Chemistry of the pyrrolo[1,2-a]benzimidazole antitumor agents: Influence of the 7-substituent on the ability to alkylate DNA and inhibit topoisomerase II

Article abstract of DOI:10.1021/jm960064d

This study addresses the influence the 7-substituent on the cytotoxicity of pyrrolo[1,2-a]benzimidazole quinones possessing a 6-aziridinyl group (PBIs) and a 6-acetamido group (APBIs). Reduction of a PBI to the aziridinyl hydroquinone results in both nucl

Full text of DOI:10.1021/jm960064d

J . Med. Chem. 1996, 39, 4321-4331
4321
Ch em istr y of th e P yr r olo[1,2-a ]ben zim id a zole An titu m or Agen ts: In flu en ce of
th e 7-Su bstitu en t on th e Ability To Alk yla te DNA a n d In h ibit Top oisom er a se II
Ru Zhou and Edward B. Skibo*
Department of Chemistry and Biochemistry, Box 871604, Arizona State University, Tempe, Arizona 85287-1604
Received J anuary 22, 1996X
This study addresses the influence the 7-substituent on the cytotoxicity of pyrrolo[1,2-a]-
benzimidazole quinones possessing a 6-aziridinyl group (PBIs) and a 6-acetamido group (APBIs).
Reduction of a PBI to the aziridinyl hydroquinone results in both nucleophile trapping
(alkylation) and 1,5-sigmatropic shift reactions. The latter process is essentially an internal
redox reaction wherein the hydroquinone causes reductive opening of the aziridinyl ring. The
7-substituent controls the fate of the aziridinyl ring by means of steric and electronic effects.
An electron-rich 7-substituent favors the 1,5-sigmatropic shift reaction. If the 7-substituent
distorts the 6-aziridinyl group from the conformation required for the 1,5-sigmatropic shift,
then nucleophile trapping occurs. The 7-methyl substituent results in significant nucleophilic
trapping, and the 7-unsubstituted and 7-methoxy substituents favor the 1,5-sigmatropic
reaction. Thus, the 7-methyl PBIs show the most cytotoxicity of the analogues studied. The
APBIs are cytotoxic only as quinones, and reduction to the hydroquinone results in loss of
activity. Consistent with this observation, the change from 7-methyl to the more electron-rich
7-methoxy results in a substantial loss of APBI cytotoxicity as well as decreased topoisomerase
II inhibition. The mechanism of inhibition is thought to involve the intercalation of only electron
deficient APBIs into DNA.
Ch a r t 1
The pyrrolo[1,2-a]benzimidazoles represent a new
class of antitumor agent exhibiting cytotoxic activity
against a variety of cancer cell lines.1-6 Since the first
reported synthesis of the pyrrolobenzimidazoles in
1990,1 two patents have been granted7,8 and several
analogues are currently undergoing in vivo trials at the
National Cancer Institute. Typically, the pyrrolo[1,2-
a]benzimidazoles possess a 7-methyl substituent and
either a 6-acetamido group (APBIs) or a 6-aziridinyl
group (PBIs), inset of Chart 1. The PBI analogues
alkylate DNA upon two-electron reduction as a result
of nucleophile-mediated opening of the protonated aziri-
dine ring.3 In contrast, the APBIs are not reductive
alkylating agents and exhibit cytotoxicity in the oxidized
form.6 In order to evaluate further the chemical and
cytotoxic properties of the pyrrolo[1,2-a]benzimidazoles,
analogues bearing 7-methoxy, 7-unsubstituted, and
7-aziridinyl substituents were prepared and studied,
1-4 in Chart 1. This report describes these synthetic
efforts along with the results of hydrolysis studies and
of cytotoxicity screening.
The results of these studies indicate that the 7-sub-
stituent greatly influences the cytotoxicity of the pyr-
rolo[1,2-a]benzimidazoles. The PBIs afford aziridinyl
hydroquinones upon reduction which undergo two com-
peting reactions: a 1,5-sigmatropic shift of hydrogen
(reaction A) and nucleophile-mediated ring opening
(reaction B), Chart 2. Only the latter reaction results
in DNA alkylation and cytotoxicity. The 7-substituent
controls the balance between the two competing reac-
tions with the 7-methyl substituent favoring the alkyl-
ation reaction. Presumably both the electronic and
steric natures of the 7-substituent control the fate of
the aziridinyl ring. The 7-substituent controls the
electronic character of the APBI and thereby the ability
to inhibit topoisomerase II. As the APBI becomes more
Ch a r t 2
electron rich, or is reduced to the hydroquinone, the
inhibition of this enzyme decreases. Thus the 7-meth-
oxy APBI 3 is much less cytotoxic than the relatively
electron deficient 7-methyl derivative.
X Abstract published in Advance ACS Abstracts, September 15, 1996.
S0022-2623(96)00064-7 CCC: $12.00 © 1996 American Chemical Society

4322 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21
Zhou and Skibo
Sch em e 1
Sch em e 2
Sch em e 3
The findings reported herein are relevant to the
design of more active pyrrolo[1,2-a]benzimidazoles as
well as to the chemistry of aziridinyl quinone antitumor
agents and to the design of topoisomerase II inhibitors.
Resu lts a n d Discu ssion
Syn th esis of th e P yr r olo[1,2-a ]ben zim id a zole
Rin g System . The presence of a substituent ortho to
an N,N-dialkylaniline gives rise to the “tert-amino
effect”.9 This effect either involves nucleophilic attack
by the tertiary nitrogen or oxidation of the tertiary
nitrogen to an iminium ion, which is then subjected to
nucleophilic attack. Cyclizations of this type have been
employed in the preparation of various benzimidazole
derivatives by Meth-Cohn and co-workers.9-11
The ring closure reactions 7 f 9a and 6 f 8 shown
in Scheme 1 both likely occur via an iminium ion
intermediate.11 Oxidation of 7 by performic acid affords
the iminium ion which is then subjected to nucleophilic
attack by the adjacent acetamido nitrogen, eventually
resulting in formation of the pyrrolo[1,2-a]benzimidazole
9a .9 On the basis of previous studies, the ring closure
of 6 to 8 is proposed to utilize a Lewis acid-catalyzed
internal redox reaction to afford the o-nitroso iminium
ion, which undergoes ring closure.11-13
followed by quinone elaboration. Displacement of the
methoxy group from quinones 18a ,b with aziridine
afforded the target PBIs 2a ,b. Reductive addition of
methoxide to 2a in the presence of air afforded the
7-methoxy derivative 3.
The preparation of 4, starting with 9a , was carried
out as outlined in Scheme 3. The synthetic steps
involved the acid hydrolysis of the 6-acetamido group
of 9a resulting in 19, which was then converted to 20
in a one pot-reaction. The conversion of 20 to 4 involved
reduction, Fremy oxidation, and substitution with aziri-
dine as previously reported.1
Completion of the synthesis of APBIs 1a ,b, starting
from the pyrrolo[1,2-a]benzimidazoles 8 and 9a , is
outlined in Scheme 1. Quinone elaboration involved
nitration and then reduction to the amino derivative and
finally Fremy oxidation of the amine.1
Preparation of the PBIs 2a ,b and 3 is outlined in
Scheme 2. The preparation of the pyrrolo[1,2-a]benz-
imidazoles 15a ,b was carried out as described above
Hyd r olytic Ch em istr y of Red u ced P BIs. The goal
of these studies was to correlate the fate of reduced PBIs
in aqueous buffer with their cytotoxicity and DNA
alkylation capability. Another goal was to obtain a
detailed mechanistic study of the aziridinyl hydro-
quinone fate. The results of this study may be pertinent
to the fate of reduced aziridinyl quinones currently used
in chemotherapy.

Chemistry of Pyrrolobenzimidazole Antitumor Agents
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21 4323
Sch em e 4
The hydrolysis of reduced 2a (23) was studied in
anaerobic buffer (µ ) 1.0, KCl) over the pH range of
2-10 at 30.0 ( 0.2 °C. Preparative anaerobic hydrolysis
of 23 followed by aerobic workup revealed the presence
of quinones 26-28 shown in Scheme 4 (see the Experi-
mental Section). The formation of these products was
pH dependent: 26 and 28 were isolated in 37% and 18%
yields, respectively, at pH 5.5 and 7.4, and only 27 was
isolated in 56% yield at pH 2.0. The quinone products
are deep blue in color (λmax ) 550 nm) due to the
presence of the aminoquinone chromophore. When
anaerobic reactions held at pH 5.5 or 7.4 were followed
by repetitive scanning in the UV-vis region, the buildup
of blue product was noted (attributed to the formation
of 26). Upon completion of the anaerobic process, the
reaction was opened to the air resulting in the formation
of additional blue product (attributed to the oxidation
of 25 to 28). When an anaerobic reaction held at pH 2
was performed, there was no buildup of blue product
indicating the formation of the hydroquinone product
24, which oxidized to 27 upon workup.
In order to assess the influence of pH on the fate of
23, the pH-rate profile for hydrolysis shown in plot A
of Figure 1 was obtained over the pH range of 2-10.
Absorbance vs time(s) measurements for hydrolysis
were made at 310 nm. All absorbance vs time(s) plots
were first-order in character, and the resulting kobsd (s-1
)
values were independent of buffer concentration but
dependent on pH. Found in plot A of Figure 1 are log
kobsd vs pH data for the hydrolysis of 23. Careful in-
spection of these data reveal the presence of three re-
gions: (i) a plateau below pH 4, (ii) a plateau between
pH 5 and 6 which presents as an inflection, and (iii) a
plateau above pH 8. The presence of three plateaus is
an indication that the three forms of 23 shown in
Scheme 4 are involved in the hydrolysis reactions. The
rate law for the hydrolysis of an equilibrating mixture
F igu r e 1. Plot of log kobsd vs pH for the hydrolysis of 23 and
its protonated forms in anaerobic aqueous buffer at µ ) 1.0
(KCl) and 30.0 ( 0.2 °C.
10-5 s-1, pKa1 ) 3.85, and pKa2 ) 6.20. This solution
was used to generate the solid line in plot A of Figure
1.
The plateau between pH 5 and 6 is difficult to see on
casual inspection of the data shown. Rather than the
diprotic rate law in eq 1, a monoprotic pH-rate law
could conceivably fit the data. Fitting the data to kobsd
) (k1aH + k2Ka1aH)/(aH + Ka1) provided the curve shown
in plot B of Figure 1. The curve in this plot misses the
high-pH data due to the inflection actually present
between pH 5 and 6.
