Synthesis of polysubstituted 3-hydroxypyridines via the revisited hetero-Diels-Alder reaction of 5-alkoxyoxazoles with dienophiles

Article abstract of DOI:10.1039/c1cc16562c

A general and single-step access to polysubstituted 3-hydroxypyridine scaffolds via hetero-Diels-Alder (HDA) reactions between readily prepared 5-ethoxyoxazoles and dienophiles is reported. The HDA reaction, run in the presence of Nd(OTf)₃ at r

Full text of DOI:10.1039/c1cc16562c

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2012, 48, 768–770
www.rsc.org/chemcomm
COMMUNICATION
Synthesis of polysubstituted 3-hydroxypyridines via the revisited
hetero-Diels–Alder reaction of 5-alkoxyoxazoles with dienophilesw
Cyrille Sabot, Emilia Oueis, Xavier Brune and Pierre-Yves Renard*
Received 21st October 2011, Accepted 8th November 2011
DOI: 10.1039/c1cc16562c
A general and single-step access to polysubstituted 3-hydroxy-
pyridine scaffolds via hetero-Diels–Alder (HDA) reactions between
readily prepared 5-ethoxyoxazoles and dienophiles is reported. The
HDA reaction, run in the presence of Nd(OTf)3 at room temperature,
was successfully applied to various 5-ethoxyoxazoles showing good
functional group tolerance, and led to a straightforward process to
obtain useful building-blocks.
3-Hydroxypyridines have attracted particular attention due to
Scheme 2 Preparation of 3-hydroxypyridine derivatives 4.
their presence in several bioactive substances such as vitamin
B6 (Scheme 1).1 They are also used as key intermediates to access
natural compounds for example pterocellin A,2 or significant
scaffolds in medicinal chemistry such as furopyridine systems.3
However, very few methods to access 3-hydroxypyridine scaffolds
are reported in the literature to date.4 They usually require
drastic reaction conditions poorly compatible with functionalized
systems.
of this reaction is under-investigated. Indeed, the scope of the
reaction remains limited and sensitive functional groups are not
compatible with the required experimental conditions (temperature
and high pressure).8 To the best of our knowledge no catalytic and
wide scope one-step procedure for this transformation, involving
mild experimental conditions, has been reported to date in the
literature.
A pioneer synthetic strategy to access functionalised 3-hydroxy-
pyridines has been proposed in the 1960’s. It is based on
Kondrat’eva’s hetero-Diels–Alder (HDA) reaction involving
5-alkoxyoxazoles 1 acting as azadienes in the presence of
electron-deficient olefinic dienophiles 2 (Scheme 2, eqn (1)).5
The unstable bicyclic cycloadduct 3 generated undergoes
aromatization via loss of alcohol, under acidic catalysis,6 to
yield substituted 3-hydroxypyridines 4. Noteworthily the
precursors 5-alkoxyoxazole 1 are readily prepared in two steps
from easily available a-amino esters 5 and carboxylic acid
derivatives 6 (eqn (2)) or via a multicomponent process
recently developed by Masson and Zhu.7 However, the use
Herein, we wish to report a practical and environmentally
friendly solvent-free route to polysubstituted pyridines from the
corresponding 5-alkoxyoxazoles, in the presence of an inexpensive
and commercially available catalyst, neodymium(III) trifluoro-
methanesulfonate Nd(OTf)3. First, we turned our attention
towards the use of rare-earth metal triflates to catalyze the
HDA reaction between 5-alkoxyoxazoles and electron-
deficient olefinic dienophiles.9 Indeed, these water-compatible
Lewis-acids should be stable in the presence of the alcohol
released during the aromatization step of the reaction and
could be used in catalytic or substoichiometric amount, contrary
to the traditional and moisture sensitive Lewis acids AlCl3, BF3
or TiCl4. In addition, they are still active in the presence of
nitrogen-, oxygen-, phosphorus- or sulfur-containing compounds
and thus the process may be compatible with a variety of
different substrates. Consequently, the model cycloaddition
reaction between the 2-methyl-4-benzyl-5-ethoxyoxazole 1a
(1 equiv.) and dimethylmaleate 2a (2 equiv.) was carried out
neat, in the presence of 25 mol% of catalyst, at rt for 24 h
(Table 1). Pleasingly, in the presence of Sc(OTf)3 the desired
3-hydroxypyridine 4a was formed, but in a low 10% yield,
along with unreacted starting materials (entry 2), whereas no
trace of the desired product was detected in the absence of
catalyst (entry 1). This encouraging result prompted us to
investigate the use of lanthanide triflates as catalysts. While
moderate yield increases were observed with Eu(III) or Yb(III)
Scheme 1 Bioactive compounds prepared from 3-hydroxypyridines.
