A two-step synthesis of cytostatically active benzo[c]phenanthridine derivatives

Article abstract of DOI:10.1002/anie.200461969

Lots, fast: Benzo[c]phenanthridine derivatives are prepared efficiently by the reaction of aldehydes and 2-methylbenzonitrile in the presence of potassium tert-butanolate, followed by dehydration (see scheme). Some of these compounds demonstrate excellent

Full text of DOI:10.1002/anie.200461969

Angewandte
Chemie
Synthetic Methods
slight variation on the synthetic sequence reported by
Gillespie et al.[8]
.
Messmer et al.[7] demonstrated the pronounced in vivo
antileukemia activity of 1, which was then verified by Stermitz
et al.[11] and in preclinical studies at the National Cancer
Institute[12]. The pharmacological and toxicological effects of
1 and other alkaloids in this group have since been tested in
many investigations. Sethi[13] and Kakiuchi et al.[14] have thus
reported the inhibition of reverse transcriptases, which, as
enzymes specific to tumor viruses, are responsible for the
transcription of viral RNA into the complementary DNA.
This DNAversion of the viral genome is incorporated into the
DNA of the host and is replicated in the cell-division process.
Compound 1 additionally demonstrates DNA-polymerase-
inhibiting properties, presumably a result of interaction with
adenosine/thymine base pairs.[15]
A Two-Step Synthesis of Cytostatically Active
Benzo[c]phenanthridine Derivatives**
Bernd Clement,* Matthias Weide, Ulrich Wolschendorf,
and Ilka Kock
The benzo[c]phenanthridine class of substances display a
variety of pharmacological properties. A large number of
naturally occurring alkaloids that contain a benzo[c]phenan-
thridine ring system and have a known spectrum of activity
are mentioned in the literature.[1,2] Aside from fagaronine (1),
the most important representative of this group of natural
products, other alkaloids with extensive pharmacological
potential are nitidine (2), chelerythrine (3), and sanguinarine
(4; Scheme 1).[3–5] The synthesis of these natural products is of
In 1983 Pezzuto et al.[16] reported that the primary
mechanism of the interaction of fagaronine with deoxyribo-
nucleic acid is DNA intercalation. It was additionally
demonstrated that binding of 1 is not—as described by
Sethi[15]—limited to adenine- and thymine-containing poly-
nucleotides. Furthermore, 1 has the ability to act as an
inhibitor of topoisomeraseI and II. In addition to restraining
topomerase activity, this leads to a stabilization of the DNA–
enzyme complex.[17]
Compound 2 exhibits a spectrum of activity similar to that
of 1.[18] As a result of the acute toxicity demonstrated by 2 in
preclinical trials, its development as a pharmaceutical was not
pursued.[19] Because the structures of 1 and 2 differ only by
one methylenedioxy group, this is thought to be linked to the
toxicity of 2.[20]
Because of the great interest in the biological activity of
these natural substances, the synthesis of benzo[c]phenan-
thridine derivatives and alkaloids in this family is an
important area of heterocyclic chemistry. The toxicological
problems that accompany some of the naturally occurring
alkaloids have led to the development of new synthetic
Scheme 1. Structural formulas of known benzo[c]phenanthridine alka-
loids.
great interest, because they can be isolated from plant
materials only in very small amounts. Cushman et al.[6]
describe yields in the 0.003 to 0.07% range for the isolation
of 2 from a series of zanthoxylum and fagara varieties.
Compound 1 was first isolated from the root of Fagara
zanthoxyloides in 1972.[7]
pathways for benzo[c]phenanthridine derivatives.[2,21,22]
A
variety of benzo[c]phenanthridine derivatives have thus
become synthetically accessible.[23–26] However, these synthe-
ses are very unwieldy and require many steps. In particular,
compounds with substituents directly flanking the endocyclic
nitrogen atom have thus far been very inconvenient to
produce.[27,28]
By raising the reaction temperature when we attempted
the synthesis of E-2-[2-(4-dimethylaminophenyl)vinyl]benzo-
nitrile (8; Scheme 3) from 4-dimethylaminobenzaldehyde (5)
and 2-methylbenzonitrile (6) as given in ref. [29], we surpris-
ingly obtained 7a as the main product (Scheme 2). Doubling
the concentration of 6 used in the reaction increased the yield
of 7a, which is formed as a racemic mixture.
