vitro receptor activities of amide
analogues were more sensitive than those of thiourea
analogues to the C1-region modifications. We further
note that appropriately substituted amides were similar
to or more potent than thiourea analogues, in contrast
to the conclusion for some other series of derivatives
with different C regions.38
ted. A conformational rationale for agonism or
antagonism of capsaicin analogues was proposed by
Walpole et al.39 based on NMR, X-ray and molecular
modeling studies. Quantitative structure activity rela-
tionships of capsaicin analogues were studied using
MULTICASE methology by Klopman et al.40 3-D-
QSAR analysis of simplified RTX analogues was inves-
tigated using the CoMFA and the COMSIA methods
by Kim.41 Nevertheless, analysis of the active con-
formation of RTX on binding to the receptor has not
been investigated despite the very potent binding affinity
of RTX itself.27 Through the conformational analysis of
RTX using NMR and computational methods,42 Van-
der Velde et al. proposed four energetically stable con-
formations of RTX in solution; unfortunately, this
approach cannot address the nature of the actual con-
formation(s) of RTX when bound to VR1.
One strategy to identify an active conformation of VR1
agonists is through correlation between the rms values
calculated from respective pharmacophoric matchings
of selected agonists on each of multiple stable RTX
conformations and their in vitro activities. Seven ago-
nists (1, 17, 40, 42 in thiourea, 3, 19, 43 in amide) were
selected by reason of containing in common a 3,4-
dimethylphenyl group as a C1-region. Their energy-
minimized structures were obtained through conform-
ational analysis by random search using the program
Sybyl 6.5 (Tripos). Five stable conformations of RTX
(RTX-AꢁE) were utilized for the comparision. The
conformations of RTX-A,B,C,Dobtained by NMR and
molecular modeling studies were imported from the
study of Vander Velde and co-workers.42 The fifth con-
formation (RTX-E) was derived by us through mol-
ecular dynamics simulation using the program Sybyl 6.5
(Tripos) and is illustrated in Figure 6.
3.2. Conformational analysis
The search for the principal pharmacophores of VR1
ligands and their active conformation is of particular
importance for the discovery of potent and selective
clinical candidates because of the difficulty in obtaining
X-ray structures of membrane proteins such as the
vanilloid receptor. To date, several structural and con-
formational analyses of VR1 ligands have been repor-
Table 1. Potencies of vanilloid agonists for binding to rat VR1 and for inducing calcium influx in CHO/VR1 cells
m
n
R
RTX binding
(Ki=nM)
Ca2+ Influx
(EC50=nM)
44.8(ꢂ3.8)
Capsaicin
Thiourea
1810 (ꢂ270)
1
2
1
1
1
1
0
0
0
1
1
2
2
1
1
2
3,4-Me2
4-t-Bu
3,4-Me2
4-t-Bu
3,4-Me2
4-t-Bu
3,4-Me2
17.4 (ꢂ4.1)
6.35 (ꢂ0.48)
25.0 (ꢂ4.4)
18.3(ꢂ5.6)
45 (ꢂ18)
1.97 (ꢂ0.56)
2.83 (ꢂ0.55)
6.3 (ꢂ2.1)
17
18
40
41
42
Amide
3
18.7 (ꢂ4.2)
19.6 (ꢂ4.9)
5.7 (ꢂ2.1)
34.6 (ꢂ9.6)
917 (ꢂ430)
2.36 (ꢂ0.39)
1
1
1
1
0
0
1
1
2
2
1
1
3,4-Me2
4-t-Bu
3,4-Me2
4-t-Bu
3,4-Me2
4-t-Bu
157 (ꢂ56)
15.0 (ꢂ3.8)
153 (ꢂ7)
12.7 (ꢂ3.5)
0.29 (ꢂ0.1)
0.78 (ꢂ0.25)
8.4 (ꢂ2.9)
482 (ꢂ74)
161 (ꢂ26)
4
19
20
43
44
36 (ꢂ12)
7000 (ꢂ4000)
910 (ꢂ210)

1060
J. Lee et al. / Bioorg. Med. Chem. 12 (2004) 1055–1069
The center points of the four pharmacophoric groups in
the lowest energy conformations of the seven VR1 ago-
nists, including the center of the vanilloid ring (A-
region), the sulfur of the thiourea or the oxygen of the
amide (B-region), the center of the phenyl ring (C1-
region) and the carbonyl of the ester (C2-region), were
matched with the center of the vanilloid ring, the C20-
ester carbonyl, the center of the orthoester phenyl, and
the C3-keto carbonyl, respectively, in the five con-
formations of RTX, as indicated in Figure 3. The
respective rms values after fitting were calculated for the
five conformations (Table 2). For initial comparision,
the ranking of the ligands based on quality of fit to each
RTX conformation was determined (see numbers in
parentheses, Table 2, lowest number indicates the best
fit). The ranking of the ligands in terms of their in vitro
activities was also indicated. The rank order of binding
affinities of the agonists correlated very well with RTX-
E; for agonistic potencies, no clear correlation with one
of the five RTX conformations was found.
Detailed analysis was performed by calculating the
correlation between the rms values in each row of RTX
conformations and in vitro activities, binding affinity or
agonistic potency, using linear regression. The best-fit
lines and regression coefficients for binding affinity and
for potency for inducing calcium influx were shown in
Figures 4 and 5, respectively. The analyses indicated
that the rms values of the agonists generally correlated
better with the corresponding binding affinities (Ki
values) than with the values for calcium influx (EC50
values). This is hardly surprising, since the RTX binding
affinity is a more robust measure of receptor interaction
than are the calcium uptake values (but see further dis-
cussion below).
Among the RTX conformations, the rms values of
RTX-E correlated best with the binding affinities of the
agonists, yielding a regression coefficient of r2=0.92
(Fig. 4e). Consequently, the RTX-E conformation
would appear from this analysis to be the most probable
Figure 3. Pharmacophoric matching between RTX and simplified RTX analogues.
Table 2. RMS values of VR agonists after overlay on five conformers of RTX
RTXꢀE
1
17
40
42
3
19
43
RTXꢀA
RTXꢀB
RTXꢀC
RTXꢀD2.368
RTXꢀE
Ki
3.000
3.222
(7)a
3.554
(6)
2.838
(4)
2.244
(3)
1.445
(1)
1.419
(1)
2.768
(6)
3.958
(7)
3.787
(7)
2.246
(4)
2.981
(4)
3.176
(6)
2.318
(5)
3.121
(5)
3.081
(5)
1.564
(1)
2.173
(2)
2.625
(2)
2.03
(2)
2.382
(3)
2.683
(3)
2.993
2.222
3.03
2.129
2.49
2.211
2.004
1.942
2.137
(7)
1.722
(1)
17.4
(1)
1.97
(2)
5.78
(3)
1.913
(2)
25
(2)
6.3
(4)
6.27
(6)
2.429
(3)
45
(3)
19.6
(6)
5.29
(5)
3.394
(7)
917
(6)
2.36
(3)
5.78
(2)
2.532
(4)
157
(5)
12.7
(5)
5.46
(1)
3.091
(5)
153
(4)
0.78
(1)
5.95
(4)
3.334
(6)
7,000
(7)
482
(7)
4.97
0.000
0.13c
0.27
Ca2+
Log Pb
6.86
a The number in the parenthesis indicates the rank of the activity in the corresponding row.
b KOWWIN Program (http://esc.syrres.com/).
c Ref 27.

J. Lee et al. / Bioorg. Med. Chem. 12 (2004) 1055–1069
1061
Figure 4. Correlation between rms and binding affinity.
active conformation of RTX on binding to the capsaicin
binding site of VR1. We therefore propose the con-
formation of RTX-E as a plausible model of the active
conformation of VR1 ligands based on the four phar-
macophore model as shown in Figure 6. In this model,
the pharmacophoric distances between the four phar-
macophores, the center of the vanilloid ring (A-region),
the C20-ester carbonyl (B-region), the center of the
orthoester phenyl (C1-region), and the C3-keto car-
bonyl (C2-region) are: d(A-B)=3.9 A, d(B-C1)=6.8
A, d(C1–C2)=8.2 A, d(C2-A)=5.9 A. Future experi-
ence will provide the ultimate measure of the utility of
this model for the design of novel VR1 ligands by the
pharmacophore-based approach.
activation of individual cells, all of which are important
for this cellular response but which complicate analysis
of receptor–ligand interactions. A further complication
in vanilloid pharmacology is that the extent of the dif-
ferences in structure activity relationships for binding
and calcium uptake strongly suggests that the two
assays detect different pools of VR1. Ultimately, of
course, the complete structure activity relations for VR1
as a function of its different co-regulators, its different
functional pools, and its different extents of agonism
and antagonism will all need to be understood. Clar-
ification of the basis for the structure activity relations
as determined for ligand binding to VR1 represents one
piece of this challenging and important enterprise.