2+
of 23, 23H+, and 23H2 is shown in eq 1:
k1a2H + k2Ka1aH + k3Ka1Ka2
kobsd
)
(1)
a2H + Ka1aH + Ka1Ka2
where the rate and equilibrium constants are those
shown in Scheme 4. Computer fitting of the data in plot
A of Figure 1 to eq 1 provided the following solution:
k1 ) 6.37 × 10-2 s-1, k2 ) 1.08 × 10-3 s-1, k3 ) 1.36 ×
The pKa assignments for the diprotic rate law in eq 1
have been made as illustrated in Scheme 4. Compound

4324 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21
Zhou and Skibo
Sch em e 5
Sch em e 6
concerted mechanism which does not require external
protons. Consistent with a concerted process, neutral
23 decomposes to 26 in dry anaerobic dimethyl sulfoxide
at approximately k3 (s-1).13 The decomposition of 23H+
could involve the prototropic process shown in the inset
of Scheme 5 where the transition state consists of an
external proton and 23. If this is the case, eq 1 must
be rewritten as eq 2, where k2 is a second-order rate
constant:
k1a2H + k2Ka1Ka2aH + k3Ka1Ka2
23 has two basic nitrogens: the aziridinyl nitrogen and
the N(4)-position. At low pH values (<3.0) both nitro-
gens are protonated to afford 23H22+. The more basic
of the two nitrogens is difficult to assess, however. A
protonated benzimidazole hydroquinone nitrogen (pro-
tonated N(4)-position of 23) generally has pKa values
in the range of 414-6.1 Typically, N-protonated benz-
imidazoles have a pKa value of 5.5.15 The aziridine
nitrogen is quite basic, and the pKa values of protonated
forms are usually >8.16 The presence of an aromatic
substituent on the aziridinyl nitrogen should decrease
the pKa to ∼6, based on the pKa values of other
protonated amino hydroquinones.17 Since the aziridinyl
nitrogen may be somewhat more basic than the N(4)-
position of 23, the structure of 23H+ shows the proton
on this nitrogen. The presence of aziridinyl alkylation
reactions by reduced PBIs near neutrality,6 in fact,
suggests the presence of a protonated aziridinyl nitro-
gen, which would be required for the nucleophile-
mediated opening of the aziridine ring to occur.18
The pH-rate profile and product studies indicate that
both 23 and 23H+ involve the formation of intermediate
A by a 1,5-sigmatropic shift of hydrogen as illustrated
in Scheme 5. A prototropic shift converts A to the
quinone 26, and oxidative dealkylation of A via inter-
mediate B affords 25. The conversion of 23 or 23H+ to
A is essentially an internal redox reaction wherein the
hydroquinone brings about the reductive opening of the
aziridine ring. Protonation of the N(4)-position will
result in an electron deficient hydroquinone incapable
kobsd
)
(2)
a2H + Ka1aH + Ka1Ka2
This equation is kinetically equivalent to eq 1 and still
fits the data shown in plot A of Figure 1. The value of
k2 obtained from the computer fit is 1733 M-1 s-1
.
Alternatively, 23H+ may spontaneously decompose at
k2 ) 1.08 × 10-3 s-1 by a sigmatropic process. The 79-
fold rate enhancement in the sigmatropic process (k2
(s-1) compared to k1 (s-1) in eq 1 for the decomposition
23H+ and 23, respectively) would be the result of the
transformation of a localized cation (23H+) to a more
stabilized delocalized cation (A). An analogy to this rate
enhancement is the oxy Cope rearrangement, where a
large change in oxygen anion stability accelerates the
reaction.19
Although the two possible decomposition mechanisms
for 23H+ cannot be rigorously distinguished, we favor
the sigmatropic shift mechanism. One reason is the
presence of this mechanism in the decomposition of 23
which suggests that 23H+ could likewise react in this
fashion. Furthermore, if an external proton were
involved (prototropic mechanism) in the reaction, gen-
eral acids should participate in the prototropic process.
However, the rate law for the decomposition of 23, eq
1, does not contain a buffer catalysis term. Finally, the
sigmatropic shift mechanism explains the differences
in alkylation reactivity observed between the various
7-substituted PBIs (see the next section).
In order to determine the influence of the 3- and
7-substituents on the course of reduced PBI decomposi-
tion, the preparative hydrolysis of reduced 2b and
reduced 3 was carried out in anaerobic Tris buffer,
Scheme 6. Reduced 2b differs from reduced 2a (23) only
with respect to the presence of a 3-acetoxy substituent.
Preparative hydrolysis of reduced 2b afforded 29 and
30 in 31% and 40% yields, respectively. An additional
2+
of reducing the aziridine ring. Therefore, 23H2 only
traps chloride nucleophile to afford 24.
The mechanism for the formation of intermediate
A will now be discussed. There are two possible mech-
anisms: the concerted, sigmatropic process already
discussed and the prototropic process shown in the inset
of Scheme 5.13 The observation that 23 decomposes in
the neutral form at k3 ) 1.36 × 10-5 s-1 favors a

Chemistry of Pyrrolobenzimidazole Antitumor Agents
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21 4325
Ta ble 1. Mean log GI50, log TGI, and log LC50 Values
Obtained from 60 Cancer Cell Line Mean Graphs for
Substituted PBIs
Ta ble 2. Mean log GI50, log TGI, and log LC50 Values from 60
Cancer Cell Line Mean Graphs for Substituted APBIs
APBI
log GI50
log TGI
log LC50
PBI
log GI50
log TGI
log LC50
APBI-A
APBI-C
1b
-5.80
-5.84
-4.73
-5.32
-5.43
-4.41
-4.73
-4.97
-4.13
PBI-C (Chart 1)
PBI-A (Chart 1)
-7.74
-7.92
-6.44
-6.57
-6.15
>-4
-6.98
-7.50
-5.87
-5.97
-5.60
>-4
-5.69
-6.61
-5.26
-5.28
-4.87
>-4
2a
2b
3
decrease in PBI cytotoxicity.4 Indeed, the 7-aziridinyl
derivative 4 was found to be inactive in all cancer cell
lines.
4
An explanation for the anomalous behavior of the
7-methyl PBIs invokes possible steric effects by the
7-substituent on the rotational conformation of the
aziridinyl group as well as the 1,5-sigmatropic shift of
hydrogen discussed in the previous section. The fate
of a reduced PBI pertains to nucleophile trapping in
competition with the 1,5-sigmatropic shift. Indeed,
hydrolytic reactions of reduced 7-methyl PBIs provide
both types of products.5,6 The 1,5-sigmatropic shift
should be sensitive to the conformation of the aziridinyl
group, which must be positioned correctly for the
pericyclic reaction involving the hydroquinone ring. If
the 7-substituent interferes with this conformation, then
the competing nucleophile-trapping reaction will be
favored. Therefore, the steric influence of the 7-methyl
substituent may be responsible for the nucleophile-
trapping reactions and hence the cytotoxicity/antitumor
activity of the PBIs. The methoxy substituent is ap-
proximately the same size as the methyl substituent,
and therefore 3 should trap nucleophiles upon reduction.
However, reduced 3 only undergoes the 1,5-sigmatropic
shift perhaps because the electron-releasing methoxy
substituent does not favor the trapping of a nucleophile.
product, 31 (8%), resulted from the ester hydrolysis of
30. Reduced 3 differs from 23 only with respect to the
presence of a 7-methoxy substituent. The products
resulting from the hydrolysis of reduced 3 are structur-
ally similar to those obtained from 23: 32 (22%), 33
(11%), and unreacted 3 (15%). The results cited above
indicate that reduced 2b and reduced 3 form hydrolysis
products structurally similar to those formed from
reduced 2a (23).
Com p a r a tive Hyd r olytic Ch em istr y a n d Cyto-
toxicity of 7-Su bstitu ted P BIs. The results cited in
the previous section indicate that the 7-unsubstituted
and the 7-methoxy PBIs share identical hydrolytic
chemistry upon reduction. It is noteworthy that neither
system traps nucleophiles at neutrality. For example,
23 did not trap water nor did it alkylate 5′-dAMP or
calf thymus DNA in anaerobic pH 7.4 Tris buffer (see
the Experimental Section). Since the electron-releasing
effect of methyl lies between that of the -H and -OCH3
substituents, the 7-methyl PBIs should behave like the
7-unsubstituted and 7-methoxy PBIs upon reduction.
Verification that the 7-methyl PBI is more electron rich
than the 7-unsubstituted derivative was obtained by
measuring the Em (NHE) for two-electron reduction of
both PBI-C (Chart 1) and 2a in pH 7.4 buffer (µ ) 1.0,
NaClO4). PBI-C possesses an Em value of -50 mV,
whereas 2a possesses an Em value of -20 mV at this
pH. Electron release by the methyl is expected to
decrease the Em value of the quinone.20
Work recently submitted for publication from this
laboratory featured the 7-n-butyl PBIs. Our steric
interpretation presented above correctly predicted that
this PBI derivative would reductively alkylate DNA.
However, the bulky 7-n-butyl substituent slowed reduc-
tive activation by DT-diaphorase and substantially
decreased the cytotoxicity of 7-n-butyl PBIs.
In contrast to the 7-unsubstituted and the 7-methoxy
PBIs, the reduced 7-methyl PBIs are alkylating agents
and trap water as well as the phosphates of 5′-dAMP
and calf thymus DNA.3,4,6 The alkylation of DNA
followed by aerobic workup affords a blue product due
to the presence of the aminoquinone chromophore. The
DNA alkylation reaction is responsible for the cytotox-
icity/antitumor activity of the 7-methyl PBIs.3 Consis-
tent with their apparent lack of alkylation activity, the
7-unsubstituted and 7-methoxy PBIs possess signifi-
cantly less cytotoxicity than the 7-methyl analogues.
Shown in Table 1 are the cytotoxicity parameters for
various 7-substituted PBI analogues. The GI50 and TGI
are the concentrations of PBI required to inhibit cancer
cell growth by 50% and 100%, respectively. The LC50
is the concentration of PBI required to cause 50% cancer
cell lethality. The GI50, TGI, and LC50 values in Table
1 are mean values obtained from 60 cancer cell line
assays carried out at the National Cancer Institute.21
Lower cytotoxicity parameters represent greater po-
tency as a cytotoxic agent. The data in Table 1 show
that the 7-methyl analogues (PBI-C and PBI-A) are
more potent than either the 7-unsubstituted PBIs (by
1 order of magnitude) or the 7-methoxy PBI (by 2 orders
of magnitude). The placement of the aziridinyl group
in the 7-position is known to result in a substantial
Com p a r a tive Cytotoxicity of th e 7-Su bstitu ted
AP BIs. The APBIs are cytotoxic as the quinones, and
cellular reduction to the hydroquinone by DT-diapho-
rase results in complete loss of activity.6 It has been
observed that the quinone derivatives intercalate DNA
and the electron-rich hydroquinones do not.22 If the
electron-rich character of the APBI is important in DNA
intercalation and cytotoxicity, then more electron-rich
APBIs (i.e., 7-methoxy derivatives) should be less cyto-
toxic than the 7-methyl derivatives. The cytotoxic
parameters of the 7-methoxy APBIs (1a ,b) are, in fact,
1 order of magnitude greater (less active) than those of
the 7-methyl analogues, Table 2.
