An intramolecular cycloaddition approach to pyrrolo[3,2-c]quinolones

Article abstract of DOI:10.1016/S0040-4039(01)02395-4

An approach to pyrrolo[3,2-c]quinolones is described which relies on an intramolecular azomethine ylide alkene cycloaddition reaction. Critical to the success of this reaction was the requirement for alkylation of the aniline nitrogen, in the absence of t

Full text of DOI:10.1016/S0040-4039(01)02395-4

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 1171–1174
An intramolecular cycloaddition approach to
pyrrolo[3,2-c]quinolones
Yong He, Hossen Mahmud, Brian R. Wayland, H. V. Rasika Dias† and Carl J. Lovely*
Department of Chemistry and Biochemistry, The University of Texas at Arlington, Arlington, TX 76019, USA
Received 15 October 2001; revised 17 December 2001; accepted 20 December 2001
Abstract—An approach to pyrrolo[3,2-c]quinolones is described which relies on an intramolecular azomethine ylide alkene
cycloaddition reaction. Critical to the success of this reaction was the requirement for alkylation of the aniline nitrogen, in the
absence of this substituent a decarboxylative-Michael addition pathway was observed. © 2002 Elsevier Science Ltd. All rights
reserved.
As part of our program focused on the synthesis of
bioactive polyguanidine natural products, we have been
NH
engaged in developing approaches to the total synthesis
HN
of the Martinella alkaloids 1 and 2.1 These relatively
N
simple natural products have attracted considerable
attention leading to the development of a number of
approaches to the heterocyclic core2 and most recently
to two total syntheses.3 In our approach to these
targets, we identified pyrroloquinolones, such as 4, as
ideal precursors en route to the natural products and
diverse analogs. It was envisioned that compounds of
this type might be accessed through a [3+2] azomethine
ylide–alkene cycloaddition (3“4). In this letter we
report on the investigation of this approach, the obser-
vation of an unexpected sequence of events, and the
development of a solution to provide the required
heterocycle.
RO2C
H
N
H
N
N
H
NH
H
N
NH2
NH
1: R (Martinelline) =
2: R (Martinellic Acid) = H
Ph
Ph
N
N
Br
Br
The requisite cyclization precursor 7 was prepared from
methyl 5-bromo-2-aminobenzoate (5) via reduction and
chemoselective acylation of the 2-amino group. The
sparingly soluble amide was converted into the benz-
aldehyde derivative by oxidation with MnO2, providing
7 in a moderate 43% yield. When 7 was reacted with
N-benzylglycine·HCl in DMF at reflux in the presence
of Et3N, the expected condensation–elimination–
cycloaddition sequence did not occur. However, a new
compound was formed in 40% isolated yield. Analysis
of the 1H NMR spectrum of the purified product
indicated that the isolated material contained an N-
N
H
O
N
H
O
4
3
methyl group, an aromatic aldehyde and a -CH2CH2-
group. Based on these data, we concluded that the
structure of the isolated product was in fact 8a, result-
ing from the net decarboxylation and conjugate addi-
tion of N-methylbenzylamine to the acrylamide
moiety.4 When the reaction was repeated in boiling
benzene, the same product was isolated in 68% yield.
Somewhat intrigued by this result, compound 7 was
reacted with N-methyl and N-(p-methoxybenz-
yl)glycine, the latter as its HCl salt. In each case
analogous conjugate addition products 8b–c were
obtained in 30 and 50% yield, respectively. These exper-
* Corresponding author. Tel.: 817-272-5446; fax: 817-272-3808;
e-mail: lovely@uta.edu
† Address correspondence to this author regarding the X-ray struc-
tures; e-mail: dias@uta.edu
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)02395-4

