Quinone methide chemistry of prekinamycins: 13C- labeling, spectral global fitting and in vitro studies

Article abstract of DOI:10.1039/b903844b

In this article, we address the presence of the prekinamycin quinone methide using the techniques of spectral global fitting and the ¹³C-labeling of the reactive centre. Two-electron reduction of a prekinamycin affords a long-lived quinone meth

Full text of DOI:10.1039/b903844b

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Quinone methide chemistry of prekinamycins: 13C- labeling, spectral global
fitting and in vitro studies†
Omar Khdour and Edward B. Skibo*
Received 25th February 2009, Accepted 12th March 2009
First published as an Advance Article on the web 7th April 2009
DOI: 10.1039/b903844b
In this article, we address the presence of the prekinamycin quinone methide using the techniques of
spectral global fitting and the13C-labeling of the reactive centre. Two-electron reduction of a
prekinamycin affords a long-lived quinone methide species that was characterised spectrally. A
correlation was made between the calculated DE (kcal/mol) values for quinone methide
tautomerisation and cytostatic activity to support the postulate that the quinone methide plays a role in
prekinamycin biological activity. We also prepared a stable quinone methide of prekinamycin and
studied its solution chemistry directly.
Introduction
The chemistry and biology of the diazobenzo[b]fluorene-based
natural products (Fig. 1) have been the subject of intense study
due to the presence of the unusual diazo functional group.1
Much of these efforts have focused on the mechanistic role
of the diazo group in exerting antitumour2 and antibacterial
activity.3 The interest in understanding diazobenzo[b]fluorene
mediated biological activity was prompted by the discovery of the
marine natural product lomaiviticin A, Fig. 1.4 This glycosylated
homodimeric diazobenzo[b]fluorene possesses potent anticancer
activity against a wide range of cancer types, as well as activity
against Gram-positive bacteria.
Recent mechanistic studies of the role of the diazo group in
biological activity provided compelling evidence of one-electron5,6
and two electron7 reductive activation of prekinamycin with
loss of nitrogen. In both mechanisms, the resulting quinone
methide species (either in radical or neutral form) is postulated
to facilitate DNA cleavage. Earlier reported mechanisms for dia-
zobenzo[b]fluorene biological activity include the metal-mediated
formation of radical species.8,9
In this article, we address the mechanistic role of the diazo
group using the techniques of spectral global fitting,10,11 and
the13C-labelling of the reactive centre.12–15From these studies,
evidence emerged of a transient reactive quinone methide species
formed upon two-electron reductive activation, Fig. 1. We also
investigate the postulate that tautomerisation of the quinone
methide to the unreactive keto form will result in loss of biological
activity. Thus, the calculation of tautomeric equilibria should
predict prekinamycin biological activity. Finally, we investigate the
postulate that the quinone methide species requires acid catalysis
to trap a weak nucleophile.
Fig. 1 Examples of diazobenzo[b]fluorene natural products and the
putative prekinamycin quinone methide.
Our postulated mechanism for prekinamycin/kinamycin bi-
ological activity involves two-electron reductive activation and
quinone methide formation by nitrogen elimination as illus-
trated in Scheme 1. The two-electron reducing enzyme DT-
diaphorase mediates the reductive activation of other quinone
antitumour agents, such as mitomycin C,16 and could likewise
reduce diazobenzo[b]fluorene-based quinones. Similarly, other
naturally occurring quinones are functionalised to permit quinone
methide formation upon two-electron reduction and leaving group
Department of Chemistry and Biochemistry, Arizona State University,
Tempe, Arizona, USA 85287-1604. E-mail: eskibo@asu.edu; Fax: +1 480
965 2747; Tel: +1 480 965 3581
† Electronic supplementary information (ESI) available: 111H-NMR, 13C-
NMR, and mass spectra of all compounds. See DOI: 10.1039/b903844b
2140 | Org. Biomol. Chem., 2009, 7, 2140–2154
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
elimination.17,18 Biological activity may pertain to nucleophile
trapping by the quinone methide species as illustrated in Scheme 1.
Results and discussion
Synthesis of prekinamycin analogues 1–6
We carried out the preparation of the 13C-labeled prekinamycins
1–5 shown in Fig. 2 to assist in the identification of both the
transient quinone methide intermediate and the complex products
that could result from quinone methide decomposition. Finally, we
placed methoxy and hydroxy groups to influence quinone methide
stabilisation by internal hydrogen bonding.
Scheme 1 Prekinamycin quinone methide formation upon quinone
reduction and the quinone methide nucleophile trapping reaction.
The prekinamycin quinone methide is in equilibrium with its
keto tautomer, Scheme 2. Since the keto-tautomer is incapable of
trapping nucleophiles, prekinamycins whose quinone methide tau-
tomeric equilibria favour the keto form should possess diminished
biological activity. In contrast, prekinamycins that possess a rel-
atively stable quinone methide should possess biological activity.
Our prekinamycins analogues possess different hydrogen-bonding
capabilities that will influence quinone methide tautomeric equi-
libria, Fig. 2. Hartree–Fock calculations19 afforded DE values
for quinone methide tautomerisation that were correlated with
cytostatic activity against five cancer lines. Indeed, we were able to
predict the most active compound now undergoing in vivo trials at
the National Cancer Institute.
Fig. 2 Target 13C-labelled prekinamycins, * shows the site of the 13C-label.
We also required a stable prekinamycin quinone methide ana-
logue to carry out kinetic studies. Stable kinamycin-like quinone
methides known to occur in nature include the stealthins23 and
the momofulvenones,24 Fig. 3. The momofulvenones were not
considered suitable because they do not possess the aromatic D-
ring of the prekinamycins. However, the stabilising influence of
the stealthin 5-amino group inspired the synthesis of 5-methoxy
prekinamycin 6, Fig. 2.
Fig. 3 Stable quinone methide-like kinamycins.
The procedures for the preparation of 13C-labeled prekinamycin
analog 1* were adapted from the literature.25,26 We used isotopi-
cally pure K13CN and chloroacetic acid to prepare 50% 13C-
labeled malonic acid as outlined in Scheme 3.27 The aldehyde
7 was prepared as previously reported28 and then subjected to
a Knoevenagel condensation29 with the labeled malonic acid to
afford the 50% of the 13C-labelled cinnamate 8*.
Scheme 4 shows the sequence of reactions that was used to
convert 8* to 1*. Quinone 10* was prepared by coupling the
cyanophthalide 930 with cinnamate 8* via the 1,4-dipole annelation
reaction.31 The quinone product 10* was reduced to the corre-
sponding hydroquinone using sodium dithionite and methylated
with dimethyl sulfate to afford 11*. This electron rich species read-
ily cyclised to the benzo[b]fluorene system 12* via an acylium ion
generated using polyphosphoric acid. The addition of the diazo
Scheme 2 Tautomeric equilibrium of the prekinamycin quinone methide
and it keto form.
In a previous study, we determined that the quinone methide
of mitomycin C requires O-protonation to react.20 The Tomasz
group provided detailed experimental evidence for this scenario
as well.21 In fact, our Hartree–Fock calculations of the mitosene-
like quinone methide show that the methide reacting centre is
neutral, but becomes positive when O-protonation occurs.22 Does
the prekinamycin quinone methide behave likewise? Hartree–Fock
calculations presented in this article indicate that the reactive
5-centre is slightly anionic. Therefore, O-protonation of the
prekinamycin quinone methide may be required for nucleophile
trapping, at least for weak nucleophiles.
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 2140–2154 | 2141
©

View Article Online
Scheme 5 Synthesis of the 13C-labelled acetanthranil 15.
Scheme 3 Synthesis of the 13C cinnamate 8*.
Scheme 6 Synthesis of 13C-labelled 2.
diazonium salt was reductively cyclised with hydroquinone to
afford 19. Demethylation of 19 with BBr3 followed by preparation
of the hydrazone and then oxidation with Fetizon’s reagent32
afforded analogue 2*. Finally, the O- methylation of 2* afforded
3*.
The unreported prekinamycin analogs 4 and 5 were prepared
by following the procedure outlined in Scheme 7. Quinone 21 was
prepared by coupling the cyanophthalide 2030 with cinnamate 8
via the 1,4-dipole annelation reaction.31 The quinone product 21
was reduced to the corresponding hydroquinone using sodium
dithionite and methylated with dimethyl sulfate to afford 22.
This electron rich species readily cyclised to the benzo[b]fluorene
system 23 via an acylium ion generated using polyphosphoric acid.
The addition of the diazo group to 11 resulting in 4 was carried
out by hydrazone synthesis followed by oxidation with Fetizon’s
reagent.32 Finally, methylation of 4 afforded 5.
Scheme 4 Synthesis of 13C-labelled 1.
group to 12* resulting in 50% 13C-labelled 1* was carried out by hy-
drazone synthesis followed by oxidation with Fetizon’s reagent.32
The synthesis of 2* outlined in Schemes 5 and 6 was adapted
from the literature.33 The preparation of 14* shown in Scheme 5
was carried out as previously described34 with copper(I) cyanide
Cu13CN was used to incorporate the 13C-label into the acetan-
thranil final product 15*.
The synthetic sequence outlined in Scheme 6 began with the
lithiation of 16 and condensation by reverse addition to acetan-
thranil 15* giving amide 17*. The amide was then subjected to
basic hydrolysis with KOH in methanol to give aniline 18* that was
diazotised in situ with isoamyl nitrite in the presence of AcOH. The
We prepared the quinone methide analogue 6* by treating the
prekinamycin 2 with rhodium(II) acetate in methanol, Scheme 8.
The processes involved in product formation upon rhodium(II)
2142 | Org. Biomol. Chem., 2009, 7, 2140–2154
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Scheme 9 Mechanism of rhodium-catalysed insertion reactions.
24* and 12* were isolated together as quinone methide hydrolysis
products.
