Structure-activity relationship and rational design of 3,4-dephostatin derivatives as protein tyrosine phosphatase inhibitors

Article abstract of DOI:10.1016/S0040-4020(99)01058-3

Several alkyl- and O-methylated-3,4-dephostatin were synthesized and evaluated for their inhibitory activity toward protein tyrosine phosphatase. Alkyl chains with a length up to that of the pentyl group gave tolerable inhibition, whereas methylation of h

Full text of DOI:10.1016/S0040-4020(99)01058-3

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 741–752
Structure–Activity Relationship and Rational Design of
3,4-Dephostatin Derivatives as Protein Tyrosine
Phosphatase Inhibitors
Takumi Watanabe,a Takayuki Suzuki,b Yoji Umezawa,a Tomio Takeuchi,a Masami Otsukac and
Kazuo Umezawab,
*
aInstitute of Microbial Chemistry, 3-14-23 Kamiosaki, Shinagawa-ku, Tokyo 141-0021, Japan
bDepartment of Applied Chemistry, Faculty of Science and Technology, Keio University, 3-14-1 Hiyoshi, Kohoku-ku,
Yokohama 223-0061, Japan
cFaculty of Pharmaceutical Sciences, Kumamoto University, 5-1 Oe-Honmachi, Kumamoto 862-0973, Japan
Received 5 November 1999; accepted 6 December 1999
Abstract—Several alkyl- and O-methylated-3,4-dephostatin were synthesized and evaluated for their inhibitory activity toward protein
tyrosine phosphatase. Alkyl chains with a length up to that of the pentyl group gave tolerable inhibition, whereas methylation of hydroxyl
groups resulted in a decrease in the activity. Based on the structure–activity relationship and X-ray crystallographic analysis of C215S
PTP1B–phosphotyrosine containing peptide complex, the mode of binding of 3,4-dephostatins to the active site was speculated with the aid
of calculation. Several hydrogen bonds and CH/p interactions were suggested to be important for inhibition of PTPase. A novel nitroso-free
methoxime compound was designed so that all the attractive interactions were maintained. The methoxime compound was synthesized and
shown to inhibit PTP1B. ᭧ 2000 Elsevier Science Ltd. All rights reserved.
Many biological phenomena including cellular prolifera-
tion, differentiation, and apoptosis are controlled by intra-
cellular signal transduction initiated by external stimuli.
Various proteins and low-molecular-weight compounds
take part in these signaling systems. Overexpression of the
specific signal proteins often causes diseases such as neo-
plasia and diabetes mellitus. Consequently, inhibitors of the
enzymes involved in such overexpression are considered to
be candidates for novel chemotherapeutic agents.
unique structural characteristics, i.e. a hydroquinone
skeleton with an N-methyl-N-nitrosoamino moiety. This
compound inhibits CD45 (receptor type PTPase) activity
comparably to sodium vanadate, and does not inhibit
serine/threonine phosphatases such as PP2A and PP2B.
However, an adequate amount of dephostatin cannot
be obtained from natural sources because of the low pro-
ductivity of microorganisms. Therefore, we accomplished
the first total synthesis of dephostatin.4 However, the
intrinsic lability of this compound hampered the practical
application of it to biological studies. To overcome this
drawback, we started to search for stable yet potent
dephostatin analogs. Substitution of the nitroso moiety to
a carboxamide or sulfonamide resulted in complete loss of
activity. The number and position of the phenolic hydroxyl
groups were also found to be important, for dephostatin lost
its activity upon the methylation or removal of hydroxyl
groups. Among the synthesized regio-isomers of dephostatin,
Protein-tyrosine kinases (PTKases) are typical signaling
molecules for cell growth, differentiation, and transforma-
tion. Therefore, protein-tyrosine phosphatases (PTPases),
enzymes having the opposite action, should be equally
important as PTKases for the regulation of these phenotypic
changes.1 However, the lack of potent and selective PTPase
inhibitor has retarded the further clarification of the
physiological function of PTPases. Currently, only sodium
vanadate, which also inhibits Ca2ϩ and Kϩ-dependent
ATPase, is frequently used as an inhibitor of PTPases.
In 1993, our group reported the isolation and structure deter-
mination of dephostatin (1) (Fig. 1)2,3 as the first naturally
occurring selective PTPase inhibitor. Dephostatin has
Keywords: enzyme inhibitors; structure–activity; molecular design;
molecular modeling/mechanics.
Figure 1. The structure of dephostatin and Me-3,4-dephostatin.
* Corresponding author; e-mail: umezawa@applc.keio.ac.jp
0040–4020/00/$ - see front matter ᭧ 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(99)01058-3

742
T. Watanabe et al. / Tetrahedron 56 (2000) 741–752
only the 3,4-dihydroxy isomer of dephostatin 2 (named
Me-3,4-dephostatin for convenience) inhibited the PTPase
activity of CD45.5 Me-3,4-dephostatin competes with the
substrate,6 and it is noteworthy that Me-3,4-dephostatin is
at least 10 times more stable than dephostatin in media used
for animal cell cultures.6
Recently, X-ray crystallographic analyses of many enzymes
and enzyme–inhibitor complexes have been reported, and it
is often possible to develop more potent inhibitors by
rational design based on the mechanistic interpretation.
X-Ray crystallography of some PTPases has already been
reported. Among them, the 3-D structure of C215S PTP1B
complexed with an O-phosphotyrosine should be useful for
speculating on the inhibition of PTPase by low-molecular-
weight compounds.32 This mutated PTP1B loses its catalytic
activity without a change in the structure of the active site.
In this paper, the roles of the N-alkyl moiety and the
phenolic hydroxyl groups of 3,4-dephostatin in its PTPase
inhibitory activity are described. Then, based on the
structure activity relationship and the 3-D structure of
PTP1B–O-phosphotyrosine complex,32 the most favorable
interaction of 3,4-dephostatin with the enzyme is calculated
computationally considering hydrogen bondings and CH/p
interactions.33,34 Finally, structurally novel and potent
PTPase inhibitors are designed and synthesized based on
such calculations.
Using Me-3,4-dephostatin, we demonstrated that PTPase
participates in nerve cell differentiation. Me-3,4-dephostatin
enhanced NGF or EGF-induced differentiation of rat
pheochromocytoma PC12 cells, possibly by sustaining the
downstream MAP kinase activity.7 PTPases such as
PTP1B8,9 and LAR10,11 are known to antagonize the effects
of insulin. PTP1B-knock out mice showed increased sensi-
tivity to insulin.12 Therefore, PTPase inhibitors may be
developed in the future as new therapeutics for neural
injury, neural diseases including Alzheimer’s and
Parkinson’s diseases, and diabetes mellitus.
Most of the PTPase inhibitors so far reported were rationally
designed based on the structure of peptides containing
phosphotyrosine. Phosphotyrosine mimetics having
a
nonhydrolyzable substituent such as phosphonate instead
of phosphate ester were frequently utilized to develop potent
and selective PTPase inhibitors. 4-Phosphono(difluoro-
methyl)-l-phenylalanine (F2Pmp) derivatives13 were
synthesized as phosphotyrosine mimetics that can bind to
the SH2 domains of various proteins such as phospha-
tidylinositol (PI) 3-kinase, Src and Grb2.14,15 F2Pmp works
as a PTPase inhibitor by taking the place of the phospho-
tyrosine residue of the peptidyl substrate. The resulting
peptides act as highly potent and selective PTPase inhibi-
tors. Mechanistic aspects of PTPase inhibitory activity of
F2Pmp-related compounds were thoroughly investi-
gated.16,17 In these studies, the fluorine atoms of the phos-
phonate moiety were shown to participate in hydrogen
bonding with the active site of PTPases. Many kinds of
F2Pmp-related substances were also synthesized.18,19
Besides F2Pmp, various phosphotyrosine mimetics includ-
ing O-malonyl tyrosine derivatives (OMT and FOMT),20,21
a thiophosphorylated tyrosine derivative,22 cinnamate
derivatives,23 and a salicylic acid carboxymethyl ether
derivative24 have been reported. In addition, some tyrosine-
based compounds inhibit PTPase activity by covalent bond
formation.25,26 Naturally occurring PTPase inhibitors such
as RK-682,27 aporphine alkaloids including anonaine and
nornuciferine,28 pulchellalactam,29 and phosphatoquinones
A and B30 have also been reported. PTPase inhibitors having
hitherto unknown structural characteristics were discovered
by the combinatorial synthesis approach.31 Although
various PTPase inhibitors as above have been reported,
sodium vanadate is still the most widely used to inhibit
PTPases in biological studies.
Synthesis of 3,4-Dephostatin Analogs
Previously we reported the synthesis of Me-3,4-dephostatin5
starting from 3,4-dimethoxyaniline. A disadvantage of this
route is that the step of trifluoroacetamide hydrolysis is
often accompanied by the simultaneous removal of the
silyl groups. The resulting fully deprotected product was
too unstable to isolate. Therefore, we looked for a more
efficient route for preparation of Me-3,4-dephostatin.
According to Wang’s synthesis of dephostatin,35 reductive
amination (HCHO and NaBH3CN) of the known aniline
derivative 336 gave monomethylated product 4. Although
the yield of this step was moderate, the unreacted starting
material could be recovered and reused. The resulting
compound 4 was identical to an intermediate of the
previously reported synthesis,5 from which the desired
Me-3,4-dephostatin can be obtained in two steps with no
difficulty (Scheme 1).
Other alkyl groups can easily be introduced by using
corresponding aldehydes. Succeeding nitrosation and
deprotection were accomplished by reported procedure to
give the requisite analogs. In this study, ethyl, butyl, pentyl,
and hexyl groups were introduced to investigate the
relationships between the bulkiness of the alkyl moiety
and the enzyme inhibitory activity (Scheme 2). Methyl
ether derivatives of Me-3,4-dephostatin (14a and 14b)
were also synthesized by the above reductive amination
from suitable starting materials. The dimethoxy compound
16 was obtained by a one step procedure (Scheme 3).
Scheme 1. Improved synthesis of Me-3,4-dephostatin.