A recent review presented evidence that the APBI
intercalation of DNA prevents topoisomerase II relax-
ation of supercoiled DNA by inhibiting the double-
strand cleavage step leading to the enzyme-DNA
complex.22 Intercalating agents are known to inhibit
topoisomerase I and II resulting in antitumor activity.23
Indeed, the activity of the topoisomerases is important
in cell division, and many types of cancer possess
elevated levels of these enzymes.24 Examples of topoi-
somerase II inhibitors that possess structural similari-
ties to the APBIs (presence of the benzimidazole ring
or the presence of a quinonoid system) are known in

4326 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21
Zhou and Skibo
the literature.25-27 Unlike the APBIs, these inhibitors
stabilize the enzyme-DNA complex formed upon double-
strand cleavage by topoisomerase II and thereby inhibit
the religation to relaxed DNA. The result is the
formation of linear DNA upon treatment of the complex
with SDS/proteinase K. There is another class of
compounds28 that behave like the APBIs and inhibit the
double-strand cleavage step leading to the DNA-
enzyme complex. The antitumor activity of these to-
poisomerase inhibitors22,28 suggests that compounds of
this type could be useful in cancer chemotherapy.
In contrast to other APBIs, which inhibit topoisomer-
ase II-mediated relaxation of SV-40 DNA at concentra-
tions g0.3 mM,22 1a caused little inhibition at concen-
trations g0.6 mM. The presence of the relatively
electron-rich PBI-A resulted in no inhibition of topo-
isomerase II-mediated relaxation of the DNA.
reactions upon hydroquinone reduction, whereas the
7-unsubstituted and the 7-methoxy analogues predomi-
nately undergo the 1,5-sigmatropic reaction. Corre-
spondingly, the 7-methyl-substituted PBIs are the most
cytotoxic of these analogues. We conclude that new
cytotoxic PBI analogues must have the 7-methyl sub-
stituent.
The influence of the 7-substituent on the cytotoxicity
of the APBI agents is also discussed. The change from
the 7-methyl to the 7-methoxy is accompanied by a 10-
fold decrease in cytotoxicity as well as a decrease in
topoisomerase II inhibition. The conclusion is that
increasing the electron-rich character of APBIs will not
lead to effective antitumor agents. Currently, electron
deficient APBIs, which cannot be reduced by DT-
diaphorase, are being developed.
These results support the proposal that electron
deficient APBIs are required for intercalation into DNA
and topoisomerase II inhibition. The electron-rich
analogues 1a and PBI-A show little or no inhibition of
topoisomerase II, while the more electron deficient
APBIs inhibit the enzyme.
Exp er im en ta l Section
All analytically pure compounds were dried under high
vacuum in a drying pistol heated with refluxing methanol.
Some compounds contained water or chloroform of crystal-
lization that was determined from the elemental analyses
1
found. The presence of chloroform was verified by an H NMR
spectrum in dimethyl sulfoxide-d6. Elemental analyses were
run at Atlantic Microlab, Inc., Norcross, Ga. Uncorrected
melting points and decomposition points were determined with
a Mel-Temp apparatus. All TLC was run with silica gel plates
with a fluorescent indicator, employing a variety of solvents.
IR spectra were taken as KBr pellets or thin films; the
strongest IR absorbances are reported. 1H NMR spectra were
obtained on a 300 MHz spectrometer, and chemical shifts are
reported relative to TMS.
Electr och em istr y. The determination of the Em value was
carried out with BAS 27 voltammograph. Measurements were
carried out in µ ) 1.0 (NaClO4) aqueous buffer at 25-26 °C
under an atmosphere of argon with a graphite mull working
electrode and with a BAS Ag/AgCl gel electrode as reference.
The electrode was calibrated against the Eo value of the
benzoquinone/hydroquinone couple (699 mV, NHE). The
midpount potential Em was determined from the average of
the anodic (Ep,a) and cathodic (Ep,c) potentials.
Kin etic Stu d ies. The kinetic studies were carried out in
buffers prepared with doubly distilled water and adjusted to
µ ) 1.0 with KCl. The following buffer systems were employed
to hold pH: acetic acid/acetate (pKa ) 4.61), phosphate
monobasic/phosphate dibasic (pKa ) 6.56), and boric acid/
borate (pKa ) 9.2). These pKa values were obtained at 30.0 (
0.2 °C in µ ) 1.0 (KCl) aqueous solutions. Measurements of
pH were made with a combination electode. The hydrolytic
studies of the hydroquinones were carried out in anaerobic
aqueous buffers employing Thunberg cuvettes. A dimethyl
sulfoxide stock of the hydroquinone to be studied was prepared
fresh, and 100 µL of this stock was added to 2.90 mL of buffer.
The hydroquinone stock solutions were prepared by dissolving
2.5 mg of quinone in 2 mL of dimethyl sulfoxide containing 5
mg of Pd on carbon. This mixture was purged with argon and
then purged with H2 gas until the solution became colorless.
The reduced solution was placed in a glovebox and filtered
through a Millex-SR filter. The absorbance vs time data were
collected on a UV-vis spectrophotometer in thermostated cells
held at 30.0 ( 0.2 °C. These data were computer-fit to a single
first-order rate law.
P r ep a r a tion of New Com p ou n d s a n d Th eir P h ysica l
P r op er ties. 5-Br om o-2,4-d in itr oa n isole (5). To 50 mL of
fuming nitric acid frozen at -78 °C was added 14.8 g (79.5
mmol) of 3-bromoanisole over 2 min, and the mixture was
allowed to warm to 10 °C over 25 min. The reaction mixture
was then poured onto ice and the resulting precipitate collected
and dried. Recrystallization from ethanol afforded 5: 13.6 g
(62%) yield; mp 83 °C; TLC (chloroform/hexane, 1:1) Rf ) 0.45;
IR (KBr pellet) 3106, 1601, 1582, 1534, 1481, 1435, 1339, 1275,
1186, 1105, 1015, 949, 903, 826 cm-1; 1H NMR (CDCl3) δ 8.63
Con clu sion s
Conclusions are drawn with respect to the influence
of the 7-substituent on the cytotoxicity of pyrrolo[1,2-
a]benzimidazole antitumor agents. These conclusions
will be of value in the design of new PBI and APBI
analogues as antitumor agents.
The ring-opening reactions of aziridines usually in-
volve N-protonation followed by nucleophilic attack.18
Aziridinyl quinone antitumor agents are known to
alkylate DNA by this process upon reduction to the
hydroquinone.29 In the present report, we show that
PBI aziridinyl hydroquinones can also undergo a 1,5-
sigmatropic transfer of hydrogen. This process is es-
sentially an internal redox reaction wherein the electron-
rich hydroquinone brings about reductive ring opening
of the aziridinyl group. This process appears to be
influenced by the adjacent 7-substituent, which may
control the favored rotational conformations of the
6-aziridinyl group as well as control the electronic
character of the hydroquinone ring. In order for the 1,5-
sigmatropic reaction to occur, the aziridinyl group must
achieve a particular conformation with respect to the
hydroquinone ring. If the 7-substituent interferes with
this conformation, then the competing nucleophile-
trapping process can occur. An electron-rich hydro-
quinone ring, as in reduced 3, would favor the 1,5-
sigmatropic shift by increasing the reducing capability
of this ring. Ongoing work in this laboratory with
aziridinyl benzoquinones related to known antitumor
agents29 showed that hydrolysis products, which could
arise by the 1,5-sigmatropic shift reaction, are formed
upon reduction. Future publications from this labora-
tory will show that hydrolysis products of this type are
actually common.
From the results of the hydrolytic study, it is apparent
that the 7-substituent of the PBI hydroquinone controls
the fate of the aziridinyl group: nucleophile trapping
(alkylation) vs the 1,5-sigmatropic shift. Therefore, the
influence of the 7-substituent is an important factor in
PBI cytotoxicity, which relies on DNA alkylation. The
7-methyl PBI analogue undergoes aziridinyl alkylation

Chemistry of Pyrrolobenzimidazole Antitumor Agents
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21 4327
and 7.27 (2H, 2s, aromatic protons), 4.10 (3H, s, methoxy); MS
(EI mode) m/ z 277 (M+). Anal. (C7H5BrN2O5) C, H, Br, N.
2,4-Din itr o-5-(N-p yr r olid in o)a n isole (6). To a solution
of 9 g (32.6 mmol) of 5 in 200 mL of ethanol was added 13.5
mL of pyrrolidine over 1 min. The mixture was allowed to sit
at room temperature for 1 h, and then the yellow precipitate
was filtered off and dried. Recrystallization from ethanol
afforded 6: 4.95 g (57%) yield; mp 151.5-152.5 °C; TLC
(chloroform) Rf ) 0.26; IR (KBr pellet) 3447, 3090, 2980, 2943,
2868, 1611, 1564, 1507, 1452, 1339, 1316, 1273, 1078, 1001,
920, 872, 804 cm-1; 1H NMR (CDCl3) δ 8.69 and 6.23 (2H, 2s,
aromatic protons), 4.00 (3H, s, methoxy), 3.34 (4H, m, meth-
ylenes adjacent to pyrrolidine nitrogen), 2.06 (4H, m, other
pyrrolidine methylenes); MS (EI mode) m/ z 267 (M+). Anal.
(C11H13N3O5) C, H, N.
evaporated in vacuo to remove the methanol. The residue was
combined with 1.0 mL of acetic anhydride and then stirred
for 30 min at room temperature. The completed reaction was
diluted with 40 mL of diethyl ether and chilled in a refrigerator
overnight resulting in crystallization of 9b. Recrystallization
from ethyl acetate/hexane afforded pure 9b: 154 mg (82%)
yield; TLC (chloroform/methanol, 9:1) Rf ) 0.33; mp 216-218
°C; IR (KBr pellet) 3314, 2936, 1744, 1670, 1591, 1539, 1478,
1414, 1368, 1256, 1227, 1207, 1105, 1017, 886, 732, 624 cm-1
;
1H NMR (CDCl3) δ 8.78 and 6.79 (2H, 2s, aromatic protons),
7.75 (1H, s, amine proton), 6.13 (1H, dd, J ) 7.5, 3.1 Hz, C(3)
proton), 4.21and 4.07 (2H, 2m, C(1) diastereomeric methylene),
3.92 (3H, s, methoxy), 3.15 and 2.63 (2H, 2m, C(2) diastereo-
meric methylene), 2.20 (3H, s, acetate methyl), 2.10 (3H, s,
acetamido methyl); mass sprectrum (EI mode) m/ z 303 (M+),
260 (M+ - acetyl), 244 (M+ - acetate), 202 (M+ - acetate and
ketene), 187 (M+ - acetate, ketene, and methyl). Anal.
(C15H17N3O4) C, H, N.
2,4-Dia ceta m id o-5-(N-p yr r olid in o)a n isole (7). A solu-
tion of 5 g (18.7 mmol) of 6 in 25 mL of acetic anhydride and
5 mL of acetic acid was shaken under 50 psi of H2 in the
presence of 200 mg of 5% Pd on carbon for 10 h. The completed
reaction was filtered through Celite and the filtrate diluted
with 500 mL of diethyl ether. The resulting white precipitate
was filtered off and dried under vacuum. Recrystallization
from chloroform/hexane gave 7: 4.2 g (78%) yield; mp 203 °C;
TLC (chloroform/methanol, 95:5) Rf ) 0.28; IR (KBr pellet)
3277, 2969, 2868, 1641, 1545, 1507, 1445, 1414, 1370, 1261,
1213, 1146, 1032, 1110, 604 cm-1; 1H NMR (dimethyl sulfoxide-
d6) δ 9.06 and 8.91 (2H, 2s, amide protons), 7.53 and 6.39 (2H,
2s, aromatic protons), 3.79 (3H, s, methoxy), 3.33 (4H, m,
methylenes adjacent to pyrrolidine nitrogen), 2.00 and 1.96
(6H, 2s, acetyl methyls), 1.83 (4H, m, other pyrrolidine
methylenes); MS (EI mode) m/ z 291 (M+), 248 (M+ - acetyl),
233 (M+ - acetamido). Anal. (C15H21N3O3‚0.2H2O) C, H, N.
