Synthesis and characterization of the aqueous solution chemistry of the food-derived carcinogen model N-acetoxy-N-(1-methyl-5H-pyrido[4,5-b]indol-3-yl)acetamide and its N-pivaloyloxy analogue

Article abstract of DOI:10.1016/j.tet.2003.08.005

We report the synthesis of N-acetoxy-N-(1-methyl-5H-pyrido[4,5-b]indol-3-yl)acetamide, 7, its N-pivaloyloxy analogue, 9, and improved synthesis of indole-2-acetonitrile, 3 (70% in five steps from indole-2-carboxylic acid), the carcinogenic amine Trp-P-2, 4 (40% from 3), and the nitro compound, 5 (40% from 4 by oxidation with H₂O₂ using Mo(CO)₆ catalyst). In aqueous solution at neutral pH, 7 primarily undergoes C-O bond cleavage to yield the hydroxamic acid, 8, but under the same conditions the sterically hindered 9 decomposes predominately by N-O bond cleavage with a pH independent rate constant that is 7.5-fold smaller than that for 7. In the pH range 0.5-7.0 three different processes for the decomposition of 9 were detected by kinetics. Only the process that dominates at neutral pH generates a nitrenium species that can be trapped by N₃^-.

Full text of DOI:10.1016/j.tet.2003.08.005

Tetrahedron 59 (2003) 8003–8010
Synthesis and characterization of the aqueous solution chemistry
of the food-derived carcinogen model N-acetoxy-N-(1-methyl-5H-
pyrido[4,5-b]indol-3-yl)acetamide and its N-pivaloyloxy analogue
*
Sridharan Rajagopal, Michael E. Brooks, Thach-Mien Nguyen and Michael Novak
Department of Chemistry and Biochemistry, Hughes Laboratory, Miami University, Oxford, OH 45056, USA
Received 7 July 2003; revised 6 August 2003; accepted 6 August 2003
Abstract—We report the synthesis of N-acetoxy-N-(1-methyl-5H-pyrido[4,5-b]indol-3-yl)acetamide, 7, its N-pivaloyloxy analogue, 9, and
improved synthesis of indole-2-acetonitrile, 3 (70% in five steps from indole-2-carboxylic acid), the carcinogenic amine Trp-P-2, 4 (40%
from 3), and the nitro compound, 5 (40% from 4 by oxidation with H2O2 using Mo(CO)6 catalyst). In aqueous solution at neutral pH, 7
primarily undergoes C–O bond cleavage to yield the hydroxamic acid, 8, but under the same conditions the sterically hindered 9 decomposes
predominately by N–O bond cleavage with a pH independent rate constant that is 7.5-fold smaller than that for 7. In the pH range 0.5–7.0
three different processes for the decomposition of 9 were detected by kinetics. Only the process that dominates at neutral pH generates a
nitrenium species that can be trapped by N23 .
q 2003 Elsevier Ltd. All rights reserved.
1. Introduction
The heterocyclic amines derived from cooked protein
containing food have been shown to be highly carcinogenic
and mutagenic.1 The heterocyclic amines are classified into
two categories: the IQ type (IQ, MeIQx, MeIQ, etc.), that
contains the imidazoquinoline or imidazoquinoxaline
structure and the non-IQ type (Glu-P-1, Trp-P-1, Trp-P-2,
AaC, etc.) that contain the 2-aminopyridine moiety. Non-IQ
Scheme 1.
amines are formed by pyrolysis of proteins and amino acid
mixtures.2 These materials have also been isolated in parts
these amines. There have been several reported syntheses of
per billion concentrations from broiled and fried meats, fish
Trp-P-2.6,7 Herein we report improvements in the yield of
and protein-rich plant material and have been shown to be
synthesis of Trp-P-2, 4, and synthesis and characterization
mutagenic to Salmonella in the presence of rat liver
of its corresponding hydroxamic acid ester derivatives 7 and
9. Surprisingly, at neutral pH, 7 does not, to any significant
extent, undergo N–O bond cleavage to generate a nitrenium
ion, the dominant neutral pH reaction of other ester
homogenates.2 These heterocyclic amines are now con-
sidered to be probable human carcinogens.3
The food-derived heterocyclic amines are promutagens/
derivatives of heterocyclic hydroxylamines and hydroxamic
procarcinogens that require oxidative metabolism for
activation. The likely activation processes are shown in
acids.5 Instead, 7 undergoes predominant C–O bond
cleavage to generate the hydroxamic acid, 8. The more
sterically hindered 9 does undergo N–O bond cleavage at
neutral pH.
Scheme 1.4 We have been studying the selectivity of the
nitrenium ions derived from these heterocyclic esters.5 The
synthesis of the hydroxylamine or hydroxamic acid ester
derivatives of these heterocyclic amines is essential for
studying the selectivity of the nitrenium ion derived from
The synthesis of 4 was previously reported by Takeda et al.7
starting from indole-2-acetonitrile, 3, in a one-pot procedure
using CH3CN and AlCl3 in 15% yield (Scheme 2). The
indole-2-acetonitrile was synthesized from indole-2-
carboxylic acid in five steps in 36% yield.8
Keywords: nitrenium ions; heterocyclic amines; food-derived carcinogens.
*
Corresponding author. Tel.: þ1-513-529-3316; fax: þ1-513-529-5715;
e-mail: novakm@muohio.edu
0040–4020/$ - see front matter q 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2003.08.005

8004
S. Rajagopal et al. / Tetrahedron 59 (2003) 8003–8010
Scheme 2. (a) SOCl2, C6H6, cat. DMF, 508C, (b) (CH3)2NH, C6H6, RT,
95% (aþb), (c) LAH/THF, 508C, 98%, (d) Mel/EtOAc, 508C, 99%,
(e) KCN/DMF, 76%, (f) CH3CN, AlCl3, 908C, 18 h, MeOH/AcOH, 40%.
