Optimisation, scope and limitations of the synthesis of 5-aminoindolizines from oxazolo[3,2-a]pyridinium salts

Article abstract of DOI:10.1016/j.tetlet.2005.11.033

In this letter, we report on our efforts to optimise the reaction conditions and to explore the scope and limitations of the preparation of 5-aminoindolizines starting from oxazolo[3,2-a]pyridinium salts and different amines to evaluate the utility for co

Full text of DOI:10.1016/j.tetlet.2005.11.033

Tetrahedron
Letters
Tetrahedron Letters 47 (2006) 261–265
Optimisation, scope and limitations of the synthesis
of 5-aminoindolizines from oxazolo[3,2-a]pyridinium salts
P. Tielmann and C. Hoenke*
Abteilung Chemische Forschung, Boehringer Ingelheim GmbH & Co. KG, Birkendorfer Str. 65, 88397 Biberach (Riss), Germany
Received 18 October 2005; revised 3 November 2005; accepted 7 November 2005
Available online 23 November 2005
Abstract—In this letter, we report on our efforts to optimise the reaction conditions and to explore the scope and limitations of the
preparation of 5-aminoindolizines starting from oxazolo[3,2-a]pyridinium salts and different amines to evaluate the utility for com-
binatorial library construction. Thus, we were able to reduce the amount of amine to 1.1 equiv by applying microwave heating, ace-
tonitrile as solvent and DMAP as additional base. Further, we observed that only aliphatic secondary amines with limited steric
demand could deliver the desired compounds.
Ó 2005 Elsevier Ltd. All rights reserved.
1. Introduction
assisted cycloisomerisations of alkynylpyridines.13 The
last method has been developed only recently, while
the other procedures have already been known for dec-
ades. Apart from these four approaches to obtain indol-
izine system, methods have been developed to
functionalise existing indolizine nuclei and thereby
expanding the scope of available indolizine derivatives.14
Chemists in academia and pharmaceutical industry have
recently become attracted to indolizines due to their
interesting and promising biological properties.
Although aromatic indolizine systems have not been
discovered in nature until now,1 they have found wide-
spread application in biological and pharmaceutical
research. Thus, several reports have appeared showing
the potential of indolizine derivatives as histamine H3
receptor antagonists,2 antimycobacterial agents,3 cal-
cium entry blockers4 and inhibitors of 15-lipoxygenase.5
In addition, various studies have shown that indolizines
offer antioxidant properties,6 delayed replicative senes-
cence of human diploid fibroblasts7 and possess antiviral
and antileishmanial activity.8
Recently, a method has been reported by Babaev et al.
to create 5-aminoindolizine systems that does not fit into
the above mentioned classification. This approach
applies a 5-methyl substituted oxazolo[3,2-a]pyridinium
moiety as stable intermediate, which can be easily pre-
pared in two steps from 2-methoxy-6-picoline and
phenacylbromides. This cationic heterocyclic system
reacts with secondary amines to form the desired indol-
izines in a rearrangement-like fashion (Scheme 1).15
Over the years, several approaches have been developed
to synthesise indolizine systems, which can be sub-
divided into four main categories: (1) Condensation
reactions of 2-alkyl-pyridines with carboxylic acid
anhydrides or a-halo ketones known as Scholtz9 or
Tschitschibabin10 reaction, respectively. (2) Reactions
of a-unsubstituted pyridines with acyl- or aryl-substi-
tuted allyl halides or esters11 and methyl propiolate.13
(3) 1,3-Dipolar cycloadditions of pyridinium N-methyl-
ides with acetylenes11 and ethylenes.11d,12 (4) Copper-
Although we believe that this approach offers an unique
entry into a broad spectrum of indolizines, the reaction
has been exploited for only one oxazolo[3,2-a]pyridi-
nium salt and three different cyclic secondary amines,
which have been applied in large excess either neat
(piperidine, morpholine) or in acetonitrile (hexamethyl-
ene imine). This restriction might be due to the obser-
vation that only 4-nitrophenyl 5-aminoindolizine
derivatives could be isolated (Scheme 1, R1 = NO2).15a
Nevertheless, we saw a considerable potential of this
chemistry for generating diversity in the synthesis of
combinatorial libraries.16 For that reason, we decided
*
Corresponding author. Tel.: +49 7351 542028; fax: +49 7351
545181; e-mail: christoph.hoenke@bc.boehringer-ingelheim.com
0040-4039/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2005.11.033

262
P. Tielmann, C. Hoenke / Tetrahedron Letters 47 (2006) 261–265
O
O
O+
Br
+
R1
i), ii)
N
N
R1
R'2
2
1
-
3
ClO4
R2
N
iii)
N
R1
4
Scheme 1. Reagents, yield: (i) acetone, 48 h, reflux, 40–60%; (ii) concd sulfuric acid, then 70% perchloric acid, ꢀ70%; (iii) neat amine, reflux, 65–80%.