No synthesis of benzo[c]phenanthridines with phenyl
substituents at C11 have thus far been described. In 1968
Devanathan et al.[30] reported attempting the synthesis of
4b,10b,11,12-tetrahydro-11-phenylbenzo[c]phenanthridine.
The synthesis failed in the cyclization of the acetamido or
benzamido derivative of 2,3-diphenyl-1,2,3,4-tetrahydro-
benzo-1-naphthylamine mediated by phosphorus oxychloride
or polyphosphoric acid to form the corresponding benzo[c]-
The first total synthesis of 1 was described by Gillespie
et al. in 1974.[8] This synthetic route gave 1 in 0.7% yield
starting from 2,3-dimethoxy-5-nitronaphthalene. Other syn-
thetic routes to 1 have been described by Ishii et al.,[9] Treus
[5]
et al.,[4] Lunch et al., and Smidrkal[10], the latter being only a
ˇ
[*] Prof. Dr. B. Clement, Dr. M. Weide, Dr. U. Wolschendorf, Dr. I. Kock
Pharmazeutisches Institut
Lehrstuhl fꢀr Pharmazeutische Chemie
Universitꢁt Kiel
Gutenbergstrasse 76, 24118 Kiel (Germany)
Fax: (+49)431-880-1352
E-mail: bclement@pharmazie.uni-kiel.de
[**] We thank the Fonds der chemischen Industrie for financial support.
Supporting information for this article (analytical data of the
compounds synthesized) is available on the WWW under http://
www.angewandte.de or from the author.
Angew. Chem. Int. Ed. 2005, 44, 635 –638
DOI: 10.1002/anie.200461969
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
635

Communications
the nucleophilic addition. Protonation could lead to 9;
however, in the presence of potassium tert-butanolate the
carbanion 10 is very likely to be present. The intramolecular
nucleophilic addition of this carbanion to the nitrile group in
the manner of a Thorpe–Ziegler reaction then leads to the
tautomeric phenylogous enaminonitrile 12 by way of the
iminonitrile 11. According to investigations carried out by
Baldwin,[31] the enaminonitrile 12 should be favored over
iminonitrile 11. Nitrile functionalities in enaminonitriles are
activated toward nucleophilic attack by the formation of
mesomeric structures.[31,32] In this way, and also supported by
the steric arrangement, the amino group of 12 adds intra-
molecularly to the nitrile group, resulting in 13. Subsequent
tautomerization finally leads to 7a.
Scheme 2. Synthesis of 7a.
phenanthridine. The synthesis of 7a from 5 and 6 reveals a
new and surprisingly simple synthetic route for the production
of benzo[c]phenanthridine derivatives. The direct reaction of
8 with 6 under the conditions described above also leads to 7a.
We can thus assume that the formation of 7a proceeds by
formation of intermediate 8 (Scheme 3).
The route used for the formation of 7a was also used to
obtain other benzo[c]phenanthridine derivatives (Table 1).
We view the following to be a probable reaction pathway.
In analogy to a vinylogous Michael reaction, a further
equivalent of 6 adds to the double bond in 8. The negative
mesomeric and inductive effect of the nitrile group facilitates
Table 1: Synthesis of 6-amino-11,12-dihydrobenzo[c]phenanthridines.
R
Product
Yield [%]
4-Me2NC6H4
Ph
4-MeOC6H4
2,4-(MeO)2C6H3
3,4-(MeO)2C6H3
H
7a
7b
7c
7d
7e
7 f
40
37
53
37
57
11
As demonstrated by the successful use of benzaldehyde, an
electron-donating group is not necessary for sufficient polar-
ization of the double bond. The phenyl derivative 7b was
obtained in similar yields to 7a. Replacement of the
dimethylamino group by a methoxy group led to higher
product yields (7c). The dimethoxy derivative 7d was
obtained in lower yields, probably because of steric hindrance.