1172
Y. He et al. / Tetrahedron Letters 43 (2002) 1171–1174
iments suggested that this pathway was independent of
the nature of the N-substituent (Scheme 1).
The isolation of these amines was completely unex-
pected as it was erroneously assumed that the electron
deficient dipolarophile would readily engage in an
intramolecular cycloaddition with the in situ generated
azomethine ylide. These observations were particularly
surprising when compared to the unactivated systems
reported previously by our group, which participate
efficiently in the cycloaddition reaction.2d,l A series of
control experiments were conducted in order to estab-
lish the sequence of events leading to these observa-
tions. The precursor alcohol 9 and the acrylated ester
10 were subjected to the cycloaddition conditions and it
was found that no reaction occurred. On the other
hand, both substrates participated in Michael reactions
with N-methylbenzylamine providing 11 (70%) and 12
(75%), respectively, suggesting that the decarboxylation
occurred prior to conjugate addition and that the alde-
hyde moiety was critical. The benzaldehyde derivative 7
participated in Michael reactions with both N-methyl-
benzylamine and dimethylamine. Taken in combina-
tion, these results suggested that the azomethine ylide
was indeed forming, but undergoing protonation and
hydrolysis prior to cycloaddition. The hydrolysis
product, an N-methylalkylamine, then adds in a
Michael fashion to 7. Support for this thesis was
obtained by running the reaction in the presence of
N-phenylmaleimide (10 equiv.) when the anticipated
intermolecular cycloadduct 13 was obtained in 19%
yield as a 3:2 mixture of stereoisomers (Scheme 2).
Scheme 2. Reagents and conditions: (a) BnNHCH2CO2H·
HCl, Et3N, PhH, reflux. (b) BnNHCH3, Et3N, PhH, reflux.
Inititially, attempts were made to alkylate 7 directly
with MeI, however, these experiments were unsuccess-
ful, therefore N-methyl isatoic anhydride was employed
as starting material. Reduction with LiAlH4 provided
the known amino alcohol,7 which was then acylated
with acroyl chloride according to the protocol of
Heaney and co-workers, providing 16 in good yield
(Scheme 3).7 MnO2 oxidation then provided the
required acrylamide in reasonable overall yield.8 When
this substrate was subjected to the cycloaddition reac-
It has been demonstrated previously that intramolecu-
lar reactions involving unsubstituted amides can be
retarded as a result of low populations of the reactive
conformation.5 However, this problem can often be
circumvented by placing additional substituents on
nitrogen, therefore this approach was explored as a
means to access compound 4.6
O
OH
b
a
O
N
NH
O
CH3
14
CH3
15
OH
O
N
O
c
O
O
Br
a
Br
X
b, c
N
NH2
N
H
O
CH3
CH3
16
17
5: X = CO2Me
6: X = CH2OH
7
R
N
6
a: R = Bn, 66%
b: R = CH3, 58%
c: R = PMB, 44%
3
O
d
d
1
Br
O
N
O
a: R = Bn, 68%
b: R = CH3, 30%
c: R = PMB, 50%
CH3
N
H
NCH3
R
18a-c
8a-c
Scheme 3. Reagents and conditions: (a) LiAlH4, THF, 0°C“
rt, 90%. (b) H2CꢀCHCOCl, CH2Cl2, NaHCO3, 0°C, 96%. (c)
MnO2, CH2Cl2, 88%. (d) RNHCH2CO2H·HCl (R=Bn,
PMB) or CH3NHCH2CO2H, Et3N, PhCH3 (R=Bn, CH3) or
DMF (R=PMB), reflux.
Scheme 1. Reagents and conditions: (a) LiAlH4, THF, 0°C,
93%. (b) H2CꢀCHCOCl, Et2O, 43%. (c) MnO2, acetone, 45%.
(d) RNHCH2CO2H·HCl (R=Bn, PMB) or CH3NHCH2-
CO2H, Et3N, PhH, reflux.