An on-going complication in the field of vanilloid
pharmacology is that the interplay between ligand
binding and gating of the VR1 channel is highly com-
plex. We use ligand binding as a relatively robust
measure of ligand–receptor interaction. This assay has
the advantage of being conducted with long incubation
times, facilitating equilibrium conditions, is not dis-
torted by spare receptors, and is relatively insensitive to
perturbation by co-regulators. The calcium influx assay
for measuring agonism integrates multiple factors, such
as rate of uptake or stability, desensitization, modula-
tion by co-factors, spare receptors, and the pattern of
4. Conclusions
The structure activity relationships of N-(3-acyloxy-2-
benzylpropyl) homovanillates and N0-(4-hydroxy-3-
methoxybenzyl) thioureas, high-affinity VR1 agonists,
have been investigated through the modification of the
C-region and have provided a series of potent agonists
with high agonism. The pharmacophoric matching
between these agonists and stable RTX conformations
suggests a putative conformation of RTX, RTX-E con-
formation, involved in binding to VR1. Finally, the data

1062
J. Lee et al. / Bioorg. Med. Chem. 12 (2004) 1055–1069
Figure 5. Correlation between rms and calcium influx.
further illustrate the conclusion that the ligand binding
and calcium uptake assays for vanilloids reflect some-
what distinct structure activity relations.
5. Experimental
5.1. General method
All chemical reagents were commercially available.
Melting points were determined on a Melting Point
Buchi B-540 apparatus and are uncorrected. Silica gel
column chromatography was performed on silica gel 60,
230–400 mesh, Merck. Proton NMR spectra were
recorded on a JEOL JNM-LA 300 at 300 MHz. Che-
mical shifts are reported in ppm units with Me4Si as a
reference standard. Mass spectra were recorded on a
VG Trio-2 GC-MS. Combustion analyses were per-
formed on an EA 1110 Automatic Elemental Analyzer,
CE Instruments, and were within 0.4% of the calculated
values unless otherwise noted.
Figure 6. Proposed active conformation of RTX (RTX-E).
matography on silica gel with EtOAc/hexanes (1:10) as
eluant to afford 2-(3,4-dimethylphenyl)acetonitrile
1
5.1.1. 4-(2-Bromoethyl)-1,2-dimethylbenzene (5). A mix-
ture of 3,4-dimethylbenzyl chloride (10 g, 64.7 mmol)
and NaCN (15.85 g, 323.3 mmol) in DMF (15 mL)
was heated at 100 ꢃC for 16 h. The reaction mixture was
cooled, diluted with H2O and extracted with CH2Cl2
several times. The combined organic layers were washed
with H2O, dried over MgSO4 and concentrated in
vacuo. The residue was purified by flash column chro-
(9.395 g, 100%) as a colorless oil.; H NMR (CDCl3) d
6.95–7.25 (m, 5H), 3.60 (s, 2H, CH2CN), 2.25 (d, 3H,
J=12.9 Hz, CH3), 2.20 (d, 3H, J=12.9 Hz, CH3).
A mixture of the above nitrile (9.395 g, 64.7 mmol) and
concentrated H2SO4 (3.5 mL, 64.7 mmol) in MeOH (25
mL) was refluxed for 3 days. The reaction mixture was
cooled, alkalinized with solid NaHCO3 and con-

J. Lee et al. / Bioorg. Med. Chem. 12 (2004) 1055–1069
1063
centrated in vacuo. The residue was partitioned between
H2O and CH2Cl2 and the aqueous layer was extracted
with CH2Cl2 several times. The combined organic layers
were washed with H2O, dried over MgSO4 and con-
centrated in vacuo. The residue was purified by flash
column chromatography on silica gel with EtOAc/hex-
anes (1:10) as eluant to afford methyl 2-(3,4-dimethy-
phenyl)acetate (11.285 g, 98%) as a yellow oil; 1H NMR
(CDCl3) d 6.95–7.1 (m, 5H), 3.66 (s, 3H, CO2CH3), 3.54
(s, 2H, CH2CO2), 2.25 (d, 3H, J=13.4 Hz, CH3), 2.20
(d, 3H, J=9.8 Hz, CH3).
Compound 6 was prepared from the above alcohol by
following the procedure described for the synthesis of 5
as a colorless oil in 90% yield; 1H NMR (CDCl3) d 7.35
(d, 2H, J=8.3 Hz), 7.14 (d, 2H, J=8.3 Hz), 3.55 (t, 2H,
J=7.3 Hz, CH2Br), 3.10 (t, 2H, J=7.8 Hz, CH2Ar),
1.31 (s, 9H, C(CH3)3).
5.2. General procedure for the synthesis of 7–8
A cooled solution of diethylmalonate (6.4 g, 40 mmol)
in DMF (20 mL) at 0 ꢃC was treated with sodium
hydride (60%, 1.92 g, 48 mmol) portionwise and stirred
for 40 min at room temperature. The reaction mixture
was treated with 5–6 (48 mmol) and stirred for 16 h at
room temperature. The mixture was diluted with H2O
and extracted with EtOAc several times. The combined
organic layers were washed with H2O and brine, dried
over MgSO4, and concentrated in vacuo. The residue
was purified by flash column chromatography on silica
gel with EtOAc/hexanes (1:10) as eluant to afford 7–8.
To a cooled suspension of lithium aluminum hydride
(4.8 g, 126.6 mmol) in ether (50 mL) at ꢀ10 ꢃC was
added dropwise a solution of the above ester (11.285 g,
63.3 mmol) in ether (30 mL) and the reaction mixture
was stirred for 30 min at room temperature. The mix-
ture was cooled in an ice-bath, quenched by successive
addition of H2O (5 mL), 15% aqueous NaOH (10 mL),
and H2O (15 mL) and stirred for 1 h at room temper-
ature. The resulting suspension was filtered after addi-
tion of EtOAc, and the combined filtrate was
concentrated in vacuo. The residue was purified by flash
column chromatography on silica gel with EtOAc/hex-
anes (1:3) as eluant to afford 2-(3,4-dimethylphenyl)-1-
ethanol (5.705 g, 60%) as a colorless oil; 1H NMR
(CDCl3) d 6.95–7.1 (m, 5H), 3.85 (m, 2H, CH2OH), 2.81
(t, 2H, J=6.6 Hz, CH2Ar), 2.25 (m, 6H, 2ꢄCH3).
5.2.1. Diethyl 2-(3,4-dimethylphenethyl)malonate (7). 76%
Yield, colorless oil; 1H NMR (CDCl3) d 6.9–7.1 (m,
3H), 4.20 (q, 4H, J=7.1 Hz, 2ꢄCO2CH2CH3), 3.34 (t,
1H, J=7.6 Hz, CH), 2.59 (t, 2H, J=6.5 Hz, CH2Ar),
2.1–2.3 (m, 6H, 2ꢄCH3 and CH2CH2Ar), 1.27 (t, 6H,
J=7.1 Hz, 2ꢄCO2CH2CH3).
5.2.2. Diethyl-2-(4-tert-butylphenethyl)malonate (8). 98%
1
A cooled solution of the above alcohol (3.285 g, 21.9
mmol) and triphenylphosphine (6.883 g, 26.2 mmol) in
THF (20 mL) at 0 ꢃC was treated with carbon tetra-
bromide (8.703 g, 26.2 mmol) and stirred for 3 h at
room temperature. The reaction mixture was diluted
with ether, filtered through a Celite pad and the filtrate
was concentrated in vacuo. The residue was purified by
flash column chromatography on silica gel with hexanes
Yield, colorless oil; H NMR (CDCl3) d 7.31 (d, 2H,
J=8.3 Hz), 7.12 (d, 2H, J=8.3 Hz), 4.19 (q, 4H, J=7.1
Hz, 2ꢄCO2CH2CH3), 3.35 (t, 1H, J=7.6 Hz, CH), 2.63
(t, 2H, J=7.3 Hz, CH2Ar), 2.21 (m, 2H, CH2CH2Ar),
1.31 (s, 9H, C(CH3)3), 1.27 (t, 6H, J=7.1 Hz,
2ꢄCO2CH2CH3).
5.3. General procedure for the synthesis of 9–10
1
as eluant to afford 5 (4.24 g, 91%) as a colorless oil; H
NMR (CDCl3) d 6.95–7.1 (m, 5H), 3.54 (t, 2H, J=7.3
Hz, CH2Br), 3.10 (t, 2H, J=7.8 Hz, CH2Ar), 2.25 (m,
6H, 2ꢄCH3).
A cooled solution of lithium aluminium hydride (3.64 g,
96 mmol) in diethyl ether (80 mL) at 0 ꢃC was treated
dropwise with a solution of 7–8 (24 mmol) in diethyl
ether (20 mL). After stirring for 3 h at room temper-
ature, the reaction mixture was cooled over an ice-bath
and treated successively by the dropwise addition of
H2O (3.5 mL), 15% NaOH solution (7 mL), and H2O
(10.5 mL). The mixture was filtered by washing with
EtOAc and the filtrate was concentrated in vacuo. The
residue was purified by flash column chromatography
on silica gel with EtOAc/hexanes (3:1) as eluant to
afford 9–10.