In contrast, the dimethoxy derivative 7e was obtained in
yields similar to those of 7c. The use of formaldehyde as the
aldehyde component resulted in a lower yield (7 f); the
reasons for this are unclear.
Dehydration with 2,3-dichloro-5,6-dicyano-p-benzoqui-
none (DDQ)[33] converted the dihydro compounds into the
corresponding 6-aminobenzo[c]phenanthridines in yields
ranging from 14 to 38% (Scheme 4). These results contradict
previous findings,[24] in which 11,12-dihydrobenzo[c]phenan-
thridine and a series of substituted 11,12-dihydrobenzo[c]-
phenanthridines could be dehydrated only by heating in the
presence of 30% palladium over activated carbon. Heating
Scheme 3. Postulated mechanism for the formation of 7a.
Scheme 4. Dehydration of 7 to form 14.
www.angewandte.org Angew. Chem. Int. Ed. 2005, 44, 635 –638
636
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Angewandte
Chemie
with DDQ in various solvents or heating with 10% palladium
over activated carbon did not lead to the desired products. In
preliminary experiments, the hydrochlorides of 7 could also
be dehydrated by heating with sulfur.[34] However, this
method demonstrated no advantages over dehydration with
DDQ, as the yields were lower. As described above,[24]
attempts at dehydration with 10% palladium over activated
carbon were not successful.
Since many natural products have substituents in posi-
tions 2 and 3, or 8 and 9, we were interested determining
whether the method described above could be used to
prepare derivatives having substituents in these positions
directly. For this, substituted 2-methylbenzonitriles are nec-
essary. Whereas electron-withdrawing groups should not be
problematic, electron-donating groups could have an adverse
effect on both the deprotonation of the methyl groups in the
ortho position and on the nucleophilic addition. The example
of easily accessible 4-methoxy-2-methylbenzonitrile[35] and its
conversion with 3,4,5-trimethoxybenzaldehyde to 7g demon-
strates the broad applicability of this synthetic route
(Scheme 5). At 71%, the yield was even surprisingly high.
The 24% overall yield obtained after dehydration of 14g was
within the same range as the other conversions.
activity against 60 human tumor cell lines from nine different
types of cancer.[36] For each compound tested, 60 dose–activity
curves, corresponding to the number of tumor cell lines, with
five test concentrations each are obtained. This makes it
possible to obtain information about the cytotoxicity as well
as the selectivity of each compound against one or more forms
of cancer. In order to characterize the antitumor activity of a
tested compound, the dose-dependent parameters GI50
(growth inhibition 50%), TGI (total growth inhibition), and
LC50 (lethal concentration 50%) are calculated. For each of
these three parameters, the meangraph midpoint (MG_MID)
is also calculated; this is the mean of the logarithms of the
GI50 values and corresponds to an average response charac-
teristic of all 60 cell lines to the test substance. This factor
allows for the characterization of the activity of a compound
in the test system and for its quantitative comparison with
other compounds. Taking the antilogarithm of this value gives
the mean of the GI50 values over all cell lines.[37]
According to these studies, a 3.39 mm concentration of 7g
and a 0.18 mm concentration of 7e halve the growth rate of the
tumor cells. Both compounds thus demonstrate a higher
activity than fagaronine (1; 9.48 mm). In addition to these data,
Table 2 contains data for other cytostatically active com-
Table 2: Cytotoxic activity of benzo[c]phenanthridine derivatives and
selected cytostatics from in vitro cell line screening by the NCI.[12]
P
GI50
Name
MG_MID (GI50)
[mm]
60
7g
7e
À5.47
À6.74
À5.00
À3.68
À4.75
À5.13
À7.65
3.39
0.18
9.48
210
17.6
7.36
0.02
fagaronine
cyclophosphamide
5-fluorouracil
6-sulfanylpurine
paclitaxel
pounds. With the exception of paclitaxel, 7e and 7g demon-
strate better cytostatic activity than established cystostatics
such as cyclophosphamide in the in vitro cell line testing of the
NCI. Further benzo[c]phenanthridine derivatives (not pre-
sented here) demonstrated similar cytostatic activities. These
results will thus initiate the further development of new
potential cancer drugs.