T. Watanabe et al. / Tetrahedron 56 (2000) 741–752
743
Scheme 2. Synthesis of alkyl-3,4-dephostatin analogs.
Inhibition of PTPase by 3,4-Dephostatin Derivatives
substrate. After the reaction, the hydrolysis product, p-nitro-
phenol, was quantified by a colorimetric method.
The PTPase inhibitory activity of the synthesized 3,4-
dephostatin analogs was examined. In our previous study,4
we used crude CD45 prepared from Jurkat cell membranes
as the enzyme source. In this study, commercially available
PTP1B was used because the 3-D structure of PTP1B has
been elucidated. This enzyme is also known to be involved
in the regulation of insulin receptor activity. The purified
enzyme was used with p-nitrophenyl phosphate as the
Inhibition of PTPases by the various 3,4-dephostatin
derivatives is shown in Table 1. The length of the alkyl
chain influenced the inhibitory activity. Up to the pentyl
chain, inhibition comparable to that of Me-3,4-dephostatin
was observed. However, introduction of hexyl group
decreased the inhibitory activity. Hydroxyl groups were
shown to be essential for the inhibitory activity. Blocking
Scheme 3. Synthesis of methylated Me-3,4-dephostatin analogs.

744
T. Watanabe et al. / Tetrahedron 56 (2000) 741–752
Table 1. PTP1B inhibition by 3,4-dephostatin analogs
the short contacts between the CH groups and p systems in
protein structures registered in the protein data bank (PDB).
This program detects XH/p interaction (XˆC, N, O and S)
bases on bond length and angle parameters.
Compound
IC50 (mg/mL)
2
7a
0.52
0.58
7b
7c
7d
14a
14b
16
0.60
0.30
1.52
The crystal structure of C215S PTP1B–O-phosphotyrosine
complex was recently reported.32 Firstly, Me-3,4-dephosta-
tin was superimposed to O-phosphotyrosine in the C215S
PTP1B–substrate complex so that the anilinic portion of
Me-3,4-dephostatin overlapped with the phenolic part of
O-phosphotyrosine, as shown in Fig. 2. Then, the energeti-
cally minimized structure was searched for, in which all of
Ͼ100
48
Ͼ100
0.30
Sodium vanadate
˚
the coordinates of atoms of PTP1B more than 7 A apart
of the hydroxyl groups by methyl group apparently
decreased the activity. These results suggest that both
hydroxyl groups of 3,4-dephostatin participate in the
interaction with PTP1B, presumably by involvement in
hydrogen bonding, as discussed later.
from the ligand were fixed. The motion of atoms between
˚
5 and 7 A was set to be movable within limit by the soft-
˚
ware, whereas that of atoms less than 5 A distant was fully
movable. Fig. 3 shows the presumed binding mode between
PTP1B and Me-3,4-dephostatin. In the optimized structure,
the oxygen atom of the nitroso moiety tightly interacted
with the guanidium residue of R 221 by hydrogen bonding.
A similar interaction was observed in phosphate oxygen of
phosphotyrosine–C215S PTP1B complex. No important
interaction was observed between the anilinic nitrogen
atom and the active site.
Computational Analysis of 3,4-Dephostatin–PTPase
Interaction
Based on mutated PTP1B–phosphotyrosine complex,32 the
most favorable interaction between PTP1B and Me-3,4-
dephostatin derivatives was obtained by calculation using
the CVFF force field parameter. In protein–ligand
complexes, ionic interaction, hydrogen bonding, and hydro-
phobic effects are considered to be the major attractive non-
covalent interactions. In particular, hydrogen bonding and
ionic interaction have been widely considered to be
involved in molecular recognition, whereas the hydrophobic
effect is a far more vague concept.
The phenolic hydroxyl group at the 3-position formed a
hydrogen bond with the side chain of glutamine. Loss of
the PTPase inhibitory activity of 3-O-methylated Me-3,4-
dephostatin clearly indicated the necessity of the hydroxyl
group. On the other hand, the 4-OH group seems not to be
responsible for binding to the active site, and is presumed to
participate only in intramolecular hydrogen bonding.
However, since 4-O-methylation of Me-3,4-dephostatin
resulted in only weak inhibitory activity, the electronic
structure of 3,4-dephostatin may also be important for the
inhibition of PTPase. In fact, monohydroxy dephostatin
analog, 3-hydroxy-N-methyl-N-nitrosoaniline, showed no
PTPase inhibitory activity.5
Recently, CH/p interaction was revealed to have a critical
role in protein–ligand complexation and folding of proteins.
CH/p interaction is a hydrogen-bond-like weak attractive
force observed between the CH hydrogen and the p electron
system. The CH hydrogens and p rings in biomolecules
often create a large network of CH/p interactions. To
estimate the contribution of CH/p interaction to PTP1B
and 3,4-dephostatin related compounds, the CHPI program
was used.37 This program was written in order to search for
The methyl group on the nitrosamine moiety was
oriented inside the active site pocket. The active site
pocket seems to be deep enough to accommodate longer
alkyl chains. Fig. 4 shows the result of calculation of
Figure 2. Superimposition of Me-3,4-dephostatin to O-phosphotyrosine in the PTP1B–substrate complex. (Me-3,4-dephostatin is shown by bold lines, and
O-phosphotyrosine by plain lines. The phenolic moiety of O-phosphotyrosine is overlapped with the portion of Me-3,4-dephostatin.)

T. Watanabe et al. / Tetrahedron 56 (2000) 741–752
745
Figure 3. Stereoview of proposed binding mode between Me-3,4-dephostatin and PTP1B. (CH/p interactions are shown by dotted lines; and hydrogen bonding
by broken lines.)
PTP1B–pentyl-3,4-dephostatin superimposition. The pentyl
derivative showed potent inhibitory activity, but the hexyl
derivative apparently showed weaker activity. The long
alkyl chain participates in the CH/p with F 182, but a larger
alkyl group cannot be accommodated without weakening
the essential hydrogen bondings.
Firstly, the tetrazole derivative (20) depicted in Scheme 4
was designed. The N–NyN bondings on the tetrazole ring
were expected to be a good substitute for the N–NyO
bondings. As shown in Scheme 4, the amino group of 1738
was converted into an isocyano group by formylation and
subsequent dehydration using triphosgene.39 The tetrazole
ring was constructed by 1,3-dipolar cycloaddition of sodium
azide to the isocyano group40 in the presence of ammonium
chloride. In this step, concomitant deacetylation occurred to
afford the desired tetrazole derivative.
Design and Synthesis of New PTPase Inhibitors
However, 20 showed no activity possibly because of an
unexpected geometrical mismatch between the N–NyN
structure of tetrazole and the N–NyO structure of the
nitrosamine. Differences in the bond angles of the five-
membered ring and the open structure would affect
their activity. In fact, simple superimposition of the
tetrazole derivative into the active site of PTP1B
showed that the NyN bond of tetrazole largely deviated
from the NyO bond of Me-3,4-dephostatin (data not
shown).
According to the possible interaction mentioned above,
new PTPase inhibitors were designed and synthesized.
Since Et-3,4-dephostatin 7a was stable compared with
Me-3,4-dephostatin, it was used in a recent biological
study in our group for the development of new antidiabetic
agents. However, Et-3,4-dephostatin turned out to be
weakly mutagenic in the Salmonella mutation assay. The
mutagenicity should come from its nitrosamine moiety.
To eliminate this disadvantage, we designed novel PTPase
inhibitors having no nitrosamine structure.
Figure 4. Stereoview of proposed binding mode between pentyl-3,4-dephostatin and the active site of PTP1B. (CH/p interactions are shown by dotted lines;
and hydrogen bonding by broken lines.)