3-Acetoxy-7-m eth oxy-6-n itr o-2,3-d ih yd r o-1H-p yr r olo-
[1,2-a ]ben zim id a zole (8). A mixture consisting 5.34 g (20
mmol) of 6, 5.44 g (40 mmol) of anhydrous ZnCl2, and 20 mL
of acetic anhydride was refluxed at 90-100 °C for 4 h. The
reaction mixture was cooled, diluted with 100 mL of water,
and then extracted with 3 × 100 mL portions of chloroform.
The dried extracts (sodium sulfate) were concentrated to a
brown residue, which was flash chromatographed on a silica
gel column using ethyl acetate/methanol (95:5) as the eluant.
Evaporation of the eluant afforded the pure product, which
was recrystallized from ethyl acetate/hexane: 940 mg (16.2%
yield); mp 164 °C; TLC (ethyl acetate/methanol, 95:5) Rf ) 0.38;
IR (KBr pellet) 3596, 3416, 3092, 2945, 1740, 1640, 1588, 526,
6-Acetam ido-7-m eth oxy-8-n itr o-2,3-dih ydr o-1H-pyr r olo-
[1,2-a ]ben zim id a zole (10a ). To a mixture of 3.5 mL of
fuming nitric acid and 0.5 mL of concentrated sulfuric acid,
chilled at -20 °C, was added 376 mg (2.0 mmol) of 9a . The
reaction mixture was stirred for 10 min and then poured into
60 mL of cold water. The mixture was neutralized carefully
to pH 7.00 using saturated sodium bicarbonate solution and
extracted with 3 × 100 mL portions of chloroform followed by
drying (sodium sulfate) of the extracts. Evaporation of the
solvent afforded a yellow solid which was a mixture of the
8-nitro isomer with a trace of the 5-nitro isomer. Chromatog-
raphy on silica gel employing chloroform/methanol (9:1) as the
eluant afforded the 8-nitro isomer 10a : 336 mg (58%) yield;
TLC (chloroform/methanol, 9:1) Rf ) 0.19; mp 254-256 °C;
IR (KBr pellet) 2986, 1663, 1530, 1470, 1414, 1366, 1262, 1049,
748 cm-1 1H NMR (CDCl3) δ 7.36 (1H, bs, amide protons),
;
6.97 (1H, s, aromatic proton), 4.11 (2H, t, J ) 7.1 Hz, C(1)
methylene), 3.91 (3H, s, methoxy), 3.08 (2H, t, J ) 7.7 Hz,
C(3) methylene), 2.75 (2H, quint, J ) 7 Hz, C(2) methylene),
2.15 (3H, s, acetamido methyl); MS (EI mode) m/ z 290 (M+),
248 (M+ - ketene), 244 (M+ - nitro), 233 (M+ - acetamido).
Anal. (C13H14N4O4‚0.85CHCl3) C, H, N.
6-Aceta m id o-3-a cetoxy-7-m eth oxy-8-n itr o-2,3-d ih yd r o-
1H-p yr r olo[1,2-a ]ben zim id a zole (10b). To a mixture of 1
mL of fuming nitric acid and 0.15 mL of sulfuric acid, chilled
in an ice-water bath (-20 °C), was added 195 mg (0.65 mmol)
of 9b. The reaction mixture was stirred in the ice-water bath
for 10 min and then poured over a mixture 20 mL of chloroform
and 20 g of ice. The pH of this mixture was adjusted to 7.0 by
adding saturated sodium bicarbonate. The chloroform layer
was separated, and the aqueous layer was then extracted with
3 × 30 mL portions of chloroform. The combined chloroform
extracts were dried (sodium sulfate) and evaporated to a
residue. Chromatography on silica gel employing chloroform/
methanol (9:1) as the eluant afforded 10b. Recrystallization
was carried out from chloroform: 132 mg (58%) yield; TLC
(ethyl acetate/methanol, 95:5) Rf ) 0.50; mp 242-244 °C; IR
(KBr pellet) 3341, 3362, 2928, 2361, 1750, 1701, 1535, 1478,
1435, 1356, 1317, 1246, 1704, 1208, 986, 837, 760 cm-1 1H
;
NMR (CDCl3) δ 8.27 and 6.92 (2H, 2s, aromatic protons), 6.16
(1H, dd, J ) 7.6, 3.4 Hz, C(3) proton), 4.29 and 4.14 (2H, 2m,
C(1) diastereomeric methylene), 3.98 (3H, s, methoxy), 3.21
and 2.70 (2H, 2m, C(2) diastereomeric methylene), 2.12 (3H,
s, acetate methyl); MS (EI mode) m/ z 291 (M+), 249 (M+
-
ketene), 234 (M+ - ketene and methyl). Anal. (C13H13N3O5)
C, H, N.
6-Aceta m id o-7-m eth oxy-2,3-d ih yd r o-1H-p yr r olo[1,2-a ]-
ben zim id a zole (9a ). A mixture of 2.5 g (8.59 mmol) of 7, 14
mL of 96% formic acid, and 7 mL of 30% hydrogen peroxide
was stirred at 50 °C for 30 min. The reaction mixture was
then diluted with water and neutralized to pH 7.00 with
concentrated ammonium hydroxide. After the neutralized
solution was chilled, the precipitated crystals were filtered off
and dried. Recrystallization from chloroform/hexane afforded
9a : 1.1 g (52.2%) yield; mp 217 °C; TLC (chloroform/methanol,
9:1) Rf ) 0.54; IR (KBr pellet) 3348, 1674, 1593, 1530, 1479,
1371, 1223, 1111, 1055, 966, 828 cm-1 1H NMR (dimethyl
;
sulfoxide-d6) δ 9.67 (1H, s, amine proton), 7.61 (1H, s, aromatic
proton), 6.11 (1H, dd, J ) 7.7, 3.5 Hz, C(3) proton), 4.25 (2H,
m, C(1) diastereomeric methylene), 3.90 (1H, s, methoxy), 3.14
and 2.63 (2H, m, C(2) diastereomeric methylene), 2.08 (1H, s,
acetate methyl), 1.99 (1H, s, acetamido methyl); MS (EI mode)
m/ z 348 (M+), 306 (M+ - ketene), 302 (M+ - nitro), 289 (M+
- acetate). Anal. (C15H16N4O6‚0.90CHCl3) C, H, N.
1460, 1441, 1254, 1204, 1105, 953, 880, 804 cm-1 1H NMR
;
(CDCl3) δ 8.68 and 6.77 (2H, 2s, aromatic protons), 7.75 (1H,
bs, amide protons), 4.06 (2H, t, J ) 7.6 Hz, C(1) methylene),
3.93 (3H, s, methoxy), 3.03 (2H, t, J ) 7.6 Hz, C(3) methylene),
2.70 (2H, quint, J ) 7.5 Hz, C(2) methylene), 2.22 (3H, s,
6-Aceta m id o-8-a m in o-7-m eth oxy-2,3-d ih yd r o-1H-p yr -
r olo[1,2-a ]ben zim id a zole (11a ). A solution of 200 mg (0.69
mmol) of 10a in 20 mL of methanol was shaken under 50 psi
of H2 in the presence of 60 mg of 5% Pd on carbon for 16 h.
The catalyst was removed by filtration through Celite, and the
solvent was removed in vacuo, 109 mg (51.8%) yield. Further
purification of the unstable amine was not attempted, and it
was carried directly to the next step.
6-Acetam ido-3-acetoxy-8-am in o-7-m eth oxy-2,3-dih ydr o-
1H-p yr r olo[1,2-a ]ben zim id a zole (11b). A suspension of
205 mg (0.59 mmol) of 10b and 50 mL of methanol was shaken
under 50 psi of H2 in the presence of 100 mg of 5% Pd on
acetamido methyl); MS (EI mode) m/ z 245(M+), 203 (M+
-
ketene), 188 (M+ - acetamido). Anal. (C13H15N3O2) C, H, N.
6-Aceta m id o-3-a cetoxy-7-m eth oxy-2,3-d ih yd r o-1H-p yr -
r olo[1,2-a ]ben zim id a zole (9b). A solution of 180 mg (0.62
mmol) of 8 in 50 mL of methanol was shaken under 50 psi of
H2 in the presence of 80 mg of 5% Pd on carbon for 5 h. The
completed reaction was filtrated through Celite into a flask
containing 0.35 mL of acetic acid, and the filtrate was

4328 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21
Zhou and Skibo
carbon for 12 h. The catalyst was removed by filtration
through Celite, and the solvent was removed in vacuo, 170
mg (90%) yield. Further purification of the unstable amine
was not attempted, and it was carried directly to the next
step.
(ethyl acetate/methanol, 95:5) Rf ) 0.62; IR (neat) 3318, 2963,
2876, 2834, 1669, 1613, 1370, 1040, 945, 862, 806, 754 cm-1
;
1H NMR (CDCl3) δ 8.550 (1H, br s, amide proton), 8.06 (1H,
d, J meta ) 2.7, Hz, aromatic C(2) proton), 7.51 and 6.50 (2H,
abx, J ortho ) 8.8 Hz, J meta ) 2.3 Hz, J para ) 0 Hz), 3.80 (3H, s,
methoxy), 2.95 (4H, m, pyrrolidine methylene adjacent to
nitrogen), 2.19 (3H, s, acetamido methyl), 1.952 (4H, m, other
pyrrolidine methylenes); MS (EI mode) m/ z 234 (M+), 219 (M+
- methyl), 191 (M+ - acetyl), 176 (M+ - acetamido). Anal.
(C13H18N2O2‚0.2H2O) C, H, N.
6-Me t h oxy-2,3-d ih yd r o-1H -p yr r olo[1,2-a ]b e n zim id -
a zole (15a ). A mixture consisting of 1.855 g (8.1 mmol) of
14, 9.5 mL of 98% formic acid, and 4.5 mL of 30% hydrogen
peroxide was stirred at 50 °C for 60 min. The reaction mixture
was then poured onto 200 mL of ice water and neutralized
with concentrated ammonia hydroxide. Extraction of the
neutralized solution with 3 × 100 mL portions of chloroform,
drying of the extracts (sodium sulfate), and then concentration
of the extracts afforded a dark solid residue. Chromatography
of this residue on silica gel, employing chloroform/methanol
as the eluant (9:1), afforded 15a as a brown solid: 1.23 g (81%)
6-Aceta m id o-7-m eth oxy-2,3-d ih yd r o-1H-p yr r olo[1,2-a ]-
ben zim id a zole-5,8-d ion e (1a ). To a suspension of 109 mg
(0.42 mmol) of 11a in 10 mL of water containing 200 mg of
potassium phosphate monobasic was added a solution of 1.0 g
of Fremy's salt in 30 mL of water containing 500 mg of
potassium phosphate monobasic (pH ) 3.0). The reaction
mixture was stirred at room temperature for 1.5 h and then
extracted with 3 × 30 mL portions of chloroform. The extracts
were dried (sodium sulfate) and concentrated to afford a brick-
red solid. Purification was carried out by chromatography on
silica gel employing chloroform/methanol (9:1) as the eluant.