to optimise all synthetic procedures at first. In particular,
the improvement of the indolizine formation step was of
primary interest for us by optimising amine equivalents,
solvent, use of an additional base and microwave heat-
ing. Further, we were interested in broadening the scope
of applicable phenacylbromides and amines in particu-
lar, that is, by varying the steric and nucleophilic charac-
ter using different primary and secondary amines.
cylpyridin-2-ones 5 (Scheme 2) were obtained in
moderate yields. The products either precipitated from
the reaction mixture or had to be purified by flash
chromatography.
The next step, dehydration/cyclisation by applying con-
centrated sulfuric acid and subsequent precipitation
with perchloric acid by slightly modifying the published
procedure15,17 gave the desired methyl oxazolo[3,2-a]-
pyridinium perchlorate salts 3 in moderate to good
yields (Scheme 1, isolated yields (R1): 3a: 78% (NO2),
3b: 81% (Cl), 3c: 40% (H), 3d: 58% (Me), 3e: 83%
(OMe)). In case of pyridone 5e carrying a methoxy sub-
stituent, 64% sulfuric acid had to be applied for cyclisa-
tion to avoid additional sulfonation of the phenyl ring.18
2. Results and discussion
To examine the scope and limitations of this method for
synthesising 5-aminoindolizines and to optimise the syn-
thetic procedures, we first tried to improve the synthesis
of the intermediate oxazolo[3,2-a]pyridinium salts.
Starting with the reaction of phenacylbromides 2 with
methoxy-picoline 1, we investigated the possibility of
reducing the reaction time by microwave heating. A sol-
vent screening to identify the best solvent for this reac-
tion was performed using acetonitrile, ethanol, DMF,
DMSO, 1,4-dioxane, acetone, THF and toluene.
Although acetonitrile was identified as the best solvent,
the reaction was still slow; the mixture had in fact to be
heated for 8 h in a microwave oven at 140 °C. Neverthe-
less, different para-phenyl-substituted 6-methyl 1-phena-
Next, we wanted to investigate the influence of the para-
substituent on the formation of 5-aminoindolizines,
which were derived from different oxazolopyridinium
salts 3a–e. By applying the faster and usually cleaner
methodology of microwave heating we were confident
to isolate pure products in contrast to earlier work
(Scheme 3).15a We were pleased to find that using excess
piperidine without solvent under microwave heating al-
lowed us to isolate all desired products in approximately
95% purity for every substituent.
O
O
O
R1:
a) NO2
b) Cl
c) H
d) Me
e) OMe
yield:
53%
56%
49%
24%
46%
O
Br
CH3CN
N
N
+
140 ˚C
R1
R1
MW, 8 h
1
2
5a-e
Scheme 2. Synthesis of 6-methyl 1-phenacylpyridin-2-ones 5 by microwave heating.
O+
R1 (yield, cmpd.):
NO2 (43%, a)
Cl (85%, b)
N
N
1
R
H
N
N
R1
110˚C
MW,
10 min
H
(28%, c)
-
ClO4
Me (28%, d)
OMe (85%, e)
4a-e
3a-e
Scheme 3. Reaction of piperidine with different oxazolo[3,2-a]pyridinium salts 3 carrying various para-substituents.