746
T. Watanabe et al. / Tetrahedron 56 (2000) 741–752
Scheme 4. Synthesis of tetrazole derivative.
Figure 5. Stereoview of expected binding mode between methoxime 22 and PTP1B. (CH/p interactions are shown by dotted lines; and hydrogen bonding by
broken lines.)
Then, we tried to design open chain alkoximes as the next
candidates. Fig. 5 shows the hypothesized mode of binding
of the designed methoxime compound (22). Both the
N–NyO bond of Me-3,4 dephostatin and the CyN–O
bond of methoxime are considered to have the character
of a partial double bond. Thus, each nitrogen–oxygen
bond should have similar electronic nature. Moreover,
Fig. 1 showed that the nitrogen atom attached to the benzene
ring of Me-3,4-dephostatin has no interaction with the
surrounding structure. Therefore, converting this nitrogen
atom to a carbon atom would likely be tolerable. The
oxygen atom of the methoxime moiety would be expected
to participate in the hydrogen bonding with the arginine side
chain. The distance from oxygen to arginine is longer than
in the case of 3,4-dephostatin, which may come from the
steric influence of the methyl group. However, their
bidentate binding would be expected to compensate for
the disadvantage in distance.
Condensation of 21 and methoxyamine hydrochloride in
the presence of sodium acetate gave the methoxime 22
(Scheme 5).41 The geometry of the resulting methoxime
was unequivocally confirmed as E by X-ray crystallographic
analysis, as shown in the Experimental section. As
expected, the methoxime inhibited PTP1B phosphatase
with an IC50 of 2.9 mg/mL, which was only five times higher
than that of Me-3,4-dephostatin (Table 2). The decrease in
activity may have come from the loss of CH/p interaction
between the methoxime group and the active site. There is
enough room for further elongation and functionalization of
the methoxy moiety of this compound. Therefore, suitable
designing of the alkoxy moiety to give an additional affinity
for the enzyme active site should potentiate its PTPase
inhibitory activity.
Conclusion
The role of the alkyl group on the nitrosamine moiety and
that of the hydroxyl groups of 3,4-dephostatin on inhibition
of PTP1B were studied. Elongation of the alkyl group newly
formed an additional CH/p interaction with the deeper side
of the active site of PTP1B. The two phenolic hydroxyl
Table 2. Inhibition of PTP1B by nitroso-free compounds
Compounds
IC50 (mg/mL)
20
22
47
2.9
Scheme 5. Synthesis of methoxime compound.

T. Watanabe et al. / Tetrahedron 56 (2000) 741–752
747
groups formed a hydrogen bond network with polar func-
tional groups of the side chains of PTP1B. Based on the
structure–activity relationship study, novel nitroso-free
PTPase inhibitors were designed and synthesized. The
synthetic procedure was simple, and a variety of alkoxime
compounds will be synthesized depending on the choice of
alkoxyamine. Design, synthesis, and evaluation of PTPase
inhibitory activity of further potent alkoximes having
additional contact with PTP1B are under way.
135 mg of NaBH3CN (2.15 mmol) in 2.5 mL of MeOH.
The solution was stirred for 24 h. After purification
(EtOAc/n-hexane 1:50), 589 mg of 5a (1.54 mmol) was
obtained as a colorless oil in 76% yield. IR (neat) nmax
2860, 1514, 1473, 1254, 1228, 839 cmϪ1
.
1H NMR
(400 MHz, CDCl3) d 6.65 (1H, d, Jˆ8.3 Hz), 6.17 (1H, d,
Jˆ2.4 Hz), 6.11 (1H, dd, Jˆ8.3, 2.4 Hz), 3.06, (2H, q,
Jˆ7.3 Hz), 1.22 (3H, t, Jˆ7.3 Hz), 0.98 (9H, s), 0.97 (9H,
s), 0.19 (6H, s), 0.15 (6H, s). 13C NMR (100 MHz, CDCl3) d
147.2, 143.3, 138.6, 121.6, 106.7, 106.0, 39.3, 26.0, 25.9,
18.5, 18.4, 15.0, Ϫ4.1, Ϫ4.2. HRFAB-MS (m/z) calcd for
C20H39NO2Si2, 381.2519; found, 381.2500 (Mϩ).
Experimental
Melting points were determined using a Yanagimoto micro
melting point apparatus and were uncorrected. IR spectra
were measured with a Horiba FT-210 spectrophotometer.
HRFAB-MS and FAB-MS were taken with a JEOL
JMS-SX 102. NMR spectra were recorded with a JEOL
JNM-EX-400 spectrometer. All reagents and solvents
were purified by standard methods. All reactions were
carried out under an argon atmosphere unless otherwise
stated.
3,4-Bis[(tert-butyldimethylsilyl)oxy]-N-butylaniline (5b).
The reaction was conducted with 738 mg of 3 (2.09 mmol),
184 mg of butyraldehyde (2.15 mmol), and 197 mg of
NaBH3CN (3.13 mmol) in 3.6 mL of THF. The solution
was stirred for 14 h. After purification (EtOAc/n-hexane
1:50), 699 mg of 5b (1.71 mmol) was obtained as a pale
yellow oil in 82% yield. IR (neat) nmax 2858, 1614, 1514,
1254, 1228, 906, 841 cmϪ1. 1H NMR (400 MHz, CDCl3) d
6.65 (1H, d, Jˆ8.8 Hz), 6.16 (1H, d, Jˆ2.4 Hz), 6.10 (1H,
dd, Jˆ8.8, 2.4 Hz), 3.01, (2H, t, Jˆ7.0 Hz), 1.57 (2H, m),
1.42 (2H, m), 0.99 (9H, s), 0.97 (9H, s), 0.95 (3H, t,
Jˆ7.3 Hz), 0.20 (6H, s), 0.15 (6H, s). 13C NMR
(100 MHz, CDCl3) d 147.2, 143.4, 138.5, 121.6, 106.7,
106.0, 44.6, 31.8, 26.1, 26.0, 20.3, 18.5, 18.4, 13.9, Ϫ4.1,
Ϫ4.2. HRFAB-MS (m/z) calcd for C22H43NO2Si2, 409.2832;
found, 409.2856 (Mϩ).
Synthesis of 3,4-dephostatin
3,4-Bis[(tert-butyldimethylsilyl)oxy]-N-methylaniline (4).
To a solution of 3 (469 mg, 1.28 mmol) in 1.6 mL of
MeOH was added 0.11 mL of formalin at 0ЊC, and the solu-
tion was stirred at the temperature for 30 min. Then,
NaBH3CN (80 mg, 1.28 mmol) was added portionwise to
the reaction mixture at 0ЊC, and the solution was stirred
for additional 3 h at room temperature. The solvent was
evaporated after TLC confirmation of cessation of the reac-
tion. The resulting residue was dissolved in CHCl3, and the
solution was washed with brine. The aqueous layer was
back extracted with CHCl3 twice. The combined organic
layers were dried over Na2SO4 and concentrated in vacuo.
The residue was purified by silica gel column chromato-
graphy by elution with EtOAc/n-hexane (1:40 to 1:10) to
give 170 mg of 4 (0.463 mmol, 35% yield) as a colorless
foam and recovered starting material 3 (0.194 mg, 0.548
mmol). The yield of 4 based on recovered 3 was 60%.
The obtained starting material could be reused for the same
reaction procedure. The physicochemical data of 4 were fully
identical to the reported one.5 Requisite 3,4-dephostatin was
afforded from 4 by the reported reaction scheme.5
3,4-Bis[(tert-butyldimethylsilyl)oxy]-N-pentylaniline
(5c). The reaction was conducted with 832 mg of 3
(2.35 mmol), 202 mg of valeraldehyde (2.35 mmol), and
222 mg of NaBH3CN (3.52 mmol) in 4.5 mL of MeOH.
The solution was stirred for 21 h. After purification
(EtOAc/n-hexane 1:40), 737 mg of 5c (1.74 mmol) was
obtained as an orange oil in 74% yield. IR (neat) nmax
2858, 1514, 1254, 1225, 906, 839 cmϪ1
.
1H NMR
(400 MHz, CDCl3) d 6.65 (1H, d, Jˆ8.3 Hz), 6.16 (1H, d,
Jˆ2.4 Hz), 6.10 (1H, dd, Jˆ8.3, 2.4 Hz), 3.00, (2H, t,
Jˆ7.1 Hz), 1.59 (2H, m), 1.40–1.30 (4H, m), 0.98 (9H, s),
0.97 (9H, s), 0.91 (3H, t, Jˆ7.1 Hz), 0.19 (6H, s), 0.15 (6H,
s). 13C NMR (100 MHz, CDCl3) d 147.2, 143.4, 138.5,
121.6, 106.7, 106.0, 44.9, 29.4, 26.1, 26.0, 22.6, 18.5,
18.4, 14.1, Ϫ4.1, Ϫ4.2. HRFAB-MS (m/z) calcd for
C23H45NO2Si2, 423.2989; found, 423.2966 (Mϩ).
3,4-Bis[(tert-butyldimethylsilyl)oxy]-N-hexylaniline (5d).
The reaction was conducted with 1.22 g of 3 (3.45 mmol),
351 mg of capronaldehyde (3.50 mmol), and 325 mg of
NaBH3CN (5.17 mmol) in 6 mL of MeOH. The solution
was stirred for 24 h. After purification (EtOAc/n-hexane
1:50), 1.17 g of 5d (2.67 mmol) was obtained as a colorless
oil in 77% yield. IR (neat) nmax 2858, 1514, 1471, 1254,
Synthesis of 3,4-dephostatin analogs with elongated alkyl
chain
Alkylation of compound 3. General procedure. Aldehyde
and 3 were dissolved in MeOH. To the solution was added
NaBH3CN at 0ЊC, and the reaction mixture was stirred at
room temperature. Then, the solution was evaporated, and
the resulting residue was dissolved in EtOAc and washed
with brine. The aqueous layer was extracted with EtOAc
twice. The combined organic layer was dried over Na2SO4
and concentrated in vacuo. The residue was purified by
silica gel column chromatography.
1
1223, 906, 839, 779 cmϪ1. H NMR (400 MHz, CDCl3) d
6.65 (1H, d, Jˆ8.3 Hz), 6.16 (1H, d, Jˆ2.9 Hz), 6.10 (1H,
dd, Jˆ8.3, 2.4 Hz), 3.00, (2H, t, Jˆ7.1 Hz), 1.62–1.55 (2H,
m), 1.34–1.30 (6H, m), 0.98 (9H, s), 0.97 (9H, s), 0.90 (3H,
t, Jˆ7.1 Hz), 0.19 (6H, s), 0.15 (6H, s). 13C NMR
(100 MHz, CDCl3) d 147.2, 143.4, 136.5, 121.5, 106.7,
106.0, 44.9, 31.7, 29.7, 26.9, 26.0, 22.6, 18.5, 18.4, 14.0,
Ϫ4.1, Ϫ4.2. HRFAB-MS (m/z) calcd for C24H47NO2Si2,
437.3145; found, 437.3167 (Mϩ).
3,4-Bis[(tert-butyldimethylsilyl)oxy]-N-ethylaniline (5a).
The reaction was conducted with 750 mg of
3
(2.15 mmol), 95 mg of acetaldehyde (2.15 mmol), and