5.1.2. 4-(2-Bromoethyl)-4-tert-butylbenzene (6). A cooled
solution of 4-tert-butylstyrene (4.81 g, 30 mmol) and
sodium borohydride (0.34 g, 9 mmol) in THF (30 mL)
was treated dropwise with a solution of boron tri-
fluoride diethyl etherate (1.84 mL, 15 mmol) in THF (10
mL) and stirred for 2 h at room temperature. The reac-
tion mixture was quenched with water (10 mL) carefully
and allowed to stand until no further hydrogen was
evolved. The solution was made alkaline by the addition
of dilute sodium hydroxide solution (15 mL), followed
by 30% hydrogen peroxide solution (15 mL) in 2–3 mL
portions. The reaction mixture was poured into ice
water and extracted with ether several times. The com-
bined organic layers were dried over MgSO4 and con-
centrated in vacuo. The residue was purified by flash
column chromatography on silica gel with EtOAc/hex-
anes (1:5) as eluant to afford 2-(4-tert-butylphenyl)-1-
ethanol (4.17 g, 78%) as a colorless oil; 1H NMR
(CDCl3) d 7.34 (d, 2H, J=8.0 Hz), 7.16 (d, 2H, J=8.0
Hz), 3.86 (dd, 2H, J=6.4, 12.7 Hz, CH2OH), 2.85 (t,
2H, J=6.6 Hz, CH2Ar), 1.31 (s, 9H, C(CH3)3).
5.3.1. 2-(3,4-Dimethylphenethyl)-1,3-propanediol (9). 74%
Yield, white solid, mp=70 ꢃC; H NMR (CD C3l) d 6.9–
1
7.1 (m, 3H), 3.85 (m, 2H, CH2OH), 3.71 (m, 2H, CH2OH),
2.60 (t, 2H, J=7.3 Hz, CH2Ar), 2.23 (dd, 6H, 2ꢄCH3),
1.82 (m, 1H, CH), 1.5–1.65 (m, 2H, CH2CH2Ar).
5.3.2. 2-(4-tert-Butylphenethyl)-1,3-propanediol (10). 86%
Yield, white solid, mp=68 ꢃC; 1H NMR (CDCl3) d 7.30
(d, 2H, J=8.3 Hz), 7.12 (d, 2H, J=8.3 Hz), 3.86 (m,
2H, CH2OH), 3.72 (m, 2H, CH2OH), 2.64 (t, 2H, J=7.3
Hz, CH2Ar), 1.82 (m, 1H, CH), 1.5–1.65 (m, 2H,
CH2CH2Ar), 1.31 (s, 9H, C(CH3)3).

1064
J. Lee et al. / Bioorg. Med. Chem. 12 (2004) 1055–1069
5.3.3. General procedure for the synthesis of 11–12. A
cooled solution of 9–10 (15 mmol) and pyridine (16.5
mmol, 1.33 mL) in CH2Cl2 (50 mL) was treated with
pivaloyl chloride (16.5 mmol, 2.02 mL) at 0 ꢃC. After
being stirred for 30 min at 0 ꢃC, the reaction mixture
was quenched with ice and extracted with EtOAc
several times. The combined organic layers were washed
with H2O and brine, dried over MgSO4 and
concentrated in vacuo. The residue was purified by col-
umn chromatography on silica gel with EtOAc/hexanes
(1:4) as eluant to afford 11–12.
hydrogen balloon for 2 h. The reaction mixture was fil-
tered and the filtrate was concentrated in vacuo to
afford the corresponding amine in a quantitative yield,
which was used for the next step without further
purification. A solution of amine (1 mmol) in CH2Cl2
(10 mL) was treated with isothiocyanate (1 mmol)
and stirred overnight at room temperature. After the
mixture was concentrated in vacuo, the residue was
purified by flash column chromatography on silica gel
with EtOAc/hexanes (1:1) as eluant to afford 15–16. A
cooled solution of 15–16 (1 mmol) in CH2Cl2 (4
mL) at 0 ꢃC was treated with trifluoroacetic acid (2
mL) and stirred for 1 h at room temperature. The
mixture was quenched with solid NaHCO3, filtered,
and the filtrate was concentrated in vacuo. The residue
was diluted with EtOAc, washed with NaHCO3, H2O
and brine, and concentrated in vacuo. The residue
was purified by flash column chromatography on
silica gel with EtOAc/hexanes (1:1) as eluant to afford
17–18.
5.3.4. 4-(3,4-Dimethylphenyl)-2-(hydroxymethyl)butyl pi-
1
valate (11). 84% Yield, colorless oil; H NMR (CDCl3)
d 6.9–7.1 (m, 3H), 4.26 (dd of AB, 1H, J=4.4, 11.4 Hz,
CH2OCO), 4.14 (dd of AB, 1H, J=6.1, 11.4 Hz,
CH2OCO), 3.56 (m, 2H, CH2OH), 2.62 (t, 2H, J=7.8
Hz, CH2Ar), 2.23 (dd, 6H, 2ꢄCH3), 2.03 (t, 1H, OH),
1.85 (m, 1H, CH), 1.55–1.75 (m, 2H, CH2CH2Ar), 1.21
(s, 9H, C(CH3)3).
5.3.5. 4-(4-tert-Butylphenyl)-2-(hydroxymethyl)butyl pi-
1
5.5.1. N-[4-(3,4-Dimethylphenyl)-2-(pivaloyloxymethyl)-
butyl] - N0 - [3 - methoxy - 4 - (methoxymethoxy)benzyl]-
thiourea (15). 85% yield, white solid, mp=42 ꢃC; 1H
NMR (CDCl3) d 6.8–7.15 (m, 6H), 6.32 (t, 1H, NH),
6.01 (bs, 1H, NH), 5.21 (s, 2H, OCH2O), 4.50 (d, 2H,
J=5.2 Hz, CSNHCH2Ar), 4.18 (dd, 1H, J=3.7, 11.5
Hz, CH2OCO), 3.95 (dd, 1H, J=5.1, 11.5 Hz,
CH2OCO), 3.86 (s, 3H, OCH3), 3.76 (m, 1H,
CHCH2NHCS), 3.50 (s, 3H, OCH3), 3.23 (m, 1H,
CHCH2NHCS), 2.65 (t, 2H, J=8.3 Hz, CH2CH2Ar),
2.15–2.3 (m, 7H, 2ꢄCH3), 1.94 (m, 1H, CH), 1.60 (m,
2H, CH2CH2Ar), 1.20 (s, 9H, COC(CH3)3).
valate (12). 80% Yield, colorless oil; H NMR (CDCl3)
d 7.33 (d, 2H, J=8.2 Hz), 7.14 (d, 2H, J=8.2 Hz), 4.29
(dd of AB, 1H, J=4.3, 11.4 Hz, CH2OCO), 4.17 (dd of
AB, 1H, J=6.0, 11.4 Hz, CH2OCO), 3.63 (m, 1H,
CH2OH), 3.55 (m, 1H, CH2OH), 2.68 (t, 2H, J=7.8 Hz,
CH2Ar), 2.08 (bs, 1H, OH), 1.90 (m, 1H, CH), 1.6–1.75
(m, 2H, CH2CH2Ar), 1.33 (s, 9H, C(CH3)3), 1.24 (s, 9H,
COC(CH3)3).
5.4. General procedure for the synthesis of 13–14
A mixture of 11–12 (10 mmol), triphenylphosphine (20
mmol, 5.25 g), diethyl azodicarboxylate (20 mmol, 3.15
mL) in THF (70 mL) was treated with diphenylpho-
sphorylazide (20 mmol, 4.32 mL) and stirred at room
temperature for 18 h. The solvent was removed in vacuo
and the residue was purified by column chromato-
graphy on silica gel with EtOAc/hexanes (1:10) as
eluant to afford 13–14.
5.5.2. N-[4-tert-Butylphenyl-2-(pivaloyloxymethyl)butyl]-
N0-[3-methoxy-4-(methoxymethoxy)benzyl]thiourea (16).
89% Yield, white solid, mp=45 ꢃC; H NMR (CDCl3)
1
d 7.30 (d, 2H, J=8.3 Hz), 7.1–7.15 (m, 3H), 6.8–6.9
(m, 2H), 6.41 (bt, 1H, NH), 6.35 (bs, 1H, NH),
6.10 (bs, 1H, NH), 5.21 (s, 2H, OCH2O), 4.51 (d,
2H, J=4.1 Hz, CSNHCH2Ar), 4.18 (dd, 1H,
J=3.7, 11.5 Hz, CH2OCO), 3.95 (dd, 1H, J=5.1,
11.5 Hz, CH2OCO), 3.86 (s, 3H, OCH3), 3.75 (m, 1H,
CHCH2NHCS), 3.50 (s, 3H, OCH3), 3.25 (m, 1H,
CHCH2NHCS), 2.69 (t, 2H, J=7.8 Hz, CH2CH2Ar),
2.03 (m, 1H, CH), 1.60 (m, 2H, CH2CH2Ar), 1.30 (s,
3H, C(CH3)3), 1.19 (s, 9H, COC(CH3)3).