Spectral global fitting
The study of prekinamycin quinone methide formation and fate
is difficult because of the presence of two transient species,
the prekinamycin hydroquinone and quinone methide. Although
these species possess different UV-visible spectra, absorbance vs.
time measurements at a single wavelength may not provide rate
constant data for both of these reacting species. To carry out
this study we used spectral global fitting, a technique that has
been used to visualise and study photochemical intermediates.10,11
Spectral global fitting involves repetitive UV-visible scanning of
reaction mixtures with time resulting in a surface showing spectral
changes associated with the reaction. Fitting the entire range of
wavelengths (global fitting) to a rate law provides accurate rate
constants and intermediate spectra associated with the reaction
under study. We recently reported the use of spectral global
fitting to study a transient intermediate involved in CC-1065 A-
ring opening37 and quinone methide formation from pyrido and
pyrroloindoles.22
Scheme 7 Preparation of prekinamycins 4 and 5.
In order to generate the hydroquinone forms of 1* and 2* in
an anaerobic atmosphere, we utilized a Thunberg cuvette with a
quinone DMSO stock solution in the top port and a buffer solution
containing suspended 5% Pd on carbon in the bottom port. The
cuvette was purged with hydrogen and sealed. The reduction
reaction was initiated by mixing the ports. Quinone reduction
to the hydroquinone by catalytic hydrogenation is preferred
over reduction with a soluble reducing agent such as sodium
dithionite. Quinone methide trapping with sulfur nucleophiles
would accompany dithionite-mediated reductive activation.14
Repetitive UV-visible scanning of the Thunberg reaction of 2*
with time afforded the absorbance vs. time vs. wavelength surface,
shown in Fig. 4. Global fitting of this surface to a three-exponential
rate law (Abs = Ae-kt + Be-k¢t + Ce-k¢¢t + D) afforded the three rate
constants associated with quinone reduction (k), quinone methide
(26) formation from the hydroquinone (k¢), and quinone methide
disappearance (k¢¢), Scheme 10. The internal hydrogen bonding
capability of quinone methide 26 may be responsible for its buildup
in solution. Indeed, the analogue 1* did not show the buildup of
an intermediate in the global fitting experiment perhaps because of
the 6-methoxy group prevents this hydrogen bonding interaction.
Scheme 8 Rhodium catalysed reactions of Prekinamycin analogues.
treatment are outlined in Scheme 9. The first step involves
rhodium-mediated carbene formation followed by carbene in-
sertion into methanol as previously described for other diazo
compounds.35,36 Tautomerisation of the insertion product affords
the observed product 6*. We also carried out the same reaction
starting with 1* and 4. The formation of 24* from 1* is attributed
to hydrolysis of the 5- methoxy group during aqueous workup. The
formation of 25 likely results from addition of a second equivalent
of methanol to the 5-methoxy quinone methide followed by air
oxidation.
Calculations presented later in this article suggest that internal
hydrogen bonding may be responsible for the resistance of 24*
to tautomerisation to its keto form 12* (Scheme 4). Indeed, both
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 2140–2154 | 2143
©

View Article Online
(2.57 0.2) ¥ 10-2, k¢¢ = (2.00 0.21) ¥ 10-2; pH 7.59, k = (3.00
0.17) ¥ 10-2, k¢ = (3.01 0.17) ¥ 10-2, k¢¢ = (3.00 0.17) ¥ 10-2;
pH 8.08, k = (1.7 0.3) ¥ 10-2, k¢ = (1.7 0.4) ¥ 10-2, k¢¢ =
(1.76 0.03) ¥ 10-2; pH 8.83, k = (2.11 0.01) ¥ 10-2, k¢ =
(2.33 0.23) ¥ 10-2, k¢¢ = (2.4 0.3) ¥ 10-3; and pH 9.43, k =
(3.6 0.5) ¥ 10-2, k¢ = (3.7 0.4) ¥ 10-2. These data indicate that
the rate of quinone reduction (k) has an average value of 2.5 ¥
10-2 min-1 over the pH range studied. The rate of quinone methide
formation (k¢) is independent of pH with an average value of
2.6 ¥ 10-2 min-1. The protonation/deprotonation reaction of 2*H2
shown in Scheme 10 would have a net neutral transition state and
rates would be independent of pH. The rate of quinone methide
disappearance (k¢¢) decreases with increasing pH due to enolate
(26-) formation. At pH 9.43, the final product spectrum is that of
the enolate 26-.
The buildup of the quinone methide species 26 observed with
global fitting correctly suggested that NMR could detect the 13
C
resonance of the 5-position of the quinone methide. The buildup
of a species with a 13C-NMR chemical shift of 94 ppm at the
5-position confirmed that this transient intermediate is actually
the quinone methide. Furthermore, we confirmed the quinone
methide structure with high-resolution mass spectroscopy. Further
evidence of the structure of 26 is its UV-visible spectrum shown
in Fig. 5. This spectrum was calculated from the surface shown in
Fig. 4. This spectrum matched that of the stable quinone methide
analogue 6.
Fig. 4 Spectral global fitting surface obtained for the reaction of 2* in
anaerobic pH 7.5 phosphate buffer in the presence of H2 and 5% Pd on
charcoal. The quinone methide QM species is 26 in Scheme 10.
Scheme 10 Reactions responsible for the surface in Fig. 4.
In Thunberg reactions of 2* held between pH 7.2 and 9.4,
surfaces were obtained similar to that shown in Fig. 4. However,
reactions held at pH > 9 with phosphate buffer exhibited a stable
quinonemethidespecies and we fitthe surface to atwo-exponential
rate law (Abs = Ae-kt + Be-k¢t + C). The stability of the quinone
methide in this pH range is due to the formation of a stable
enolate 26-, see Scheme 11. Quinone methides can afford enolates
at pH values > 8 resulting in an electron rich system resistant to
nucleophilic attack.24,38
Fig. 5 Spectra obtained from the surface shown in Fig. 4: gray, spectrum
of quinone 2 at zero time and black, the spectrum of the quinone methide
intermediate 26, structure shown in Scheme 10.
Prekinamycin product studies using 13C NMR and isolation
We carried out these studies to confirm the presence of the quinone
methide intermediate and to determine the products arising from
this intermediate. The use of enriched 13C-NMR at the reactive
methide center rapidly provided a snapshot of the number and
structure of these products before further reactions occurred.
Eventually the NMR resonances of these products disappear due
to oxygen-mediated radical formation, a phenomenon previously
reported by Gould.39 In summary, we observed reactions that
are typical of quinone methides: nucleophile trapping, proton
trapping, and dimerization.38,40–44
Scheme 11 Ionic forms of quinone methide 26.
Rate data obtained from global fits at various pH values are
summarised below in min-1: pH 7.22, k = (2.2 0.6) ¥ 10-2, k¢ =
2144 | Org. Biomol. Chem., 2009, 7, 2140–2154
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Fig. 6 Enriched 13C NMR spectrum of the reaction of reduced 2*.
The 13C NMR snapshot of the reaction of reduced 2* in
anaerobic buffer followed by aerobic workup is shown in Fig. 6.
The identity of the species giving rise to the resonances in this
figure were determined based on their 13C chemical shift, isolation,
and high resolution MS, see Supplementary Information. The
structures determined from these studies, 26*–29*, are shown in
Scheme 12.
structural assignment, although tautomerisation to its keto form
may have occurred during workup.
The formation of 27* results from the trapping of the quinone
methide by water followed by oxidation and tautomerisation,
Scheme 13. The C-5 chemical shift of 27* is also shielded due to
enolate formation, 125 ppm rather than 164 ppm for the neutral
quinone methide. Another tautomerisation reaction converts 27*
to its keto form 28* that was readily isolated and identified without
decomposition.
Scheme 13 Mechanism of formation of 27* and 28* from quinone
methide 26.
The appearance of both 27* and 28* in the 13C-NMR spectrum
shown in Fig. 6 does not seen to agree with the DE values discussed
later in this article. Given the thermodynamic favourability of the
conversion of 27* to 28*, only the latter should be observed in
this spectrum. Actually, the spectrum shown in Fig. 6 does not
represent an equilibrium condition but rather a snap shot in time.
We isolated the crude reaction mixture by ethyl acetate extraction
and ran a 13C NMR in DMSO-d6. These aprotic solvents would
slow the tautomerisation process. Thus, the spectrum in Fig. 6
shows both transient species (quinone methide and enol tautomer)
as well as final products (keto tautomer and dimer). Eventually,
the resonances for the transient species disappear from solution.
The dimeric species 29* was isolated as a green solid, which
is 11H-NMR-silent due to radical formation in the presence
of oxygen. Gould39 also observed 11H-NMR silent radicals
in structurally-related benzo[b]fluorenone natural products. The
formation of dimer 29* is a classic quinone methide reaction that
was observed upon reductive activation of anthracyclines40 and
Scheme 12 Products resulting from the reaction of reduced 2* followed
by aerobic workup. We show the chemical shifts for the 13C-enriched C-5
centre.
Quinone methides can afford enolates at pH values > 7 resulting
in an electron rich system. The 13C-NMR chemical shifts for
enolates were calculated using the Hartree–Fock 3-21G basis set
using Spartan 2006 software. The quinone methide species 26*,
13C-NMR shown in Fig. 6, has a C-5 chemical shift of 94 ppm due
to enolate formation (predicted chemical shift of C-5 of the enolate
is 104 ppm). On the other hand, the neutral quinone methide C-5
centre is predicted to have a 13C chemical shift of 130 ppm. The
high resolution MS of this species supports the quinone methide
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 2140–2154 | 2145
©

View Article Online
Fig. 7 Enriched 13C NMR spectrum of the reaction of reduced 1*.
mitomycin C.45 We also observed quinone methide dimerisations
upon reductive activation of two other systems.12,13 The mechanism
of dimer formation could be either a radical or ionic one as
postulated previously for the prekinamycin quinone methide.6
Although, compounds 26*–29* were isolated and identified (see
Supplementary Information), the mass balance was only ~50%
due to decomposition to uncharacterised polymeric material that
could not be eluted off a silica gel chromatography column.