2. Results and discussion
We improved the yield of 3 by modifications of the original
procedure.8 By completely removing the benzene-thionyl
chloride solution prior to treatment with dimethylamine gas
in the first step, instead of reducing to 1/3 volume as
reported, the yield of 1 was increased from 60 to 95% and
the formation of the 3-chlororindole-2-carboxamide by-
product was eliminated. During the transformation of the
quaternary ammonium iodide 2 to the nitrile 3 by SN2
displacement, the replacement of the reported solvent
MeOH by DMF improved the yield. Prior to this
modification, the methanol also acted as a nucleophile and
gave the corresponding ether compound in addition to the
desired nitrile 3 in equal amounts. The use of DMF resulted
in a single product without need of column chromatography.
The overall yield of 3 was 70% starting from indole-2-
carboxylic acid compared to 36% previously reported.8
Scheme 3. (a) Mo(CO)6, CF3COOH, (CF3CO)2O, 30% H2O2, 40%, (b) 5%
Pd/C, NH2NH2·H2O, THF, 2208C, 1 h, 96%, (c) 2.2 equiv. Et3N, 2.2 equiv.
AcCl, THF, 08C, 30 min, 98%, (d) Et3N, (CH3)3CCOCN, THF, 2208C,
30 min, 37%, (f) 1.3 equiv. (CH3)3CCOCl, 1.3 equiv. Et3N, THF, 08C, 1 h,
91%.
corresponding hydroxylamine, 6.11 Although synthesis of
6 has been reported from the nitro compound it was
characterized only by its mass spectrum.9 We report more
detailed characterization here. The attempted synthesis of
the ultimate carcinogenic ester, 10, with pivaloyl cyanide12
was unsuccessful, as the ester decomposed spontaneously
during workup. The acyl cyanide procedure does provide
isolated N-arylhydroxylamine esters in cases in which these
compounds are less reactive.5a,b The hydroxylamine, 6, was
treated with 2.2 equiv. of acetyl chloride and triethylamine
at 08C in THF for 30 min to give the hydroxamic acid ester,
7, after workup. The ester, 7, was 99% pure according to
HPLC and NMR (Scheme 3). The pivalic acid ester, 9, was
made in modest yield by successive addition of acetyl
chloride and pivaloyl chloride. The intermediate
hydroxamic acid, 8, was isolated and characterized.
The yield of 4 from 3 was also improved by modification of
the reported procedure.7 After the reaction, the excess AlCl3
was quenched with water, the pH of the solution was
adjusted to 4–5 and the reaction mixture was extracted with
Et2O to remove the neutral reaction products. The water
layer was then made basic with 2 M K2CO3 until pH.9.
The product was isolated from the basic aqueous solution
and purified as described in the literature.7 The Et2O layer
gave unreacted 3 in ca. 50% yield. Longer reaction time
(72 h) did not increase the yield of 4. Based on the isolated
yield of the recovered unreacted starting material 3, the
yield of 4 was increased from the reported 15 to 40%
(Scheme 2). The overall yield of 4 starting from indole-2-
carboxylic acid was 28%.
Kinetics of the decomposition of 7 and 9 were examined by
UV and HPLC methods at 208C in 5 vol% CH3CN–H2O,
m¼0.5 (NaClO4) in NaH2PO4/Na2HPO4, AcOH/NaOAc,
and HCO2H/NaHCO2 buffers (0.02–0.06 M) at pH.2.5
and in HClO4 solutions at pH#2.5. For both compounds the
initial decomposition of the ester at neutral and acidic pH
was followed by a much slower decomposition of an initial
reaction product that complicated the determination of rate
constants by the UV method. This subsequent reaction was
more pronounced for 7. This was not a problem for kinetics
followed by the HPLC method because HPLC made it
possible to directly monitor the concentration of the ester
without interference from subsequent reactions. In spite of
this complication, it was possible to determine reliable rate
constants for the decomposition of 7 and 9 by UV methods
either by fitting the absorbance vs time data to a
consecutive first-order rate equation or by monitoring the
The synthesis of 7 requires the oxidation of 4 to the
corresponding nitro compound, 5. The reported synthesis of
5 from 4 requires the use of Na2WO4, (CF3CO)2O, and
H2O2 with a yield of 12%.9 Attempts were made to increase
the yield of this procedure without any success. However,
the reported oxidation of Glu-P-1 and Glu-P-2 to the
corresponding nitro compound using Mo(CO)6 gave 60–
70% yield.10 The oxidation of 4 using Mo(CO)6 and H2O2
following the reported procedure for Glu-P-1 and Glu-P-2
gave 5 in 40% yield after 1 h (Scheme 3).
A selective and controlled reduction of the nitro compound
with 5% Pd/C and hydrazine monohydrate gave the

S. Rajagopal et al. / Tetrahedron 59 (2003) 8003–8010
8005
Figure 1. log kobs vs pH for 7 and 9 at 208C. 7: V. 9: O (HPLC), X (UV). Data were fit to Eq. (1) for 9 and to the last term of Eq. (1) for 7 by non-linear least-
squares methods.
disappearance of the ester at an isosbestic point for the
subsequent reaction. Rate constants for decomposition of 7
and 9, kobs, were buffer independent, but did exhibit pH-
dependence. A plot of log kobs vs pH is shown in Figure 1 for
both compounds.