5.4.1. 2-(Azidomethyl)-4-(3,4-dimethylphenyl)butyl piva-
1
late (13). 82% Yield, colorless oil; H NMR (CDCl3) d
6.9–7.1 (m, 3H), 4.10 (m, 2H, CH2OCO), 3.38 (d, 2H,
J=5.8 Hz, CH2N3), 2.60 (t, 2H, J=7.4 Hz, CH2Ar),
2.23 (dd, 6H, 2ꢄCH3), 1.95 (m, 1H, CH), 1.6–1.7 (m,
2H, CH2CH2Ar), 1.21 (s, 9H, C(CH3)3).
5.5.3. N-[4-(3,4-Dimethylphenyl)-2-(pivaloyloxymethyl)-
butyl]-N0-[4-hydroxy-3-methoxybenzyl]thiourea
5.4.2. 2-(Azidomethyl)-4-(4-tert-butylphenyl)butyl piva-
1
(17).
1
late (14). 84% Yield, colorless oil; H NMR (CDCl3) d
73% Yield, yellow oil; H NMR (CDCl3) d 6.8–7.1 (m,
6H), 6.29 (bt, 1H, NH), 6.00 (bs, 1H, NH), 5.60 (s, 1H,
OH), 4.48 (bs, 2H, CSNHCH2Ar), 4.20 (dd, 1H, J=3.7,
11.5 Hz, CH2OCO), 3.96 (dd, 1H, J=5.1, 11.5 Hz,
CH2OCO), 3.87 (s, 3H, OCH3), 3.78 (bs, 1H,
CHCH2NHCS), 3.25 (m, 1H, CHCH2NHCS), 2.65
(t, 2H, J=8.3 Hz, CH2CH2Ar), 2.2–2.3 (m, 6H,
2ꢄCH3), 2.00 (m, 1H, CH), 1.58 (m, 2H,
CH2CH2Ar), 1.19 (s, 9H, COC(CH3)3); IR (neat)
3355, 2962, 1715, 1557, 1516, 1274, 1156 cmꢀ1; MS
(FAB) m/z 487 (MH+). Anal. calcd for C27H38N2O4S:
C, 66.63; H, 7.87; N, 5.76; S, 6.59. Found: C, 66.83; H,
7.90; N, 5.74; S, 6.57.
7.32 (d, 2H, J=8.5 Hz), 7.11 (d, 2H, J=8.5 Hz), 4.12
(dd of AB, 1H, J=4.9, 11.2 Hz, CH2OCO), 4.08
(dd of AB, 1H, J=4.4, 11.4 Hz, CH2OCO), 3.38
(d, 2H, J=5.8 Hz, CH2N3), 2.64 (t, 2H, J=7.3 Hz,
CH2Ar), 1.95 (m, 1H, CH), 1.65–1.75 (m, 2H,
CH2CH2Ar), 1.31 (s, 9H, C(CH3)3), 1.21 (s, 9H,
COC(CH3)3).
5.5. General procedure for the synthesis of 15–18
A suspension of 13–14 (1 mmol) and Lindler’s catalyst
(100 mg) in EtOH (10 mL) was hydrogenated under a

J. Lee et al. / Bioorg. Med. Chem. 12 (2004) 1055–1069
1065
5.5.4. N-[4-tert-Butylphenyl-2-(pivaloyloxymethyl)butyl]-
N0 - [4 - hydroxy - 3 - methoxybenzyl]thiourea (18). 78%
Yield, yellow oil; 1H NMR (CDCl3) d 7.33 (d, 2H,
J=8.2 Hz), 7.14 (d, 2H, J=8.2 Hz), 6.8–6.9 (m, 3H),
6.34 (bt, 1H, NH), 6.10 (bs, 1H, NH), 5.63 (s, 1H, OH),
4.50 (bs, 2H, CSNHCH2Ar), 4.20 (dd, 1H, J=3.7, 11.6
Hz, CH2OCO), 3.97 (dd, 1H, J=5.1, 11.6 Hz,
CH2OCO), 3.89 (s, 3H, OCH3), 3.79 (bs, 1H,
CHCH2NHCS), 3.28 (m, 1H, CHCH2NHCS), 2.71 (t,
2H, J=8.1 Hz, CH2CH2Ar), 2.04 (m, 1H, CH), 1.62 (m,
2H, CH2CH2Ar), 1.33 (s, 3H, C(CH3)3), 1.21 (s, 9H,
COC(CH3)3; IR (neat) 3355, 2962, 1715, 1557, 1516,
1274, 1156 cmꢀ1; MS (FAB) m/z 515 (MH+). Anal.
calcd for C29H42N2O4S: C, 67.67; H, 8.22; N, 5.44; S,
6.23. Found: C, 67.90; H, 8.24; N, 5.43; S, 6.20.
mmol, 505 mg) in dimethylsulfoxide (5 mL) was treated
with benzyl halide (3 mmol) at 0 ꢃC and stirred for 30
min at room temperature. The reaction mixture was
neutralized with 1 N hydrochloric acid and extracted
with EtOAc several times. The combined organic layers
were washed with water and brine, dried over MgSO4
and concentrated in vacuo. The residue was purified by
flash column chromatography on silica gel with EtOAc/
hexanes (1:10) as eluant to afford 22–24.
5.8. Ethyl 3-(3,4-dimethylphenyl) - 2 - [(diphenylmethyle-
ne)amino]propanoate (22)
1
42% Yield, yellow oil; H NMR (CDCl3) d 7.55 (m,
2H), 7.2–7.45 (m, 6H), 6.6–7.0 (m, 5H), 4.1–4.3 (m, 3H,
NCHCO2CH2), 3.0–3.4 (m, 2H, CH2Ar), 2.1–2.2 (m,
6H, 2ꢄCH3),1.26 (m, 3H, CO2CH2CH3).
5.6. General procedure for the synthesis of 19–20
A suspension of 13–14 (1 mmol) and Lindler’s catalyst (100
mg) in EtOH (10 mL) was hydrogenated under a hydrogen
balloon for 2 h. The reaction mixture was filtered and the
filtrate was concentrated in vacuo to afford the corres-
ponding amine in a quantitative yield, which was used for
the next step without further purification. A solution of
amine (1 mmol) was treated with pentafluorophenyl ester
(1 mmol) in CH2Cl2 (10 mL), stirred overnight at room
temperature and concentrated in vacuo. The residue was
purified by flash column chromatography on silica gel with
EtOAc/hexanes (1:1) as eluant to afford 19–20.
5.8.1. Ethyl 3-(4-tert-butylphenyl)-2-[(diphenylmethyle-
ne)amino]propanoate (23). 70% yield, white solid,
mp=44 ꢃC; H NMR (CDCl3) d 7.58 (d, 2H, J=8.3
1
Hz), 7.15–7.4 (m, 10H), 6.94 (d, 2H, J=8.0 Hz), 4.1–
4.25 (m, 3H, NCHCO2CH2), 3.24 (dd of AB, 1H,
J=4.1, 13.2 Hz, CH2Ar), 3.12 (dd of AB, 1H, J=9.2,
13.2 Hz, CH2Ar), 1.30 (s, 9H, C(CH3)3), 1.25 (t, 3H,
J=7.0 Hz, CO2CH2CH3).
5.8.2. Ethyl 4-(3,4-dimethylphenyl)-2-[(diphenylmethyle-
ne)amino]butanoate (24). 65% Yield, yellow oil; 1H
NMR (CDCl3) d 7.66 (m, 2H), 7.35–7.5 (m, 6H), 7.14 (m,
2H), 6.85–7.0 (m, 3H), 4.0–4.25 (m, 3H, NCHCO2CH2),
2.4–2.6 (m, 2H, CH2Ar), 2.15–2.3 (m, 8H, 2ꢄCH3 and
NCHCH2),1.26 (t, 3H, J=5.0 Hz, CO2CH2CH3).
5.6.1. N - [4 - (3,4 - Dimethylphenyl) - 2 - (pivaloyloxy-
methyl)butyl]-2-[4-hydroxy-3-methoxyphenyl]acetamide
1
(19). 59% Yield, pink oil; H NMR (CDCl3) d 6.7–7.05
(m, 6H), 5.76 (bt, 1H, NH), 5.63 (s, 1H, OH), 4.02 (dd,
1H, J=4.4, 11.5 Hz, CH2OCO), 3.95 (dd, 1H, J=5.4,
11.5 Hz, CH2OCO), 3.86 (s, 3H, OCH3), 3.49 (s, 2H,
OCCH2Ar), 3.29 (m, 1H, CH2NH), 3.13 (m, 1H,
CH2NH), 2.58 (m, 2H, CH2CH2Ar), 2.1–2.3 (m, 6H,
2ꢄCH3), 1.84 (m, 1H, CH), 1.54 (m, 2H, CH2CH2Ar),
1.17 (s, 9H, COC(CH3)3); IR (neat) 3445, 2962, 1725,
1716, 1650, 1556, 1538, 1516 cmꢀ1; MS (FAB) m/z 456
(MH+). Anal. calcd for C27H37NO5: C, 71.18; H, 8.19;
N, 3.07. Found: C, 71.39; H, 8.20; N, 3.05.