We also studied the reductive activation of the O-methylated
analogue 1* with the expectation that aerobic radical formation
will not be a factor. However, quinone methide formation from 1*
was not observed as with 2* possibly because in quinone methide
(26) stabilisation by internal hydrogen bonding is absent, see
Scheme 10. The 13C NMR snapshot of the reaction of reduced 1* in
anaerobic buffer followed by aerobic workup is shown in Fig. 7 and
the structures of the reaction products are shown in Scheme 14.
Both 12* and 24* were isolated in 80% yield without formation
of radicals (i.e. no broadening of resonances in the 11H-NMR).
Compound 12* is identical to the synthetic intermediate shown in
Scheme 4. The methoxy groups of 1 result in electron rich transient
species that do not afford enolates at the pH 7.4. Consequently,
the 13C chemical shifts shown in Fig. 7 and Scheme 14 are not
shielded due to anion formation.
observe its tautomer 30 in non-isolatable quantities. These product
studies indicate that the fate of the quinone methide derived
from 1* predominately involves nucleophile trapping by water.
The propensity of this quinone methide to trap nucleophiles is
consistent with the cytostatic activity exhibited by 1.
Kinetic study of quinone methide fate
A hydrolytic study of the stable quinone methide 6 was carried out
to provide insights into the reactivity of the prekinamycin quinone
methide. These studies were carried out in m = 1 (KCl) aerobic
borate and phosphate buffers at 0.05 to 0.2 M using spectral global
fitting from 250 to 600 nm to measure first-order rate constants.
◦
All reactions were carried out at 30.0 C and the proton activity
was measured with a pH meter. The results of these hydrolytic
studies supported the mechanism shown in Scheme 15.
Scheme 15 Mechanism for the hydrolysis of 6.
Spectral global fitting studies indicated that quinone methide
6 was converted to ketone 28 by a single first-order process.
Reactions held at pH values greater than 7 exhibited buffer
catalysis and a series of buffer dilutions were carried out at
constant pH values to determine lyate-only (water, hydroxide and
hydronium) rate constants. Fig. 8 shows buffer dilution plots with
boric acid buffers held at various pH values. The increasing slope
of these plots with decreasing pH values clearly indicates that
general acid catalysis is present. The y-intercept values of these
plots represent general acid catalysis by water (kH2O) as well as
specific- acid catalysis (kH) at lower pH values. Plotting the log of
Scheme 14 Products resulting from the reaction of reduced 1* followed
be aerobic workup. We show the chemical shifts for the 13C-enriched C-5
centre.
The absence of aerobic oxidation is a possible explanation of
the high mass balance of this reaction. The putative quinone
methide species was not observed in this reaction, but we did
2146 | Org. Biomol. Chem., 2009, 7, 2140–2154
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
neutral quinone methide with water acting as a general acid (kH2O
=
0.0009 min-1). The presence of general acid catalysis at pH > 8
occurs because equilibrium protonation of 6 to afford the cation
6H+ is not possible at these pH values. Rather, protonation by a
buffer acid occurs in concert with attack by water, see structure
shown in the inset of Scheme 15.
Our kinetic studies indicate that the quinone methide species
must be protonated, either by a specific or general acid, to
afford a cation before it can trap the water nucleophile. On the
other hand, the strongly nucleophilic sulfur anion could add to
the prekinamycin quinone methide without prior protonation.6
Further studies will be required to assess the role of the quinone
methide cation in biological alkylations.
Fig. 8 Buffer dilution plots for the hydrolysis of 6 in borate buffer.
lyate rate constants versus pH generated the pH rate profile shown
in Fig. 9. These data were computer fit to the rate law shown in
this figure to afford the solid line.
The thermodynamic cycle for quinone methide protonation and
tautomerisation
Previously, we reported the pKa for the mitomycin C carbocation-
quinone methide equilibrium to be 7.1,20 a value similar to the
pKa value we report above. The main reason for these high pKa
values is the formation of a high free energy quinone methide
species that is close in energy to its cation. Once the quinone
methide tautomerises to the ketone, the pKa of the protonated
ketone should shift to a lower value. In fact, the titration curve
shown in Fig. 11 reveals that the protonated ketone possesses a
negative pKa value.
Fig. 9 pH-Rate profile for the hydrolysis of 6.
The pH-rate profile in Fig. 9 is consistent with equilibrium
protonation of the quinone methide 6 to afford a cation 6H+
(kinetic pKa = 6.66). The pKa obtained from kinetic measure-
ments (kinetic pKa) may not be the same as the actual value
(thermodynamic pKa) because the former may include rate and
equilibrium constants.46 Since 6H+ persists in solution, we were
able to obtain a thermodynamic pKa (6.46) by spectrophotometric
titration, Fig. 10. The good agreement between the kinetic and
thermodynamic pKa values supports the mechanism in Scheme 15,
where both 6 and 6H+ react without intervening equilibrium
reactions. These findings indicate that the pH rate profile shown
in Fig. 9 is the result of an acid dissociation rather than another
process such as a rate determining step change.
Fig. 11 Titration of 28 at 500 nm using aqueous and Ho buffers.47
With acid dissociation constants measured for the cation-
quinone methide equilibrium (Fig. 9) and for the protonated
ketone tautomer (Fig. 10), we utilised a thermodynamic cycle to
determine the energy difference between the quinone methide and
the ketone tautomer, Scheme 16. The similar electronic effects of
the 5-hydroxy and 5-methoxy substituents suggest that the pKa of
28H+ would be similar to that measured for 6H+ (Fig. 10). This
cycle indicates that the ketone tautomer is more stable than the
quinone methide by -10.2 kcal/mol. We calculated the following
energy differences for quinone methide tautomerisation to the
ketone: -8.4 kcal/mol and -11.8 kcal/mol using Hartree–Fock
3-21G and STO-3G basis sets respectively. The average energy
difference of -10.1 kcal/mol (inset of Scheme 16) is in good
agreement with the measured value of -10.2 kcal/mol. Hartree–
Fock calculations in the following section utilised the 3-21G basis
set, which provided an accurate tautomerisation value in a minimal
calculation time.
Fig. 10 Titration of the quinone methide 6 at 500 nm.
The results enumerated above indicate that the pH < 7 plateau
is associated with water trapping by the cation (kH = 0.066 min-1)
and the pH > 8 plateau is associated with water trapping of the
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 2140–2154 | 2147
©

View Article Online
Scheme 16 Determination of the DE for tautomerisation by thermody-
namic cycle and calculation (inset).
Tautomerisation and prekinamycin cytostatic/cytotoxic activity
The trapping of cellular nucleophiles by the prekinamycin quinone
methide could result in cytostatic and perhaps cytotoxic activity.
If the prekinamycin quinone methide readily tautomerises to
its keto form, nucleophile trapping will not be possible and
there should be little or no activity against cells. To test this
hypothesis, we correlated biological activity against five human
cancer cell lines with the calculated DE (kcal/mol) values for
quinone methide tautomerisation. The results described below
indicate that compounds with relatively stable keto forms are less
active than those with relatively stable the quinone methide (enol)
forms.
Scheme 16 shows the prekinamycin analogues that were studied
and their corresponding enol/keto forms along with the calculated
DE (kcal/mol) values. Table 1 shows the concentrations of 1–5
that result in 50% growth inhibition (GI50) of five human cancer
cell lines. We judged the inactivity of the prekinamycins based on
the number of cell lines with a GI50 value too high to measure.
Conversely, an active prekinamycin would have measurable GI50
values in most if not all cell lines. Inspection of the data in Table 1
reveals that the level of cytostatic activity of 1, 3–5 depends on
the thermodynamic favourability of the quinone methide species
compared to the corresponding keto form.
Scheme 17 Quinone and keto tautomers derived from prekinamycins
1–5.
and 36 respectively in Scheme 17. The exception is prekinamycin 2,
which is cytostatic and possesses a relatively stable keto tautomer
32 compared to its quinone methide 26. The activity of 2 may
pertain to the kinetic stability of 26 due to internal hydrogen
bonding, see Scheme 10.
At this time, the National Cancer Institute has completed
screening prekinamycins 1–5 against a 60-cell human cancer
panel.48 The results merely reflect the data shown in Table 1.
Prekinamycin 1 is currently undergoing in vivo screens based on
its activity against this panel. In contrast, the NCI categorised
perkinamycin 4 as inactive.
The results enumerated above suggest that the quinone methide
species plays an important role in prekinamycin biological activity.
In fact, the quinone methide analogue resistant to tautomerisation
6 exhibited cytostatic activity, Table 1, and shows cytotoxic activity
against the 60-cell human cancer panel. Other factors beside
the presence of a quinone methide will control prekinamycin
The most cytostatic prekinamycins 1 and 5 are associated with
thermodynamically stable quinone methides 31 and 37 respectively
in Scheme 17. In contrast, the less active prekinamycins 3 and 4
are associated with thermodynamically stable keto tautomers 34
Table 1 Concentrations in mg per mL that cause 50% growth inhibition (GI50). GI50 values marked as > 10 mg per mL were too high to measure.
Highlighted compounds 3 and 4 are considered inactive
Parameter GI50
BxPC-3 Pancreas
MCF-7 Breast
SF-268 CNS
H460 NSC Lung
KM20L2 Colon
DU-145 Prostrate
1
2
3
4
5
6
0.52
1.3
> 10
4.2
0.43
1.1
> 10
7.3
0.53
2.8
> 10
> 10
0.89
1.2
0.6
> 10
> 10
0.59
3.3
0.46
3.1
> 10
> 10
1.3
4.4
0.6
1.6
> 10
0.66
1.5
0.5
2
0.41
1.6
2.2
1.2
2148 | Org. Biomol. Chem., 2009, 7, 2140–2154
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
activity against cancer cell lines. These include the compound’s
bioavailability as well as the presence of a cellular reducing enzyme
that will convert quinone to hydroquinone. The DT-diaphorase-
mediated two-electron bioreductive activation process occurs in
quinone-based antitumor agents such as mitomycin C49 and may
apply to the kinamycins as well.