(Scheme 4) becomes the major product. In the presence of
N23 the yield of 11, but not 8, decreases. The N23 -adduct, 12,
is formed in place of 11 (Fig. 2). This occurs without a
change in kobs at [N23 ] sufficiently high to completely trap
out 11 (kobs¼(3.4^0.1)£1025 s21 measured by HPLC in
the presence of 10 mM N23 in pH 6.95 0.02 M H2PO24 /
HPO422 buffer compared to (3.1^0.1)£1025 s21 in the same
kobs ¼ ðkHð102pHÞ2 þ kA 102pH þ kBKaÞ=ðKa þ 102pH
Þ ð1Þ
The pH dependence of the decomposition of 9 was fit quite
well by Eq. (1). Only the last term (kBKa/(Kaþ102pH)) was
necessary to fit the data taken over a smaller pH range for 7.
All of these terms have been observed for other esters of
heterocyclic hydroxylamines and hydroxamic acids.5 For 9,
the first term of Eq. (1) is consistent with acid-catalyzed
decomposition of the conjugate acid 9Hþ, and the second
term is consistent with spontaneous decomposition of 9Hþ
(Scheme 4).5 The last term of Eq. (1), common to both 7 and
9, is consistent with spontaneous decomposition of the
neutral ester (Scheme 4).5 Parameters obtained from the
kinetic fits are provided in Table 1. The kinetically
determined pKas were equivalent, within experimental
error, to those obtained by spectrophotometric titration of
7 and 9.
The last term of Eq. (1) is consistent either with spontaneous
N–O bond cleavage to generate a nitrenium ion, or with an
uncatalyzed acyl-transfer reaction to H2O (Scheme 4).5
Examination of reaction products for the decomposition of 7
at pH 7.0 showed that the hydroxamic acid, 8, accounted for
ca. 90% of reaction products. Slow decomposition of this
product accounts for the subsequent reaction observed by
UV spectroscopy. Clearly, C–O bond cleavage dominates
for this compound. This is somewhat surprising since all
previously examined carboxylic acid esters of heterocyclic
hydroxylamines or hydroxamic acids undergo predominant
N–O bond cleavage at neutral pH.5
The more sterically hindered 9 decomposes 7.5-fold more
slowly than 7 at neutral pH, and the yield of 8 is reduced
to 40% for this compound. In its place, the amide, 11,
Scheme 4.

8006
S. Rajagopal et al. / Tetrahedron 59 (2003) 8003–8010
Table 1. Kinetic and titration parameters at 208C
about 2 orders of magnitude more selective for reaction with
N23 than are the N-acetylAaC intermediates.
Ester
pKa
(kinetics/titration)
105kH
(M21 21
106kA
(s21
105kB
(s21
s
)
)
)
7
7
9
9
3.92^0.04
4.15^0.06
4.12^0.07
4.11^0.03
24^1
5.8^0.6
1.5^0.2
3.2^0.2
buffer in the absence of N23 ).13 The mechanism of Scheme 4
is consistent with these observations.
Trapping of the nitrenium ion, 13, and/or its conjugate base,
14, by N23 can explain the decreasing yield of 11 in the
presence of N23 provided that 11 is also generated from 13
and/or 14. Previously we have shown that the conjugate
acid–base pairs of several heterocyclic nitrenium ions,
capable of deprotonation to form a quinoid-like conjugate
base such as 14, are subject to rapid reduction.5b,d The yield
of 8 is not affected because the N23 trapping occurs in a step
subsequent to the decomposition of 9 via the competing
N–O and C–O bond cleavage processes. The trapping ratio,
kaz/kc, in pH 6.95 0.02–0.06 M H2PO24 /HPO422 buffer is
buffer concentration-independent with an average value of
(3.9^0.8)£104 M21. More extensive trapping studies at
different pH will be required to determine if this is the
trapping ratio for 13 or 14, or an average value intermediate
between the two. The results do show that at neutral pH the
N-acetylated Trp-P-2 intermediates are more selectively
trapped by N23 than are the corresponding intermediates
from AaC, 15 (kaz/kc¼1.2£103 M21), and N-acetylAaC, 16
(kaz/kc¼4.5£102 M21).5b,d We have previously shown that
the mutagenicity of the parent amines to Salmonella
typhimurium TA 98 or TA 100 correlates positively with
the selectivity of the corresponding nitrenium species for
reaction with N23 at neutral pH.14 Trp-P-2 is about 2 orders
of magnitude more mutagenic than AaC in both of these
Salmonella strains.15 This is in accord with our observation
that at neutral pH the N-acetylTrp-P-2 nitrenium species are
Small amounts of 12 can be detected by HPLC when 7
undergoes decomposition at neutral pH in the presence of
N23 . This shows that N–O bond cleavage is a minor
component of the decomposition of 7 (ca. 5%). Rate
constants for nucleophilic attack on the carboxyl carbon of
pivalic acid esters are typically 1 to 2 orders of magnitude
smaller than those for acetic acid esters.16 Taking into
account that C–O bond cleavage represents ca. 40% of the
decomposition rate of 9 and 95% of the decomposition rate
of 7 at neutral pH, the replacement of Me by tert-Bu in the
acyloxy component of the molecule slows the acyl-transfer
by a factor of ca. 18-fold. This allows the N–O bond
cleavage to become the dominant, although not exclusive,
process in 9.
It is possible that the spontaneous decomposition of 7 into 8
via the kB pathway involves intramolecular nucleophilic
attack of the pyridyl N in a 5-member ring transition state.