5.9. General procedure for the synthesis of 25–27
A solution of 22–24 (1 mmol) in tetrahydrofuran (10
mL) was adjusted to pH 4 with 1 N hydrochloric acid
and stirred for 30 min. The mixture was neutralized
with 1 N sodium hydroxide solution and extracted with
ethyl acetate several times. The combined organic layers
were washed with water and brine, dried over MgSO4
and concentrated in vacuo. The residue was purified by
flash column chromatography on silica gel with EtOAc/
hexanes (1:2) as eluant to afford 25–27.
5.6.2. N - [4 - (4 - tert - Butylphenyl) - 2 - (pivaloyloxy-
methyl)butyl]-2-[4-hydroxy-3-methoxyphenyl]acetamide
1
(20). 55% yield, yellow oil; H NMR (CDCl3) d 7.28
5.9.1. Ethyl 2-amino-3-(3,4-dimethylphenyl)propanoate
1
(d, 2H, J=8.2 Hz), 7.06 (d, 2H, J=8.3 Hz), 6.7-6.9 (m,
3H), 5.75 (bt, 1H, NH), 5.57 (s, 1H, OH), 4.04 (dd, 1H,
J=4.6, 11.6 Hz, CH2OCO), 3.94 (dd, 1H, J=5.1, 11.6
Hz, CH2OCO), 3.86 (s, 3H, OCH3), 3.49 (s, 2H,
OCCH2Ar), 3.28 (m, 1H, CH2NH), 3.14 (m, 1H,
CH2NH), 2.60 (dd, 2H, J=6.3, 9.5 Hz CH2CH2Ar),
1.86 (m, 1H, CH), 1.56 (m, 2H, CH2CH2Ar), 1.30 (s, 3H,
C(CH3)3), 1.17 (s, 9H, COC(CH3)3); IR (neat) 3444, 2960,
1726, 1716, 1651, 1557, 1539, 1516 cmꢀ1; MS (FAB) m/z
484 (MH+). Anal. calcd for C29H41NO5: C, 72.02; H,
8.54; N, 2.90. Found: C, 72.24; H, 8.56; N, 2.88.
(25). 61% yield, yellow oil; H NMR (CDCl3) d 6.9–7.1
(m, 3H), 4.1–4.25 (m, 2H, CO2CH2), 3.70 (m, 1H,
NH2CH), 3.0–3.2 (m, 1H, CH2Ar), 2.80 (m, 1H, CH2Ar),
2.2–2.3 (m, 6H, 2ꢄCH3), 1.25 (m, 3H, CO2CH2CH3).
5.9.2. Ethyl 2-amino-3-(4-tert-butylphenyl)propanoate (26).
1
81% yield, yellow oil; H NMR (CDCl3) d 7.32 (d, 2H,
J=8.3 Hz), 7.13 (d, 2H, J=8.3 Hz), 4.17 (q, 2H, J=7.1
Hz, CO2CH2), 3.70 (dd, 1H, J=5.2, 8.0 Hz, NH2CH),
3.05 (dd of AB, 1H, J=5.2, 13.5 Hz, CH2Ar), 3.82 (dd
of AB, 1H, J=8.0, 13.5 Hz, CH2Ar), 1.30 (s, 9H,
C(CH3)3), 1.23 (t, 3H, J=7.1 Hz, CO2CH2CH3).
5.7. General procedure for the synthesis of 22–24
5.9.3. Ethyl 2-amino-4-(3,4-dimethylphenyl)butanoate (27).
1
A mixture of N-(diphenylmethylene)glycine ethyl ester
(21) (3 mmol, 800 mg) and potassium hydroxide (9
82% Yield, yellow oil; H NMR (CDCl3) d 6.9–7.1 (m,
3H), 4.18 (q, 2H, J=7.1 Hz, CO2CH2), 3.45 (dd, 1H,

1066
J. Lee et al. / Bioorg. Med. Chem. 12 (2004) 1055–1069
J=5.4, 7.8 Hz, NH2CH), 2.6–2.7 (m, 2H, CH2Ar), 2.3–
2.4 (m, 6H, 2ꢄCH3), 2.05 (m, 1H, CH2CH2Ar), 1.82 (m,
1H, CH2CH2Ar), 1.28 (m, 3H, J=7.1 Hz, CO2CH2CH3).
Hz, CH2OH), 3.55 (dd, 1H, J=5.6, 11.0 Hz, CH2OH),
2.80 (d, 2H, J=7.3 Hz, CH2Ar), 1.41 (s, 9H,
CO2C(CH3)3), 1.30 (s, 9H, C(CH3)3).
5.11.3. tert-Butyl N-[2-(3,4-dimethylphenyl)-1- (hydroxy-
methyl)ethyl]carbamate (33). 96% Yield, colorless oil;
1H NMR (CDCl3) d 6.9–7.1 (m, 3H), 4.70 (bs, 1H, NH),
3.82 (bs, 1H, NHCH), 3.5–3.7 (m, 2H, CH2OH), 2.7–2.9
(m, 2H, CH2Ar), 2.2–2.3 (m, 6H, 2ꢄCH3), 2.05 (m, 1H,
CH2CH2Ar), 1.85 (m, 1H, CH2CH2Ar), 1.42 (s, 9H,
CO2C(CH3)3).
5.10. General procedure for the synthesis of 28–30
A mixture of 25–27 (1 mmol) and di-tert-butyl dicarbo-
nate (1.2 mmol) in tetrahydrofuran (5 mL) was stirred
for 16 h and concentrated in vacuo. The residue was
purified by flash column chromatography on silica gel
with EtOAc/hexanes (1:4) as eluant to afford 28–30.
5.10.1. Ethyl 2-[(tert-butoxycarbonyl)amino]-3-(3,4-dime-
1
5.12. General procedure for the synthesis of 34–36
thylphenyl)propanoate (28). 98% Yield, colorless oil; H
NMR (CDCl3) d 6.85–7.1 (m, 3H), 4.96 (m, 1H, NH),
4.50 (m, 1H, NHCH), 4.1–4.2 (m, 2H, CO2CH2), 3.0–
3.2 (m, 2H, CH2Ar), 2.2–2.35 (m, 6H, 2ꢄCH3), 1.41 (d,
9H, J=6.8 Hz, CO2C(CH3)3), 1.2–1.3 (t, 3H,
CO2CH2CH3).
A cooled solution of 31-33 (1 mmol), triethylamine (3
mmol) and dimethylaminopyridine (0.1 mmol) at 0 ꢃC
in CH2Cl2 (10 mL) was treated with pivaloyl chloride
(1.5 mmol) and stirred for 1 h at room temperature. The
mixture was concentrated in vacuo and the residue was
purified by flash column chromatography on silica gel
with EtOAc/hexanes (1:4) as eluant to afford 34–36.
5.10.2. Ethyl 2-[(tert-butoxycarbonyl)amino]-3-(4-tert-bu-
1
tylphenyl)propanoate (29). 98% Yield, colorless oil; H
NMR (CDCl3) d 7.30 (d, 2H, J=8.3 Hz), 7.06 (d, 2H,
J=8.3 Hz), 4.96 (bd, 1H, NH), 4.56 (m, 1H, NHCH),
4.16 (q, 2H, J=7.1 Hz, CO2CH2), 3.04 (d, 2H, CH2Ar),
1.42 (s, 9H, CO2C(CH3)3), 1.30 (s, 9H, C(CH3)3), 1.22
(t, 3H, J=7.1 Hz, CO2CH2CH3).
5.12.1. 2-[(tert-Butoxycarbonyl)amino]-3-(3,4-dimethyl-
1
phenyl)propyl pivalate (34). 99% Yield, colorless oil; H
NMR (CDCl3) d 6.9–7.1 (m, 3H), 4.61 (d, 1H, NH),
3.9–4.2 (m, 3H, CH2OCO and NHCH), 2.6–2.8 (m, 2H,
CH2Ar), 2.2–2.3 (m, 6H, 2ꢄCH3), 1.41 (s, 9H,
CO2C(CH3)3), 1.23 (s, 9H, COC(CH3)3).
5.10.3. Ethyl 2-[(tert-butoxycarbonyl)amino]-4-(3,4-dime-
1
thylphenyl)butanoate (30). 94% Yield, colorless oil; H
5.12.2. 2-[(tert-Butoxycarbonyl)amino]-3-(4-tert-butyl-
phenyl)propyl pivalate (35). 99% Yield, white solid,
NMR (CDCl3) d 6.90–7.1 (m, 3H), 5.08 (d, J=7.5 Hz,
1H, NH), 4.32 (m, 1H, NHCH), 4.16 (q, 2H, J=7.1 Hz,
CO2CH2), 2.5–2.7 (m, 2H, CH2Ar), 2.2–2.35 (m, 6H,
2ꢄCH3), 2.10 (m, 1H, CH2CH2Ar), 1.90 (m, 1H,
CH2CH2Ar), 1.45 (s, 9H, CO2C(CH3)3), 1.23 (t, 3H,
J=7.1 Hz, CO2CH2CH3).
mp=58 ꢃC; H NMR (CDCl3) d 7.31 (d, 2H, J=8.3
1
Hz), 7.10 (d, 2H, J=8.3 Hz), 4.57 (bs, 1H, NH), 4.0–4.2
(m, 3H, CH2OCO and NHCH), 2.75–2.9 (m, 2H,
CH2Ar), 1.41 (s, 9H, CO2C(CH3)3), 1.30 (s, 9H,
C(CH3)3), 1.23 (s, 9H, COC(CH3)3).