-12 kcal/mol, then the parent prekinamycin’s cytostatic activity
was absent in most cell lines (i.e. > 10 mg/mL). On the other
hand, prekinamycin’s with quinone methide tautomerisation DE
values ~8 kcal/mol (i.e. less tendency to tautomerise) were potent
cytostatic agents. An exception to this rule is the cytotoxic
compound 2, whose quinone methide ketonization reaction has
a DE of -9.9 kcal/mol. The activity of 2 may pertain to the kinetic
stability of its quinone methide 26 in solution (see Fig. 4) due to
an internal hydrogen bonding interaction.
The results of DE calculations shown in Scheme 17 show
that internal hydrogen bonds may influence the thermodynamics
of quinone methide tautomerisation in some instances. For the
prekinamycin quinone methide without internal hydrogen bonds
(33), the keto tautomer is favoured by -11.9 kcal/mol. The
negative DE values for tautomerisation of quinone methides
26 and 37 likewise indicate that the corresponding keto forms
are favoured. Internal hydrogen bonding is present in both of
these quinone methides as well as in their corresponding keto
tautomers balancing out any stabilising interactions. In contrast,
the prekinamycin quinone methides 31 and 37 possess an internal
hydrogen bonding interaction involving the 1-methoxy group, but
none in the corresponding keto forms. Furthermore, the angle of
this internal hydrogen bond (162◦) is close to the co-linearity (180◦)
of the ideal hydrogen bond. It is noteworthy that the resistance of
24* to tautomerisation to its 5-one derivative 12* in aqueous media
(see Scheme 14) is likely due to this internal hydrogen bonding
interaction.
Experimental
General
All TLCs were performed on silica gel plates using a variety
of solvent systems and a fluorescent indicator for visualisation.
IR spectra were taken as KBr pellets and only the strongest
absorbances were reported. 1H and 13C NMR spectra were
obtained with a 300 or 500 MHz spectrometer. All chemical
shifts are reported relative to TMS. High Resolution mass spectra
Nebraska Center for Mass Spectrometry, University of Nebraska-
Lincoln.
Known compounds that were prepared with a 13C-label
(Schemes 3–6) were characterised with 11H-NMR, enriched 13C-
NMR, and high resolution mass spectroscopy. New compounds
(Scheme 7) were fully characterised. Unstable compounds pro-
duced in the course of the reductive activation of 1 and 2 (26, 27, 29,
and 30) could not be purified and were characterised only by 13C-
NMR and high resolution mass spectroscopy. New compounds
from these reactions that could be purified (24* and 28*) were
fully characterised. The Supplementary Information Section has
the 13C-NMR and high resolution mass spectra of all compounds.
Conclusion
Previous studies in the kinamycin area have postulated that
a quinone methide species is involved in kinamycin biological
activity.5–7 In this article, we use the methods of spectral global
fitting and 13C incorporation at the methide centre to provide
evidence of this species. Upon two-electron reduction of the
prekinamycin to afford its hydroquinone form, proton transfer
to the 5-carbon centre is followed by elimination of molecular
nitrogen to afford the quinone methide species.
Kinetic studies of quinone methide formation and fate. The
stock solution of 2 was prepared by diluting 4 to 5 mg in a 5 mL
volumetric flask with dry dimethyl sulfoxide (DMSO) to afford
solutions with concentrations of 2.5 to 3.3 mM. All kinetic studies
were carried out with buffers prepared from doubly distilled water.
Final concentrations of the 2 in reaction mixtures ranged from
83 mM to 110 mM. DMSO concentrations in reaction mixtures
were kept < 2% to minimise solvent effects. The following buffer
systems were used to hold pH: 1M acetate buffer (pH = 4 to 6.5),
0.2 M phosphate buffer (pH = 6.5 to 8.5), and 0.2 M borate buffer
(pH 7.5 to 10). All buffers were maintained at an ionic strength (m)
of 1.0 with KCl. The temperature of all the kinetic measurements
were maintained at 30.0 0.3 ◦C with circulating water from a
thermostatic water bath.
The reactions were carried out in Thunberg cuvettes in the
following way: To the bottom port of the Thunberg was added
2.8 mL of buffer and 100 ml of a sonicated suspension consisting
of 50 mg of 5% Pd on carbon in 10 mL of water. The amount of
catalyst added to the bottom port was 0.5 mg. To the top port
of the Thunberg cuvette, we added 100 mL of the DMSO stock
of 2. The cuvette was sealed with Apiazon L grease and Teflon
tubes from a hydrogen source threaded into the top and bottom
ports. After purging the ports with hydrogen gas for 5 min, the
cuvette was sealed and equilibrated at 30 ◦C in the thermostated
spectrometer. We started the reaction by combining the top and
bottom ports.
This study give credence to a two-electron bioreductive acti-
vation process that would afford the quinone methide species.
The DT-diaphorase-mediated two-electron bioreductive activa-
tion process occurs in quinone-based antitumor agents such as
mitomycin C49 and may apply to the kinamycins as well. Hypoxia
due to low blood flow50 and/or the unusually high expression of
the quinone 2-electron reducing enzyme DT-diaphorase in some
histological cancer types49,51–54 contribute to the tumour’s tendency
to reduce quinones. This conclusion does not dismiss the single-
electron reductive activation process documented in an earlier
publication.6 Cytochromes could mediate single-electron reductive
activation to afford quinone methide radicals in some cell lines.
Our kinetic studies of the hydrolysis of the quinone methide
analogue 6 revealed that O-protonation of this species results in a
titratable and observable cation. The pKa value of this protonated
quinone methide species is near neutrality due to the high energy
of the quinone methide species compared to its keto tautomer, see
the thermodynamic cycle in Scheme 16. Previously, we observed
similarly high pKa values for the protonated quinone methide of
mitomycin C20 and of pyrroloindole and pyridoindole analogs.22
A correlation was made between the calculated DE (kcal/mol)
values for quinone methide tautomerisation and cytostatic ac-
tivity against five human cancer cell lines. If quinone methide
tautomerisation to the ketone has a calculated DE of -11 to
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 2140–2154 | 2149
©

View Article Online
Reactions were monitored between 240 nm and 600 nm with an
OLIS UV-visible spectrophotometer equipped with spectral global
fitting software. The reaction times varied from 400 to 600 min.
depending on the pH of the reaction. When the reactions were
complete the pH of the samples were taken and recorded for use
on the pH-rate profile.
Methyl-1,4,8-trimethoxy-3-(2-methoxy-4-methylphenyl)-naph-
thalene-2-carboxylate-50%-13CO2 (11*). The procedure for
preparing this intermediate was adapted from the literature:29 yield
90%; 1H NMR (CDCl3) d 7.77 (1 H, d, J = 8.0 Hz), 7.46 (1 H, t,
J = 8.0 Hz), 7.13 (1 H, d, J = 8.0 Hz), 6.92 (1 H, d, J = 8 Hz),
6.81 (1 H, d, J = 8.0 Hz), 6.78 (1 H, s), 4.00 (3 H, s), 3.88 (3 H, s),
3.71 (3 H, s), 3.57 (3 H, t, 13COOCH3, J = 1.8 Hz), 3.55 (3 H, s);
13C NMR (CDCl3) d 167.85 (13COOCH3); HREI MS m/z (50%
13C enriched) calcd for 12C2213C1H24O6 397.1606, found 397.1604,
calcd for C23H24O6 396.1574, found 396.1574.
Malonic acid-50%-1-13C. The procedure for preparing 50%
13C-labeled malonic acid from K13CN and chloroacetic acid was
adapted from the literature synthesis of 14C-malonic acid:27 80%
yield; 1H NMR (DMSO-d6) d 3.25 (2H, d, J = 7.8 Hz, long range
13C splitting); 13C NMR (DMSO-d6) d 169.35 (13COOH).
6,11-Dihydroxy-1,7-dimethoxy-3-methyl-5H-benzo[b]fluoren-5-
one-50%-5-13C (12*). The procedure for preparing this interme-
diate was adapted from the literature:26 yield 75%; mp = 225–
226 ◦C; IR (KBr), 3295, 2915, 2840, 1770, 1662, 1630, 1605, 1453,
1332, 1305 cm-1; 1H NMR (CDCl3) d 10.54 (1 H, bs), 9.37 (1 H,
s), 7.79 (1 H, d, J = 8.4 Hz), 7.47 (1 H, t, J = 8.1 Hz), 7.23 (1 H,
s), 6.88 (1 H, t, J = 7.9 Hz), 4.06 (3 H, s), 4.02 (3 H, s), 2.37 (3 H,
s); 13C NMR (CDCl3) d 192.49 (13C-ketone); HREI MS m/z (50%
13C enriched) calcd for 12C1913C1H16O5 337.1031, found 337.1030,
calcd for 12C20H16O5 336.0998, found 336.0998.
Methyl 2-methoxy-4-methylcinnamate-50%-13CO2(8*). The
preparation was carried out from the 13C labeled malonic acid
by the two-step procedure outlined below. The procedure for
preparing 2-methoxy-4-methylcinnamic acid was adapted from
the literature,29 but with an altered workup. A solution of
13C-labeled malonic acid (1.0 g, 9.6 mmol), aldehyde 7 (1.7 g,
11.3 mmol), 50 ml of piperidine and 1.0 mL of pyridine was
heated at reflux for 4 h. The reaction mixture was acidified with
concentrated HCl followed by adding 15 mL of diethyl ether to
dissolve the resulting precipitate. The organic phase was extracted
with 3 ¥ 10 mL of 1 M sodium hydroxide. The combined aqueous
fractions were then acidified with concentrated HCl resulting
in crystallization of the cinnamic acid derivative. The white
crystals were filtered off and dried to afford 1.1 g of product.