Scheme 5. Synthesis of 7g and 14g.
The products obtained can be converted into other
benzo[c]phenanthridine derivatives, especially by substitution
in the 6-position. In particular, removal of the amino group
from 14 f should make the basic ring system of this class of
compounds easily accessible. The use of 4,5-dimethoxy-2-
methylbenzonitrile and paraformaldehyde should allow for
the straightforward formation of a fagaronine precursor.
The yields of 7 and 14 are not outstanding; however, the
substances can be prepared in large amounts and with
analytical purity (for pharmacological studies, for example),
without chromatographic purification. Initial investigations of
compounds of type 14 and to some extent type 7 demon-
strated a very high cytotoxic potential. Studies of 7g and 14g
by the U.S. National Cancer Institute (NCI) serve to illustrate
this. NCI uses in vitro cell line screening to test compounds for
Experimental Section
7a–d: General synthetic protocol: A solution of the appropriate
aldehyde (10 mmol) and 2-methylbenzonitrile (2.34 g, 20 mmol) in
5 mL 1,3-dimethyltetrahydro-2-pyrimidone (DMPU) was added
dropwise under nitrogen to a solution of potassium tert-butanolate
(2.47 g, 22 mmol) in 20 mL DMPU. The reaction mixture was stirred
under nitrogen for 5 h at 358C, then poured into a solution of
ammonium chloride (2.2 g, 40 mmol) in 100 mL ice water, and
extracted three times with 100 mL dichloromethane. The organic
phase was filtered through cotton if necessary and concentrated to a
volume of about 100 mL on a rotary evaporator. The organic phase
was vigorously shaken with 3n hydrochloric acid; the resulting
precipitate was filtered, washed with dichloromethane, and dried. The
corresponding 11-substituted 6-amino-11,12-dihydro-benzo[c]phe-
Angew. Chem. Int. Ed. 2005, 44, 635 –638
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
637

Communications
nanthridinium chlorides were obtained after recrystallization. The
[5] M. A. Lunch, O. Duval, P. Pochet, R. D. Waigh, Bull. Soc. Chim.
Fr. 1994, 131, 718 – 722.
[6] M. Cushman, L. Cheng, J. Org. Chem. 1978, 43, 286 – 288.
[7] W. M. Messmer, M. Tin-Wa, H. H. S. Fong, C. Bevelle, N. R.
Farnsworth, D. J. Abraham, J. Trojanek, J. Pharm. Sci. 1972, 61,
1858 – 1859.
free amines were released in diethyl ether with dilute ammonia.
Addition of petroleum ether and removal of the diethyl ether on a
rotary evaporator provided the solid products.
7e: As described above, with the following changes: 9.86 g
(88 mmol) potassium tert-butanolate in 90 mL DMPU; 6.65 g
(40 mmol) 3,4-dimethoxybenzaldehyde, 9.36 g (80 mmol) 2-methyl-
benzonitrile in 40 mL DMPU; dropwise addition at 408C; stirring for
4 h at 35–408C; 8.8 g (80 mmol) ammonium chloride in 400 mL ice
water; extracted three time with 150 mL dichlormethane; vigorously
stirred overnight with 20 mL 5n hydrochloric acid.
[8] J. P. Gillespie, L. G. Amoros, F. R. Stermiz, J. Org. Chem. 1974,
39, 3239 – 3241.
[9] H. Ishii, I. S. Cheng, T. Ishikawa, J. Chem. Soc. Perkin Trans. 1
1987, 671 – 676.