748
T. Watanabe et al. / Tetrahedron 56 (2000) 741–752
Nitrosation of alkyl aniline compounds
obtained as a pale yellow oil in 93% yield. IR (neat) nmax
1
2860, 1514, 1464, 1255, 902, 841, 783 cmϪ1. H NMR
(400 MHz, CDCl3) d 7.03 (1H, d, Jˆ1.4 Hz), 6.92 (1H,
dd, Jˆ8.8, 1.4 Hz), 6.89 (1H, d, Jˆ8.8 Hz), 3.95, (2H, t,
Jˆ7.8 Hz), 1.53–1.47 (4H, m), 1.32–1.22 (4H, m), 1.00
(9H, s), 0.99 (9H, s), 0.86 (3H, t, Jˆ7.3 Hz), 0.23 (6H, s),
0.22 (6H, s). 13C NMR (100 MHz, CDCl3) d 147.5, 146.6,
135.4, 121.2, 113.6, 113.1, 44.5, 31.3, 26.7, 26.4, 25.9, 22.5,
18.5, 14.0, Ϫ4.1. HRFAB-MS (m/z) calcd for
C24H47N2O3Si2, 467.3125; found, 467.3151 (MHϩ).
General procedure. To a THF solution of 5 were added 1 N
HCl and NaNO2 at 0ЊC successively, the solution was then
stirred at that temperature for 30 min. Next, THF was
evaporated, and the resulting aqueous mixture was extracted
with EtOAc three times. The combined organic layers were
washed with brine, dried over Na2SO4, and concentrated in
vacuo. The residue was purified by silica gel column
chromatography.
3,4-Bis[(tert-butyldimethylsilyl)oxy]-N-ethyl-N-nitroso-
aniline (6a). The reaction was conducted with 248 mg of 5a
(0.651 mmol), 2.6 mL of 1 N HCl, and 49.4 mg of NaNO2
(0.716 mmol) in 13 mL of THF. After purification (EtOAc/
n-hexane 1:10), 267 mg of 6a (0.651 mmol) was obtained as
a pale yellow oil in quantitative yield. IR (neat) nmax 2860,
Final deprotection
General procedure. To a solution of nitrosated compound
was added HF–NaF buffer (pH 4.73) at 0ЊC. The solution
was stirred at room temperature. After confirming the dis-
appearance of the substrate, THF was evaporated. The
resulting mixture was diluted with brine and extracted
with CHCl3. The combined organic layers were dried over
Na2SO4 and concentrated in vacuo. The residue was purified
by silica gel column chromatography or preparative silica
gel chromatography.
1
1514, 1471, 1298, 1255, 901 cmϪ1. H NMR (400 MHz,
CDCl3) d 7.05 (1H, d, Jˆ2.4 Hz), 6.92 (1H, dd, Jˆ8.3,
2.4 Hz), 6.89 (1H, d, Jˆ8.3 Hz), 4.01, (2H, q, Jˆ7.3 Hz),
1.15 (3H, t, Jˆ7.3 Hz), 1.01 (9H, s), 1.00 (9H, s), 0.23 (6H,
s), 0.22 (6H, s). 13C NMR (100 MHz, CDCl3) d 147.5,
146.6, 135.1, 121.2, 113.5, 112.9, 39.7, 25.9, 18.4, 11.7,
Ϫ4.1. HRFAB-MS (m/z) calcd for C20H39N2O3Si2,
411.2499; found, 411.2509 (MHϩ).
3,4-Dihydroxy-N-ethyl-N-nitrosoaniline (7a). The reac-
tion was conducted with 2.30 g of 6a (5.60 mmol) and
25 mL of buffer in 100 mL of THF for 20 h. After purifica-
tion (EtOAc/n-hexane 4:1), 479 mg of 7a (2.63 mmol) was
obtained as a light brown powder in 47% yield: mp 124–
126ЊC. IR (KBr) nmax 1604, 1533, 1463, 1248, 891 cmϪ1. 1H
NMR (400 MHz, acetone-d6) d 8.23 (2H, br), 7.10 (1H, d,
Jˆ2.4 Hz), 6.95 (1H, d, Jˆ8.8 Hz), 6.89 (1H, dd, Jˆ8.8,
2.4 Hz), 4.01, (2H, q, Jˆ7.3 Hz), 1.08 (3H, t, Jˆ7.3 Hz).
13C NMR (100 MHz, acetone-d6) d 147.8, 145.7, 135.1,
116.3, 113.0, 109.3, 40.1, 11.9. HRFAB-MS (m/z) calcd
for C8H11N2O3, 183.0770; found, 183.0792 (MHϩ).
3,4-Bis[(tert-butyldimethylsilyl)oxy]-N-butyl-N-nitroso-
aniline (6b). The reaction was conducted with 123 mg of 5b
(0.307 mmol), 2.6 mL of 1 N HCl, and 23.5 mg of NaNO2
(0.341 mmol) in 14 mL of THF. After purification (EtOAc/
n-hexane 1:50), 118 mg of 6b (0.269 mmol) was obtained as
a pale yellow oil in 87% yield. IR (neat) nmax 2860, 1514,
1471, 1255, 904, 840 cmϪ1. 1H NMR (400 MHz, CDCl3) d
7.04 (1H, d, Jˆ2.4 Hz), 6.92 (1H, dd, Jˆ8.8, 2.4 Hz), 6.89
(1H, d, Jˆ8.8 Hz), 3.96, (2H, t, Jˆ7.3 Hz), 1.51 (2H, m),
1.30 (2H, m), 1.00 (9H, s), 0.99 (9H, s), 0.90 (3H, t,
Jˆ7.3 Hz), 0.24 (6H, s), 0.23 (6H, s). 13C NMR
(100 MHz, CDCl3) d 147.5, 146.6, 135.3, 121.1, 113.6,
113.0, 48.6, 44.2, 28.5, 25.9, 20.3, 18.4, 13.6, Ϫ4.1.
HRFAB-MS (m/z) calcd for C22H43N2O3Si2, 439.2812;
found, 439.2785 (MHϩ).
3,4-Dihydroxy-N-butyl-N-nitrosoaniline (7b). The reac-
tion was conducted with 104 mg of 6b (0.237 mmol) and
1.8 mL of buffer in 4.2 mL of THF for 24 h. After purifica-
tion (EtOAc/n-hexane 1:3 to 1:1), 28.1 mg of 7b
(0.134 mmol) was obtained as a gray powder in 56%
yield: mp 100–102ЊC. IR (KBr) nma1x 1618, 1537, 1467,
1271, 1174, 1101, 872, 795 cmϪ1. H NMR (400 MHz,
CDCl3) d 9.71 (1H, br), 7.10 (1H, d, Jˆ2.4 Hz), 6.98 (1H,
d, Jˆ8.8 Hz), 6.78 (1H, dd, Jˆ8.8, 2.4 Hz), 5.89 (1H, br),
4.11, (2H, t, Jˆ7.3 Hz), 1.59 (2H, m), 1.34 (2H, m), 0.92
(3H, t, Jˆ7.3 Hz). 13C NMR (100 MHz, CDCl3) d 145.4,
145.0, 132.3, 114.5, 110.7, 107.9, 45.3, 28.3, 20.3, 13.5.
HRFAB-MS (m/z) calcd for C10H15N2O3, 211.1083; found,
211.1091 (MHϩ).
3,4-Bis[(tert-butyldimethylsilyl)oxy]-N-pentyl-N-nitroso-
aniline (6c). The reaction was conducted with 89.9 mg of 5c
(0.189 mmol), 1.5 mL of 1 N HCl, and 14.3 mg of NaNO2
(0.207 mmol) in 9 mL of THF. After purification (EtOAc/n-
hexane 1:3), 76.1 mg of 6c (0.168 mmol) was obtained as an
orange amorphous in 89% yield. IR (neat) nmax 2858, 1514,
1
1471, 1255, 1095, 903, 838 cmϪ1. H NMR (400 MHz,
CDCl3) d 7.03 (1H, d, Jˆ2.4 Hz), 6.92 (1H, dd, Jˆ8.3,
2.4 Hz), 6.89 (1H, d, Jˆ8.3 Hz), 3.95, (2H, t, Jˆ7.5 Hz),
1.53 (2H, m), 1.35–1.22 (4H, m), 1.00 (9H, s), 0.99 (9H, s),
0.87 (3H, t, Jˆ7.1 Hz), 0.22 (12H, s). 13C NMR (100 MHz,
CDCl3) d 147.5, 146.6, 135.4, 121.2, 113.6, 113.0, 44.5,
29.2, 26.1, 25.9, 25.8, 22.2, 18.5, 13.8, Ϫ4.08, Ϫ4.13.
HRFAB-MS (m/z) calcd for C23H45N2O3Si2, 453.2969;
found, 453.2975 (MHϩ).
3,4-Hydroxy-N-pentyl-N-nitrosoaniline (7c). The reac-
tion was conducted with 75.0 mg of 7c (0.166 mmol) and
1.3 mL of buffer in 3.5 mL of THF for 61 h. After purifica-
tion (EtOAc/n-hexane 1:3), 20.6 mg of 7c (91.7 mmol) was
obtained as a white powder in 55% yield: mp 101–103ЊC.
IR (KBr) nmax 1616, 1535, 1209, 1176, 1105, 872,
1
794 cmϪ1. H NMR (400 MHz, CDCl3) d 7.46 (1H, d,
3,4-Bis[(tert-butyldimethylsilyl)oxy]-N-hexyl-N-nitroso-
aniline (6d). The reaction was conducted with 96.9 mg of
5d (0.221 mmol), 2.3 mL of 1 N HCl, and 21.3 mg of
NaNO2 (0.309 mmol) in 15 mL of THF. After purification
(EtOAc/n-hexane 1:10), 95.7 mg of 6d (0.205 mmol) was
Jˆ2.9 Hz), 6.97 (1H, d, Jˆ8.8 Hz), 6.80 (1H, dd, Jˆ8.8,
2.9 Hz), 4.08, (2H, t, Jˆ7.8 Hz), 1.60 (2H, m), 1.31 (4H, m),
0.88 (3H, t, Jˆ6.8 Hz). 13C NMR (100 MHz, CDCl3) d
145.1, 144.9, 133.9, 114.6, 110.9, 107.9, 45.3, 29.1, 26.0,