Recrystallization from chloroform/hexane afforded 34 mg (18%)
yield: mp 182-184 °C; TLC (chloroform/methanol, 9:1) Rf )
0.43; IR (KBr pellet) 3403, 3156, 2986, 1663, 1530, 1470, 1418,
1366, 1262, 1227, 1184, 1111, 1049, 748 cm-1; 1H NMR (CDCl3)
δ 7.14 (1H, s, amide proton), 4.23 (2H, t, J ) 7.2 Hz, C(1)
methylene), 4.09 (3H, s, methoxy), 2.95 (2H, t, J ) 7.5, C(3)
methylene), 2.71 (2H, quint, J ) 7 Hz, C(2) methylene), 2.19
(3H, s, acetamido methyl); MS (EI mode) m/ z 275 (M+), 233
(M+ - ketene), 218 (M+ - ketene and methyl). Anal.
(C13H13N3O4) C, H, N.
yield; TLC (chloroform/methanol, 95:5) Rf
) 0.30; mp 128-
129 °C; IR (KBr pellet) 3424, 3046, 2959, 2899, 2839, 1586,
1528, 1489, 1447, 1300, 1242, 1198, 1150, 1103, 1034, 941, 891,
806, 719 cm-1
; 1H NMR (CDCl3) δ 7.20 (1H, d, J meta ) 2.3 Hz,
C(8) aromatic proton), 7.18 and 6.86 (2H, abx, J ortho ) 8.6 Hz,
J meta ) 2.3 Hz, C-5 and C-6 aromatic protons, respectively),
4.08 (2H, t, J ) 7.0 Hz C(1) methylene), 3.86 (3H, s, methoxy),
3.05 (2H, t, J ) 7.6 Hz, C(3) methylene), 2.69 (2H, quint, J )
7.5 Hz C(2) methylene); MS (EI mode) m/ z 188 (M+), 173 (M+
- methyl), 157 (M+ - methoxy). Anal. (C11H12N2O‚0.2H2O)
C, H, N.
6-Aceta m id o-3-a cetoxy-7-m eth oxy-2,3-d ih yd r o-1H-p yr -
r olo[1,2-a ]ben zim id a zole-5,8-d ion e (1b). To a suspension
of 11b, 170 mg (0.53 mmol), in 10 mL of water containing 200
mg of potassium phosphate monobasic was added a solution
of 1.0 g of Fremy’s salt in 30 mL of water containing 500 mg
of potassium phosphate monobasic (pH ) 3.0). The reaction
mixture was stirred for 5 h and then extracted with 3 × 50
mL portions of chloroform. The extracts were dried (sodium
sulfate) and concentrated, and the residue was chromato-
graphed on silica gel using chloroform/methanol (9:1) as the
eluant. Recrystallization from chloroform/ hexane afforded
pure 1b: 42 mg (21.4%) yield; TLC (chloroform/methanol, 9:1)
Rf ) 0.48; mp 166-168 °C; IR (KBr pellet) 3459, 3225, 3005,
2961, 1736, 1687, 1655, 1595, 1530, 1370, 1327, 1236, 1047,
3-Acet oxy-6-m et h oxy-2,3-d ih yd r o-1H -p yr r olo[1,2-a ]-
ben zim id a zole (15b). A solution consisting of 3.49 g (16
mmol) of 14 and 4.35 g (32 mmol) of anhydrous ZnCl2 in 16
mL of acetic anhydride was refluxed at 90-100 °C for 5 h.
The reaction mixture was cooled to room temperature, diluted
with 100 mL of water, and then extracted with 3 × 100 mL
portions of chloroform. The dried extracts (sodium sulfate)
were concentrated to a brown residue, which was flash
chromatographed on a silica gel column using ethyl acetate/
methanol (95:5) as the eluant. Evaporation of the eluants
afforded the pure product, which was recrystallized from ethyl
acetate/hexane: 520 mg (13% yield); mp 155-156 °C; TLC
(ethyl acetate) Rf ) 0.19; IR (KBr pellet) 2961, 1746, 1626,
748, 725, 594 cm-1 1H NMR (CDCl3) δ 7.17 (1H, s, amide
;
proton), 6.06 (1H, dd, J ) 7.7, 3.0 Hz, C(3) proton), 4.31 (2H,
m, C(1) diastereomeric methylene), 4.10 (3H, s, methoxy), 3.15
and 2.63 (2H, 2m, C(2) diastereomeric methylene), 2.19 (3H,
s, acetate methyl), 2.08 (3H, s, acetamido methyl); MS (EI
mode) m/ z 333 (M+), 291 (M+ - ketene), 276 (M+ - ketene
and methyl). Anal. (C15H15N3O6) C, H, N.
1582, 1530, 1495, 1423, 1370, 1246, 1146, 1080, 1028, 947, 814
1
cm-1; H NMR (dimethyl sulfoxide-d6) δ 7.45 (1H, d, J ortho
)
)
3-Nitr o-4-(N-p yr r olid in o)a n isole (13). A solution con-
sisting of 4.0 g (17.3 mmol) of 4-bromo-3-nitroanisole (12) and
6 mL of pyrrolidine (69.2 mol) in 10 mL of ethanol was heated
at reflux for 30 h. The solvent was then removed under
vacuum, and the residue was poured over 200 g of cracked
ice. The resulting mixture was extracted with 3 × 50 mL
portions of chloroform. The dried extracts (sodium sulfate)
were concentrated to an oily residue, which was placed on a
silica gel flash column. The product was eluted with ethyl
acetate/methanol (95:5). Evaporation of the eluant afforded
a red oil which slowly solidified in the refrigerator: 3.67 g
(96%) yield; TLC (ethyl acetate/methanol, 95:5) Rf ) 0.65; mp
53-54 °C; IR (KBr pellet) 2988, 2947, 2859, 1561, 1518, 1489,
1462, 1364, 1333, 1269, 1221, 1159, 1069, 1044, 918, 872, 808,
743, 631 cm-1; 1H NMR (CDCl3) δ 7.31 (1H, d, J meta ) 3.1 Hz,
aromatic C(2) proton), 7.14 and 6.83 (2H, abx, J ortho ) 9.2 Hz,
J meta ) 3.1 Hz, J para ) 0 Hz, C(5) and C(6) protons), 3.79 (3H,
s methoxy), 3.18 (4H, m, pyrrolidine methylene adjacent to
nitrogen), 1.98 (4H, m, other pyrrolidine methylenes); MS (EI
mode) m/ z 222 (M+), 205 (M+ - OH), 175 (M+ - OH and NO).
Anal. (C11H14N2O3‚0.2H2O) C, H, N.
8.8 Hz, C(8) aromatic proton), 7.72 and 6.62 (2H, abx, J ortho
8.8 Hz, J meta ) 2.3 Hz, J para ) 0, C(8) and C(7) protons,
respectively), 6.10 (1H, dd, J ) 7.6, 3.2 Hz, C(3) proton), 4.16
(2H, m, C(1) diastereomeric methylene), 3.78 (3H, s, methoxy),
3.13 (3H, s, acetate methyl), 3.11 and 2.59 (2H, 2m, C(2)
diastereomeric methylene); MS (EI mode) m/ z 246 (M
(M+ - acetyl), 187 (M+ - acetate). Anal. (C13H14N2O3) C, H,
N.
+), 203
6-Meth oxy-5-n itr o-2,3-d ih yd r o-1H-p yr r olo[1,2-a ]ben z-
im id a zole (16a ). To 4.2 mL of fuming nitric acid, chilled in
a 2-propanol-dry ice bath, was added 450 mg (2.4 mmol) of
15a . The reaction mixture was then warmed by placing it in
a salt-ice water bath for 5 min. The reaction mixture was
poured over 100 g of cracked ice and resulting solution adjusted
to pH 7 with saturated sodium bicarbonate solution. The
neutralized solution was extracted three times with 100 mL
portions of chloroform, and the dried extracts (sodium sulfate)
were concentrated to a yellow solid. Purification of this solid
was carried out by flash chromatography on silica gel using
ethyl acetate/methanol (95:5) as the eluant followed by re-
crystallization from chloroform/hexane: 212 mg (38%) yield;
mp 182 °C; TLC (ethyl acetate/methanol, 95:5) Rf ) 0.17; IR
(KBr pellet) 3432, 2953, 2915, 2367, 2346, 1584, 1524, 1449,
3-Aceta m id o-4-(N-p yr r olid in o)a n isole (14). A mixture
consisting of 3.2 g (14.4 mmol) of 13, 750 mg of 5% Pd on
carbon, 25 mL of acetic anhydride, and 3.2 mL of acetic acid
was shaken under 50 psi of H2 for 2 h. The completed reaction
was filtered through Celite. Concentration of the filtrate in
vacuo afforded a brownish oil. Chromatography of this oil on
silica gel employing ethyl acetate/methanol (95:5) as the eluant
afforded the product as a yellow oil: 3.3 g (98%) yield; TLC
1408, 1373, 1333, 1298, 1269, 1209, 1088, 955, 874, 799 cm-1
;
1H NMR (CDCl3) δ 7.57 and 6.76 (2H, abq, J ortho ) 8.9 Hz,
C(8) and C(7) aromatic proton, respectively), 4.15 (2H, t, J )
7.1 Hz, C(1) methylene), 3.97 (3H, s, methoxy), 3.13 (2H, t, J
) 7.6 Hz, C(3) methylene), 2.76 (2H, quint, J ) 7.5 Hz, C(2)
methylene); MS (EI mode) m/ z 233 (M+), 218 (M+ - methyl),

Chemistry of Pyrrolobenzimidazole Antitumor Agents
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21 4329
203 (M+ - NO), 186 (M+ - NO and OH). Anal. (C11H11N3O3‚
0.1H2O) C, H, N.
276 (M+), 233 (M+ - acetyl), 217 (M+ - acetate). Anal.
(C13H12N2O5) C, H, N.
3-Acetoxy-6-m eth oxy-5-n itr o-2,3-d ih yd r o-1H-p yr r olo-
[1,2-a ]ben zim id a zole (16b). To 0.7 mL of fuming nitric acid
and 0.7 mL of acetic anhydride cooled in an ice bath was added
98 mg (0.40 mmol) of 15b. The reaction mixture was stired
for about 5 min and then poured over 50 mg of cracked ice.