ˇ
[10] J. Smidrkal, Collect. Czech. Chem. Commun. 1988, 53, 1384 –
7 f: As described above, with the following changes: 2.47 g
(22 mmol) potassium tert-butanolate in 20 mL DMPU; 0.3 g
(10 mmol) paraformaldehyde; 2.34 g (20 mmol) 2-methylbenzonitrile
in 12 mL DMPU; dropwise addition in 2 mL portions every
15 minutes; stirring for 6 h at 358C. The organic phase was
concentrated on a rotary evaporator until vigorous precipitation
occurred and was then stored overnight in a refrigerator; recrystal-
lization from methanol/dichloromethane.
1392.
[11] F. R. Stermiz, J. P. Gillespie, L. G. Amoros, R. Romero, T. A.
Stermiz, J. Med. Chem. 1975, 18, 708 – 713.
[12] National Cancer Institute, Cancer Screen 10/2002 Data; http://
dtp.nci.nih.gov.
[13] M. L. Sethi, J. Nat. Prod. 1979, 42, 187 – 196.
[14] N. Kakiuchi, M. Hattori, H. Ischi, T. Namba, Planta Med. 1987,
22 – 27.
7g:
A
solution of 4-methoxy-2-methylbenzonitrile (8.83 g,
[15] V. S. Sethi, Cancer Res. 1976, 36, 2390 – 2395.
[16] J. N. Pezzuto, S. K. Antosiak, W. M. Messmer, M. B. Salytor,
G. R. Honig, Chem.-Biol. Interact. 1983, 43, 323 – 339.
[17] A. K. Larsen, L. Gronard, J. Couprie, B. Desoize, L. Comoe, J.-
C. Jardillier, J.-F. Riou, Biochem. Pharmacol. 1993, 46, 1403 –
1412.
[18] S. D. Phillips, R. N. Castle, J. Heterocycl. Chem. 1981, 18, 223 –
232.
[19] Y. L. Janin, E. Bisagni, Tetrahedron 1993, 49, 10305 – 10316.
[20] H. Ishii, J.-I. Ichikawa, E. Kawabane, M. Ishikawa, T. Ishikawa,
K. Kuretani, M. Inomata, A. Hoshi, Chem. Pharm. Bull. 1995,
33, 4139 – 4151.
60 mmol) and 3,4,5-trimethoxybenzaldehyde (5.89 g, 30 mmol) in
30 mL DMPU was added dropwise under nitrogen to a solution of
potassium tert-butanolate (7.41 g, 66 mmol) in 70 mL DMPU at 308C.
The reaction mixture was subsequently stirred for 4 h at 35–408C and
then carefully hydrolyzed in a solution of ammonium chloride (6.54 g,
120 mmol) in 300 mL ice water. The aqueous phase was extracted
three times with 150 mL dichloromethane, and the combined organic
phases were then dried over sodium sulfate and concentrated on a
rotary evaporator. Precipitation of 7g occurred after the organic
phase had been stirred vigorously overnight with 10 mL concentrated
hydrochloric acid and 10 mL water. The precipitate was filtered out,
washed with a small amount of dichloromethane, recrystallized from
methanol, and dried for 24 h under vacuum (oil pump).
[21] I. Ninomiya, T. Naito, Rec. Dev. Chem. Nat. Carbon Compd.
1984, 10, 11 – 90.
14b–g: General synthetic protocol: A solution of DDQ in
dioxane was added to a solution of 7a–g in dioxane and heated to
reflux. The cooled reaction mixture was hydrolyzed by pouring it into
a saturated sodium hydrogen carbonate solution, and the reaction
mixture was then extracted with diethyl ether. The ether phase was
washed once with dilute sodium hydrogen carbonate solution and
three times with water. The organic phase was dried over sodium
sulfate and concentrated on a rotary evaporator. The 6-amino-
benzo[c]phenanthridiniumperchlorate was precipitated by vigorous
stirring of the organic phase overnight with 70% perchloric acid. The
precipitate was filtered out and recrystallized from methanol. 14b:
0.5 g (1.6 mmol) 7b in 25 mL dioxane, 0.63 (2.8 mmol) DDQ in 25 mL
dioxane, 4 h reflux; 14c: 1.5 g (4.3 mmol) 7c in 100 mL dioxane, 3.8 g
(16.7 mmol) DDQ in 100 mL dioxane, 8 h reflux; 14d: 0.5 g
(1.3 mmol) 7d in 10 mL dioxane, 0.54 g (2.3 mmol) DDQ in 35 mL
dioxane, 4 h reflux; 14e: 2.0 g (5.2 mmol) 7e in 50 mL dioxane, 4.4 g
(21.0 mmol) DDQ in 100 mL dioxane, 16 h reflux; 14 f: 0.25 g
(1.0 mmol) 7 f in 15 mL dioxane, 0.40 (1.7 mmol) DDQ in 35 mL
dioxane, 4 h reflux; 14g: 1.62 g (3.4 mmol) 7g in 90 mL dioxane,
3.11 g (13.7 mmol) DDQ in 70 mL dioxane, 9 h reflux.