P. Tielmann, C. Hoenke / Tetrahedron Letters 47 (2006) 261–265
263
The differences in isolated yields (Scheme 3) were not
due to a differing reactivity but caused by the work up
procedure. Typically, the reaction mixture was poured
into water followed by filtration or centrifugation to iso-
late the precipitated product. In some cases, this proved
to be very difficult due to the finely dispersed precipitate.
effect (entries 2/3, 6/7 and 10/11), while no conclusion
can be drawn for the temperature (entries 10/11). Even
without an additional base, the reaction could be run
to completion by using three amine equivalents. Thus,
the best result was obtained by employing 1.1 equiv of
piperidine and 2 equiv of DMAP at 100 °C for 1 h.20,21
The fact that the desired 5-aminoindolizines were
obtained in almost pure form in all cases might not only
be due to the heating by microwave irradiation, but also
to the simplified work up. Since in earlier work the crude
products were always purified by column chromatogra-
phy, we believe that the acidic nature of silica gel might
be responsible for the described instability.
After optimising the synthetic procedure, we finally per-
formed a small amine screen to explore the scope and
limitations of the formation of 5-aminoindolizines
addressing the impact on reactivity of some primary
and secondary amines with different steric and electronic
character (Scheme 5).
Sterically less demanding aliphatic secondary amines
6–8 deliver the desired indolizine system, in two out of
three cases even in quantitative yield as determined by
LC/MS.22 The low yield for diethylamine (8) is most
likely due to its high volatility. Further, 4-amino-piper-
idine (7) carrying both a primary and a secondary amine
reacts only with the more nucleophilic secondary
nitrogen delivering the product in 100% regioselectivity.
Sterically hindered aliphatic (i.e., 9) and the less nucleo-
philic aromatic amines (i.e., 10 and 11) do not react.
As we were interested in the construction of a combina-
torial library of 5-aminoindolizines by applying this
reaction sequence, we found it highly desirable to use
the amine just in stoichiometric quantities and to per-
form the reaction in a common organic solvent. For that
reason, we tried to optimise the synthetic procedure
with regard to amine equivalents, additional base and
reaction conditions by using piperidine and 5-methyl-
2-(p-chlorophenyl)-oxazolo[3,2-a]pyridinium salt 3b as
substrate (Scheme 4). Selected results of this screening
process are presented in Table 1.
With all primary amines 12–16, no 5-aminoindolizine
derivatives could be detected. Aromatic amines 12–14
remained unchanged. However, with aliphatic amines
like 15 and 16, a different product was detected in quan-
titative yield. After isolation and analytical investigation
of these compounds, they were identified to be imi-
dazo[3,2-a]pyridinium salts (Scheme 6a). In these cases
cyclisation does not occur via the methyl, but via the
amino group. To the best of our knowledge, this type
We were able to reduce the amount of amine to
1.1 equiv by using an additional base. Results by
employing DMAP were slightly superior compared to
triethylamine (entries 9–11 and 5–8). This is of impor-
tance for library construction so that a large and diverse
set of amines could be applied in stoichiometric
amounts. In contrast, the reaction time had only a small
O+
N
N
N
H
Cl
N
Cl
CH3CN,
add. base,
MW
-
ClO4
4b
3b
Scheme 4. Test reaction to screen for best reaction conditions and amine equivalents.
Table 1. Selected results of the optimisation process towards the synthesis of 5-aminoindolizines
Entry
Amine equivalents
Additional base
Temperature (°C)
Reaction time (h)
Yielda (%)
1
2
3
4
5
6
7
8
4
3
3
1.1
3
2
2
1.1
2
1.1
1.1
—
—
—
—
100
100
100
100
100
100
100
100
100
100
140
0.5
0.5
1
100
89
97
70
100
91
100
76
100
94
3
2 equiv triethylamine
2 equiv triethylamine
2 equiv triethylamine
2 equiv triethylamine
2 equiv DMAP
2 equiv DMAP
2 equiv DMAP
0.5
0.5
1
0.5
1
9
10
11
1
3
97
a Yields were determined by LC/MS (DAD purity), reaction conditions: solvent: acetonitrile, microwave heating (max 300 W).

264
P. Tielmann, C. Hoenke / Tetrahedron Letters 47 (2006) 261–265
sec. amines:
HN
(100%)
N
H
HN
NH2
N
H
6
7
(100%)
8
(8%)
9
(0%)
N
N
H
H
11 (0%)
10 (0%)
OMe
NH2
N
prim. amines:
NH2
NH2
NH2
12 (0%)
NH2
13 (0%)
16
15
(0%/100%)
14
(0%/9%)
(0% /100%)
Scheme 5. Set of amines applied in the synthesis of 5-aminoindolizines to explore the scope and limitations. The yields of 5-aminoindolizines as
determined by LC/MS are given in parentheses. For primary amines the yield of side product (see text below) is also given (substrates 14–16, second
percentage value in parentheses).
with primary aliphatic amines imidazopyridinium salts
are formed.