T. Watanabe et al. / Tetrahedron 56 (2000) 741–752
749
22.2, 13.8. HRFAB-MS (m/z) calcd for C11H17N2O3,
3-tert-Butyldimethylsilyloxy-4-methoxy-N-methyl-N-
nitrosoaniline (13a). To a solution of 839 mg of 12a
(3.14 mmol) in 135 mL of THF were successively added
27 mL of 1N HCl and 237 mg of NaNO2 (3.43 mmol) at
0ЊC. The solution was stirred for 30 min at 0ЊC. Then,
THF was evaporated, and the resulting mixture was
extracted with EtOAc three times. The combined organic
layers were washed with brine, dried over Na2SO4, and
concentrated in vacuo. The residue was purified by silica
gel column chromatography (EtOAc/n-hexane 1:10) to give
931 mg of 13a (3.14 mmol) as a yellow powder in quanti-
tative yield: mp 56–57ЊC. IR (KBr) nmax 1593, 1515, 1425,
225.1239; found, 225.1218 (MHϩ).
3,4-Dihydoxy-N-hexyl-N-nitrosoaniline (7d). The reac-
tion was conducted with 94.0 mg of 6d (0.201 mmol) and
1.5 mL of buffer in 4.5 mL of THF. After purification
(EtOAc/n-hexane 1:1), 23.8 mg of 7d (94.3 mmol) was
obtained as a light brown powder in 47% yield: mp 104–
106ЊC. IR (KBr) nmax 1616, 1537, 1466, 1377, 1273, 1174,
1
1105, 871, 793 cmϪ1. H NMR (400 MHz, CDCl3) d 7.47
(1H, d, Jˆ2.4 Hz), 6.97 (1H, d, Jˆ8.8 Hz), 6.89 (1H, dd,
Jˆ8.8, 2.4 Hz), 4.09 (2H, t, Jˆ7.8 Hz), 1.65–1.55 (2H, m),
1.35–1.25 (6H, m), 0.87 (3H, t, Jˆ7.3 Hz). 13C NMR
(100 MHz, CDCl3) d 145.3, 145.0, 133.7, 114.6, 110.8,
107.9, 45.5, 31.2, 26.7, 26.3, 22.4, 13.9. HRFAB-MS
(m/z) calcd for C12H19N2O3, 239.1396; found, 239.1413
(MHϩ).
1
1267 cmϪ1. H NMR (400 MHz, CDCl3) d 7.10 (1H, d,
Jˆ2.9 Hz), 7.02 (1H, dd, Jˆ8.8, 2.9 Hz), 6.91 (1H, d,
Jˆ8.8 Hz), 3.85 (3H, s), 3.42 (3H, s), 1.01 (9H, s), 0.19
(6H, s). 13C NMR (100 MHz, CDCl3) d 150.5, 145.7,
135.8, 113.0, 112.7, 112.0, 55.7, 32.0, 25.6, 18.4, Ϫ4.7.
HRFAB-MS (m/z) calcd for C14H25N2O3Si, 297.1634;
found, 297.1638 (MHϩ).
Synthesis of methoxy derivatives
3-Hydroxy-4-methoxy-N-methyl-N-nitrosoaniline (14a).
To a solution of 305 mg of 13a (1.03 mmol) in 20 mL of
THF was added 3.9 mL of HF–NaF buffer (pH 4.73) at
0ЊC. The solution was stirred for 16 h at room temperature.
Then, THF was evaporated, and the resulting mixture was
extracted with EtOAc. The aqueous layer was extracted
with EtOAc twice. The combined organic layers were
washed with brine, dried over Na2SO4, and concentrated
in vacuo. The residue was purified by silica gel column
chromatography (EtOAc/n-hexane 1:3) to give 188 mg of
14a (1.03 mmol) as a yellow powder in quantitative yield:
mp 108–109ЊC. IR (KBr) nmax 1601, 1535, 1425, 1367,
3-tert-Butyldimethylsilyloxy-4-methoxyaniline (11a). To
a solution of 5.14 g of 8 (36.9 mmol) and 10.55 g of
imidazole (155 mmol) in 120 mL of DMF was added
8.01 g of TBSCl (73.7 mmol) at 0ЊC. The solution was
stirred for 5 h at room temperature. Then, brine was poured
onto the solution, and the resulting mixture was extracted
with toluene three times. The combined organic layers were
dried over Na2SO4 and concentrated in vacuo. The residue
was purified by silica gel column chromatography (EtOAc/
n-hexane 1:6) to give 9.35 g of 11a (36.9 mmol) as a color-
less oil in quantitative yield. IR (neat) nmax 1728, 1601,
1
1
1425, 1363, 1219, 1144, 1026 cmϪ1. H NMR (400 MHz,
1219, 1144, 1026 cmϪ1. H NMR (400 MHz, CDCl3) d
CDCl3) d 6.68 (1H, d, Jˆ8.3 Hz), 6.28 (1H, d, Jˆ2.9 Hz),
6.25 (1H, dd, Jˆ8.3, 2.4 Hz), 3.72 (3H, s), 3.37 (2H, br),
0.99 (9H, s), 0.15 (6H, s). 13C NMR (100 MHz, CDCl3) d
146.0, 144.3, 140.6, 114.3, 109.5, 108.2, 56.6, 25.7, 18.4,
Ϫ4.7. HRFAB-MS (m/z) calcd for C13H23NO2Si, 253.1498;
found, 253.1516 (Mϩ).
7.14 (1H, d, Jˆ2.4 Hz), 7.01 (1H, dd, Jˆ8.8, 2.4 Hz),
6.92 (1H, d, Jˆ8.8 Hz), 5.78 (1H, s), 3.95 (3H, s), 3.42
(3H, s). 13C NMR (100 MHz, CDCl3) d 146.3, 146.0,
136.4, 111.2, 110.1, 106.8, 56.3, 32.0. HRFAB-MS (m/z)
calcd for C8H11N2O3, 183.0770; found, 183.0772 (MHϩ).
4-tert-Butyldimethylsilyloxy-3-methoxy-nitrobenzene
(10). To a solution of 2.00 g of 9 (11.8 mmol) and 2.41 g of
imidazole (35.4 mmol) in 50 mL of DMF was added 2.67 g
of TBSCl (17.7 mmol) at 0ЊC. The solution was stirred for
19 h at room temperature. Then, brine was poured onto the
solution, and the resulting mixture was extracted with
toluene three times. The combined organic layers were
dried over Na2SO4 and concentrated in vacuo. The residue
was purified by silica gel column chromatography (EtOAc/
n-hexane 1:40) to give 3.16 g of 10 (11.8 mmol) as a pale
yellow powder in quantitative yield: mp 41–43ЊC. IR (KBr)
3-tert-Butyldimethylsilyloxy-4-methoxy-N-methylaniline
(12a). To a solution of 1.79 g of 11a (7.06 mmol) in 7.5 mL
of DMF was added 0.64 mL of formalin (7.59 mmol) at 0ЊC.
The solution was stirred for 30 min at 0ЊC. Then, 477 mg of
NaBH3CN (7.59 mmol) was added to the solution, and the
mixture was stirred for 3 h at room temperature. After
confirmation of the cessation of the reaction, MeOH was
evaporated, and the resulting mixture was extracted with
EtOAc. The organic layer was washed with brine. The
aqueous layer was extracted with EtOAc twice. The
combined organic layers were dried over Na2SO4 and
concentrated in vacuo. The residue was purified by silica
gel column chromatography (EtOAc/n-hexane 1:15 to 1:5)
to give 548 mg of 12a (1.85 mmol) and 840 mg of unreacted
substrate 11a (3.14 mmol), the former obtained as a color-
less foam in 30% yield (53% yield based on recovered 11a).
nmax 1585, 1508, 1466, 1352, 1298, 1233, 1097, 1026 cmϪ1
.
1H NMR (400 MHz, CDCl3) d 7.81 (1H, dd, Jˆ8.8, 2.9 Hz),
7.74 (1H, d, Jˆ2.9 Hz), 6.89 (1H, d, Jˆ8.8 Hz), 3.89 (3H,
s), 1.01 (9H, s), 0.20 (6H, s). 13C NMR (100 MHz, CDCl3) d
151.5, 150.8, 142.1, 120.1, 117.6, 107.3, 55.7, 25.6, 18.5,
Ϫ4.5. HRFAB-MS (m/z) calcd for C13H22NO4Si, 284.1318;
found, 284.1324 (MHϩ).
1
IR (neat) nmax 1616, 1589, 1515, 1471 cmϪ1. H NMR
(400 MHz, CDCl3) d 6.74 (1H, d, Jˆ8.3 Hz), 6.22 (1H, d,
Jˆ2.9 Hz), 6.19 (1H, dd, Jˆ8.3, 2.4 Hz), 3.73 (3H, s), 2.78
(3H, s), 1.00 (9H, s), 0.16 (6H, s). 13C NMR (100 MHz,
CDCl3) d 146.3, 144.4, 143.5, 114.7, 106.9, 105.1, 56.8,
31.4, 25.7, 18.4, Ϫ4.7. HRFAB-MS (m/z) calcd for
C14H25NO2Si, 267.1655; found, 267.1649 (MHϩ).
4-tert-Butyldimethylsilyloxy-4-methoxyaniline (11b). To
a solution of 3.12 g of 10 (11.7 mmol) was added 750 mg of
10% Pd/C and the mixture was stirred under H2 atmosphere
for 21 h at room temperature. Catalyst was filtered out
through celite and the filtrate was concentrated in vacuo.