The resulting mixture was adjusted to pH 7 with saturated
sodium bicarbonate solution and then extracted three times
with 50 mL portions of chloroform. Drying the combined
chloroform extracts (sodium sulfate) followed by concentration
afforded a yellow solid, which was flash chromatographed on
silica gel using ethyl acetate as the eluant. The purified
product was recrystallized from chloroform/hexane: 72 mg
(68%) yield; TLC (ethyl acetate) Rf ) 0.28; mp 144-146 °C;
IR (KBr pellet) 2947, 1746, 1626, 1530, 1450, 1443, 1370, 1246,
6-(N-Azir id in yl)-2,3-d ih yd r o-1H -p yr r olo[1,2-a ]b en z-
im id a zole-5,8-d ion e (2a ). To a solution of 25 mg (0.11 mmol)
of 18a in 4 mL of dry methanol was added 0.5 mL of
ethylenimine, and the resulting mixture was stirred at room
temperature for 2 h. The solvent was removed in vacuo and
the red residue purified by flash chromatography on silica gel
employing chloroform/methanol (95:5) as the eluant. The
purified product was recrystallized from chloroform/hexane:
22 mg (83.5%) yield; TLC (chloroform/methanol, 95:5) Rf )
0.31; mp 206 °C; IR (KBr pellet) 3435, 2998, 2961, 2928, 2375,
1676, 1640, 1574, 1476, 1402, 1362, 1304, 1265, 1130, 1067,
1
993, 868 cm-1; H NMR (CDCl3) δ 5.83 (1H, s, C(7) aromatic
proton), 4.22 (2H, t, J ) 7.2 Hz, C(1) methylene), 2.96 (2H, t,
J ) 7.7 Hz, C(3) methylene), 2.70 (2H, quint, J ) 7.5 Hz, C(2)
methylene), 2.26 (4H, s, aziridine protons); MS (EI mode) m/ z
229 (M+). Anal. (C12H11N3O2) C, H, N.
1
1080, 1028, 843 cm-1; H NMR (CDCl3) δ 7.68 and 6.87 (2H,
abq, J ) 8.9 Hz, C(8) and C(7) aromatic protons, respectively),
6.16 (2H, dd, J ) 7.5, 3.2 Hz, C(3) proton), 4.32 and 4.18 (2H,
2m, C(1) diastereomeric methylene), 3.98 (3H, s, methoxy),
3.25 and 2.68 (2H, 2m, C(2) diastereomeric methylene), 2.13
(3H, s, acetate methyl); MS (EI mode) m/ z 291 (M+), 248
(M+ - acetyl), 232 (M+ - acetate). Anal. (C13H13N3O5) C, H,
N.
5-Am in o-6-m e t h oxy-2,3-d ih yd r o-1H -p yr r olo[1,2-a ]-
ben zim id a zole (17a ). A suspension of 50 mg (0.21 mmol) of
16a in 20 mL of methanol containing 70 mg of 5% Pd on carbon
was shaken under 50 psi of H2 for 3 h. The reaction mixture
was filtered through Celite and the filter cake washed with
methanol. Evaporation of the solvent afforded crude 17a . The
resulting solid was unstable and used in the Fremy oxidation
without further purification.
3-Acetoxy-5-a m in o-6-m eth oxy-2,3-d ih yd r o-1H-p yr r olo-
[1,2-a ]ben zim id a zole (17b). A suspension of 136 mg (0.47
mmol) of 16b in 20 mL of methanol containing 100 mg of 5%
Pd on carbon was shaken under 50 psi of H2 for 2 h. The
reaction mixture was filtered through Celite and the filter cake
washed with methanol. Evaporation of solvent afforded crude
17b which was used without further purification.
3-Acetoxy-6-(N-a zir id in yl)-2,3-d ih yd r o-1H-p yr r olo[1,2-
a ]ben zim id a zole-5,8-d ion e (2b). To a solution of 42 mg
(0.15 mmol) of 18b in 5 mL of dry methanol was added 0.2
mL of ethylenimine. The reaction mixture was then stirred
at room temperature for 2 h. The solvent was removed in
vacuo and the red residue flash chromatographed on silica gel
employing chloroform/methanol (95:5) as the eluant. The
purified product was recrystallized from chloroform/hexane:
33 mg (78%) yield; TLC (chloroform/methanol, 95:5) Rf ) 0.23;
mp 164-166 °C; IR (KBr pellet) 2992, 1748, 1678, 1572, 1510,
1364, 1257, 1235, 1064, 1024, 997, 893 cm-1; 1H NMR (CDCl3)
δ 6.08 (1H, dd, J ) 7.5, 3.0 Hz, C(3) proton), 5.90 (1H, s, C(7)
aromatic proton), 4.32 (2H, m, C(1) diastereomeric methylene),
3.15 and 2.64 (2H, 2m, C(2) diastereomeric methylene), 2.28
(4H, s, aziridinyl protons), 2.10 (3H, s, acetate methyl); MS
(EI mode) m/ z 287 (M+), 244 (M+ - acetyl), 230 (M+ - acetate).
Anal. (C14H13N3O4‚0.2H2O) C, H, N.
6-(N-Azir id in yl)-7-m et h oxy-2,3-d ih yd r o-1H -p yr r olo-
[1,2-a ]ben zim id a zole-5,8-d ion e (3). To a mixture of sodium
methoxide prepared by reacting 40 mg of sodium metal in 5
mL of methanol was added 12.3 mg (0.054 mmol) of 2a . The
mixture was stirred at room temperature for 10 min and then
diluted with 10 mL of water. The resulting mixture was
extracted three times with 20 mL portions of chloroform. The
dried extracts (sodium sulfate) were concentrated and chro-
matographed on silica gel employing chloroform/methanol as
the eluant. Recrystallization from chloroform/hexane afforded
8 mg (57% yield) of 3: mp 156-158 °C dec; TLC (chloroform/
methanol, 95:5) Rf) 0.44; IR (KBr pellet) 1701, 1645, 1580,
6-Meth oxy-2,3-dih ydr o-1H-pyr r olo[1,2-a ]ben zim idazole-
5,8-d ion e (18a ). To a suspension of the crude 17a in 10 mL
of water, containing 80 mg of monobasic potassium phosphate,
was added a solution of 400 mg of Fremy's salt in 30 mL of
water containing 200 mg of monobasic potassium phosphate.
The reaction mixture was stirred at room temperature for 2 h
and then extracted three times with 50 mL portions of
chloroform. The dried extracts (sodium sulfate) were concen-
trated and then purified by chromatography on silica gel
employing chloroform/methanol (95:5) as the eluant. Recrys-
tallization from chloroform/hexane afforded 18a in 29 mg
(63.3%) yield: TLC (chloroform/methanol, 95:5) Rf ) 0.40; mp
221 °C; IR (KBr pellet) 3490, 3054, 2980, 2942, 1692, 1647,
1
1514, 1343, 1306, 1148, 1067, 1018 cm-1; H NMR (CDCl3) δ
4.22 (2H, t, J ) 7.2 Hz, C(1) methylene), 3.93 (3H, s, methoxy),
2.94 (2H, t, J ) 7.7 Hz, C(3) methylene), 2.70 (2H, quint,
J ) 7.5 Hz, C(2) methylene), 2.33 (4H, s, aziridinyl protons);
MS (EI mode) m/ z 259 (M+), 244 (M+ - methyl). Anal.
(C13H13N3O3‚1.5CHCl3) C, H, N.
1
1589, 1476, 1298, 1244, 1134, 1084, 1040, 868, 723 cm-1; H
NMR (CDCl3) δ 5.70 (1H, s, C(7) aromatic proton), 4.25 (2H,
t, J ) 7.2 Hz, C(1) methylene), 3.86 (3H, s, methoxy), 2.97 (2H,
t, J ) 7.6 Hz, C(3) methylene), 2.72 (2H, quint, J ) 7.5 Hz,
6-Am in o-7-m e t h oxy-2,3-d ih yd r o-1H -p yr r olo[1,2-a ]-
ben zim id a zole (19). A solution of 326 mg (1.9 mmol) of 9a
in 8 mL of concentrated sulfuric acid and 8 mL of water was
warmed at 80 °C for 8 h. The mixture was poured onto ice
and adjusted to pH 8.0 with concentrated ammonium hydrox-
ide solution. Extraction of the cooled solution with 3 × 50 mL
of chloroform followed by drying (sodium sulfate) of the
extracts and evaporation of the solvent afforded 19 as a crude
solid: 370 mg (96%) yield; TLC (chlorform/methanol, 9:1) Rf
) 0.51; mp 205-207 °C; 1H NMR (CDCl3) δ 7.06 and 6.73 (2H,
2s, aromatic protons), 4.04 (2H, t, J ) 7.0 Hz, C(1) methylene),
3.90 (3H, s, methoxy), 3.01 (2H, t, J ) 7.6 Hz, C(3) methylene),
2.69 (2H, quint, J ) 7.2 Hz, C(2) methylene); MS (EI mode)
m/ z 203 (M+), 172 (M+ - methoxy).
7-Meth oxy-5-n itr o-2,3-d ih yd r o-1H-p yr r olo[1,2-a ]ben z-
im id a zole (20). A mixture of 370 mg (1.83 mmol) of 19 in 12
mL of concentrated sulfuric acid and 5 mL of acetic acid was
cooled to -10 °C, and then 0.88 g of sodium nitrite in 4 mL of
water was added while maintaining the reaction temperature
at less than 10 °C. After the addition was complete, the
reaction mixture was stirred at 0 °C for 45 min. The resulting
mixture was added to 12 mL of 50% hypophosphous acid while
maintaining the temperature at less than 5 °C. The reaction
C(2) methylene); MS (EI mode) m/ z 218 (M+), 203 (M+
methyl). Anal. (C11H10N2O3‚0.1H2O) C, H, N.
-
3-Acet oxy-6-m et h oxy-2,3-d ih yd r o-1H -p yr r olo[1,2-a ]-
ben zim id a zole-5,8-d ion e (18b). To a suspension of the
crude 17b in 10 mL of water, containing 200 mg of monobasic
potassium phosphate, was added a solution of 1.5 g of Fremy's
salt in 50 mL of water containing 500 mg of monobasic
potassium phosphate. The reaction mixture was stirred at
room temperature for 2 h and extracted with 3 × 50 mL
portions of chloroform. The dried extracts (sodium sulfate)
were concentrated to a residue, which was then chromato-
graphed over silica gel employing chloroform/methanol (95:5)
as eluant. Recrystallization from chloroform/hexane afforded
52 mg (39%) yield: TLC (chloroform/methanol, 95:5) Rf ) 0.56;
mp 189-191 °C dec; IR (KBr pellet) 2994, 1748, 1638, 1582,
1530, 1371, 1275, 1213, 799 cm-1; 1H NMR (CDCl3) δ 6.09 (1H,
dd, J ) 7.7, 3.2 Hz, C(3) proton), 5.78 (1H, s, C(7) aromatic
proton), 4.35 (2H, m, C(1) diastereomeric methylene), 3.88 (3H,
s, methoxy), 3.16 and 2.66 (2H, 2m, C(2) diastereomeric
methylene), 2.11 (3H, s, acetate methyl); MS (EI mode) m/ z

4330 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21
Zhou and Skibo
mixture was then stirred at room temperature for 90 min, after
which it was diluted with 25 mL of water and neutralized to
pH 7.00 with concentrated ammonia hydroxide solution. The
neutralized mixture was then extracted three times with 50
mL portions of chloroform, and the extracts were dried (sodium
sulfate) and concentrated. Chromatography on silica gel
employing chloroform/methanol (9:1) as the eluant afforded
pure 20, which was recrystallized from chloroform/hexane: 67
mg (15.7%) yield; mp 157-158 °C; TLC (chloroform/methanol,
95:5) Rf ) 0.49; IR (KBr pellet) 2986, 1626, 1576, 1508, 1443,
1393, 1285, 1252, 1192, 1088, 995, 824 cm-1; 1H NMR (CDCl3)
δ 7.69 and 7.08 (2H, 2d, J ) 2.4 Hz, aromatic proton), 4.13
(2H, t, J ) 7.1 Hz, C(1) methylene), 3.89 (3H, s, methoxy),
3.14 (2H, t, J ) 7.7 Hz, C(3) methylene), 2.76 (2H, quint, J )
7.4 Hz, C(2) methylene); MS (EI mode) m/ z 233 (M+), 188 (M+
- NO2). Anal. (C11H11N3O3) C, H, N.