The aqueous liquor was extracted with 3 ¥ 20 mL of diethyl
ether, the ether extracts were dried (Na2SO4), and concentrated to
5-Diazo-1,7-dimethoxy-3-methyl-5H -benzo[b]fluoren-6,11-
dione-50%-5-13C(1*). The procedure for preparing this final
1
product was adapted from the literature:25 yield 60%; H NMR
(CDCl3) (500 MHz) d 7.94 (1 H, d, J = 8 Hz), 7.67 (1 H, t,
J = 8 Hz), 7.24 (1 H, d, J = 8 Hz), 6.89 (1 H, s), 6.64 (1 H,
s), 4.03 (6 H, s), 2.44 (3 H, s); 13C NMR (500 MHz) (CDCl3)
d 180.66, 177.84, 159.77, 156.97, 139.28, 137.96, 136.68, 136.03,
135.05, 126.95, 120.56, 120.14, 117.64, 116.53, 111.14, 109.33,
1
give an additional 0.4 g of product: overall yield 81%; H NMR
69.99 (diazo), 56.43, 56.01, 53.40 (CH2Cl2), 22.02; HREI MS m/z
(DMSO-d6) d 12.21 (1H, s), 7.65 (1H, d, J = 5.5 Hz), 7.56 (1H,
d, J = 8.1 Hz), 6.91 (1H, s), 6.81 (1H, d, J = 8.1 Hz), 6.47 (1H,
d, J = 5.5 Hz), 3.85 (3H, s), 2.33 (3H, s); 13C NMR (DMSO-d6)
d 168.91 (13COOH); HREI MS m/z (50% 13C enriched) calcd for
12
(50% 13C enriched) calcd for
C
13C1H14 N2O4 347.0987, found
19
347.0987, calcd for 12C20H16O5 346.0954, found 346.0954.
2-Nitrobenzonitrile-13CN (13*). The procedure for preparing
this intermediate from Cu13CN and 2-bromonitrobenzene was
adapted from the literature:34 yield 90%; 1H NMR (CDCl3) d 8.33
(1 H, m), 7.93(1 H, m), 7.84 (2 H, t); 13C NMR (CDCl3) d 114.85;
HREI MS m/z (100% 13CN enriched) calcd for 12C613C1H4 N2O2
149.0306, found 149.0303.
12
12
C
10
13C1H12O3 193.0820, found 193.0823; calcd for C11H12O3
192.0784, found 192.078.
A solution of the cinnamic acid obtained as described above
(1 g, 5.12 mmol) and concentrated sulfuric acid (1 mL) in methanol
(60 mL) was heated at reflux for 12 h. The reaction was then cooled
and diluted with water and extracted with methylene chloride (5 ¥
50 ml), the combined extracts were washed with saturated sodium
bicarbonate solution (100 ml), dried Na2SO4, and concentrated to
an oil. The oil solidified overnight to afford a white solid (970 mg)
of 8*: 90% yield; 1H NMR (CDCl3) d 7.96 (1H, dt, J = 16.5 Hz),
7.38 (1H, d, J = 7.8 Hz), 6.76 (1H, d, J = 7.8 Hz), 6.70 (1H, s),
6.49 (1H, dt, J = 16.5 Hz), 3.85 (3H, s), 3.77 (3H, s), 2.35 (1H, s);
13C NMR (CDCl3) d 168.13 (13COOCH3); HREI MS m/z (50%
13C enriched) calcd for 12C1113C1H14O3 207.0976, found 207.0974,
calcd for 12C12H14O3 206.0943, found 206.0942.
2-Nitrobenzoic acid-13COOH (14*). The procedure for prepar-
ing this compound was adapted from the literature:34 yield 70%;
1H NMR (CDCl3) d 8.33 (1 H, m), 7.93 (1 H, m), 7.84 (2 H, t);
13C NMR (DMSO-d6) d 168.02; HREI MS m/z (100% 13COOH
enriched) calcd for 12C613C1H5 NO4 168.0252, found 168.0254.
2-Methyl-3,1-benzoxazin-4-one-4–13C (15*). This compound
was prepared in two steps: A mixture of 10*(1.0 g, 6 mmol), 500 mg
of 5% Pd on carbon, and 100 mL of methanol, was reduced for
1.5 h under 50 psi H2. The catalyst was removed by filtration
through Celite, and the mixture was concentrated to a residue.
No attempt was made to purify the resulting anthranilic acid; this
product was dissolved in 15 mL of acetic anhydride and refluxed
for 2 h. The reaction mixture was cooled to room temperature and
dried under vacuum to afford crude product, which was extracted
(4 ¥ 20 mL) with diethyl ether. The extracts were dried (Na2SO4),
concentrated, and crystallised from methylene chloride/hexane:
700 mg (72%) yield; mp 181–182 ◦C; IR (KBr) cm-1: 3349, 1701,
Methyl 3-(2-methoxy-4-methylphenyl)-8-methoxy-1,4-naphtho-
quinone-2-carboxylate-50%-13CO2(10*). The procedure for
preparing this intermediate was adapted from the literature:29
1
yield 50%; H NMR (CDCl3) d 7.78 (1H, dd, J = 8,1 Hz), 7.71
(1H, t, J = 8.0 Hz), 7.32 (1H, dd, J = 8,1 Hz), 7.07 (1H, d, J =
8 Hz), 6.79 (1H, d, J = 8.0 Hz), 6.75 (1H, s), 4.00 (3H, s), 3.73
(3H, s) 3.69 (3H, s); 13C NMR (CDCl3) d 164.92 (13COOCH3);
12
HREI MS m/z (50% 13C enriched) calcd for
C
13C1H18O6
20
12
1
367.1137, found 367.1141, calcd for C21H18O6 366.1103, found
366.1103.
1640; H NMR (DMSO-d6) d 8.11 (1 H, m), 7.93 (1 H, m), 7.56
(2 H, t), 2.40 (3 H, s); 13C NMR (DMSO-d6) d 160.05 (ester);
2150 | Org. Biomol. Chem., 2009, 7, 2140–2154
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
HREI MS m/z (100% 4-13C): calcd for 12C813C1H7 NO2, 162.0510;
found, 162.0505.
2030 (0.397 g, 2.5 mmol) in dry THF (10 mL) was added to a
solution of lithium diisopropylamine (LDA). LDA was prepared
from diisopropylamine (4.25 mL, 3 mmol), and butyllithium
(1.75 mL, 2.81 mmol) in dry THF (25 mL) at -78 ◦C. The mixture
was stirred for 1 h, and a yellow precipitate appeared. A solution
of the cinnamate 8 (0.515 g, 2.5 mmol) in dry THF was added at
-78 ◦C. The mixture was warmed to room temperature over 3 h and
stirred overnight. A solution of p-toluenesulfonic acid (0.475 g,
2.5 mmol) in dry THF (5 mL) was added, and the resulted mixture
was poured into iced water and brought to pH 2.0 using 1 N HCl.
The resulting solution was extracted with ethyl acetate (4 ¥ 50 mL).
The combined organic layers were dried over Na2SO4, filtered,
and evaporated to give orange solid. The solid was dissolved in
0.5 M KOH (15 mL) and THF (15 mL), and air was bubbled
through the solution for 2 h. The resulted precipitate was filtered
and extracted with ethyl acetate (3 ¥ 50 mL) and the combined
organic layers were dried over Na2SO4, filtered, and evaporated to
give an orange solid. Crude product was purified by flash silica
gel chromatography (hexane/ethyl acetate 50:50) as the eluent.
The product was recrystallised from methylene chloride/hexane
to afford 21 as an orange crystalline solid: 560 mg (65%) yield; mp
167–169 ◦C; TLC (chloroform) Rf = 0.38; IR (KBr), 3076, 2956,
2847, 1765, 1700, 1290, 1126, 1033, 705 cm-1; 1H NMR (DMSO-
d6) d 8.11 (4 H, m), 7.16 (1 H, d, J = 7.8 Hz), 6.94 (1 H, s), 6.84
(1 H, d, J = 7.8 Hz), 3.68 (3 H, s), 3.84 (3 H, s), 2.36 (3 H, s);
HREI MS m/z calcd for C20H16O5 336.0998, found 336.0981.
3-(2¢-N-Acetaminobenzoyl-13C-CO)-1,4,5-trimethoxynaphtha-
lene (17*). The procedure for preparing this intermediate was
adapted from the literature:33 yield 50%; 1H NMR (CDCl3) d 11.65
(1 H, bs), 8.76 (1 H, d, J = 8.7 Hz), 7.93 (1 H, d, J = 9 Hz), 7.56
(3 H, m), 6.99 (2 H, m), 6.62 (1 H, d, J = 3.9 Hz), 3.98 (3 H, s), 3.94
(3 H, s), 3.71 (3 H, s), 2.30 (3 H, s); 13C NMR (CDCl3) d 201.31
(13C-5-ketone); HREI MS m/z (100% 13C-ketone enriched): calcd
for 12C2113C1H21NO5, 380.1453; found, 380.1461.
3-(2¢-Aminobenzoyl-13C-CO)-1,4,5-trimethoxynaphthalene
(18*). The procedure for preparing this intermediate was
adapted from the literature:33 yield 98%; 1H NMR (CDCl3) d 7.91
(1 H, dd, J = 7.2, 1.2 Hz), 7.46 (1 H, t, J = 7.5 Hz), 7.34 (1 H, m),
7.25 (1 H, t, J = 7.2 Hz), 6.97 (1 H, d, J = 6.9 Hz), 6.73 (1 H, d,
J = 8.4 Hz), 6.65 (1 H, d, J = 3.9 Hz), 6.53 (1 H, t, J = 7.8 Hz),
3.98 (3 H, s), 3.93 (3 H, s), 3.74 (3 H, s); 13C NMR (CDCl3) d
199.44 (13C-ketone); HREI MS m/z (100% 13C-ketone enriched):
calcd for 12C1913C1H19NO4, 338.1348; found, 338.1359.