5.12.3. 2-[(tert-Butoxycarbonyl)amino]-4-(3,4-dimethyl-
phenyl)butyl pivalate (36). 87% Yield, white solid,
5.11. General procedure for the synthesis of 31–33
mp=52 ꢃC; H NMR (CDCl3) d 6.9–7.1 (m, 3H), 4.52
1
A cooled suspension of lithium aluminium hydride (2
mmol) in diethyl ether (10 mL) at 0 ꢃC was treated
dropwise with 28–30 (1 mmol) in diethyl ether (10 mL).
After being stirred for 1 h at room temperature, the
mixture was cooled and quenched with H2O (1 mL),
15% NaOH solution (2 mL) and H2O (3 mL) succes-
sively. The suspension was filtered with washing of
EtOAc and the filtrate was concentrated in vacuo. The
residue was purified by flash column chromatography
on silica gel with EtOAc/hexanes (4:1) as eluant to
afford 31–33.
(d, 1H, NH), 3.9–4.2 (m, 3H, CH2OCO and NHCH),
2.6–2.8 (m, 2H, CH2Ar), 2.2–2.3 (m, 6H, 2ꢄCH3), 1.6–
1.8 (m, 2H, CH2CH2Ar), 1.45 (s, 9H, CO2C(CH3)3),
1.20 (s, 9H, COC(CH3)3).
5.13. General procedure for the synthesis of 37–42
A cooled solution of 34–36 (1 mmol) in CH2Cl2 (8 mL)
at 0 ꢃC was treated dropwise with trifluoroacetic acid (2
mL) and stirred for 3 h at room temperature. The mix-
ture was concentrated in vacuo to afford the amine salt
which was used for the next step without further pur-
ification. A solution of the amine salt (1 mmol) and
triethylamine (1.2 mmol) in CH2Cl2 (10 mL) was treated
with isothiocyanate (1 mmol) and stirred overnight at
room temperature. After the mixture was concentrated
in vacuo, the residue was purified by flash column
chromatography on silica gel with EtOAc/hexanes (1:1)
as eluant to afford 37–39. A cooled solution of 37–39 (1
mmol) in CH2Cl2 (4 mL) at 0 ꢃC was treated with tri-
fluoroacetic acid (2 mL) and stirred for 1 h at room
temperature. The mixture was quenched with solid
NaHCO3, filtered, and the filtrate was concentrated in
5.11.1. tert-Butyl N-[3-(3,4-dimethylphenyl)-1- (hydroxy-
methyl)propyl]carbamate (31). 85% Yield, colorless oil;
1H NMR (CDCl3) d&#32;6.9–7.1 (m, 3H), 4.70 (bs, 1H,
NH), 3.82 (bs, 1H, NHCH), 3.5–3.7 (m, 2H, CH2OH),
2.7–2.9 (m, 2H, CH2Ar), 2.2–2.3 (m, 6H, 2ꢄCH3), 1.42
(s, 9H, CO2C(CH3)3).
5.11.2. tert-Butyl N - [3 - (4 - tert - butylphenyl)-1-
(hydroxymethyl)propyl]carbamate (32). 80% Yield,
white solid, mp=48 ꢃC; H NMR (CDCl3) d 7.32 (d,
1
2H, J=8.3 Hz), 7.13 (d, 2H, J=8.3 Hz), 4.70 (bs, 1H,
NH), 3.86 (bs, 1H, NHCH), 3.67 (dd, 1H, J=3.4, 11.0

J. Lee et al. / Bioorg. Med. Chem. 12 (2004) 1055–1069
1067
vacuo. The residue was diluted with EtOAc, washed
with NaHCO3, H2O and brine, and concentrated in
vacuo. The residue was purified by flash column chro-
matography on silica gel with EtOAc/hexanes (1:1) as
eluant to afford 40–42.
J=5.1, 11.5 Hz, CH2OCO), 3.99 (dd, 1H, J=4.4, 11.4
Hz, CH2OCO), 3.87 (s, 3H, OCH3), 3.00 (dd, 1H,
J=4.9, 13.4 Hz, CH2Ar), 2.72 (dd, 1H, J=8.0, 13.4 Hz,
CH2Ar), 1.30 (s, 9H, C(CH3)3), 1.18 (s, 9H,
COC(CH3)3); IR (KBr) 3353, 2960, 1729, 1514, 1462,
1363, 1274, 1154, 1034 cmꢀ1; MS (FAB) m/z 487 (MH+).
Anal. calcd for C27H38N2O4S: C, 66.63; H, 7.87; N, 5.76;
S, 6.59. Found: C, 66.84; H, 7.89; N, 5.72; S, 6.56.
5.13.1. N-[3-(3,4-Dimethylphenyl)-1-pivaloyloxy-2-pro-
pyl]-N0-[3-methoxy-4-(methoxymethoxy)benzyl]thiourea
(37). 97% Yield, yellow oil; 1H NMR (CDCl3) d 6.7–7.1
(m, 6H), 6.36 (bs, 1H, NH), 6.03 (d, 1H, J=8.0 Hz,
NH), 5.21 (s, 2H, OCH2O), 4.70 (m, 1H, CH), 4.46 (bs,
2H, NHCH2Ar), 4.15 (dd, 1H, J=5.1, 11.5 Hz,
CH2OCO), 3.98 (dd, 1H, J=4.7, 11.5 Hz, CH2OCO),
3.85 (s, 3H, OCH3), 3.50 (s, 3H, OCH3), 2.98 (dd, 1H,
J=5.1, 13.4 Hz, CH2Ar), 2.68 (dd, 1H, J=8.3, 13.4 Hz,
CH2Ar), 2.2–2.3 (m, 6H, 2ꢄCH3), 1.19 (s, 9H,
COC(CH3)3).
5.13.6. N-[4-(3,4-Dimethylphenyl)-1-pivaloyloxy-2-butyl]
- N0 - [4 - hydroxy - 3 - methoxybenzyl]thiourea (42):. 76%
yield, white solid, mp=38-40 ꢃC; H NMR (CDCl3) d
1
6.7–7.0 (m, 6H), 6.53 (bs, 1H, NH), 5.92 (bs, 1H, NH),
5.69 (s, 1H, OH), 4.46 (bs, 2H, NHCH2Ar), 4.25 (dd,
1H, J=4.6, 11.5 Hz, CH2OCO), 3.96 (bs, 1H,
CH2OCO), 3.85 (s, 3H, OCH3), 2.57 (t, 2H, J=7.6 Hz,
CH2Ar), 2.1-2.3 (m, 6H, 2ꢄCH3), 1.80 (m, 2H,
CH2CH2Ar), 1.14 (s, 9H, COC(CH3)3); IR (KBr) 3357,
2968, 1727, 1613, 1537, 1514, 1480, 1462, 1432, 1367,
1278, 1237, 1155, 1034 cmꢀ1; MS (FAB) m/z 473 (MH+).
Anal. calcd for C26H36N2O4S: C, 66.07; H, 7.68; N, 5.93;
S, 6.78. Found: C, 66.30; H, 7.70; N, 5.90; S, 6.75.
5.13.2. N-[3-(4-tert-Butylphenyl)-1-pivaloyloxy-2-pro-
pyl]-N0 -[3-methoxy-4-(methoxymethoxy)benzyl]thiourea
1
(38). 52% Yield, yellow oil; H NMR (CDCl3) d 7.31
(d, 2H, J=8.3 Hz), 7.12 (d, 2H, J=8.3 Hz), 7.10 (d, 1H,
J=8 Hz), 6.88 (d, 1H, J=1.7 Hz), 6.80 (dd, 1H, J=1.7,
8 Hz), 6.42 (bs, 1H, NH), 6.04 (d, 1H, J=7.6 Hz, NH),
5.21 (s, 2H, OCH2O), 4.70 (bs, 1H, CH), 4.48 (bs, 2H,
NHCH2Ar), 4.16 (dd, 1H, J=6.3, 11.4 Hz, CH2OCO),
3.98 (dd, 1H, J=4.6, 11.4 Hz, CH2OCO), 3.86 (s, 3H,
OCH3), 3.50 (s, 3H, OCH3), 3.01 (dd, 1H, J=4.6, 13.4
Hz, CH2Ar), 2.73 (dd, 1H, J=8.0, 13.4 Hz, CH2Ar),
1.30 (s, 9H, C(CH3)3), 1.18 (s, 9H, COC(CH3)3).