An intramolecular general base catalysed attack of H2O
involving a 7-member ring appears less likely. If the
intramolecular process does occur, the decreased rate of
decomposition of 9 via the acyl-transfer path may be due to
the steric bulk of the tert-Bu group hindering the
participation of the pyridyl N in a cyclic transition
state much as is the case in an intermolecular process.16
Our current data cannot distinguish these mechanistic
Figure 2. Yields of 8 (O), 11 (X), and 12 (P) as a function of [N23 ] in 0.06 M pH 6.95 phosphate buffer. Data for 11 and 12 were fit to standard equations,
described elsewhere,5 that provide kaz/kc¼(3.9^0.5)£104 M21. The line for 8 is the average value of its yield throughout the experiment.

S. Rajagopal et al. / Tetrahedron 59 (2003) 8003–8010
8007
possibilities, but the intramolecular nucleophilic process
may explain why 7Hþ is less reactive than 7 (Fig. 1) even
though the leaving group in 7Hþ is presumably better.
spontaneous decomposition of 9Hþ and by acid-catalyzed
hydrolysis of 9Hþ.
The reactions that 9 undergoes under acidic conditions were
also examined. HPLC analysis shows that at both pH 2.0
and 0.7 the initial product of decomposition of 9 is 8, and
that 8 subsequently decomposes into 6. Similar to other
heterocyclic hydroxylamines we have investigated, 6 is not
subject to rapid Bamberger rearrangement under moderately
acidic conditions, so it is reasonably stable in the acidic pH
range.5c,d The acid-catalysed decomposition of 9Hþ to
generate 8 that is governed by kH[Hþ] dominates at pH 0.7,
accounting for ca. 90% of the observed rate of decompo-
sition of 9 at this pH. This reaction is similar to that
previously observed for the Glu-P derivatives 17a and
17b.5c
4. Experimental
4.1. General
All solvents were distilled before use. All reagents were
used as received. N-(1H-indole-2-ylmethyl)-N,N-dimethyl-
amine and N-(1H-indole-2-ylmethyl)-N,N,N-trimethyl-
ammonium iodide (2) were synthesized following the
reported procedure.8
4.1.1. N,N-Dimethylindole-2-carboxamide (1). This com-
pound was synthesized following the procedure published
by Schindler8 with a few modifications. A few drops of
DMF were added to allow the indole-2-carboxylic acid to
completely dissolve into benzene instead of heating the
solution. The excess thionyl chloride was removed by
evaporating the entire benzene solution instead of reducing
it to 1/3 volume. Dimethylamine was added directly into the
acyl halide solution by bubbling dimethylamine gas into the
solution instead of adding the dimethylamine as a benzene
solution. The product was obtained in 95% yield: mp 177–
1808C (lit.8 180–1828C).
Spontaneous decomposition of the conjugate acid of the
ester (kA) has been observed previously for the N-acetyl-
AaC derivative 18, but in that case the hydroxylamine was
generated directly from 18 with no intermediate hydroxamic
acid.5d In the present study HPLC data show that the
hydroxamic acid, 8, is definitely the initial product of the
decomposition of 9 at pH 2.0, a pH at which the kA term is
the dominant kinetic term, accounting for 70% of the overall
reaction. Decomposition of the initially formed 8 into the
hydroxylamine, 6, occurs with the same rate constant
observed for authentic 8 at this pH (6.6£1027 s21). The first
step of this process for 9 is apparently attack of H2O on the
ester carboxyl of 9Hþ to generate 8 via acyl-transfer.
4.1.2. 2-(Cyanomethyl)indole (3). This compound was
synthesized following the procedure published by
Schindler8 with one modification. DMF was chosen as the
solvent instead of CH3OH. The product was obtained in
76% yield: mp 93–958C (lit.8 96–988C).
It is now clear that the propensity for C–O vs N–O bond
cleavage in ester derivatives of heterocyclic N-acetyl-N-
arylhydroxamic acids at neutral pH is strongly dependent on
structure. While 17a and 17b decompose exclusively by
N–O bond cleavage,5c 18 exhibits ca. 15% C–O bond
cleavage,5d and 7 decomposes with ca. 95% C–O bond
cleavage under the same conditions. The more sterically
hindered ester 9 still undergoes 40% C–O bond cleavage.
This variation in reaction type is likely controlled by the
relative stability of the respective nitrenium ions, the
leaving group ability of the hydroxamates, and possibly
the efficiency of intramolecular nucleophilic attack by the
pyridyl N. Since both reactions exhibit the same kinetics at
neutral pH it is necessary to examine product distributions
to determine which reaction is occurring in individual cases.
4.1.3. 1-Methyl-5H-pyrido[4,5-b]indol-3-ylammonium
acetate (4). This compound was synthesized following the
published procedure by Takeda, et al.7 with minor
modifications. To a solution of 3 (1.5 g, 9.6 mmol)
dissolved in 50 mL of CH3CN was added AlCl3 (15 g,
0.11 mol) portion-wise at 08C over 30 min. After the
addition, the mixture was refluxed for 18 h. Then water
was added to quench the excess AlCl3 and the pH of the
solution was adjusted to 4–5. It was then extracted with
Et2O (3£25 mL). The Et2O extracts were combined and
dried over anhydrous Na2SO4 and evaporated under reduced
pressure to give 0.75 g (50%) of 3. The aqueous layer was
made basic with 2 M K2CO3. To the aqueous solution
5–10% (w/v) of Rochelle’s salt was added and the solution
extracted with CH2Cl2 (5£100 mL) and EtOAc
(5£100 mL). The combined organic extract was dried
over anhydrous Na2SO4. The organic solution was evapor-
ated under reduced pressure. The resulting crude product
was dissolved in a small amount of MeOH, and a few
millilitres of EtOAc and a few drops of AcOH were added.