(b)
(a)
HO
N
N+
N
Acknowledgements
Cl
Cl
N
We are thankful to Professor Dr. Volkhard Austel and
Dr. Stefanie Weyler for helpful discussions and critical
proofreading of the manuscript.
O
Scheme 6. Proposed structures of (a) imidazo[3,2-a]pyridinium salts
and (b) chinoid oxidation products.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.tetlet.
2005.11.033.
of compound has so far not been prepared starting from
oxazolo[3,2-a]pyridinium salts.
Finally, we performed stability measurements to ensure
the utility of 5-aminoindolizines for library construc-
tion. For that reason, we stored some 5-aminoindolizine
derivatives in DMSO at rt in the presence of air. Unfor-
tunately, the products slowly oxidised to form chinoid
structures as evidenced by high resolution mass spec-
trometry (Scheme 6b). This instability towards oxida-
tion has already been shown for 8-aminoindolizine
derivatives.19
References and notes
1. Two alkaloids containing an indolizine nucleus within a
fused ring system have been reported: Flitsch, W. In
Comprehensive Heterocyclic Chemistry; Katritzky, A. R.,
Ress, C. W., Eds.; Pergamon: Oxford, 1984; Vol. 4, p 476.
2. Chai, W.; Breitenbucher, J. G.; Kwok, A.; Li, X.; Wong,
V.; Carruthers, N. I.; Lovenberg, T. W.; Mazur, C.;
Wilson, S. J.; Axe, F. U.; Jones, T. K. Bioorg. Med. Chem.
Lett. 2003, 13, 1767.
In conclusion, we have demonstrated the utility of
oxazolo[3,2-a]pyridinium salts to create a set of diverse
5-aminoindolizines. Microwave heating in conjunction
with the use of an auxiliary base and a common organic
solvent made it possible to reduce the necessary amount
of amines to 1.1 equiv, which makes the procedure suit-
able for the production of highly diverse combinatorial
libraries. Further, we were able to extend the reported
procedure15 to a wider range of applicable bromo-
ketones. More critical is the observation that these indo-
lizines are slowly oxidised in solution and may therefore
cause storage problems. The procedure is however lim-
ited to sterically less demanding secondary amines, while
3. Gundersen, L.-L.; Negussie, A. H.; Østby, O. B. Arch.
Pharm. Pharm. Med. Chem. 2003, 336, 191.
4. Gupta, S. P.; Mathur, A. N.; Nagappa, A. N.; Kumar, D.;
Kumaran, S. Eur. J. Med. Chem. 2003, 38, 867.
5. (a) Gundersen, L.-L.; Malterud, K. E.; Negussie, A. H.;
Rise, F. R.; Teklu, S.; Østby, O. B. Bioorg. Med. Chem.
2003, 11, 5409; (b) Teklu, S.; Gundersen, L.-L.; Larsen, T.;
Malterud, K. E.; Rise, F. Bioorg. Med. Chem. 2005, 13,
3127.
6. Østby, O. B.; Dalhus, B.; Gundersen, L.-L.; Rise, F.; Bast,
A.; Haenen, G. R. M. M. Eur. J. Org. Chem. 2000, 3763.
7. Wang, P.; Zhang, Z.; Ma, X.; Huang, Y.; Liu, X.; Tu, P.;
Tong, T. Mech. Ageing Dev. 2003, 124, 1025.

P. Tielmann, C. Hoenke / Tetrahedron Letters 47 (2006) 261–265
265
8. (a) Medda, S.; Jaisankar, P.; Manna, R. K.; Pal, B.; Giri,
V. S.; Basu, M. K. J. Drug Target 2003, 11, 123; (b) Bolle,
L. D.; Andrei, G.; Snoeck, R.; Zhang, Y.; Lommel, A. V.;
Otto, M.; Bousseau, A.; Roy, C.; Clercq, E. D.; Naesens,
L. Biochem. Pharmacol. 2004, 67, 325.
9. (a) Scholtz, M. Ber. Dtsch. Chem. Ges. 1912, 45, 734; (b)
Boekelheide, V.; Windgassen, R. J., Jr. J. Am. Chem. Soc.