750
T. Watanabe et al. / Tetrahedron 56 (2000) 741–752
The residue was purified by silica gel column chromato-
graphy (EtOAc/n-hexane 1:4) to give 2.36 g of 11b
(9.31 mmol) as an amber powder in 80% yield: mp 39–
40ЊC. IR (KBr) nmax 1632, 1593. 1514, 1236, 1169, 1122,
washed with brine, dried over Na2SO4, and concentrated
in vacuo. The residue was purified by silica gel column
chromatography (EtOAc/n-hexane 1:3) to give 61.4 mg of
14a (0.337 mmol) as a yellow powder in quantitative yield:
mp 42–44ЊC. IR (KBr) nmax 1618, 1516, 1259, 1122,
1
1032 cmϪ1. H NMR (400 MHz, CDCl3) d 6.65 (1H, d,
1
Jˆ8.3 Hz), 6.26 (1H, d, Jˆ2.4 Hz), 6.16 (1H, dd, Jˆ8.3,
2.9 Hz), 3.75 (3H, s), 3.42 (2H, br), 0.98 (9H, s), 0.12 (6H,
s). 13C NMR (100 MHz, CDCl3) d 151.4, 140.9, 137.6,
121.3, 107.0, 100.9, 53.4, 25.8, 18.4, Ϫ4.4. HRFAB-MS
(m/z) calcd for C13H23NO2Si, 253.1498; found, 253.1515
(Mϩ).
1093 cmϪ1. H NMR (400 MHz, CDCl3) d 7.21 (1H, d,
Jˆ2.4 Hz), 7.00 (1H, d, Jˆ8.8 Hz), 6.92 (1H, dd, Jˆ8.8,
2.4 Hz), 5.68 (1H, s), 3.95 (3H, s), 3.45 (3H, s). 13C NMR
(100 MHz, CDCl3) d 147.0, 145.2, 135.4, 114.4, 112.2,
103.4, 56.2, 32.2. HRFAB-MS (m/z) calcd for C8H11N2O3,
183.0770; found, 183.0776 (MHϩ).
4-tert-Butyldimethylsilyloxy-3-methoxy-N-methylaniline
(12b). To a solution of 2.34 g of 11b (9.23 mmol) in 10 mL
of MeOH was added 0.88 mL of formalin (9.23 mmol) at
0ЊC. The solution was stirred for 30 min at 0ЊC. Then,
659 mg of NaBH3CN (10.5 mmol) was added to the
solution, and the mixture was stirred for 2.5 h at room
temperature. After confirmation of the cessation of the reac-
tion, MeOH was evaporated, and the resulting mixture was
extracted with EtOAc. The organic layer was washed with
brine. The aqueous layer was extracted with EtOAc twice.
The combined organic layers were dried over Na2SO4 and
concentrated in vacuo. The residue was purified by silica gel
column chromatography (EtOAc/n-hexane 1:10 to 1:3) to
give 738 mg of 12b (2.76 mmol) and 1.19 g of unreacted
substrate 11b (4.70 mmol), the former obtained as a color-
less foam in 30% yield (61% yield based on recovered 11b).
IR (neat) nmax 1616, 1589, 1514, 1471, 1236, 1120,
3,4-Dimethoxy-N-methyl-N-nitrosoaniline (16). To a
solution of 635 mg of 1542 (2.57 mmol) in 150 mL of THF
were successively added 31 mL of 1 N HCl and 287 mg of
NaNO2 (4.16 mmol) at 0ЊC. The solution was stirred for
30 min at 0ЊC. Then, THF was evaporated, and the resulting
mixture was extracted with EtOAc three times. The
combined organic layers were washed with brine, dried
over Na2SO4, and concentrated in vacuo. The residue was
purified by silica gel column chromatography (EtOAc/n-
hexane 1:6) to give 711 mg of 16 (3.62 mmol) as a yellow
powder in 95% yield: mp 83–84ЊC. IR (KBr) nmax 1597,
1
1518, 1471, 1280, 1255, 1234, 1180, 1040, 1022 cmϪ1. H
NMR (400 MHz, CDCl3) d 7.22 (2H, d, Jˆ1.5 Hz), 6.94
(1H, d, Jˆ1.5 Hz), 3.93 (3H, s), 3.93 (3H, s), 3.45 (3H, s).
13C NMR (100 MHz, CDCl3) d 149.6, 148.5, 135.9, 111.1,
111.1, 103.8, 56.1, 56.1, 31.9. HRFAB-MS (m/z) calcd for
C9H13N2O3, 197.0926; found, 197.0931 (MHϩ).
1
1036 cmϪ1. H NMR (400 MHz, CDCl3) d 6.70 (1H, d,
Jˆ8.3 Hz), 6.19 (1H, d, Jˆ2.4 Hz), 6.09 (1H, dd, Jˆ8.3,
2.4 Hz), 3.77 (3H, s), 2.80 (3H, s), 0.98 (9H, s), 0.12 (6H, s).
13C NMR (100 MHz, CDCl3) d 151.5, 144.6, 136.9, 121.2,
103.8, 98.5, 55.4, 31.4, 25.7, 18.4, Ϫ4.8. HRFAB-MS (m/z)
calcd for C14H25NO2Si, 267.1655; found, 267.1670 (Mϩ).
PTP1B inhibition assay
This PTP1B assay was carried out with p-nitrophenyl phos-
phate (pNPP) as the substrate in 96-well plates at 37ЊC. The
assay mixture containing 5 mL of 40 mM NiCl2 in H2O,
5 mL of a bovine serum albumin solution (5 mg/mL in
H2O), 10 mL of PTP1B-agarose (Upstate Biotechnology,
Lake Placid, NY), 53 mL of 50 mM Tris–HCl buffer (pH
7.0)/0.1 mM CaCl2, and 2 mL of various concentrations of
inhibitor in assay buffer (50 mM Tris–HCl, 0.1 mM CaCl2,
pH 7.0) (Upstate Biotechnology, Lake Placid, NY), was pre-
incubated for 15 min. The enzyme reaction was started by
the addition of 125 mL of pNPP solution (1.5 mg/mL in
50 mM Tris–HCl buffer). After 10 min incubation, the reac-
tion was stopped by adding 20 mL of 13% (w/v) K2HPO4
solution, and the absorbance at 405 nm was then measured.
4-tert-Butyldimethylsilyloxy-3-methoxy-N-methyl-N-
nitrosoaniline (13b). To a solution of 650 mg of 12b
(2.57 mmol) in 120 mL of THF were successively added
21 mL of 1 N HCl and 194 mg of NaNO2 (2.81 mmol) at
0ЊC. The solution was stirred for 30 min at 0ЊC. Then, THF
was evaporated, and the resulting mixture was extracted
with EtOAc three times. The combined organic layers
were washed with brine, dried over Na2SO4, and concen-
trated in vacuo. The residue was purified by silica gel
column chromatography (EtOAc/n-hexane 1:9) to give
590 mg of 13a (1.99 mmol) as an orange oil in 77% yield.
IR (neat) nmax 1594, 1514, 1471, 1392, 1342, 1280, 1255,
1180, 1124, 1090, 1034 cmϪ1. 1H NMR (400 MHz, CDCl3)
d 7.18 (1H, d, Jˆ2.4 Hz), 6.92 (1H, d, Jˆ8.3 Hz), 6.85 (1H,
dd, Jˆ8.3, 2.4 Hz), 3.86 (3H, s), 3.44 (3H, s), 1.01 (9H, s),
0.19 (6H, s). 13C NMR (100 MHz, CDCl3) d 151.5, 144.6,
136.5, 120.8, 111.3, 104.2, 55.5, 31.9, 25.6, 18.4, Ϫ4.7.
HRFAB-MS (m/z) calcd for C14H25N2O3Si, 297.1634;
found, 297.1635 (MHϩ).
Molecular modeling study
Initial structures of PTP1B–ligand complexes were gener-
ated by superimposition of the ligand benzene ring and the
benzene ring of phosphotyrosine reported from X-ray
analysis of mutated PTP1B-phosphotyrosine-containing
substrate.32 Coordinates of the nitroso and oxime moiety
of the ligand were fixed on the plane of its benzene ring.
The optimization was carried out by molecular mechanics
energy minimization with program Insight II 99.0/Discover
3.0. The force-field parameters of CVFF were used, with
coordinates of all heavy atoms of PTP1B fixed. CH/p inter-
actions were detected by the CHPI program as reported
previously.43
4-Hydroxy-3-methoxy-N-methyl-N-nitrosoaniline (14b).
To a solution of 100 mg of 13b (0.337 mmol) in 6 mL of
THF was added 1.4 mL of HF–NaF buffer (pH 4.73) at 0ЊC.
The solution was stirred for 14 h at room temperature. Then,
THF was evaporated, and the resulting mixture was
extracted with EtOAc. The aqueous layer was extracted
with EtOAc twice. The combined organic layers were