[22] S. P. Mackay, O. Meth-Cohn, R. D. Waigh, Adv. Heterocycl.
Chem. 1987, 25, 345 – 389.
[23] R. K.-Y. Zee-Cheng, C. C. Cheng, J. Med. Chem. 1975, 18, 66 –
71.
[24] R. Beugelmans, J. Chastanet, H. Ginsburg, L. Quintero-Cortez,
G. Roussi, J. Org. Chem. 1985, 50, 4933 – 4938.
[25] S. P. Mackay, L. Comoe, B. Desoize, O. Duvall, J.-C. Jardellier,
R. D. Waigh, Anticancer Drug Des. 1998, 13, 797 – 813.
[26] T. Ishikawa, Med. Res. Rev. 2001, 21, 61 – 72.
[27] S. V. Kessar, Y. P. Gupta, P. Balakrishnan, K. K. Sawal, T.
Mohammed, M. Dutt, J. Org. Chem. 1988, 53, 1708 – 1713.
[28] Y. L. Janin, A. Croisy, J.-F. Riu, E. Bisagni, J. Med. Chem. 1993,
36, 3686 – 3692.
[29] K. Takhaschi, T. Okamoto, K. Yamada, H. Iida, Synthesis 1977,
58 – 59.
[30] V. C. Devanathan, V. Kesavan, N. Arumugam, Indian J. Chem.
1968, 7, 124 – 126.
[31] S. Baldwin, J. Org. Chem. 1961, 26, 3288 – 3295.
[32] I. F. Barnard, J. A. Elvidge, J. Chem. Soc. Perkin Trans. 1 1983,
1137 – 1140.
[33] D. Walker, J. D. Hiebert, J. Org. Chem. 1967, 32, 153 – 196.
[34] P. A. Plattner, Angew. Chem. 1942, 55, 131 – 146; P. A. Plattner,
Angew. Chem. 1942, 55, 154 – 158.
Received: September 13, 2004
Published online: December 13, 2004
[35] J. L. Neumeyer, K. K. Weinhardt, J. Med. Chem. 1970, 13, 613 –
616.
[36] A. Monks, D. Scudiero, P. Skehan, R. Shoemaker, K. Paull, D.
Vistica, C. Hose, J. Langley, P. Cronise, A. Vaigro-Wolff, M.
Gray-Goodric, H. Campell, J. Mayo, M. Boyd, J. Natl. Cancer
Inst. 1991, 83, 757 – 766.
Keywords: alkaloids · antitumor agents ·
benzophenanthridines · synthetic methods
.
[1] S. Simeon, J. L. Rios, A. Villar, Pharmazie 1989, 44, 593 – 597.
[2] T. Ishikawa, H. Ishii, Heterocycles 1999, 50, 627 – 639.
[3] S. D. Phillips, R. N. Castle, J. Heterocycl. Chem. 1981, 18, 223 –
232.
[37] M. R. Boyd, K. D. Paull, Drug Dev. Res. 1995, 34, 91 – 109.
[4] M. Treus, J. C. Estevez, L. Castedo, R. J. Esteves, Tetrahedron
Lett. 2002, 43, 5323 – 5325.
638
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 635 –638