5.13.7. General procedure for the synthesis of 43–44. A
cooled solution of 34–36 (1 mmol) in CH2Cl2 (8 mL) at
0 ꢃC was treated dropwise with trifluoroacetic acid (2
mL) and stirred for 3 h at room temperature. The mix-
ture was concentrated in vacuo to afford the amine salt
which was used for the next step without further pur-
ification. A solution of amine salt (1 mmol) and trie-
thylamine (1.2 mmol) ) in CH2Cl2 (10 mL) was treated
with pentafluorophenyl ester (1 mmol) in CH2Cl2 (10
mL) and stirred overnight at room temperature. After
the mixture was concentrated in vacuo, the residue was
purified by flash column chromatography on silica gel
with EtOAc/hexanes (1:1) as eluant to afford 43–44.
5.13.3. N-[4-(3,4-Dimethylphenyl)-1-pivaloyloxy-2-butyl]-
N0-[3-methoxy-4-(methoxymethoxy)benzyl]thiourea (39).
64% Yield, white solid, mp=62 ꢃC; H NMR (CDCl3)
1
d 6.8–7.1 (m, 6H), 6.52 (bs, 1H, NH), 5.94 (bs, 1H, NH),
5.20 (s, 2H, OCH2O), 4.52 (bs, 2H, NHCH2Ar), 4.26
(dd, 1H, J=4.4, 11.5 Hz, CH2OCO), 3.96 (bs, 1H,
CH2OCO), 3.85 (s, 3H, OCH3), 3.49 (s, 3H, OCH3),
2.5–2.7 (m, 2H, CH2Ar), 2.1–2.3 (m, 6H, 2ꢄCH3), 1.80
(m, 2H, CH2CH2Ar), 1.14 (s, 9H, COC(CH3)3).
5.13.8. N-[3-(3,4-Dimethylphenyl)-1-pivaloyloxy-2-pro-
pyl]-2-[4-hydroxy-3-methoxyphenyl]acetamide (43). 78%
1
yield, yellow oil; H NMR (CDCl3) d 6.6-7.05 (m, 6H),
5.58 (d, 1H, J=8.5 Hz, NH), 4.34 (m, 1H, CH), 3.97
(dd of AB, 2H, CH2OCO), 3.85 (s, 3H, OCH3), 3.46 (s,
2H, OCCH2Ar), 2.68 (ddd of AB, 2H, CH2Ar), 2.2–2.3
(m, 6H, 2ꢄCH3), 1.14 (s, 9H, COC(CH3)3); IR (neat)
3295, 2969, 1730, 1650, 1514, 1455, 1366, 1279, 1236,
1155, 1035 cmꢀ1; MS (FAB) m/z 428 (MH+). Anal.
calcd for C25H33NO5: C, 70.23; H, 7.78; N, 3.28. Found:
C, 70.47; H, 7.81; N, 3.26.
5.13.4. N-[3-(3,4-Dimethylphenyl)-1-pivaloyloxy-2-pro-
pyl]-N0-[4-hydroxy-3-methoxybenzyl]thiourea (40). 61%
Yield, white solid, mp=43–48 ꢃC; H NMR (CDCl3) d
1
6.7–7.05 (m, 6H), 6.35 (bs, 1H, NH), 6.00 (t, 1H, NH),
5.68 (bs, 1H, OH), 4.70 (bs, 1H, CHNH), 4.40 (bs, 2H,
NHCH2Ar), 4.13 (m, 1H, CH2OCO), 4.00 (m, 1H,
CH2OCO), 3.86 (s, 3H, OCH3), 3.00 (ddd, 1H, CH2Ar),
2.73 (ddd, 1H, CH2Ar), 2.2–2.3 (m, 6H, 2ꢄCH3), 1.19
(s, 9H, COC(CH3)3); IR (KBr) 3352, 2960, 1728, 1515,
1461, 1364, 1273, 1154, 1033 cmꢀ1; MS (FAB) m/z 459
(MH+). Anal. calcd for C25H34N2O4S: C, 65.47; H,
7.47; N, 6.11; S, 6.99. Found: C, 65.66; H, 7.49; N, 6.09;
S, 6.96.
5.13.9. N-[3-(4-tert-Butylphenyl)-1-pivaloyloxy-2-pro-
pyl]-2-[4-hydroxy-3-methoxyphenyl]acetamide (44). 79%
yield, yellow oil; 1H NMR (CDCl3) d 7.25 (d, 2H,
J=8.3 Hz), 6.94 (d, 2H, J=8.3 Hz), 6.88 (d, 1H, J=8
Hz, H-5), 6.69 (d, 1H, J=2 Hz, H-2), 6.65 (dd, 1H,
J=8, 2 Hz, H-6), 5.57 (d, 1H, J=8.5 Hz, NH), 5.30 (s,
1H, OH), 4.40 (m, 1H, CH), 3.95 (ddd of AB, 2H,
CH2OCO), 3.86 (s, 3H, OCH3), 3.47 (s, 2H, OCCH2Ar),
2.72 (ddd of AB, 2H, CH2Ar), 1.29 (s, 9H, C(CH3)3),
1.13 (s, 9H, COC(CH3)3); IR (neat) 3296, 2963, 1729,
5.13.5. N-[3-(4-tert-Butylphenyl)-1-pivaloyloxy-2-propyl]
- N0 - [4 - hydroxy - 3 - methoxybenzyl]thiourea (41). 62%
Yield, white solid, mp=57–60 ꢃC; H NMR (CDCl3) d
1
7.30 (d, 2H, J=8.3 Hz), 7.10 (d, 2H, J=8.3 Hz), 6.87
(d, 1H, J=8 Hz, H-5), 6.83 (d, 1H, J=1.7 Hz, H-2),
6.77 (dd, 1H, J=1.7, 8 Hz, H-6), 6.36 (bs, 1H, NH),
5.99 (d, 1H, J=8.0 Hz, NH), 5.65 (bs, 1H, OH), 4.71
(bs, 1H, CH), 4.42 (bs, 2H, NHCH2Ar), 4.16 (dd, 1H,
1650, 1514, 1462, 1364, 1279, 1236, 1155, 1035 cmꢀ1
;
MS (FAB) m/z 456 (MH+). Anal. calcd for
C27H37NO5: C, 71.18; H, 8.19; N, 3.07. Found: C,
71.42; H, 8.22; N, 3.05.

1068
J. Lee et al. / Bioorg. Med. Chem. 12 (2004) 1055–1069
3. Caterina, M. J.; Schumacher, M. A.; Tominaga, M.;
5.14. Molecular modeling
Rosen, T. A.; Levine, J. D.; Julius, D. Nature 1997, 389,
816.
4. Walpole, C. S. J.; Wrigglesworth, R. Capsaicin in the
Study of Pain; Academic Press: San Diego, CA, 1993; 63-
82.
5. Appendino, G.; Szallasi, A. Life Sci. 1997, 60, 681.
6. Zygmunt, P. M.; Petersson, J.; Andersson, D. A.;
Chuang, H.-H.; Sorgard, M.; Di Marzo, V.; Julius, D.;
Hogestatt, E. D. Nature 1999, 400, 452.
The structures of all chemicals, resiniferatoxin (RTX)
and thiourea and amide analogues, were built using the
Sybyl molecular modeling program (Tripos, Inc.), and
then the geometries were fully optimized using the
Tripos force field with the following non-default
options (method: conjugate gradient, termination:
.
gradient 0.005 kcal/mol A, and max iterations: 10,000).
The partial atomic charges were calculated by the Gas-
teiger-Huckel method in the Sybyl program. The avail-
able conformational space of RTX was explored using
molecular dynamic (MD) simulation. The MD simu-
lation was accomplished using the Sybyl Dynamic pro-
gram. The active conformation of RTX was generated
by the following protocol as previously described:28 The
system was brought to equilibrium, over 1 ps at 300 K,
before a molecular simulation study spanning, a further
500 ps (at 300 K) was undertaken. Snap shots taken at 1
ps intervals were minimized using the optimization cri-
teria outlined above. The conformational analyses of
thiourea and amide analogues (1, 3, 17, 19, 40, 42, and
43) were performed using the Sybyl Random Search
method (Maximum Cycles: 2,000 and Chirality: R con-
figuration). As a final step, global minima were found
for each chemical and used in the further analysis. The
pharmacophore model has two hydrophobic ring sys-
tems and two hydrogen bond acceptors. The center of
vanilloid ring, the center of orthoester phenyl ring, the
C3-keto carbonyl, and the C20-ester carbonyl in RTX
were superimposed with the center of vanilloid ring, the
center of phenyl ring, the carbonyl of ester, and the
sulfur atom of the thiourea or the oxygen atom of the
amide in the vanilloid agonist templates. All calcula-
tions for thiourea and amide analogues were performed
on a Silicon Graphics O2 R10000 workstation.
7. Hwang, S. W.; Cho, H.; Kwak, J.; Lee, S. Y.; Kang, C. J.;
Jung, J.; Cho, S.; Min, K. H.; Suh, Y. G.; Kim, D.; Oh,
U. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 6155.