1959, 81, 1456.
10. (a) Tschitschibabin, A. E. Ber. Dtsch. Chem. Ges. 1927, 60,
1606; (b) Hurst, J.; Melton, T.; Wibberley, D. G. J. Chem.
Soc. 1965, 2948; (c) Jones, G.; Stanyer, J. J. Chem. Soc. C
1969, 901.
16. The construction of indolizine libraries has recently been
reported: (a) Chen, Z.; Yue, G.; Lu, C.; Yang, G. Synlett
2004, 1231; (b) Yue, G.; Wan, Y.; Song, S.; Yang, G.;
Chen, Z. Bioorg. Med. Chem. Lett. 2005, 15, 453.
17. The reaction mixture was poured onto ice and not diluted
with water and the product was precipitated at room
temperature.
18. This type of reactivity has already been observed: Brad-
sher, C. K.; Zinn, M. F. J. Heterocycl. Chem. 1967, 4, 66.
19. Terenin, V. I.; Kabanova, E.; Feoktistova, E.; BundelÕ, Y.
G. Chem. Heterocycl. Compd. 1992, 28, 766.
20. Since the effect of applying higher temperature and longer
reaction times on the yield was marginal, we decided to
choose the milder and faster conditions as ÔbestÕ result.
21. Typical procedure: 100 mg (291 lmol) of 3b were dis-
solved in 2 ml acetonitrile, whereafter 71 mg (581 lmol)
DMAP and 31.6 ll (320 lmol) of piperidine were added.
The reaction mixture was heated for 1 h to 100°C in a
microwave oven (CEM DiscoverTM, max 300 W, for details
see Supplementary material) and poured into 10 ml of
water. Most of the yellow product precipitated after
standing at rt overnight so that the supernatant could be
decanted. The residue was carefully washed with water
and dissolved in ethyl acetate. The supernatant was
centrifuged for 1 h at 3000 rpm, whereafter some more
product could be obtained after decantation and washing
with water. This second crop was also dissolved in ethyl
acetate and the combined organic layer was dried with
sodium sulfate. After removal of the solvent the product
was isolated in approx. 95% purity. Yield: 58 mg (61%).
22. For analytical data of all isolated 5-aminoindolizines see
Supplementary material.
11. (a) Miki, Y.; Hachiken, H.; Takemura, S. Heterocycles
1984, 22, 701; (b) Padwa, A.; Austin, D. J.; Precedo, L.;
Zhi, L. J. Org. Chem. 1993, 58, 1144; (c) De Bue, G.;
Nasielski, J. Bull. Soc. Chim. Belg. 1997, 106, 97; (d)
Katritzky, A. R.; Qiu, G.; Yang, B.; Ye, H.-Y. J. Org.
Chem. 1999, 64, 7618; (e) Sarkunam, K.; Nallu, M. J.
Heterocycl. Chem. 2005, 42, 5.
12. (a) Wei, X.; Hu, Y.; Li, T.; Hu, H. J. Chem. Soc., Perkin
Trans. 1 1993, 2487; (b) Fang, X.; Wu, Y.-M.; Deng, J.;
Wang, S. W. Tetrahedron 2004, 60, 5487.
13. (a) KelÕin, A. V.; Sromek, A. W.; Gevorgyan, V. J. Am.
Chem. Soc. 2001, 123, 2074; (b) Kim, J. T.; Gevorgyan, V.
Org. Lett. 2002, 4, 4697.
14. Park, C.-H.; Ryabova, V.; Seregin, I. V.; Sromek, A. W.;
Gevorgyan, V. Org. Lett. 2004, 6, 1159.
15. (a) Babaev, E. V.; Efimov, A. V. Chem. Heterocycl.
Compd. 1997, 33, 875; (b) Babaev, E. V.; Efimov, A. V.;
Maiboroda, D. A.; Jug, K. Eur. J. Org. Chem. 1998, 193;
(c) Babaev, E. V.; Efimov, A. V.; Zhukov, S. G.; Rybakov,
V. Chem. Heterocycl. Compd. 1998, 34, 852.