6,7,11-Trimethoxy-5H-benzo[b]fluoren-5-one-5-13C (19*). The
procedure for preparing this intermediate was adapted directly
from the literature:33 yield 70%; 1H NMR (CDCl3) d 7.98 (1 H, d,
J = 7.8 Hz), 7.76 (1 H, dd, J = 5.1, 2.7 Hz), 7.70 (1 H, d, J =
8.4 Hz), 7.58 (1 H, m), 7.36 (1 H, t, J = 7.2 Hz), 7.25 (1 H, s), 6.92
(1 H, d, J = 8.4 Hz), 4.07 (3 H, s), 4.00 (3 H, s), 3.98 (3 H, s); 13
C
NMR (CDCl3) d 190.29 (13C-ketone).
Methyl-1,4-dimethoxy-3-(2-methoxy-4-methylphenyl)-naphtha-
lene-2-carboxylate (22). The procedure for preparing this in-
termediate was adapted from the literature.29 A mixture of the
quinone 21 (0.5 g 1.48 mmol), dimethyl sulfate (5.0 mL), anhydrous
K2CO3 (8.0 g), and sodium dithionite (1.3 g) in acetone (70 mL)
under nitrogen was heated at reflux for 20 h. Water (1.5 mL)
was added dropwise, and the mixture was heated at reflux an
additional 3 h and then cooled to room temperature. Acetone
was evaporated from the mixture and water (50 mL) and 10% aq
NaOH (5 ml) were added. After stirring for 2 h, the mixture was
extracted with ether (3 ¥ 50 mL). The combined organic layers
were dried over Na2SO4, filtered, and evaporated to give a yellow
oil. The product was purified by flash silica gel chromatography
(hexane/ethyl acetate 60:40) as the eluent. Compound 10 was
crystallised from methylene chloride/hexane to afford the prod◦uct
as a white crystalline solid: 500 mg (91%) yield; mp 126–127 C;
TLC (chloroform) Rf = 0.28; IR (KBr) 3000, 2945, 2836, 1732,
1437, 1355, 1301, 1066, 779 cm-1; 1H NMR (CDCl3) d 8.17 (2 H,
m), 7.56 (2 H, m), 7.14 (1 H, d, J = 7.2 Hz), 6.83 (1 H, d, J =
6.7 Hz), 6.78 (1 H, s), 4.02 (3 H, s), 3.73 (3 H, s), 3.58 (3 H, s),
3.56 (3 H, s), 2.40 (3 H, s); HREI MS m/z calcd for C22H22O5,
366.1467; found, 366.1479.
5-Diazo-4-hydroxy-5H-benzo[b]fluoren-6,11-dione-5-13C (2*).
The procedure for preparing this intermediate was adapted from
the literature:33 yield 70%; 1H NMR (CDCl3) d 12.10 (1 H, s), 8.53
(1 H, m), 7.75 (1 H, d, J = 7.2 Hz), 7.62 (1 H, d, J = 8.7 Hz), 7.57
(1 H, d, J = 6 Hz), 7.45 (2 H, m), 7.21 (1 H, d, J = 7.8 Hz); 13
C
NMR (CDCl3) d 68.50 (13C-diazo); HREI MS m/z (100% 5-13C
enriched) calcd for 12C1613C1H8N2O3 289.0569, found 289.0576.
5-Diazo-4-methoxy-5H-benzo[b]fluoren-6,11-dione-5-13C (3*).
A mixture of 2* (25 mg, 0.086 mmol), CH3I (1 mL) and K2CO3
(1.5 g) in dry DMF (10 mL) was stirred at room temperature
for 24 hours. The K2CO3 was removed by filtration and the
solution was diluted with water and EtOAc. The layers were
separated and the organic phase was washed successively with
water and brine. The solvent was dried (Na2SO4), filtered and
removed under vacuum, where the crude product purified by flash
chromatography to give the title compound 3* (20 mg): yield 76%;
mp dec. ≥ 207 ◦C; TLC (chloroform) Rf = 0.32; IR (KBr), 3049,
2940, 2847, 2098, 1643, 1577, 1526, 1432, 1272, 1201, 1131, 1038,
1
755 cm-1; H NMR (CDCl3) d 8.49 (1 H, m), 7.92 (1 H, d, J =
6.9 Hz), 7.70 (1 H, t, J = 6 Hz), 7.55 (1 H, m), 7.42 (2 H,
m), 7.28 (1 H, d, J = 6.2 Hz); 13C NMR (CDCl3) d 180.85,
179.73, 160.31, 137.12, 135.24, 134.14,133,47, 130.23, 130.11,
6-Hydroxy-1,11-dimethoxy-3-methyl-5H-benzo[b]fluoren-5-one
(23). The procedure for preparing this intermediate was adapted
from the literature.26 A mixture of 10 (0.5 g, 1.36 mmol) and 20 g of
polyphosphoric acid was heated at 90 ◦C for 1 h and then poured
over ice. The resulting slurry (reddish precipitate) was stirred for
20 min, diluted with water, and extracted with ethyl acetate (3 ¥
50 mL). The combined organic phase was washed with water and
brine, dried (Na2SO4), filtered and concentrated in vacuo to give 23
as a red solid that was purified by flash silica gel chromatography
using methylene chloride as the eluant: 320 mg (73%) yield; mp
126.98, 126.63, 125.80, 120.57, 119.83, 118.37, 117.28, 69.27 (13C-
12
diazo), 56.46; LRFAB pos ion m/z calcd for
C
13C1H12N2O3
17
(M + 2, hydroquinone) 305.088, found 305.1, HRFAB pos ion
12
m/z (100% 13C-diazo enriched) calcd for
C
13C1H12O3 (M + 2,
17
hydroquinone) 305.0881, found 305.0862.
Methyl-3-(2-methoxy-4-methylphenyl)-1,4-naphthoquinone-2-
carboxylate (21). The procedure for preparing this intermediate
was adapted from the literature.29 A solution of the cyanophthalide
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 2140–2154 | 2151
©

View Article Online
248–249 ◦C; TLC (dichloromethane/methanol 95:5) Rf = 0.56; IR
(KBr) 3251, 2994, 2940, 2836, 1705, 1590, 1345, 1033, 782 cm-1;1H
NMR (CDCl3) d 9.78 (1 H, s), 8.20 (2 H, m), 7.58 (2 H, m), 6.87
(1 H, s), 4.17 (3 H, s), 4.09 (3 H, s), 2.38 (3 H, s); 13C NMR (CDCl3)
d 190.06, 151.35, 151.30, 142.87, 139.87, 137.59, 130.90, 130.44,
128.95, 128.82, 127.16, 124.63, 123.54, 120.07, 118.69, 117.82,
115.26, 62.76, 56.88, 21.54; HREI MS m/z calcd for C20H16O4,
320.1049; found, 320.1047.
taken up in ethyl acetate. The organic layer was washed with
water and brine, dried over Na2SO4, filtered and concentrated. The
crude product was subjected to preparative silica gel thin-layer
chromatographic separation using hexane/ethyl acetate (70:30)
as eluent followed by recrystallisation from chloroform/hexane
to afford 6* as a major product: 30 mg (74%) yield; mp, dec.
◦
≥ 154 C; TLC (dichloromethane/methanol 95:5) Rf = 0.20; IR
(KBr), 3399, 3049, 2923, 2858, 1645, 1585, 1514, 1454, 1323, 1290,
1
1224, 1164, 1039, 1022, 705 cm-1; H NMR (DMSO-d6) d 10.47
5-Diazo-1-hydroxy-3-methyl-5H-benzo[b]fluoren-6,11-dione (4).
The procedure for preparing this prekinamycin analogue was
adapted from the literature.33 Compound 23 (100 mg, 0.31 mmol)
was dissolved in 20 mL of dry methylene chloride under nitrogen
at -78 ◦C. To this solution BBr3 (1 M in methylene chloride,
0.630 mL) was added dropwise at -78 ◦C. The mixture was stirred
overnight, allowing the reaction to warm to room temperature.
The product mixture was quenched with ice water and then diluted
with ethyl acetate (50 mL). The layers were separated and the
organic phase dried over Na2SO4. The solvent was removed under
vacuum and the crude product carried over to next step.
The above product was immediately dissolved in 15 mL of
ethanol with anhydrous hydrazine (0.6 mL). The reaction mixture
was refluxed for 0.5 h under nitrogen. The reaction was cooled to
room temperature and the solvent was removed under reduced
pressure to afford the crude hydrazone. The crude hydrazone
was suspended in 10 mL of methylene chloride with 0.4 mL
of triethylamine. Fetizon’s reagent (3 g) was added at once and
the reaction mixture was stirred at room temperature for 5 min.
The suspension was filtered and the filtrate was evaporated under
reduced pressure. Chromatography of the residue on silica gel
(ethyl acetate:methylene chloride, 20:80) afforded pure 4: 65 mg
(69%) yield; mp dec. ≥ 202 ◦C; TLC, (dichloromethane/methanol
95:5) Rf = 0.68; IR (KBr) 3420, 3044, 2929, 2093 (diazo), 1645,
1574, 1443, 1257, 1050, 711 cm-1; 1H NMR (CDCl3) d 12.07 (1 H,
s), 8.50 (1 H, m), 7.73 (1 H, dd, J = 8.7, 6.6 Hz), 7.60 (2 H, m),
7.43 (2 H, m), 7.19 (1 H, dd, J = 8.7, 6.6 Hz), 1.52 (3 H, s); HREI
MS m/z calcd for C18H10N2O3, 302.0691; found, 302.0596.