Y. He et al. / Tetrahedron Letters 43 (2002) 1171–1174
1173
was confirmed independently through X-ray crystallog-
raphy. The bromine substituent was converted via a
Pd-catalyzed carbonylation into a carboxymethyl moi-
ety,13 followed by a chemoselective removal of the
pyrrole benzyl group to provide the polar product in
72% yield for two steps (Scheme 4).
tion with N-benzyl glycine·HCl in refluxing toluene, we
were delighted to find that it had undergone the desired
cycloaddition to provide the cis pyrroloquinolone 18a
in 66% yield. In addition to the cis fused pyrrolo-
quinolone (J2a,5a=5.5 Hz), a small quantity of the trans
product (J2a,5a=13.8 Hz), ca. 5% was isolated.9 Simi-
larly,
with
sarcosine
and
N-(p-methoxyben-
In summary, a short, stereocontrolled intramolecular
cycloaddition approach to pyrrolo[3,2-c]quinol-2-ones
has been investigated, providing adducts that can be
readily functionalized and that should be suitable for
elaboration into the Martinella alkaloids and/or con-
geners. Critical to the success of this cycloaddition is
the need for N-substitution on the acrylamide
precursors.
zyl)glycine·HCl, cis pyrroloquinolones 18b and c were
obtained in 58 and 44% yields, respectively.10
In order to establish whether this approach might be
viable en route to the Martinella alkaloids, the N-ben-
zyl acrylamide derivative was constructed through a
largely similar sequence of reactions to those previously
employed. Thus, benzoylation of anthranilate deriva-
tive 5 gave 19, which was reduced efficiently to provide
the N-benzyl alcohol 20.11 Acroylation and oxidation
afforded the cyclization substrate 21. When 21 was
subjected to the cycloaddition reaction in toluene at
reflux, the desired cis pyrroloquinolone 22 was obtained
in 51% yield. The H5a benzylic proton appeared as a
doublet in the 1H NMR spectrum, with a coupling
constant of 5.0 Hz, which suggested that the ring fusion
was cis. This assignment was subsequently confirmed
through an X-ray structure determination on this
adduct, which unequivocally demonstrated that the
pyrrole–quinoline ring fusion was cis. In addition to the
major cis adduct, a small quantity (7%) of the trans
isomer was isolated (J2a,5a=13.8 Hz).12 This assignment
Acknowledgements
This work was supported by the Robert A. Welch
Foundation and The University of Texas at Arlington
(Start-up Funds). The NSF (CHE-9601771) is thanked
for partial funding of the purchase of a 500 MHz NMR
spectrometer.
References
1. Witherup, K. M.; Ransom, R. M.; Graham, A. C.;
Bernard, A. M.; Salvatore, M. J.; Lumma, W. C.; Ander-
son, P. S.; Pitzenberger, S. M.; Varga, S. L. J. Am. Chem.
Soc. 1995, 117, 6682.
2. (a) Gurjar, M. K.; Pal, S.; Rao, A. V. R. Heterocycles
1997, 45, 231; (b) Ho, T. C. T.; Jones, K. Tetrahedron
1997, 53, 8287; (c) Hadden, M.; Stevenson, P. J. Tetra-
hedron Lett. 1999, 40, 1215; (d) Lovely, C. J.; Mahmud,
H. Tetrahedron Lett. 1999, 40, 2079; (e) Snider, B. B.;
Ahn, Y.; Foxman, B. M. Tetrahedron Lett. 1999, 40,
3339; (f) Batey, R. A.; Simoncic, P. D.; Lin, D.; Smyj, R.
P.; Lough, A. J. Chem. Commun. 1999, 651; (g) Frank, K.
E.; Aube, J. J. Org. Chem. 2000, 65, 655; (h) Nieman, J.
A.; Ennis, M. D. Org. Lett. 2000, 2, 1395; (i) Nyerges,
M.; Fejes, I.; Toke, L. Tetrahedron Lett. 2000, 41, 7951;
(j) Escolano, C.; Jones, K. Tetrahedron Lett. 2000, 41,
8951; (k) Snider, B. B.; O’Hare, S. M. Tetrahedron Lett.
2001, 42, 2455; (l) Mahmud, H.; Lovely, C. J.; Dias, H.
V. R. Tetrahedron 2001, 57, 4095; (m) Hadden, M.;
Nieuwenhuyzen, M.; Potts, D.; Stevenson, P. J.; Thomp-
son, N. Tetrahedron 2001, 57, 5615; (n) Hadden, M.;
Nieuwenhuyzen, M.; Osborne, D.; Stevenson, P. J.;
Thompson, N. Tetrahedron Lett. 2001, 42, 6417; (o)
Batey, R. A.; Powell, D. A. Chem. Commun. 2001, 2362–
2363.
3. (a) Ma, D.; Xia, C.; Jiang, J.; Zhang, J. Org. Lett. 2001,
3, 2189; (b) Snider, B. B.; Ahn, Y.; O’Hare, S. M. Org.
Lett. ASAP.
Scheme 4. Reagents and conditions: (a) BzCl, NaHCO3,
CH2Cl2, 95%. (b) LiAlH4, THF, 0°C, 93%. (c)
H2CꢀCHCOCl, CH2Cl2, NaHCO3, 89%. (d) MnO2, CH2Cl2,
85%. (e) BnNHCH2CO2H·HCl, Et3N, PhCH3, reflux, 51%
(cis). (f) Pd(OAc)2, PPh3, 75 psi CO, MeOH, NaOAc, DMF,
110°C, 76%. (g) Pd(OH)2, H2, MeOH, HCl, 95%.
4. A similar observation has been made by Tsuge and
co-workers, however, in their studies the cycloadditions
were intermolecular. Tsuge, O.; Kanemasa, S.; Ohe, M.;
Takenaka, S. Bull. Chem. Soc. Jpn. 1987, 60, 4079.
5. Iwamatsu, S.-I.; Matsubara, K.; Nagashima, H. J. Org.
Chem. 1999, 64, 9625.

1174
Y. He et al. / Tetrahedron Letters 43 (2002) 1171–1174
6.6–6.8 Hz) for related pyrroloquinolones (see Ref. 2a).
6. Jones, K.; McCarthy, C. Tetrahedron Lett. 1989, 30,
10. The corresponding trans isomers were isolated in 4%
yield.
11. Bhalay, G.; Blaney, P.; Palmer, V. H.; Baxter, A. D.
Tetrahedron Lett. 1997, 38, 8375.
12. We believe that the formation of cis and trans isomers is
as kinetic phenomenon since subjection of either isomer
to the reaction conditions does not result in their inter-
conversion.
13. Unlike the carbonylations that we had performed in the
pyrroloquinoline series (see Ref. 2l), we found that the
yield was much lower. This was due to the formation of
22 via a base-induced retro Michael reaction.
2657.
7. Heaney, F.; Bourke, S.; Cunningham, D.; McArdle, P. J.
Chem. Soc., Perkin Trans. 2 1998, 547.
8. It was found that good yields for this oxidation were
obtained on small scales (0.3 mmol) but on larger scales
(8.2 mmol) the yield diminished to 55%.
9. We were not able to unequivocally establish the stereo-
chemistry through NOE experiments since the signal due
to H2a was obscured by another absorption. However,
the magnitude of the coupling constant is consistent with
that found for other cis pyrroloquinolines (J=4.0–8.3
Hz) and in particular with that reported by Gurjar (J=