The crystalline precipitate was collected and washed with
EtOAc to give 400 mg of pure 4. Chromatography (silica
gel, 5% MeOH/EtOAc) of the mother liquor yielded an
additional 100 mg of 4. The yield of the reaction is 40%
3. Conclusion
We were able to increase the yield of the synthesis of indole-
2-acetonitrile, 3, from the previously reported 36 to 70%.
The overall yields of the carcinogenic amine, 4, and the
nitro derivative, 5, starting from indole-2-carboxylic acid,
were also improved to 28 and 11%, respectively from the
previously reported overall yields of 5 and ,1%,
respectively. The model carcinogen, 7, undergoes predomi-
nant C–O bond cleavage at neutral pH to generate the
hydroxamic acid, 8. This reaction is suppressed in the more
sterically hindered pivalic acid ester, 9, so that N–O bond
cleavage to generate nitrenium species competes with acyl-
transfer. Under acidic conditions 9 decomposes to 8 via
1
(based on the recovered 3 from the Et2O layer). H NMR
(300 MHz, DMSO-d6) d: 10.98 (1H, s), 7.82 (1H, d,
J¼7.6 Hz), 7.30 (1H, d, J¼7.6 Hz), 7.22 (1H, t, J¼7.9 Hz),
7.09 (1H, t, J¼7.9 Hz), 6.21 (1H, s), 5.65 (2H, br), 2.67 (3H,
s), 1.90 (3H, s).

8008
S. Rajagopal et al. / Tetrahedron 59 (2003) 8003–8010
4.1.4. 3-Nitro-1-methyl-5H-pyrido[4,5-b]indole (5). To a
mixture of 4.4 mL of (CF3CO)2O and 0.7 mL of CF3COOH
was added 1.35 mL (12 mmol) of 30% H2O2 in a dropwise
fashion at 08C. To the cold mixture was added 0.135 g,
(0.51 mmol) of Mo(CO)6 in small portions. A solution of
Trp-P-2-acetate (4), 0.25 g (0.97 mmol) in 15 mL of
CH2Cl2 and 0.7 mL of CF3COOH, also cooled to 08C,
was added in a dropwise fashion over 20–30 min to the
stirred oxidizing media. The reaction mixture was cooled in
an ice bath throughout the reaction. Reaction progress was
monitored by TLC with precoated silica gel plates (EtOAc).
The reaction was complete within 1 h of the last addition of
4 based on TLC. The solution was made basic with ca.
20 mL of cold 2 M K2CO3, ca. 40 mL of brine was added,
and the resulting mixture was extracted with CH2Cl2
(3£25 mL) and EtOAc (3£25 mL). The combined organic
extracts were evaporated under reduced pressure without
drying, and the crude product was subjected to column
chromatography on silica gel (EtOAc) to give 5 that exhibits
no impurities in its 1H NMR spectrum, yield: 0.088 g (40%),
1H NMR (300 MHz, DMSO-d6) d: 12.40 (1H, br), 8.28–
8.26 (2H, m), 7.72 (d, 1H, J¼8.2 Hz), 7.61 (t, 1H,
J¼7.1 Hz), 7.40 (t, 1H, J¼7.9 Hz), 3.02 (3H, s); 13C
NMR (75 MHz, DMSO-d6) d: 152.4 (C), 151. 5 (C), 144.3
(C), 141.6 (C), 127.9 (CH), 123.0 (CH), 121.2 (CH), 121.1
(C), 120.2 (C), 112.2 (CH), 99.8 (CH), 23.2 (CH3); LC/MS
(ESI, positive ion mode): C12H10N3O2 (MþH)þ requires
m/e 228.07, found 228.00.
J¼7.8 Hz), 7.58 (d, 1H, J¼7.3 Hz), 7.55 (1H, s), 7.47 (t, 1H,
J¼6.9 Hz), 7.29 (t, 1H, J¼7.9 Hz), 2.88 (3H, s), 2.31 (3H,
s), 2.17 (3H, s); 13C NMR (75 MHz, DMSO-d6) d: 169.0
(C), 151.1 (C), 148.6 (C), 146.1 (C), 145.3 (C), 140.4 (C),
126.3 (CH), 122.1 (CH), 120.9 (C), 120.4 (CH), 116.5 (C),
111.4 (CH), 97.4 (CH), 23.4 (CH3), 22.4 (CH3), 18.1 (CH3);
LC/MS (ESI, positive ion mode): C14H14N3O2 (MþH–
C2H2O)þ requires m/e 256.2, found 256.1; high-resolution
MS (ES): C16H16N3O3 (MþH)þ requires m/e 298.1191,
found 298.1194.