(1 H, s), 8.05 (1 H, d, J = 7.8 Hz), 7.74 (1 H, d, J = 7.8 Hz), 7.53
(3 H, m), 7.28 (1H, t, J = 7.8 Hz), 6.90 (1 H, d, J = 7.8 Hz), 4.51
12
(3H, d, J = 4.2 Hz); HREI MS m/z calcd for
C
13C1H12O4
17
12
293.0769, found 293.0765. Anal. Calcd for:
74.97; H, 4.14. Found: C, 74.78; H, 4.08.
C
13C1H12O4: C,
17
5,11-Dihydroxy-3-methyl-1,7-dimethoxy-6H-benzo[b]fluoren-6-
one (24). To a suspension of compound 1 (40 mg, 0.115 mmol)
in 20 mL of methanol was added 3 mg of Rh2(OAc)4. This
mixture was heated at 80 ◦C overnight. The methanol solvent
was evaporated and the residue was taken up in ethyl acetate.
The organic layer was washed with water and brine, dried
over Na2SO4, filtered and concentrated. The crude product was
subjected to preparative silica gel thin-layer chromatographic
separation using chloroform/methanol (96:4) as eluent followed
by recrystallisation from chloroform/hexane to afford 24 as a
major product: 26 mg (67%) yield; mp, dec. ≥ 209 ◦C; IR (KBr),
3437, 3202, 2923, 2852, 1629, 1579, 1465, 1295, 1197, 1110,
1
1066 cm-1; H NMR (400 MHz) (CDCl3) d 10.64 (2H, s), 7.60
(1H, d, J = 8 Hz), 7.49 (1H, t, J = 8 Hz), 6.97 (1H, d, J = 8.4 Hz),
6.63 (2H, d, J = 6.8 Hz), 4.09 (3H, s), 3.82 (3H, s), 2.21 (3H, s);
13C NMR (400 MHz) (CDCl3) d 180.8, 161.1, 151.6, 150.8, 150.1,
142.7, 141.0, 138.3, 136.7, 133.8, 121.2, 120.1, 118.3, 117.3, 115.0,
114.4, 111.1, 56.5, 56.1, 21.7; HREI MS m/z calcd for C20H16O5
336.0998, found 336.1006. Anal. calcd for C20H16O5 C, 71.42; H,
4.79. Found: C, 71.38; H, 4.83.
1-Hydroxy-3-methyl-5,5-dimethoxy-5H-benzo[b]fluorene-6,11-
dione (25). To a suspension of compound 4 (10 mg, 0.033 mmol)
in 10 mL of methanol was added 0.5 mg of Rh2(OAc)4. The
mixture was heated at 80 ◦C overnight. Methanol was evaporated
and the residue was taken up in ethyl acetate. The organic layer
was washed with water and brine, dried over Na2SO4, filtered
and concentrated. The crude product was subjected to preparative
silica gel thin-layer chromatographic separation using methylene
chloride as eluent followed by recrystallisation from methylene
chloride/hexane to a◦fford the title compound 25: 6.5 mg (58%)
yield; mp dec. ≥ 150 C; TLC (dichloromethane/methanol 95:5)
Rf = 0.57; IR (KBr), 3432, 3200, 2925, 2851, 1630, 1580, 1467 cm-1;
1H NMR (400 MHz) (CDCl3) d 10.90 (1H, s), 8.25 (2H, m), 7.87
(1H, m), 7.83 (1H, m), 6.92 (1H, s), 6.84 (1H, s), 3.43 (6H, s),
2.40 (3H. s); HREI MS m/z calcd for C20H16O5 336.0998, found
336.1006.
5-Diazo-1-methoxy-3-methylbenzo[b]fluorene-6,11-dione
(5).
A mixture of 4 (30 mg, 0.1 mmol), CH3I (1 mL) and K2CO3
(1.5 g) in dry DMF (10 mL) was stirred at room temperature
for 24 h. The K2CO3 was removed by filtration and the solution
was diluted with water (20 mL) and ethyl acetate (25 mL).
The layers were separated and the organic phase was washed
successively with water and brine, and then dried over (Na2SO4).
The crude product, obtained by removal of Na2SO4 and then
concentration of the organic phase, was purified by flash silica
gel chromatography using methylene chloride as the eluent.
Crystallisation of purified 5 from methylene chloride/hexane
afforded red crystals: 26 mg (84%) yield; mp dec. ≥ 189 ◦C; TLC
(dichloromethane/methanol 95:5) Rf = 0.60; IR (KBr) 2994,
1
2923, 2093 (diazo), 1650, 1492, 1437, 1246, 826 cm-1; H NMR
(CDCl3) d 8.26 (1 H, d, J = 7.2 Hz), 8.10 (1 H, d, J = 7.2 Hz), 7.71
(2 H, m), 6.93 (1 H, s), 6.66 (1 H, s), 4.04 (3 H, s), 2.46 (3 H, s);
HREI MS m/z calcd for C19H12N2O3 316.0848, found 316.0853.
Hydrolysis of 2*H2 in anaerobic aqueous buffer. A solution
consisting of 10 mL DMSO and 15 mg (0.052 mmol) of 2*
(13C- enriched) was added to 30 mL of phosphate buffer pH 7.4
containing 1 M KCl. To this solution was added 5 mg of 5% Pd
on carbon and the mixture was then degassed with nitrogen for
30 min, followed by bubbling hydrogen gas for 10 min, and finally
bubbling with nitrogen for 30 min to remove the excess hydrogen.
7,11-Dihydroxy-5-methoxy-6H-benzo[b]fluoren-6-one (6*). To
a suspension of compound 2* (40 mg, 0.138 mmol) in 20 mL of
methanol was added 3 mg of Rh2(OAc)4. The mixture was heated
at 80 ◦C overnight. Methanol was evaporated and the residue
2152 | Org. Biomol. Chem., 2009, 7, 2140–2154
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
◦
The reaction mixture was incubated at 30 C for 24 h and then
8 Hz), 7.49 (1H, t, J = 8 Hz), 6.97 (1H, d, J = 8.4 Hz), 6.63
(2H, d, J = 6.8 Hz), 4.09 (3H, s), 3.82 (3H. s), 2.21 (3H, s); 13C
NMR (400 MHz) (CDCl3) d 180.8, 161.1, 151.6, 150.8, 150.1,
142.7, 141.0, 138.3, 136.7, 133.8, 121.2, 120.1, 118.3, 117.3, 115.0,
114.4, 111.1, 56.5, 56.1, 21.7; HREI MS m/z calcd for C20H16O5
336.0998, found 336.1006. Anal. calcd for C20H16O5 C, 71.42; H,
4.79. Found: C, 71.38; H, 4.83. Physical properties of labelled 20*:
HREI MS m/z calcd for 12C1913C1H16O5 337.1031, found 337.1036,
calcd for 12C20H16O5 336.0998, found 336.1006.
opened to the air. The catalyst was filtered off and the filtrate
extracted 3¥ with 20 mL portions of ethyl acetate. The extracts
were dried (Na2SO4) and concentrated to a red solid, which
was subjected to preparative silica gel thin-layer chromatographic
separation using ethyl acetate/methanol (95:5) as eluent. The
physical properties of hydrolysis products are provided below:
7,11-Dihydroxy-6H-benzo[b]fluoren-6-one-5-13C (26*). Trace
yield; 13C-NMR (400 MHz) (DMSO-d6) d 94.21; HREI MS m/z
calcd for 13C1 C16H10O3 263.0663, found 263.0674.
6,11-Dihydroxy-1,7-dimethoxy-3-methyl-5H-benzo[b]fluoren-5-
one-50%-5-13C (12*). 7.4 mg (25%) yield; the physical properties
for 12* were the same as reported above.
12
5,7,11-Trihydroxy-6H-benzo[b]fluoren-6-one-5-13C
(27*).
Trace yield; 13C NMR (400 MHz) (DMSO-d6) d 125.3; IR (KBr)
3431, 3200, 2922, 1629, 1580, 1463, 1293, 1195, 1111 cm-1; HREI
Calculations
MS m/z calcd for 13C1 C16H12O4 279.0613, found 279.0784.
12
Electrostatic potential maps, tautomerisation equilibrium con-
stants, and 13C-NMR chemical shifts were obtained with Hartree–
Fock calculations (Spartan 2006 software) that were carried out
using the 3-21G basis set.19
6,7,11-Trihydroxy-5H-benzo[b]fluoren-5-one-5-13C (28*). Ma-
jor product 7 mg (48%) yield; mp, dec. ≥ 217 ◦C; TLC
(dichloromethane : methanol, 95:5), Rf = 0.21; IR (KBr), 3404,
3060, 1678, 1596, 1530, 1465, 1383, 1350, 1312, 1230, 1082, 989,
880, 766 cm-1; 1H NMR (500 MHz) (DMSO-d6) d 10.42 (1H, s),
9.61 (1H, s), 8.09 (1H, d, J = 8 Hz), 7.70 (1H, d, J = 8.5 Hz),
7.58 (1H, d, J = 7.5 Hz), 7.54 (1H, t, J = 7.5 Hz), 7.43 (1H, t,
In vitro screening procedure
We tested compounds 1–6 against select human cancer cell lines.
All cell lines are inoculated onto a series of standard 96-well
microtitre plates that consist of cell suspensions that were diluted
according to the particular cell type (expected target cell density
5000–40 000 cells per well based on cell growth characteristics).
Inoculates were preincubated for a period of 24 h at 37 ◦C in the
absence of any test compound. Each test compound was added in
100 mL aliquots to the micro-titer plate wells and evaluated at five
10-fold dilutions with a high concentration of 10-4 M. Incubation
with test compounds lasted 48 hours in an atmosphere of 5%
carbon dioxide with 100% humidity. The cells are assayed by
the sulforhodamine B procedure. Afterwards, a plate reader was
used to read the optical densities, and a computer processed the
optical densities into the concentration parameters GI50, TGI and
LC50.55
J = 8 Hz), 7.28 (1H, t, J = 7.5 Hz), 6.89 (1H, d, J = 7.5 Hz); 13
C
NMR (500 MHz) (DMSO-d6/TFA (10%)) d 190.4, 157.6, 151.6,
143.1, 141.3, 136.3, 135.0, 134.3, 130.7, 128.0, 125.1, 123.3, 120.9,
116.2, 115.5, 112.7, 112.5; HREI MS m/z calcd for 13C1 C16H10O4
12
279.0613, found 279.0606.