T. Watanabe et al. / Tetrahedron 56 (2000) 741–752
751
Synthesis of tetrazole compound
HRFAB-MS (m/z) calcd for C9H9N4O2, 179.0569; found,
179.0569 (MHϩ).
N-(3,4-Diacetoxyphenyl)-formamide (18). To a solution of
1.20 mL of formic acid (32.0 mmol) in 60 mL of CH2Cl2
was added 2.74 g of 1-(3-dimethylaminopropyl)-3-ethyl-
carbodiimide hydrochloride (EDCI) (16.0 mmol) at 0ЊC,
and the solution was stirred for 30 min at that temperature.
Then, 2.76 mL of diisopropyl ethylamine (8.00 mmol) and a
solution of 1.67 g of 17 (8.00 mmol) in 60 mL of CH2Cl2
were successively added to the reaction mixture at 0ЊC, and
the solution was stirred for 4 h at room temperature. Brine
was poured onto the solution and the aqueous layer was
extracted with CHCl3 twice. The combined organic layers
were dried over Na2SO4 and concentrated in vacuo. The
residue was purified by silica gel column chromatography
(EtOAc/n-hexane 1:9) to give 1.43 g of 18 (6.03 mmol) as a
colorless oil in 76% yield. This compound was obtained as a
mixture of cis- and trans-isomers. IR (KBr) nmax 1772,
1745, 1697, 1685, 1541, 1525, 1427, 1234, 1195,
Synthesis of methoxime compound
N-Methoxy-3,4-dihydroxybenzaldimine (22). To a solu-
tion of 275 mg of 21 (1.99 mmol) in 1.8 mL of EtOH–H2O
(2:1) were added 264 mg of sodium acetate and 183 mg of
O-methylhydroxylamine hydrochloride, and the solution
was stirred for 20 h at room temperature. Then, EtOH was
evaporated, and the resulting mixture was extracted by
CHCl3 three times. The combined organic layers were
dried over Na2SO4 and concentrated in vacuo. The residue
was purified by silica gel column chromatography (EtOAc/
n-hexane 2:1) to give 226 mg of 22 (1.35 mmol) as a yellow
powder in 68% yield: mp 83–84ЊC. IR (KBr) nmax 1608,
1595, 1525, 1438, 1273, 1200, 1057, 820, 764 cmϪ1. H
1
NMR (400 MHz, acetone-d6) d 7.95 (1H, s), 7.20 (1H, d,
Jˆ2.0 Hz), 6.98 (1H, dd, Jˆ8.3, 2.0 Hz), 6.86 (1H, d,
Jˆ8.3 Hz), 5.61 (1H, br), 5.48 (1H, br), 3.94 (3H, s). 13C
NMR (100 MHz, CDCl3) d 114.4, 145.7, 143.6, 125.2,
121.6, 115.4, 112.8, 61.8. HRFAB-MS (m/z) calcd for
C8H10NO3, 168.0661; found, 168.0663 (MHϩ) (Fig. 6).
1
1168 cmϪ1. H NMR (400 MHz, CDCl3) d 8.60 (0.35H, d,
Jˆ11.2 Hz), 8.20 (0.35H, br), 8.16 (1H, d, Jˆ1.5 Hz), 7.84
(1H, br), 7.58 (1H, d, Jˆ2.4 Hz), 7.17 (1H, dd, Jˆ8.8,
2.4 Hz), 7.14 (0.35H, d, Jˆ1.5 Hz), 7.06 (1H, d,
Jˆ8.8 Hz), 6.93 (0.7H, m), 2.30 (2.1H, s), 2.29 (3H, s),
2.29 (2.1H, s), 2.28 (3H, s). 13C NMR (100 MHz, CDCl3)
d 168.7, 168.5, 168.1, 162.3, 159.5, 159.3, 142.8, 142.0,
139.3, 138.3, 135.6, 135.2, 124.5, 123.5, 117.8, 116.9,
115.3, 114.3, 20.7, 20.6, 20.6, 20.4. HRFAB-MS (m/z)
calcd for C11H12NO5, 238.0715; found, 238.0712 (MHϩ).
3,4-Diacetoxybenzoisocyanide (19). To a solution of
70.0 mg of 18 (0.295 mmol) in 0.8 mL of CH2Cl2 were
added 83.0 mL of Et3N (0.590 mmol) and 26.3 mg of
triphosgene (90.4 mmol) at 0ЊC. The solution was stirred
for 3 h at room temperature. The precipitate was removed
by passage through celite, and the filtrate was diluted with
EtOAc. The organic layer was washed with brine, and the
aqueous layer was extracted with EtOAc twice. The
combined organic layers were washed with brine, dried
over Na2SO4, and concentrated in vacuo. The residue was
purified by silica gel column chromatography (EtOAc/n-
hexane 1:2) to give 61.4 mg of 19 (0.225 mmol) as a yellow
oil in 85% yield based on triphosgene. IR (neat) nmax 2129,
1
1776, 1500, 1371, 1201, 1171, 1041, 1012 cmϪ1. H NMR
(400 MHz, CDCl3) d 7.30–7.22 (3H, m), 2.31 (3H, s), 2.30
(3H, s). 13C NMR (100 MHz, CDCl3) d 167.6, 167.6, 165.5,
142.9, 142.5, 124.7, 124.4, 122.0, 20.6, 20.5. HRFAB-MS
(m/z) calcd for C11H10NO4, 220.0610; found, 220.0604
(MHϩ).
Figure 6. ORTEP drawing of crystal structure of 22.
3-(3,4-Dihydroxyphenyl)-tetrazole (20). A solution of
150 mg of 19 (1.11 mmol), 722 mg of NaN3 (11.1 mmol)
and 594 mg of NH4Cl in 6 mL of DMF was stirred for 3 h at
80ЊC. The precipitate was removed by passage through
celite, and the filtrate was concentrated in vacuo. Addition
of Et2O to the residue afforded a precipitate. The solid was
washed with acetone and CHCl3 to give 196 mg of 20
(1.10 mmol) as a brown powder in 99% yield: mp
Ͼ300ЊC. IR (KBr) nma1x 1616, 1527, 1489, 1346, 1306,
1213, 1107, 816 cmϪ1. H NMR (400 MHz, acetone-d6) d
9.53 (1H, s), 7.34 (1H, d, Jˆ2.4 Hz), 7.21 (1H, dd, Jˆ8.8,
2.4 Hz), 7.04 (1H, d, Jˆ8.8 Hz). 13C NMR (100 MHz,
acetone-d6) d 147.4, 147.9, 142.3, 127.7, 116.7, 110.0.
Acknowledgements
We thank Drs H. Naganawa and R. Sawa, Institute of
Microbial Chemistry, for spectroscopic analysis, and Ms
H. Nakamura, Institute of Microbial Chemistry, for X-ray
crystallographic analysis. This work was financially
supported in part by the Science Research Promotion
Fund of the Promotion and Mutual Aid Corporation for
Private Schools of Japan, and by the Special Coordination
Funds for Promotions of Science and Technology from the
Science and Technology Agency, Japan.