4.1.7. N-Hydroxy-N-(1-methyl-5H-pyrido[4,5-b]indol-3-
yl)acetamide (8). 16 mg, (0.075 mmol) of 6 was dissolved
in 5 mL of dry THF and the solution was stirred under N2 at
2208C. 11.6 mL, (0.083 mmol) of triethylamine was added
followed by the stepwise addition of 5.8 mL, (0.083 mmol)
of acetyl chloride over a 5 min period. The solution was
stirred for 30 min during which time the mixture was
allowed to warm to 08C. The reaction was then quenched
with 15 mL of brine. The mixture was extracted thoroughly
with EtOAc (3£15 mL). The combined organic extract was
dried with anhydrous Na2SO4, filtered and evaporated under
vacuum. The residue was taken up into a minimum volume
of EtOAc, and subjected to column chromatography on
silica gel (EtOAc) to yield 7 mg of 8 (37%). IR (KBr): 3425,
1
3280, 1659, 1602, 1581, 1378 cm21; H NMR (300 MHz,
DMSO-d6) d: 11.74 (1H, br), 10.57 (1H, br), 8.10 (1H, d,
J¼7.9 Hz), 7.55–7.53 (1H, m) 7.55 (1H, s), 7.44 (1H, t,
J¼7.5 Hz), 7.27 (1H, t, J¼7.2 Hz), 2.91 (3H, s), 2.23 (3H,
s); 13C NMR (75 MHz, DMSO-d6) d: 169.0 (C), 150.7 (C),
148.6 (C), 145.3 (C), 140.3 (C), 126.0 (CH), 122.0 (CH),
121.1 (C), 120.2 (CH), 115.7 (C), 111.2 (CH), 97.4 (CH),
23.2 (CH3), 22.7 (CH3); high-resolution MS (ES):
C14H14N3O2 (MþH)þ requires m/e 256.1086, found
256.1080.
4.1.5. 3-Hydroxyamino-1-methyl-5H-pyrido[4,5-
b]indole (6). 21.4 mg (0.095 mmol) of 5 was dissolved in
15 mL of dry THF. To this solution was added 21 mg of 5%
Pd/C. The solution was cooled to 2208C while stirring
under N2. To the solution was added 20 mL of hydrazine
monohydrate. The reaction was monitored by HPLC (C8
column, 254 nm, 6/4 MeOH/H2O, flow rate¼1 mL/min).
When the peak area for 5 was below 10% of the starting
concentration, the Pd/C was immediately filtered by vacuum
filtration. The Pd/C was washed thoroughly with THF. The
combined THF was evaporated under vacuum to dryness to
give 19.3 mg of the hydroxylamine 6 (96% yield). 1H NMR
(300 MHz, DMSO-d6) d: 11.31 (s, 1H), 8.49 (br, 2H,
exchangeable), 7.92 (d, 1H, J¼7.7 Hz), 7.41 (d, 1H, J¼
7.7 Hz), 7.30 (t, 1H, J¼7.2 Hz), 7.16 (t, 1H, J¼7.7 Hz), 6.69
(s, 1H), 2.75 (s, 3H); 13C NMR (75 MHz, DMSO-d6) d:
161.0 (C), 150.5 (C), 147.0 (C), 139.6 (C), 124.2 (CH),
122.3 (C), 120.8 (CH), 119.6 (CH), 111.9 (C), 110.5 (CH),
85.2 (CH), 23.0 (CH3); LC/MS (ESI, positive ion mode):
C12H12N3O (MþH)þ requires m/e 214.24, found 214.10.
4.1.8. N-Pivaloxy-N-(1-methyl-5H-pyrido[4,5-b]indol-3-
yl)acetamide (9). 10 mg, (0.039 mmol) of 8 was dissolved
in 8 mL of dry THF and the solution was stirred under N2 at
08C. 6.0 mL, (0.043 mmol) of triethylamine was added
followed by the stepwise addition of 5.3 mL, (0.043 mmol)
of pivaloyl chloride over a 5 min period. The reaction
progress was monitored by HPLC (C8 column, 254 nm, 75/
25 MeOH/H2O containing 50 mg/L deferoxamine mesylate,
buffered with 0.05 M 1:1 KOAc/AcOH, flow rate of 1.0 mL/
min). Additional triethylamine and pivaloyl chloride (ca.
20% of the original volumes) were added to bring the
reaction to completion. The reaction was quenched by
addition of 15 mL of brine followed by 1 mL of 5% aqueous
NaHCO3. The mixture was extracted thoroughly with
EtOAc (3£15 mL). The organic layer was dried over
anhydrous Na2SO4, filtered and evaporated under vacuum
to yield 12.1 mg of 9 (91%). IR (ATR): 3285, 2922, 1780,
1674, 1610, 1579, 1379 cm21; 1H NMR (200 MHz, DMSO-
d6) d: 11.79 (1H, br), 8.09 (1H, d, J¼7.7 Hz), 7.68 (1H, s),
7.55 (1H, d, J¼8.1 Hz), 7.45 (1H, t, J¼7.0 Hz), 7.28 (1H, t,
J¼7.9 Hz), 2.86 (3H, s), 2.19 (3H, s), 1.36 (9H, s); 13C
NMR (75 MHz, DMSO-d6) d: 176.0 (C), 169.0 (C), 151.6
(C), 146.8 (C), 146.2 (C), 141.2 (C), 127.0 (CH), 122.8
(CH), 121.8 (C), 121.45 (CH), 116.8 (CH), 112.2 (C), 97.1
(CH), 38.1 (C), 27.9 (CH3), 24.0 (CH3), 23.4 (CH3); high-
resolution MS (ES): C19H22N3O3 (MþH)þ requires m/e
340.1661, found 340.1661.