Equipe de Chimie Bio-Organique, CNRS UMR 6014-COBRA & FR
`
3038, IRCOF, rue Lucien Tesniere, 76131 Mont-Saint-Aignan Cedex,
France. E-mail: pierre-yves.renard@univ-rouen.fr;
Fax: +33 (0)2 35 52 29 71; Tel: +33 (0)2 35 52 24 76
w Electronic supplementary information (ESI) available. See DOI:
10.1039/c1cc16562c
c
768 Chem. Commun., 2012, 48, 768–770
This journal is The Royal Society of Chemistry 2012

View Article Online
Table 2 Scope of the HDA reaction from various dienes 1a
Table 1 Lewis acids screening for synthesis of 3-hydroxypyridinesa
Entry
Catalyst
Yieldb (%)
1
2
3
4
5
6
7
8
None
0
10
31
38
Sc(OTf)3
Eu(OTf)3
Yb(OTf)3
La(OTf)3
Sm(OTf)3
Nd(OTf)3
TfOH
57
58
59 (72)c
0
9
10
Hf(OTf)4ÁH2O
o5
Cu(OTf)2
0
a
All reactions were carried out with 1.0 equiv. of oxazole 1a, 2.0 equiv. of
dimethylmaleate 2a, 25 mol% of catalyst, under solvent-free conditions
at rt for 24 h. Yield determined by 1H NMR of the crude reaction
c
mixture relative to 1a. Isolated yield using 40 mol% of Nd(OTf)3.
b
(31, 38%, entries 3 and 4), the use of La(III), Sm(III) or Nd(III)
notably improved the reaction efficiency to almost 60% yield
of 4a (entries 5–7). Optimized conditions (40 mol% of
Nd(OTf)3) afforded the 3-hydroxypyridine 4a in 72% isolated
yield (entry 7). Meanwhile, we also established that triflic acid
(possibly generated in the reaction medium while using
RE(OTf)3 as catalyst) could not catalyze the reaction by itself
(0%, entry 8).
Finally, the ability of transition metal complexes to catalyze
the [4+2] cycloaddition reaction was unsuccessfully explored
since neither the transition metal Hf(IV) nor Cu(II) allowed
efficient formation of the pyridine 4a (o5 and 0% yields,
entries 9 and 10).
a
All reactions were carried out with 1.0 equiv. of oxazole, 2.0 equiv. of
dimethylmaleate, 40 mol% of Nd(OTf)3, under solvent-free conditions
Next, the scope and limitation of the [4+2] cycloaddition
reaction between a family of 5-ethoxyoxazoles 1 and dimethyl-
maleate 2a, under the previously optimized conditions (e.g.