26: 1H NMR (CDCl3) δ 6.00 (1H, bs, amino proton), 5.20 (1H,
s, aromatic proton), 4.24 (2H, t, J ) 7.2 Hz, C(1) methylene),
3.19 (2H, m, methylene of ethyl), 2.93 (2H, t, J ) 7.5 Hz, C(3)
methylene), 2.70 (2H, quint, J ) 7.3 Hz, C(2) methylene), 1.32
(3H, t, J ) 7.2 Hz, methyl of ethyl); MS (EI mode) m/ z 289
(M+), 246 (M+ - acetyl), 230 (M+ - acetate).
27: 1H NMR (CDCl3) δ 6.28 (1H, bs, amine), 5.25 (1H, s,
aromatic), 4.25 (2H, t, J ) 7.2 Hz, C (1) methylene), 3.72 (2H,
t, J ) 5.7 Hz, methylene next to chloro), 3.53 (2H, q, J ) 6.0
Hz, methylene next to amino), 2.95 (2H, t, J ) 7.5 Hz, C(3)-
methylene), 2.70 (2H, quint, J ) 7.2 Hz, C(2)methylene); MS
(EI mode) m/ z 265 (M+), 230 (M+ - Cl), 216(M+ - chlorom-
ethyl).
28: 1H NMR (CDCl3) δ 5.50 (1H, s, aromatic proton), 5.21
(2H, bs, amino proton), 4.23 (2H, t, J ) 7.8 Hz, C(1) methyl-
ene), 2.94 (2H, t, J ) 7.8 Hz, C(3) methylene), 2.70 (2H, quint,
J ) 6.9 Hz, C(2) methylene); MS (EI mode) m/ z 261 (M+), 218
(M+ - acetyl).
Hyd r olysis of Red u ced 2b. The reduction of 2b in 0.05
M, pH 7.4, Tris buffer and workup of the hydrolysis products
were carried out as described for 2a . Silica gel chromatogra-
phy of the chloroform extracts employing chloroform/methanol
(95:5) as the eluant separated products 29 and 30. The yields,
which were obtained upon evaporation of collected fractions,
were 31% and 40%, respectively. The aqueous layer which
had been extracted with chloroform was added to a 20 g
Bakerbond phenyl (40 µm) reverse-phase column prepared
with water. The column was eluted with water resulting in
removal of a purple product (33), 8% yield upon concentration
of the product fraction. Physical properties of 29-31 are
provided below.
5-Am in o-7-m e t h oxy-2,3-d ih yd r o-1H -p yr r olo[1,2-a ]-
ben zim id a zole (21). A solution of 63 mg (0.27 mmol) of 20
in 20 mL of methanol was shaken under 50 psi of H2 in the
presence of 50 mg of 5% Pd on carbon for 4 h. The catalyst
was removed by filtration through Celite, and the solvent was
removed under vacuum. Evaporation of the solvent afforded
crude 21 which was used without further purification.
7-Meth oxy-2,3-dih ydr o-1H-pyr r olo[1,2-a ]ben zim idazole-
5,8-d ion e (22). To a suspension of crude 21 in 10 mL of water
containing 200 mg of monobasic potassium phosphate was
added a solution of 0.5 g of Fremy's salt in 30 mL of water
containing 500 mg of potassium phosphate monobasic. The
reaction mixture was stirred at room temperature for 2 h and
then extracted with 3 × 30 mL portions of chloroform. Drying
of the extracts (sodium sulfate) and concentration afforded a
yellow solid, which was purified by chromatography on silica
gel employing chloroform/methanol (9:1) as the eluant. The
purified product was recrystallized from chloroform/hexane:
18 mg (31%) yield; mp 214-216 °C; TLC (chloroform/methanol,
95:5) Rf ) 0.51; IR (KBr pellet) 3102, 1672, 1645, 1589, 1512,
1335, 1252, 1184, 1163, 1090, 1030, 872 cm-1; 1H NMR (CDCl3)
δ 5.78 (1H, s, aromatic proton), 4.28 (2H, t, J ) 7.2 Hz, C(1)
methylene), 3.85 (3H, s, methoxy), 2.99 (2H, t, J ) 7.6 Hz,
C(3) methylene), 2.74 (2H, quint, J ) 7.3 Hz, C(2) methylene);
MS (EI mode) m/ z 218 (M+). Anal. (C11H10N2O3) C, H, N.
7-(N-Azir id in yl)-2,3-d ih yd r o-1H -p yr r olo[1,2-a ]b en z-
im id a zole-5,8-d ion e (4). To a solution of 10 mg (0.046 mmol)
of 22 in 2 mL of methanol, chilled at 0 °C, was added 0.2 mL
of ethylenimine. After stirring at 0 °C for 2 h, the solvent was
removed and the orange residue chromatographed on silica
gel employing ethyl acetate/methanol (9:1) as the eluant. Final
purification was carried out by recrystallization from ethyl
acetate/hexane: 3.5 mg (33%) yield; mp 142 °C; TLC (chloro-
form/methanol, 95:5) Rf ) 0.28; IR (KBr pellet) 2926, 2233,
29: 1H NMR (CDCl3) δ 6.08 (1H, dd, J ) 7.5, 3.0 Hz, C(3)
proton), 6.05 (1H, bs, amine proton), 5.28 (1H, s, aromatic
proton), 4.35 (2H, m, C(1) diastereomeric methylene), 3.21 (2H,
m, methylene of ethyl), 3.19 and 2.63 (2H, 2m, C(2) diaster-
eomeric methylene), 2.11 (3H, s, acetate methyl), 1.33 (3H, t,
J ) 7.2 Hz, methyl of ethyl); MS (EI mode) m/ z 298 (M+), 246
(M+ - acetyl), 230 (M+ - acetate).
30: 1H NMR (CDCl3) δ 6.08 (1H, dd, J ) 7.5, 3.0 Hz, C(3)
proton), 5.57 (1H, s, aromatic proton), 5.29 (2H, bs, amino
proton), 4.34 (2H, m, C(1) diastereomeric methylene), 3.15 and
2.64 (2H, 2m, C(2) diastereomeric methylene), 2.11 (3H, s,
acetate methyl); MS (EI mode) m/ z 261 (M+), 218 (M+
acetyl), 202 (M+ - acetate).
-
31: 1H NMR (dimethyl sulfoxide-d6) δ 7.28 (2H, bs, amino
proton), 5.85 (1H, m, C(3) proton), 5.34 (1H, s, aromatic
proton), 4.96 (1H, m, hydroxyl proton), 4.19 and 4.07 (2H, 2m,
C(1) diastereomeric methylene), 2.88 and 2.35 (2H, 2m, C(2)
diastereomeric methylene); MS (EI mode) m/ z 219 (M+), 202
(M+ - OH).
1
1666, 1643, 2579, 1512, 1359, 1163, 985, 732 cm-1; H NMR
Hyd r olysis of Red u ced 3. The reduction of 3 in 0.05 M,
pH 7.4, Tris buffer and workup of the hydrolysis products were
carried out as described for 2a . The concentrated chloroform
extracts were subjected to preparative thin layer chromato-
graphic separation employing chloroform/methanol (95:5) as
the eluant. The isolated compounds included 32 (22%), 33
(11%), and unreacted 3 (15%). Physical properties of 32 and
33 are provided below.
(CDCl3) δ 5.87 (1H, s, aromatic proton), 4.23 (2H, t, J ) 7.2
Hz, C(1) methylene), 2.94 (2H, t, J ) 7.6 Hz, C(3) methylene),
2.69 (2H, quint, J ) 7.3 Hz, C(2) methylene), 2.18 (4H, s,
aziridinyl proton); MS (EI mode) m/ z 229 (M+), 187 (M+
aziridine).
-
Hyd r olysis of Red u ced 2a (23). To 10 mL of the buffer
were added 11 mg (0.04 mmol) of 3 in 2 mL of dimethyl
sulfoxide and 10 mg of 5% Pd on carbon. The mixture was
purged with argon for 10 min, and then hydrogen gas was
passed through the deaerated mixture for 15 min, resulting
in the formation of a colorless solution. Excess hydrogen was
removed from this solution by purging with argon for ∼5 min.
Incubation of the reaction mixture for 18 h at 30 °C resulted
in the formation of a purple solution, which was opened to the
air and filtered through Celite. The filtrate was extracted with
3 × 20 mL portions of chloroform. The extracts were washed
two times with water, dried (sodium sulfate), and then
concentrated to a residue. Silica gel chromatography employ-
ing chloroform/methanol (95:5) as the eluant successfully
separated hydrolysis products.
32: 1H NMR (CDCl3) δ 5.60 (1H, bs, amine proton), 4.23
(2H, t, J ) 7.2 Hz, C(1) methylene), 3.78 (3H, s, methoxy),
3.63 (2H, quint, J ) 6.6 Hz, methylene of ethyl), 2.91 (2H, t,
J ) 7.2 Hz, C(2) methylene), 2.69 (2H, quint, J ) 7.2 Hz, C(3)
methylene), 1.27 (3H, t, J ) 7.2 Hz, methyl of ethyl); MS (EI
mode) m/ z 261 (M+), 246 (M+ - methyl).
33: 1H NMR (CDCl3) δ 5.03 (2H, bs, amino proton), 4.22
(2H, t, J ) 7.2 Hz, C(1) methylene), 3.89 (3H, s, methoxy),
2.92 (2H, t, J ) 7.5 Hz, C(3) methylene), 2.69 (2H, quint, J )
7.2 Hz, C(2) methylene); MS (EI mode) m/ z 233 (M+), 218 (M+
- methyl).
Ca lf Th ym u s DNA Tr ea ted w ith Red u ced 2a . The
reaction mixture consisted of the following components: 7.18
mg of sonicated calf thymus DNA dissolved in 5 mL of 0.05
M, pH 7.4, Tris buffer, 4.87 mg (0.0213 mmol) of 2a dissolved
in 2 mL of dimethyl sulfoxide, and 10 mg of 5% Pd on carbon.
The combination of these components, deaeration, and cata-
lytic reduction were carried out as described under Hydrolysis
When 0.05 M, pH 7.4 and 5.5, Tris buffers were employed,
26 and 28 were obtained in 37% and 18% yields, respectively.
When 1.0 M, pH 2.0, formate buffer (µ ) 1.0, KCl) was
employed, the only product was 27 (56% isolated yield).
Physical properties of 26-28 are provided below.