7,11,7¢,11¢-Tetrahydroxy-[5,5¢]bi[benzo[b]fluorenyl]-6,6¢-dione
(29*). Trace yield; 13C NMR (400 MHz) (DMSO-d6) d 155.81;
IR (KBr) 3415, 3060, 1618, 1579, 1525, 1454, 1377, 1306, 1263,
1213, 1159, 766 cm-1; HREI MS m/z calcd for 13C2 C32H18O6
12
524.1171, found 524.1179.
Hydrolysis of 1*H2 in anaerobic aqueous buffer. A solution
consisting of 20 mL DMSO and 30 mg (0.087 mmol) of 1* (13C-
enriched) was added to 50 mL of 0.05 M phosphate buffer pH 7.4.
To this solution was added 5 mg of 5% Pd on carbon and the
mixture was then degassed with nitrogen for 30 min, followed
by bubbling hydrogen gas for 10 min, and finally bubbling with
nitrogen for 30 min to remov◦e the excess hydrogen. The reaction
mixture was incubated at 30 C for 24 h and then opened to the
air. The catalyst was filtered off and the filtrate extracted 3¥ with
20 mL portions of ethyl acetate. The extracts were dried (Na2SO4)
and concentrated to a red solid, which was subjected to preparative
silica gel thin-layer chromatographic separation using methylene
chloride/methanol (95:5) as eluent. The physical properties of
hydrolysis products are provided below:
Acknowledgements
We wish to thank the Arizona Biomedical Research Commission
for their generous support.
References
1 J. Marco-Contelles and M. T. Molina, Curr. Org. Chem., 2003, 7, 1433.
2 B. B. Hasinoff, X. Wu, J. C. Yalowich, V. Goodfellow, R. S. Laufer, O.
Adedayo and G. I. Dmitrienko, Anti-Cancer Drugs, 2006, 17, 825.
3 S. Omura, A. Nakagawa, H. Yamada, T. Hata and A. Furusaki, Chem.
Pharm. Bull., 1973, 21, 931.
4 H. He, W.-D. Ding, V. S. Bernan, A. D. Richardson, C. M. Ireland, M.
Greenstein, G. A. Ellestad and G. T. Carter, J. Am. Chem. Soc., 2001,
123, 5362.
5 K. S. Feldman and K. J. Eastman, J. Am. Chem. Soc., 2005, 127, 15344.
6 K. S. Feldman and K. J. Eastman, J. Am. Chem. Soc., 2006, 128, 12562.
7 T. E. Ballard and C. Melander, Tetrahedron Lett., 2008, 49, 3157.
8 D. P. Arya and D. J. Jebaratnam, J. Org. Chem., 1995, 60, 3268.
9 W. Zeng, T. E. Ballard, A. G. Tkachenko, V. A. Burns, D. L. Feldheim
and C. Melander, Bioorg. Med. Chem. Lett., 2006, 16, 5148.
10 J. M. Beecham and L. Brand, Photochem. Photobiol., 1986, 44, 323.
11 I. B. C. Matheson, Anal. Instrum., 1987, 16, 345.
1-Methoxy-3-methyl-5-benzo[b]fluorene-6,11-dione-50%-5-13C
(30*). Trace yield; 13C NMR (400 MHz) (DMSO-d6) d 34.1;
HREI MS m/z calcd for 13C1 C19H16O4 321.1082, found 321.1076,
12
calcd for 12C20H16O4 320.1049, found 320.1047.
5,11-Dihydroxy-3-methyl-1,7-dimethoxy-6H-benzo[b]fluoren-6-
one-50%-5-13C (24*). 15.5 mg (53.5%) yield. Physical prop-
erties of unlabelled analogue: mp, dec.
≥
209 ◦C; TLC
(dichloromethane:methanol, 95:5), Rf = 0.58; IR (KBr), 3437,
3202, 2923, 2852, 1629, 1579, 1465, 1295, 1197, 1110, 1066 cm-1;
1H NMR (400 MHz) (CDCl3) d 10.64 (2H, s), 7.60 (1H, d, J =
12 O. Khdour, O. Ouyang and E. B. Skibo, J. Org. Chem., 2006, 71, 5855.
13 E. B. Skibo, C. Xing and T. Groy, Bioorg. Med. Chem., 2001, 9, 2445.
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 2140–2154 | 2153
©

View Article Online
14 A. Ouyang and E. B. Skibo, Biochemistry, 2000, 39, 5817.
15 A. Ouyang and E. B. Skibo, J. Org. Chem., 1998, 63, 1893.
16 D. Siegel, H. Beall, C. Senekowitsch, M. Kasai, H. Arai, N. W. Gibson
and D. Ross, Biochemistry, 1992, 31, 7879.
17 H. W. Moore and R. Czerniak, Med. Res. Rev., 1981, 1, 249.
18 H. W. Moore, Science (Washington, D.C.), 1977, 197, 527.
19 J. R. Durig, R. J. Berry, D. T. Durig, J. F. Sullivan and T. S. Little,
Teubner-Texte zur Physik, 1988, 20, 54.
34 N. K. Chaudhuri, S. V. Potdar and T. J. Ball, J. Labelled Compd.
Radiopharm., 1982, 19, 75.
35 R. Paulissen, H. Reimlinger, E. Hayez, A. J. Hubert and P. Teyssie,
Tetrahedron Lett., 1973, 14, 2233.
36 F. M. Wong, J. Wang, A. C. Hengge and W. Wu, Org. Lett., 2007, 9,
1663.
37 D. V. LaBarbera and E. B. Skibo, J. Am. Chem. Soc., 2006, 128, 3722.
38 E. B. Skibo, J. Org. Chem., 1992, 57, 5874.
20 R. C. Boruah and E. B. Skibo, J. Org. Chem., 1995, 60, 2232.
21 G. S. Kumar, R. Lipman, J. Cummings and M. Tomasz, Biochemistry,
1997, 36, 14128.
22 O. Khdour and E. B. Skibo, J. Org. Chem., 2007, 72, 8636.
23 S. J. Gould, C. R. Melville, M. C. Cone, J. Chen and J. R. Carney,
J. Org. Chem., 1997, 62, 320.
24 C. Volkmann, E. Rossner, M. Metzler, H. Zahner and A. Zeeck, Liebigs
Ann., 1995, 1995, 1169.
25 F. M. Hauser and M. Zhou, J. Org. Chem., 1996, 61, 5722.
26 S. J. Gould, J. Chen, M. C. Cone, M. P. Gore, C. R. Melville and N.
Tamayo, J. Org. Chem., 1996, 61, 5720.
27 W. Augustyniak, J. Bukowski, J. Jemielity, R. Kanski and M. Kanska,
J. Radioanal. Nucl. Chem., 2001, 247, 371.
28 F. M. Hauser and S. R. Ellenberger, Synthesis, 1987, 723.
29 M. P. Gore, S. J. Gould and D. D. Weller, J. Org. Chem., 1992, 57, 2774.
30 J. N. Freskos, G. W. Morrow and J. S. Swenton, J. Org. Chem., 1985,
50, 805.
39 S. J. Gould and C. R. Melville, Tetrahedron Lett., 1997, 38, 1473.
40 J. Fisher, K. Ramakrishnan and J. E. Becvar, Biochemistry, 1983, 22,
1347.
41 E. B. Skibo, J. Org. Chem., 1986, 51, 522.
42 R. L. Lemus and E. B. Skibo, J. Org. Chem., 1988, 53, 6099.
43 S. R. Angle, J. D. Rainier and C. Woytowicz, J. Org. Chem., 1997, 62,
5884.
44 S. E. Rokita, J. H. Yang, P. Pande and W. A. Greenberg, J. Org. Chem.,
1997, 62, 3010.
45 I. Han, D. J. Russell and H. Kohn, J. Org. Chem., 1992, 57, 1799.
46 T. C. Bruice and G. L. Schimir, J. Am. Chem. Soc., 1959, 81, 4552.
47 M. A. Paul and F. A. Long, Chem. Rev., 1957, 57, 1.
48 A. Monks, D. A. Scudiero, G. S. Johnson, K. D. Paull and E. A.
Sausville, Anti-Cancer Drug Des., 1997, 12, 533.
49 A. M. Rauth, Z. Goldberg and V. Misra, Oncol. Res., 1997, 9, 339.
50 A. J. Lin, L. A. Cosby, C. W. Shansky and A. C. Sartorelli, J. Med.
Chem., 1972, 15, 1247.
31 B. L. Chenard, M. G. Dolson, A. D. Sercel and J. S. Swenton, J. Org.
Chem., 1984, 49, 318.
32 M. Fetizon, M. Golfier, R. Milcent and I. Papadakis, Tetrahedron,
1975, 31, 165.
33 W. Williams, X. Sun and D. Jebaratnam, J. Org. Chem., 1997, 62, 4364.
51 P. L. Gutierrez, Free Radical Biol. Med., 2000, 29, 263.
52 M. D. Garrett and P. Workman, Eur. J. Cancer, 1999, 35, 2010.
53 I. J. Stratford and P. Workman, Anti-Cancer Drug. Des., 1998, 13, 519.
54 P. Workman, Oncol. Res., 1994, 6, 461.
55 M. R. Boyd and K. D. Paull, Drug Dev. Res., 1995, 34, 91.
2154 | Org. Biomol. Chem., 2009, 7, 2140–2154
This journal is
The Royal Society of Chemistry 2009
©