40 mol% of Nd(OTf)3, neat, rt, 24 h), was examined. Results
obtained are summarized in Table 2. Oxazoles with alkyl
substituents at positions 2 and 4 generally led to the corres-
ponding 5-hydroxypyridines 4a–e in good yields (70–85%). To
our satisfaction, the reaction outcome was not significantly
dependent on the steric hindrance of the substituent at position
2 of the oxazoles. Reactions of 2-unsubstituted oxazole 1d
and 2-iso-propyl-oxazole 1e led to the pyridines 4d and 4e
respectively, in high yields (85 and 75%). However, small size
substituents such as a methyl group at position 5 of the
oxazole afforded the pyridine 4f in a lower yield of 41%, due
to partial hydrolysis of the unstable intermediate cycloadduct.8d
2-Aryl substituted 5-alkoxyoxazole afforded the pyridine 4g in a
moderate yield of 55%, probably due to the greater aromaticity
of the oxazole ring. Indeed, the loss of conjugation required for
the cycloaddition step is then more difficult. It is worthy of note
that in this case, pyridine 4g is yet obtained in 90% yield based
on recovered starting material. Moreover, we focused our
attention on 2- and 4-functionalized 5-ethoxyoxazoles. Oxazole
1h bearing both an ether and a thermally unstable azide group
gave 4h in 66% yield while only degradation of the starting
b
for 24 h (isolated yields). Yield based on recovered starting oxazole.
oxazole was observed upon running the reaction under usual
thermal conditions.10
In addition, the conditions were compatible with thioether
and monosubstituted alkyne function. Indeed, oxazoles 1i and
1j delivered the corresponding pyridines 4i and 4j in 52
and 61% yield respectively. To our knowledge these results
constitute a rare report of [4+2] cycloadditions of functionalized
5-alkoxyoxazoles.7
Eventually, the dienophile scope was also evaluated. Besides
dimethylmaleate, acryloyl-type dienophiles were also examined
and the results are summarized in Table 3. The reaction of the
5-ethoxyoxazole 1a and 1b with methyl acrylate gave the
corresponding pyridine derivatives 4k and 4l in high yields of
85 and 74%, respectively, under previously optimized conditions
(e.g. 40 mol% of Nd(OTf)3, neat, rt, 24 h). It is noteworthy that
the reaction run between 1b and methyl acrylate in the absence
of catalyst led to less than 10% yield of the desired pyridine 4l.
In the case of the methyl vinyl ketone, we noticed that only
5 mol% of Nd(OTf)3 were sufficient to afford the formation of
the pyridines 4m and 4n in 75% and 81% yield respectively.
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 768–770 769

View Article Online
Table 3 Scope of the HDA reaction with electrophilesa
polyheterocycle 8 in 56% (unoptimized) yield, readily functio-
nalised for further transformations.
In conclusion, we reported an efficient one-step process for
synthesis of polysubstituted 3-hydroxypyridine derivatives.
The mild reaction conditions involved enable the HDA
reaction with unprecedented sensitive functional groups,
leading to building-blocks of biological importance. The rapid
polyfunctionalisation generated by this process is currently
under investigation for the ligation of biomolecules.
This work was supported by la Region Haute-Normandie
´
via the CRUNCh program (CPER 2007–2013) and a PhD
fellowship to EO, and ERDF funding (ISCE-Chem-Interreg IV4
program). We thank A. Marcual (CNRS) and A. Leboisselier
(INSA de Rouen) for HRMS and elemental analyses,
respectively, and Dr A. Romieu (Universite de Rouen) for
´
HPLC analyses.
Notes and references
1 For an example, see: S. Sharif, D. Schagen, M. D. Toney and
H.-H. Limbach, J. Am. Chem. Soc., 2007, 129, 4440.
2 M. M. O’Malley, F. Damkaci and T. R. Kelly, Org. Lett., 2006,
8, 2651.
3 (a) T. Cailly, S. Lemaıtre, F. Fabis and S. Rault, Synthesis, 2007, 3247;
(b) S. Price, K. William, P.-P. Savy, H. Dyke, J. G. Montana,
M. S. Stanley and L. Bao, WO2008024725, 2008; (c) A. J.
Buckmelter, J. P. Lyssikatos, G. Miknis, L. Ren and S. M.
Wenglowsky, WO2008028141, 2008; (d) M. P. Bell, C. R. O’Dowd,
J. S. S. Rountree, G. P. Tevitt, T. Harrison and M. M. Mcfarland,
WO2011055115, 2011; (e) A. J. Buckmelter, L. Ren, E. R. Laird,
B. Rast, G. Miknis, S. Wenglowsky, S. Schlachter, M. Welch,
E. Tarlton, J. Grina, J. Lyssikatos, B. J. Brandhuber, T. Morales,
N. Randolph, G. Vigers, M. Martinson and M. Callejo, Bioorg. Med.
Chem. Lett., 2011, 21, 1248.