Chemistry of Pyrrolobenzimidazole Antitumor Agents
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21 4331
(11) Grantham, R. K.; Meth-Cohn, O. The Formation of Benzimid-
azolones and Quinoxalines from o-Nitrophenyldialkylanilines:
A Re-Investigation. J . Chem. Soc. C 1969, 70-74.
(12) The chemistry of benzimidazole N-oxides is consistent with these
reactions: see: (a) Takahashi, S.; Kanoj, H. Benzimidazole
N-Oxides. VII. The Reactivity of 1,2-Dimethylbenzimidazole
3-Oxide. Chem. Pharm. Bull. 1966, 14, 1219-1227. (b) Taka-
haski, S.; Kanoj, H. Benzimidazole N-Oxides. III. The Reactivity
of 1-Methylbenzimidazole 3-Oxide. Chem. Pharm. Bull. 1964,
12, 783-788.
(13) Skibo, E. B.; Islam, I.; Schulz, W. G.; Zhou, R.; Bess, L.; Boruah,
R. The Organic Chemistry of the Pyrrolo[1,2-a]benzimidazole
Antitumor Agents. An Example of Rational Drug Design. Synlett
1996, 4, 197-309.
(14) Skibo, E. B. Formation and Fate of Benzimidazole-Based
Quinone Methides. Influence of pH on Quinone Methide Fate.
J . Org. Chem. 1992, 57, 5874-5878.
(15) Preston, P. N. Benzimidazoles. In Benzimidazoles and Conge-
neric Tricyclic Compounds, Part 1; Preston, P. N., Ed.; J ohn
Wiley: New York, 1981; Chapter 1, p 80.
of Reduced 2a . The anaerobic incubation of the reaction was
carried out at 30 °C for 24 h. The completed reaction was
opened to the air and filtered throught Celite to remove the
catalyst. The filtrate was extracted three times with 50 mL
portions of chloroform to remove hydrolysis products. These
products were separated and yields obtained as described
under Hydrolysis of Reduced 2a . The aqueous layer was
adjusted to 0.3 M sodium acetate with 3.0 M, pH 5.1, sodium
acetate stock and then diluted with 3 vol of ethanol. This
mixture was chilled at -20 °C for 24 h and then centrifuged
at 5000g for 15 min. The DNA pellet was washed by suspend-
ing in ethanol and centrifuging. Weight of the vacuum-dried
DNA pellet was 5.87 mg, and the yields of hydrolysis products
were 26, 35%, and 28, 10%.
Top oisom er a se II Assa ys. The topoisomerase relaxation
reactions were carried out with 0.50 µg of SV-40 supercoiled
(form I) DNA and 20 units of Drosophila melanogaster topo-
isomerase II (United States Biochemical) in 20 λ of 100 mM
Tris‚HCl (pH 7.9) containing 500 mM NaCl, 500 mM KCl, 50
mM MgCl2, 1 mM EDTA, 150 µg/mL BSA, and 10 mM ATP.
The reactions were run for 30 min at 37 °C and then combined
with 1 λ of a solution of proteinase K and SDS (3.75 and 75
mg, respectively, in 1.5 mL of H2O). After incubation at 37
°C for 45 min further, the reaction mixtures were combined
with 4 λ of 6× gel loading solution (0.25% bromophenol blue,
0.25% xylene cyanol FF, 30% glycerol in water). A 1.3%
agarose gel run for 18 h in 0.045 M Tris borate buffer with 1
mM EDTA was employed to assay the relaxation reactions.
The DNA bands were visualized with ethidium bromide.
(16) Dermer, O. C.; Ham, G. E. Ethyleneimines and Other Aziridines;
Academic Press: New York, 1969; p 108.
Values for
(17) Boruah, R. C.; Skibo, E. B. Determination of the pKa
the Mitomycin C Redox Couple by Titration, pH Rate Profiles,
and Nernst-Clark Fits. Studies of Methanol Elimination, Car-
bocation Formation, and the Carbocation/Quinone Methide
Equilibrium. J . Org. Chem. 1995, 60, 2232-2243.
(18) Reference 16, pp 205-303.
(19) Evans, D. A.; Golob, A. M. [3,3]Sigmatropic Rearrangement of
1,5-Dione Alkoxides. The Powerful Accelerating Effects of the
Alkoxide Substituent. J . Am. Chem. Soc. 1975, 97, 4765-4766.
(20) Electron-rich quinones possess low redox potentials: Becker, H.-
D. Quinones as Oxidants and Dehydrogenating Agents. In The
Chemistry of the Quinonoid Compounds; Patai, S., Ed.; Wiley:
New York, 1979; Chapter 7.
(21) Boyd, M. R. Status of the NCI Preclinical Antitumor Drug
Discovery Screen. Princ. Pract. Oncol. (PPO Updates) 1989, 3
(No. 10).
(22) Skibo, E. B. The Discovery of the Pyrrolo[1,2-a]benzimidazole
Antitumor AgentssThe Design of Selective Antitumor Agents.
Curr. Med. Chem. 1996, 2, 900-931.
(23) Liu, L. F. DNA Topoisomerase Poisons as Antitumor Drugs.
Annu. Rev. Biochem. 1989, 58, 351-375.
Ack n ow led gm en t. This research was supported by
awards from the American Cancer Society and the
National Science Foundation.
Refer en ces
(1) Islam, I.; Skibo, E. B. Synthesis and Physical Studies of
Azamitosene and Iminoazamitosene Reductive Alkylating Agents.
Iminoquinone Hydrolytic Stability, Syn/Anti Isomerization, and
Electrochemistry. J . Org. Chem. 1990, 55, 3195-3205.
(2) Islam, I.; Skibo, E. B.; Dorr, R. T.; Alberts, D. S. Structure-
Activity Studies of Antitumor Agents Based on Pyrrolo-
[1,2-a]benzimidazoles: New Reductive Alkylating DNA Cleaving
Agents. J . Med. Chem. 1991, 34, 2954-2961.
(3) (a) Skibo, E. B.; Schulz, W. G. Pyrrolo[1,2-a]benzimidazole-Based
Aziridinyl Quinones. A New Class of DNA Cleaving Agent
Exhibiting G and A Base Specificity. J . Med. Chem. 1993, 36,
3050-3055. (b) Schulz, W. G.; Nieman, R. A.; Skibo, E. B.
Evidence for DNA Phosphate Backbone Alkylation and Cleavage
by Pyrrolo[1,2-a]benzimidazoles: Small Molecules Capable of
Causing Base-Pair-Specific Phosphodiester Bond Hydrolysis.
Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 11854-11858.
(4) Skibo, E. B.; Islam, I.; Heileman, M. J .; Schulz, W. G. Structure-
Activity Studies of Benzimidazole-Based DNA-Cleaving Agents.
Comparison of Benzimidazole, Pyrrolobenzimidazole and Tet-
rahydropyrido benzimidazole Analogues. J . Med. Chem. 1994,
37, 78-92.
(5) Boruah, R. C.; Skibo, E. B. A Comparison of the Cytotoxic and
Physical Properties of Aziridinyl Quinone Derivatives Based on
the Pyrrolo [1,2-a]benzimidazole and Pyrrolo[1,2-a]indole Ring
Systems. J . Med. Chem. 1994, 37, 1625-1631.
(6) Schulz, W. G.; Islam, E.; Skibo, E. B. Pyrrolo[1,2-a]benzimid-
azole-Based Quinones and Iminoquinones. The Role of the 3-Sub-
stituent on Cytotoxicity. J . Med. Chem. 1995, 38, 109-118.
(7) Skibo, E. B.; Islam, I. Synthesis and Elucidation of Azamitosene
and Iminoazamitosene. U.S. Patent 5,015,742, 1991.
(8) Skibo, E. B.; Islam, I.; Alberts, D. S. Antineoplastic Compounds
and Methods of Using the Same. U.S. Patent 5,246,955, 1993.
(9) Meth-Cohn, O.; Suschitzky, H. Heterocycles by Ring Closure of
Ortho-Substituted t-Anilines (The t-Amino Effect). Adv. Hetero-
cycl. Chem. 1972, 14, 211-278.
(10) Meth-Cohn, O.; Suschitzky, H. Syntheses of Heterocyclic Com-
pounds. Part IV. Oxidative Cyclization of Aromatic Amines and
Their N-Acyl Derivatives. J . Chem. Soc. 1963, 4666-4669.
(24) (a) Holden, J . A.; Rolfson, D. H.; Wittwer, C. T. Human DNA
Topoisomerase II: Evaluation of Enzyme Activity in Normal
and Neoplastic Tissue. Biochemistry 1990, 29, 2127-2134. (b)
McLeod, H. L.; Douglas, F.; Oates, M.; Symonds, R. P.; Prakash,
D.; van der Zee, A. G. J .; Kaye, S. B.; Brown, R.; Keith, W. N.
Topoisomerase I and II Activity in Human Breast, Cervix, Lung
and Colon Cancer. Int. J . Cancer 1994, 59, 607-611.
(25) Corbett, A. H.; Guerry, P.; Pflieger, P.; Osheroff, N. A pyrimido-
[1,6-a]benzimidazole That Enhances DNA Cleavage Mediated
by Eukaryotic Topoisomerase II: a Novel Class of Topoisomerase
IIsTargeted Drugs with Cytotoxic Potential. Antimicrob. Agents
Chemother. 1993, 37, 2599-2605.
(26) Radisky, D. C.; Radisky, E. S.; Barrows, L. R.; Copp, B. R.;
Kramer, R. A.; Ireland, C. M. Novel Cytotoxic Topoisomerase II
Inhibiting Pyrroloiminoquinones from Fijian Sponges of the
Genus Zyzzya. J . Am. Chem. Soc. 1993, 115, 1632-1638.
(27) Fujii, N.; Yamashita, Y.; Arima, Y.; Nagashima, M.; Nakano,
H. Induction of Topoisomerase II-Mediated DNA Cleavage by
the Plant Naphthoquinones Plumbagin and Shikonin. Antimi-
crob. Agents Chemother. 1992, 36, 2589-2594.
(28) Permana, P. A.; Snapka, R. M.; Shen, L. L.; Chu, D. T. W.;
Clement, J . J .; Plattner, J . J . Quinobenoxazines: A Class of
Novel Antitumor Quinolones and Potent Mammalian DNA
Topoisomerase II Catalytic Inhibitors. Biochemistry 1994, 33,
11333-11339.
(29) (a) Hartley, J . A.; Berardini, M.; Ponti, M.; Gibson, N. W.;
Thompson, A. S.; Thurston, D. E.; Hoey, B. M.; Butler, J . DNA
Cross-Linking and Sequence Selectivity of Aziridinyl-
benzoquinones: A Unique Reaction at 5'-GC-3' Sequences with
2,5-Diaziridinyl-1,4-benzoquinone upon Reduction. Biochemistry
1991, 30, 11719-11724. (b) Lee, C.-S.; Hartley, J . A.; Berardini,
M. D.; Butler, J .; Siegel, D.; Ross, D.; Gibson, N. W. Alteration
in DNA Cross-Linking and Sequence Selectivity of a Series of
Aziridinylbenzoquinones and Enzymatic Reduction by DT-
Diaphorase. Biochemistry 1992, 31, 3019-3025.
J M960064D