4.1.6. N-Acetoxy-N-(1-methyl-5H-pyrido[4,5-b]indol-3-
yl)acetamide (7). 19.3 mg, (0.091 mmol) of 6 was dis-
solved in 15 mL of dry THF and the resulting solution was
stirred under N2 at 08C. 28 mL, (0.20 mmol) of triethyl-
amine was added followed by the addition of 14 mL,
(0.20 mmol) of acetyl chloride. The solution was stirred for
30 min, and then quenched with 10 mL of 5% aqueous
NaHCO3. The mixture was extracted thoroughly with
EtOAc (3£15 mL). The organic layer was washed with
brine, dried with anhydrous Na2SO4, filtered and evaporated
under vacuum to yield 0.026 g of product (98% yield). IR
(ATR): 3285, 1790, 1685, 1610, 1380 cm21 1H NMR
;
(300 MHz, DMSO-d6) d: 11.85 (1H, br), 8.11 (d, 1H,

S. Rajagopal et al. / Tetrahedron 59 (2003) 8003–8010
8009
4.1.9. N-(1-Methyl-5H-pyrido[4,5-b]indol-3-yl)acet-
amide (11). This compound was made from 4 in the same
manner that 8 was made from 6 except that 2 equiv. of
triethylamine were utilized. IR (KBr): 3250, 1672, 1610,
standard first-order rate equation to obtain kobs. More
detailed procedures are described elsewhere.5
Reaction products were monitored by HPLC with UV
detection at 254 nm. HPLC conditions were: C8 reverse-
phase analytical column, 75/25 or 50/50 MeOH/H2O eluent,
containing 50 mg/L deferoxamine mesylate, buffered with
0.05 M 1:1 KOAc/AcOH, flow rate of 1.0 mL/min.
1
1378 cm21; H NMR (300 MHz, DMSO-d6) d: 11.58 (1H,
br), 10.42 (1H, br), 8.08 (1H, s), 8.02 (1H, d, J¼7.8 Hz) 7.47
(1H, d, J¼7.9 Hz), 7.38 (1H, t, J¼7.4 Hz), 7.22 (1H, t,
J¼7.4 Hz), 2.84 (3H, s), 2.09 (3H, s); 13C NMR (75 MHz,
DMSO-d6) d: 169.0 (C), 150.7 (C), 148.0 (C), 146.0 (C),
140.0 (C), 125.2 (CH), 121.6 (C), 121.4 (CH), 120.0 (CH),
114.1 (C), 111.0 (CH), 92.4 (CH), 24.0 (CH3), 22.8 (CH3);
high-resolution MS (ES): C14H15N3O (MþH)þ requires m/e
240.1137, found 240.1133.
Acknowledgements
This work was supported by a grant from the American
Cancer Society (RPG-96-078-03-CNE).
4.1.10. N-(1-Methyl-4-azido-5H-pyrido[4,5,b]indol-3-
yl)acetamide (12). The pivalic acid ester 9 (11 mg,
0.032 mmol) was dissolved in 2.5 mL of dry, distilled
DMF. This solution was added in a stepwise fashion
(0.5 mL at 6 h intervals) to 100 mL of a pH 7.0, 5 vol%
CH3CN–H2O, m¼0.5 (NaClO4), 0.04 M NaH2PO4/
Na2HPO4 buffer containing 2.0 mM NaN3 that was main-
tained at 208C. The reaction mixture was extracted with
EtOAc (4£20 mL) 48 h after the last addition of 9. The
combined EtOAc extracts were washed with ice-cold 20%
NaOH (2£15 mL) and then with brine (1£15 mL). The
EtOAc solution was then dried over anhydrous Na2SO4 and
evaporated under vacuum to yield 12 contaminated with ca.
30% of a mixture of 8 and 11. Since 12 decomposes upon
attempted purification, spectral analysis was performed on
the crude isolated sample. IR (ATR): 3248, 2127,
References
1. (a) Felton, J. S.; Jagerstad, M.; Knize, M. G.; Skog, K.;
Wakabayashi, K. In Food Borne Carcinogens: Heterocyclic
Amines; Nagao, M., Sugimura, T., Eds. Wiley: New York,
2000; pp 31–72. (b) Nagao, M. In Food Borne Carcinogens:
Heterocyclic Amines; Nagao, M., Sugimura, T., Eds. 2000;
pp 163–196. (c) Yamazoe, Y.; Nagata, K. In Food Borne
Carcinogens: Heterocyclic Amines; Nagao, M., Sugimura, T.,
Eds. 2000; pp 74–89. (d) King, R. S.; Kadlubar, F. F.;
Turesky, R. J. In Food Borne Carcinogens: Heterocyclic
Amines; Nagao, M., Sugimura, T., Eds. 2000; pp 90–111.
2. (a) Hatch, F. T.; Knize, M. G.; Felton, J. S. Environ. Mol.
Mutagen. 1991, 17, 419. (b) Felton, J. S.; Knize, M. G.; Shen,
N. H.; Andersen, B. D.; Bjeldanes, L. F.; Hatch, F. T. Environ.
Health Perspect. 1986, 67, 17–24. (c) Sugimura, T. Environ.
Health Perspect. 1986, 67, 5–10.
1
1664 cm21; H NMR (300 MHz, CD2Cl2) d: 9.13 (1H, s),
8.2 (1H, brs), 7.95 (1H, d, J¼9.0 Hz), 7.47–7.42 (2H, m),
7.33–7.27 (1H, m), 2.68 (3H, s), 2.32 (3H, s); 13C NMR
(75 MHz, CD2Cl2) d: 126.8 (CH), 122.3 (CH), 121.5 (CH),
111.8 (CH), 30.0 (CH3), 27.4 (CH3); LC-MS (ESI, positive
ion mode): C14H13N4O (M2N2þH)þ requires m/e 253.1,
found 253.2.
3. (a) Sugimura, T. Mutat. Res. 1985, 150, 33–41. (b) Ohgaki,
H.; Matsukura, N.; Morino, K.; Kanachi, T.; Sugimura, T.;
Takayama, S. Carcinogenesis 1984, 5, 815–819.
(c) Eisenbrand, O.; Tang, W. Toxicology 1993, 84, 1–82.