8. Hayes, P.; Meadows, H. J.; Gunthorpe, M. J.; Harries,
M. H.; Duckworth, D. M.; Cairns, W.; Harrision, D. C.;
Clarke, C. E.; Ellington, K.; Prinjha, R. K.; Barton,
A. J. L.; Medhurst, A. D.; Smith, G. D.; Topp, S.; Mur-
dock, P.; Sanger, G. J.; Terrett, J.; Jenkins, O.; Benham,
C. D.; Randall, A. D.; Gloger, I. S.; Davis, J. B. Pain
2000, 88, 205.
9. Jordt, S.-E.; Julius, D. Cell 2002, 108, 421.
10. Savidge, J.; Davis, C.; Shah, K.; Colley, S.; Phillips, E.;
Ranasinghe, S.; Winter, J.; Kotsonis, P.; Rang, H.;
McIntyre, P. Neuropharmacology 2002, 43, 450.
11. Montell, C.; Birnbaumer, L.; Flockerzi, V.; Bindels, R. J.;
Bruford, E. A.; Caterina, M. J.; Clapham, D. E.; Harte-
neck, C.; Heller, S.; Julius, D.; Kojima, I.; Mori, Y.; Pen-
ner, R.; Prawitt, D.; Scharenberg, A. M.; Schultz, G.;
Shimizu, N.; Zhu, M. X. Mol. Cell 2002, 9, 229.
12. Bhave, G.; Zhu, W.; Wang, H.; Brasier, D. J.; Oxford,
G. S.; Gereau, R. W., IV Neuron 2002, 35, 721.
13. Numazaki, M.; Tominaga, T.; Toyooka, H.; Tominaga,
M. J. Biol. Chem. 2002, 277, 13375.
14. Welch, J. M.; Simon, S. A.; Reinhart, P. H. Proc. Natl.
Acad. Sci. U.S.A. 2000, 97, 13889.
15. Jordt, S. E.; Tominaga, M.; Julius, D. Proc. Natl. Acad.
Sci. U.S.A. 2000, 97, 8134.
16. Prescott, E. D.; Julius, D. Science 2003, 300, 1284.
17. Caterina, M. J.; Leffler, A.; Malmberg, A. B.; Martin,
W. J.; Traton, J.; Petersen-Zeitz, K. R.; Koltzenburg, M.;
Basbaum, A. I.; Julius, D. Science 2000, 288, 306.
18. Davis, J. B.; Gray, J.; Gunthorpe, M. J.; Hatcher, J. P.;
Davey, P. T.; Overend, P.; Harries, M. H.; Latcham, J.;
Clapham, C.; Atkinson, K.; Hughes, S. A.; Rance, K.;
Grau, E.; Harper, A. J.; Pugh, P. L.; Rogers, D. C.;
Bingham, S.; Randall, A.; Sheardown, S. A. Nature 2000,
405, 183.
19. Pomonis, J. D.; Harrison, J. E.; Mark, L.; Bristol, D. R.;
Valenzano, K. J.; Walker, K. J. Pharmacol. Exp. Ther.
2003, 306, 387.
20. Szallasi, A.; Blumberg, P. M. Adv. Pharmacol. 1993, 24,
123.
5.15. Statistical data analysis
The regression equations of the best-fit line and the
values of the correlation coefficient R2 were measured
with Microsoft Excel. Since the data of binding affinity
(Kis) and calcium influx (EC50s) have a wide range of
values each data point was log-transformed.
5.16. VR1 binding and 45Ca2+ uptake assay
The methods were described in previous reports.27,33
21. Szallasi, A.; Blumberg, P. M. Neuroscience 1989, 30, 515.
22. Szallasi, A. Drugs Aging 2001, 18, 561.
23. Park, N. S.; Seong, C. M.; Jung, Y. S.; Kim, W. B.; Kim,
S. H. Drugs of the Future 2000, 25, 1131.
Acknowledgements
24. Urban, L.; Campbell, E. A.; Panesar, M.; Patel, S.;
Chaudhry, N.; Kane, S.; Buchheit, K.-H.; Sandells, B.;
James, I. F. Pain 2000, 89, 65.
We would like to thank Dr. David G. Vander Velde,
University of Kansas, for providing the coordinates of
RTX conformations A,B,C,D. This research was sup-
ported by a KOSEF grant 2001-21500-001-1.
25. Detailed progress is represented in http://www.afferon.com
26. Sterner, O.; Szallasi, A. TIPS 1999, 20, 459.
27. Szallasi, A.; Blumberg, P. M.; Annicelli, L. L.; Krause,
J. E.; Cortright, D. N. Mol. Pharmacol. 1999, 56, 581.
28. Lee, J.; Lee, J.; Kim, J.; Kim, S. Y.; Chun, M. W.; Cho,
H.; Hwang, S. W.; Oh, U.; Park, Y. H.; Marquez, V. E.;
Beheshti, M.; Szabo, T.; Blumberg, P. M. Bioorg. Med.
Chem. 2001, 9, 19.
References and notes
1. Szallasi, A.; Blumberg, P. M. Pharmacol. Rev. 1999, 51, 159.
2. Tominaga, M.; Caterina, M. J.; Malmberg, A. B.; Rosen,
T. A.; Gilbert, H.; Skinner, K.; Raumann, B. E.; Bas-
baum, A. I.; Julius, D. Neuron 1998, 21, 531.
29. Lee, J.; Park, S.-U.; Kim, J.-Y.; Kim, J.-K.; Lee, J.; Oh,
U.; Marquez, V. E.; Beheshti, M.; Wang, Q. J.; Modarres,

J. Lee et al. / Bioorg. Med. Chem. 12 (2004) 1055–1069
1069
S.; Blumberg, P. M. Bioorg. Med. Chem. Lett. 1999, 9,
2909.
30. Lee, J.; Lee, J.; Kang, M.-S.; Kim, K.-P.; Chung, S.-J.;
Blumberg, P. M.; Yi, J.-B.; Park, Y. H. Bioorg. Med.
Chem. 2002, 10, 1171.
U.; Blumberg, P. M.; Lee, J. Mol. Pharmacol. 2002, 62,
947 (Published erratum appears in Mol. Pharmacol 63,
2003, 958.).
35. Wang, Y.; Toth, A.; Tran, R.; Szabo, T.; Welter, J. D.;
Blumberg, P. M.; Lee, J.; Kang, S.-U.; Lim, J.-O.; Lee, J.
Mol. Pharmacol. 2003, 64, 325.
36. Lal, B.; Pramanik, B. N.; Manhas, M. S.; Bose, A. K. Tet.
Lett. 1977, 23, 1977.
31. Walpole, C. S. J.; Bevan, S.; Bloomfield, G.; Breck-
enridge, R.; James, I. F.; Ritchie, T.; Szallasi, A.; Winter,
J.; Wrigglesworth, R. J. Med. Chem. 1996, 39, 2939.
32. This labelling is based on previous subdivision of capsai-
cin described in following reference: Walpole, C. S. J.;
Wrigglesworth, R.; Bevan, S.; Campbell, E. A.; Dray, A.;
James, I. F.; Perkins, M. N.; Reid, D. J.; Winter, J. J.
Med. Chem. 1993, 36, 2362.
33. Lee, J.; Lee, J.; Kang, M.; Shin, M.-Y.; Kim, J.-M.;
Kang, S.-U.; Lim, J.-O.; Choi, H.-K.; Suh, Y.-G.; Park,
H.-G.; Oh, U.; Kim, H.-D.; Park, Y.-H.; Ha, H.-J.; Kim,
Y.-H.; Toth, A.; Wang, Y.; Tran, R.; Pearce, L. V.;
Lundberg, D. J.; Blumberg, P. M. J. Med. Chem. 2003,
46, 3116.
37. Acs, G.; Lee, J.; Marquez, V. E.; Blumberg, P. M. Mol.
Brain Res. 1996, 35, 173.
38. Walpole, C. S. J.; Wrigglesworth, R.; Bevan, S.; Camp-
bell, E. A.; Dray, A.; James, I. F.; Masdin, K. J.; Perkins,
M. N.; Winter, J. J. Med. Chem. 1993, 36, 2373.
39. Walpole, C. S. J.; Bevan, S.; Bovermann, G.; Boelsterli,
J. J.; Breckenridge, R.; Davies, J. W.; Hughes, G. A.;
James, I.; Oberer, L.; Winter, J.; Wrigglesworth, R. J.
Med. Chem. 1994, 37, 1942.
40. Klopman, G.; Li, J.-Y. J. Comput.-Aid. Mol. Des 1995, 9,
283.
34. Wang, Y.; Szabo, T.; Welter, J. D.; Toth, A.; Tran, R.;
Lee, J.; Kang, S. U.; Lee, Y.-S.; Min, K. H.; Suh, Y.-G.;
Park, M.-K.; Park, H.-G.; Park, Y.-H.; Kim, H.-D.; Oh,
41. Kim, K. H. Bioorg. Med. Chem. 2002, 10, 1367.
42. Victory, S. F.; Appendino, G.; Vander Velde, D. G.
Bioorg. Med. Chem. 1998, 6, 223.