4 (a) M. D. Fletcher, T. E. Hurst, T. J. Miles and C. J. Moody,
Tetrahedron, 2006, 62, 5454; (b) J.-Y. Lu and H.-D. Arndt, J. Org.
Chem., 2007, 72, 4205; (c) J.-Y. Lu, J. A. Keith, W.-Z. Shen,
M. Schuermann, H. Preut, T. Jacob and H.-D. Arndt, J. Am.
Chem. Soc., 2008, 130, 13219; (d) K. Yoshida, F. Kawagoe,
K. Hayashi, S. Horiuchi, T. Imamoto and A. Yanagisawa, Org.
Lett., 2009, 11, 515.
5 (a) G. Ya. Kondrat’eva, Khim. Nauka Prom-st., 1957, 2, 666;
(b) D. L. Boger, Chem. Rev., 1986, 86, 781; (c) E. E. Harris,
R. A. Firestone, K. Pfister, R. R. Boettcher, F. J. Cross,
R. B. Currie, M. Monaco, E. R. Peterson and W. Reuter,
J. Org. Chem., 1962, 27, 2705.
6 B. A. Johnsen and K. Undheim, Acta Chem., Scand., 1983, 37, 127.
7 C. Lalli, M. J. Bouma, D. Bonne, G. Masson and J. Zhu,
Chem.–Eur. J., 2011, 17, 880.
a
All reactions were carried out with 1.0 equiv. of oxazole, 2.0 equiv.
of dimethylmaleate, 40 mol% of Nd(OTf)3, under solvent-free condi-
b
tions, rt for 24 h (isolated yields). Without catalyst. 5 mol% of
d
Nd(OTf)3. THF was used as solvent.
c
The reaction with acrylonitrile or acrylamide as dienophile led
to the formation of the pyridine system 4o and 4p in 75% and
52% yield respectively. The final attempt to form compound 4q
from oxazole 1a and the phenyl vinyl sulfone was unsuccessful.
With these promising results in hand, we wanted to evaluate
whether this reaction could be applied to the synthesis of
highly functionalised 3-hydroxypyridines of interest. Furopyridine
derivatives (Scheme 1) have shown anti-cancer and/or anti-
inflammatory activity through MEK kinase activity inhibition.3b–e
Thus, we explored the possibility to generate a functionalized
furopyridine derivative from the pyridine 4a, previously
prepared via HAD reaction (Scheme 3). Under Mitsunobu
conditions, the pyridine 7 was formed in 79% yield. Final
treatment of 7 with DBU in THF afforded the desired
8 (a) R. A. Firestone, E. E. Harris and W. Reuter, Tetrahedron, 1967,
23, 943; (b) N. D. Doktorova, L. V. Ionova, M. Y. Karpeisky,
N. S. Padyukova, K. F. Turchin and V. L. Florentiev, Tetrahedron,
1969, 25, 3527; (c) B. A. Johnsen and K. Undheim, Acta
Chem., Scand., 1983, 37, 907; (d) G. Sandford, I. Wilson and
C. M. Timperley, J. Fluorine Chem., 2004, 125, 1425; (e) S. Sharif,
D. Schagen, M. D. Toney and H.-H. Limbach, J. Am. Chem. Soc.,
2007, 129, 4440.
9 For Lewis acid-catalyzed intramolecular HDA reactions with
5-alkyloxazoles, see: (a) J. I. Levin, Tetrahedron Lett., 1989,
30, 2355; (b) M. Ohba and R. Izuta, Heterocycles, 2001, 55, 823.
10 Thermal reaction conditions: 1.0 equiv. of oxazole, 2.0 equiv. of
dimethylmaleate, under solvent-free conditions, 110 1C for 24 h.
Scheme 3 Preparation of the functionalized furopyridine derivative 8.
c
770 Chem. Commun., 2012, 48, 768–770
This journal is The Royal Society of Chemistry 2012