752
T. Watanabe et al. / Tetrahedron 56 (2000) 741–752
Kole, H. K.; Wolf, G.; Shoelson, S. E.; Roller, P. P. J. Med. Chem.
References
1996, 39, 1021.
21. Roller, P. P.; Wu, L.; Zhang, Z.-Y.; Burke Jr., T. R. Bioorg.
Med. Chem. Lett. 1998, 8, 2149.
1. Tonks, N. K.; Neel, B. G. Cell 1996, 87, 365.
2. Imoto, M.; Kakeya, T.; Sawa, T.; Hayashi, C.; Hamada, M.;
Takeuchi, T.; Umezawa, K. J. Antibiot. 1993, 46, 1342.
3. Kakeya, H.; Imoto, M.; Takahashi, Y.; Naganawa, H.;
Takeuchi, T.; Umezawa, K. J. Antibiot. 1993, 46, 1716.
4. Watanabe, T.; Takeuchi, T.; Otsuka, M.; Umezawa, K. J. Chem.
Soc., Chem. Commun. 1994, 437.
5. Watanabe, T.; Takeuchi, T.; Otsuka, M.; Tanaka, S.; Umezawa,
K. J. Antibiot. 1995, 48, 1460.
6. Umezawa, K. Recent Res. Devel. Agric. Food Chem. 1997, 1,
257.
7. Fujiwara, S.; Watanabe, T.; Nagatsu, T.; Gohda, J.; Imoto, M.;
Umezawa, K. Biochem. Biophys. Res. Commun. 1997, 238, 213.
8. Lammers, R.; Bossenmaier, B.; Cool, D. E.; Tonks, N. K.;
Schlessinger, J.; Fischer, E. H.; Ullrich, A. J. Biol. Chem. 1993,
268, 22 456.
9. Ahmad, F.; Azevedo, J. L.; Cortright, R.; Dohm, G. L.;
Goldstein, B. J. J. Clin. Invest. 1997, 100, 449.
10. Hashimoto, N.; Feener, E. P.; Zang, W. R.; Goldstein, B. J.
J. Biol. Chem. 1992, 267, 13 811.
11. Zang, W.-R.; Li, P.-M.; Oswald, M. A.; Goldstein, B. J. Mol.
Endocrinol. 1996, 10, 575.
12. Elchembly, M.; Payette, P.; Michalliszyn, E.; Cromlish, W.;
Collins, S.; Loy, A. L.; Normandin, D.; Cheng, A.; Himms-Hagen,
J.; Chan, C. C.; Ramachandran, C.; Gresser, M. J.; Tremblay, M. L.;
Kennedy, B. P. Science 1999, 283, 1544.
13. Burke Jr., T. R.; Kole, H. K.; Roller, P. P. Biochem. Biophys.
Res. Commun. 1994, 204, 129.
14. Burke Jr., T. R.; Smyth, M. S.; Nomizu, M.; Otaka, A.; Roller,
P. P. J. Org. Chem. 1993, 58, 1336.
22. Hiriyanna, K. T.; Baedke, D.; Beak, K.-H.; Forney, B. A.;
Kordiyak, G.; Ingebritsen, T. S. Anal. Biochem. 1994, 223, 51.
23. Cao, X.; Moran, E. J.; Siev, D.; Lio, A.; Ohashi, C.; Mjalli,
A. M. M. Bioorg. Med. Chem. Lett. 1995, 5, 2953.
24. Burke Jr., T. R.; Yao, Z.-J.; Zhao, H.; Milne, G. W. A.; Wu, L.;
Zhang, Z.-Y.; Voigt, J. H. Tetrahedron 1998, 54, 9981.
25. Wang, Q.; Dechert, U.; Jirik, F.; Withers, S. G. Biochem.
Biophys. Res. Commun. 1994, 200, 577.
26. Arabaci, G.; Guo, X.-C.; Beebe, K. D.; Coggeshall, K. M.; Pei,
D. J. Am. Chem. Soc. 1999, 121, 5085.
27. Hamaguchi, T.; Sudo, T.; Osada, H. FEBS Lett. 1995, 372, 54.
28. Miski, M.; Shen, X.; Cooper, R.; Gillum, A. M.; Fisher, D. K.;
Miller, R. W.; Higgins, T. J. Bioorg. Med. Chem. Lett. 1995, 5,
1519.
29. Alvi, K. A.; Casey, A.; Nair, B. G. J. Antibiot. 1998, 51, 515.
30. Kagamizono, T.; Hamaguchi, T.; Ando, T.; Sugawara, K.;
Adachi, T.; Osada, H. J. Antibiot. 1999, 52, 75.
31. Rice, R. L.; Rusnak, J. M.; Yokokawa, F.; Yokokawa, S.;
Messner, D. J.; Boynton, A. L.; Wipf, P.; Lazo, J. S. Biochemistry
1997, 36, 15 965.
32. Puius, Y. A.; Zhao, Y.; Sullivan, M.; Lawrence, D. S.; Almo,
S. C.; Zhang, Z.-Y. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 13 420.
33. Nishio, M.; Hirota, M. Tetrahedron 1989, 45, 7201.
34. Nishio, M.; Hirota, M.; Umezawa, Y. The CH/p Interaction.
Evidence, Nature, and Consequences: Wiley: New York, 1998.
35. Yu, L.; McGill, A.; Ramirez, J.; Wang, P. G.; Zang, Z.-Y.
Biomed. Chem. Lett. 1995, 5, 1003.
36. Sugawara, M.; Sasaki, A.; Yamaga, H.; Shinagawa, H.;
Fukasawa, M.; Sumita, Y. EP 472,062, 1992; JP 212,102, 1990;
Chem. Abstr. 1980, 117 7729.
15. Burke Jr., T. R.; Smyth, M. S.; Otaka, A.; Nomizu, M.; Roller,
P. P.; Wolf, G.; Case, R.; Shoelson, S. E. Biochemistry 1994, 33,
6490.
37. Nishio, M.; Umezawa, Y.; Hirota, M.; Takeuchi, T. Tetra-
hedron 1995, 51, 8665.
16. Chen, L.; Wu, L.; Otaka, A.; Smyth, M. S.; Roller, P. P.; Burke
Jr., T. R.; den Hertog, J.; Zhang, Z.-Y. Biochem. Biophys. Res.
Commun. 1995, 216, 976.
17. Burke Jr., T. R.; Ye, B.; Yan, X.; Wang, S.; Jia, Z.; Chen, L.;
Zhang, Z.-Y.; Barford, D. Biochemistry 1996, 35, 15 989.
18. Wang, Q.; Huang, Z.; Ramachandran, C.; Dinaut, A. N.;
Taylor, S. D. Bioorg. Med. Chem. Lett. 1998, 8, 345.
19. Yokomatsu, T.; Murano, T.; Umesue, I.; Soeda, S.; Shimeno,
H.; Shibuya, S. Bioorg. Med. Chem. Lett. 1999, 9, 529.
20. Burke Jr., T. R.; Ye, B.; Akamatsu, M.; Ford Jr., H.; Yan, X.;
38. Mascagna, D.; Ghanem, G.; Morandini, R.; D’Ischia, M.;
Misuraca, G.; Lejeune, F.; Prota, G. Melanoma Res. 1992, 2, 25.
39. Eckert, H.; Forster, B. Angew. Chem., Int. Ed. Engl. 1987, 26,
894.
40. Fallon, F. G.; Herbst, R. M. J. Org. Chem. 1957, 22, 933.
41. Burankevich, J. V.; Love, A. M.; Volpp, G. P. J. Org. Chem.
1971, 36, 1.
42. Gangjee, A.; Zhu, Y.; Queener, S. F.; Francom, P.; Broom, A.
D. J. Med. Chem. 1996, 39, 1836.
43. Umezawa, Y.; Nishio, M. Bioorg. Med. Chem. 1998, 6, 493.