4 (a) Yamazoe, Y.; Shimada, M.; Kamataki, T.; Kato, R. Cancer
Res. 1983, 43, 5768–5774. (b) Yamazoe, Y.; Tada, M.;
Kamataki, T.; Kato, R. Biochem. Biophys. Res. Commun. 1981,
102, 432–439. (c) Gilissen, R. A. H.; J, .; Bamforth, K. J.;
Stavenuiter, J. F. C.; Coughtrie, M. W. H.; Meerman, J. H. N.
Carcinogenesis 1994, 15, 39–45. (d) Shinohara, A.; Saito, K.;
Yamazoe, Y.; Kamataki, T.; Kato, R. Cancer Res. 1986, 46,
4362–4367. (e) Yamazoe, Y.; Shimada, M.; Kamataki, T.;
Kato, R. Biochem. Biophys. Res. Commun. 1982, 107,
165–172. (f) Snyderwine, E. G.; Wirth, P. J.; Roller, P. P.;
Adamson, R. H.; Sato, S.; Thorgeirsson, S. S. Carcinogenesis
1988, 9, 411–418. (g) Meerman, J. H. N.; Ringer, D. P.;
Coughtrie, M. W. H.; Bamforth, K. J.; Gilissen, R. A. H. J.
Chem.-Biol. Interact. 1994, 92, 321–328.
4.2. Kinetics and product studies
All kinetics were performed in 5 vol% CH3CN–H2O,
m¼0.5 (NaClO4) at 208C. Buffers used to maintain pH
were 0.02–0.06 M NaH2PO4/Na2HPO4, AcOH/NaOAc,
and HCO2H/NaHCO2 at pH.2.5 and HClO4 solutions at
pH#2.5. NaN3 solutions were prepared at high concen-
tration (0.45 M for NaN3), m¼0.5, at pH 7.00 and diluted
into the working range with the phosphate buffers of the
same pH and ionic strength. Detailed procedures and
solvent purifications can be found elsewhere.5
Stock solutions of 7 and 9 were made in DMF at
concentrations of ca. 1.0£1022 M. Injection of 15 mL of
these stock solutions into 3.0 mL of solution generated
reaction solutions of initial concentrations of ca. 5.0£
1025 M. Wavelengths monitored for kinetics studies were
245, 255, 265, and 320 nm at all pH for both 7 and 9. Initial
absorbance measurements for 7 for the spectrophotometric
titration were made at 254 nm. Initial absorbance measure-
ments for 9 for the spectrophotometric titration were made
at 255 and 265 nm. Absorbance vs time data were fit to a
standard first-order rate equation or consecutive first-order
rate equation to provide kobs. Peak area data for 9 obtained
by HPLC methods described below were also fit to a
5. (a) Novak, M.; Xu, L.; Wolf, R. A. J. Am. Chem. Soc. 1998,
120, 1643–1644. (b) Novak, M.; Kazerani, S. J. Am. Chem.
Soc. 2000, 122, 3606–3616. (c) Novak, M.; Toth, K.;
Rajagopal, S.; Brooks, M. E.; Hott, L. L.; Moslener, M.
J. Am. Chem. Soc. 2002, 124, 7972–7981. (d) Novak, M.;
Nguyen, T.-M. J. Org. Chem. 2003, in press. (e) Novak, M.;
Rajagopal, S. Advances in Physical Organic Chemistry;
Richard, J. P., Tidwell, T. T., Eds. Academic: New York,
2001; Vol. 36, pp 167–254.
6. (a) Molina, P.; Almendros, P.; Fresneda, P. M. Tetrahedron
Lett. 1993, 34, 4701–4704. (b) Akimoto, H.; Kawai, A.;

8010
S. Rajagopal et al. / Tetrahedron 59 (2003) 8003–8010
Nomura, H.; Nagao, M.; Kawachi, T.; Sugimura, T. Chem.
11. Novak, M.; Kazerani, S. J. Org. Chem. 1998, 63, 895–897.
12. Famulok, M.; Bosold, F.; Boche, G. Tetrahedron Lett. 1989,
30, 321–324.
13. At higher N23 concentrations, and at lower pH where 9Hþ
concentrations are significant, a second-order reaction with
N23 can be detected, but this reaction does not significantly
compete with first-order decomposition of 9 at neutral pH at
[N23 ]#10 mM.
Lett. 1977, 1061–1064. (c) Hibino, S.; Sugino, E.; Kuwada,
T.; Ogura, N.; Sato, K.; Choshi, T. J. Org. Chem. 1992, 57,
5917–5921.
7. Takeda, K.; Ohta, T.; Shudo, K.; Okamoto, T.; Tsuji, K.;
Kosuge, T. Chem. Pharm. Bull. 1977, 25, 2145–2146.
8. Schindler, W. Helv. Chim. Acta 1957, 40, 2156–2159.
9. (a) Saito, K.; Yamazoe, Y.; Kamataki, T.; Kato, R.
Carcinogenesis 1983, 4, 1547–1550. (b) Hashimoto, Y.;
Shudo, K.; Okamoto, T. Biochem. Biophy. Res. Commun.
1980, 96, 355–362.
14. Novak, M.; Rajagopal, S. Chem. Res. Toxicol. 2002, 15,
1495–1503.
15. Sugimura, T. Mutat. Res. 1985, 150, 33–41.
16. (a) Taft, R. W. J. Am. Chem. Soc. 1952, 74, 2729–2732.
(b) Revis, C. L.; Rajamaki, M.; Fishbein, J. C. J. Org. Chem.
1995, 7733, 77381.
10. (a) Hashimoto, Y.; Shudo, K.; Okamoto, T. J. Am. Chem. Soc.
1982, 104, 7636–7640. (b) Toth, K. PhD Thesis, Miami
University, 2001.