Synthesis of fused dihydropyrano(furano)pyridines via [4 + 2]-cycloaddition of 5-alkenoxy substituted oxazoles

Article abstract of DOI:10.1021/ol300351v

A three-step procedure to access fused pyridines has been developed utilizing inexpensive amino acids and alkenols to form the key oxazole precursors. Yields are good to excellent and provide a rapid and inexpensive route to a range of pharmacologically a

Full text of DOI:10.1021/ol300351v

ORGANIC
LETTERS
2012
Vol. 14, No. 6
1604–1607
Synthesis of Fused
Dihydropyrano(furano)pyridines
via [4 þ 2]-Cycloaddition of 5-Alkenoxy
Substituted Oxazoles
Steven N. Goodman, Douglas M. Mans,* Joseph Sisko, and Hao Yin
Chemical Development - Synthetic Chemistry, GlaxoSmithKline R&D,
709 Swedeland Road, King of Prussia, Pennsylvania 19406, United States
douglas.m.mans@gsk.com
Received February 13, 2012
ABSTRACT
A three-step procedure to access fused pyridines has been developed utilizing inexpensive amino acids and alkenols to form the key oxazole
precursors. Yields are good to excellent and provide a rapid and inexpensive route to a range of pharmacologically and biologically valuable fused
pyridines with difficult to access substitution patterns.
The pyridine moiety is pervasive throughout nature as
well as biologically active synthetic compounds. In parti-
cular, fused pyridine ring systems have been granted special
status in the pharmaceutical arena. Having said this, the
syntheses of fused pyridines have often been rather cumber-
some and lengthy even for the simplest fused pyridines. An
example is the dihydropyrano[2,3-c]pyridine 1.
Despite the relatively simplistic looking ring system of 1,
our initial synthesis required 12 steps starting from Kojic
acid. Besides the obvious length, the synthesis was plagued
with several low-yielding steps, long cycle times, and cost
of goods issues.
Similarly long synthetic routes to seemingly innocuous
highly substituted and fused pyridines are found through-
out the literature. For example, the rather simple 5-methoxy-
4-methylpicolinaldehyde still required seven discrete steps
from Kojic acid for its synthesis.1 A second example, also
starting from Kojic acid, assembled 2,3-dihydrofuro[2,3-
c]pyridine-5-carbaldehyde in 10 steps and required the use
of several transition metal mediated steps which can prove
to be problematic to optimize should another functionality
be incorporated.2
Tasked with developing a new route addressing these
problems, we were drawn to the possibility of constructing
the pyridine moiety via a [4 þ 2] cycloaddition. A survey of
the literature revealed a plethora of examples involving
inter- and intramolecular Kondrat’eva pyridine syntheses.3
However, we were surprised to find a dearth of examples
involving intramolecular 5-alkenoxy tethered substrates. In
addition, literature precedence involving 5-(alkenyloxy)-2-
phenyloxazoles failed to undergo the desired [4 þ 2]
cycloaddition.4 This was rather surprising as it is known
that 5- (or 2)-alkoxy-substitution on the oxazole has an
accelerating effect on the rate of reaction due to the
increased electron density of the oxazole.3
(3) (a) Kozikowski, A. P.; Hasan, N. M. J. Org. Chem. 1977, 42, 2039.
(b) Levin, J. I.; Weinreb, S. M. J. Am. Chem. Soc. 1983, 105, 1397. (c)
Levin, J. I.; Weinreb, S. M. J. Org. Chem. 1984, 49, 4325. (d)
Subramanyam, C.; Noguchi, M.; Weinreb, S. M. J. Org. Chem. 1989,
54, 5580. (e) Levin, J. I.; Laakso, L. M., Oxazoles: Synthesis, Reactivity,
and Spectroscopy; Palmer, D. C., Ed.; John Wiley & Sons, Inc.: Hoboken,
NJ, 2003; Vol. 60, Chapter 3, pp 417. (f) Karpeiskii, M. Y.; Florent’ev,
V. L. Russ. Chem. Rev. 1969, 38, 540.
(4) We are aware of only one failed report concerning oxazole
cycloadditions involving substrates with the dienophile tether being
incoroporated at the 5-alkoxy position; see: Padwa, A.; Cohen, L. A.
J. Org. Chem. 1984, 49, 399.
(1) Kiyoto, T.; Ando, J.; Tanaka, T.; Tsutsui, Y.; Yokotani, M.;
Noguchi, T.; Ushiyama, F.; Urabe, H.; Horikiri, H. WO2007138974A1,
2007.
(2) Brooks, G.; Giordano, I.; Hennessy, A. J.; Pearson, N. D.
WO2007122258A1, 2007.
r
10.1021/ol300351v
Published on Web 03/07/2012
2012 American Chemical Society

Undeterred by the lack of precedence and knowing the
accelerating effect of a 5-alkoxy group, we set out to
explore the feasibility of utilizing a 5-alkenyloxy tethered
oxazole to give the desired dihydropyrano[2,3-c]pyridine
ring system 1 directly.
Scheme 1. Synthesis of Dihydropyranopyridine 1
Our retrosynthesis of 1led to the 5-(pent-4-en-1-yloxy)oxazole
2(Figure1).We reasoned that oxazole 2 would arise from
a cyclodehydration of oxamate 3 through the use of a
suitable dehydrating agent. The oxazole precursor 3
would arise ultimately from N-Boc-glycine (4) and
4-penten-1-ol (5).
first step we were hoping to replace triflic anhydride as the
dehydrating agent due to cost, safety, and environmental
concerns. As with the intramolecular 5-alkenoxy substi-
tuted cycloaddition, only a sparse account of dehydrating
agents used to successfully form 5-alkoxy substituted
oxazoles was found.7 A screen of dehydrating agents was
performed using model oxamate 7 (eq 1).
Figure 1. Retrosynthesis of dihydropyranopyridine 1.
In the forward sense, synthesis of 3 commenced with the
CDI-mediated esterification of N-Boc-glycine 4 with
4-penten-1-ol 5 to provide intermediate glycine ester 6
which was carried on directly without isolation (Scheme 1).
Boc removal with methanesulfonic acid followed by
treatment with either dimethyl oxalate or methyl chloro-
oxalate provided oxazole precursor 3 in 56ꢀ80% yields.5
Treatment of 3 with triflic anhydride/pyridine in DCM led
to efficient cyclodehydration to oxazole 2.6 Based on
literature precedent we had expected to isolate and purify
the oxazole 2 prior to subjecting it to the typical thermal
cycloaddition conditions.3
Of the reagents examined only Tf2O/pyridine was found
to give high yields of oxazole 8 in a reasonable time frame
and mild conditions. Other previously reported desic-
cants gave no reaction, decomposition, or extremely low
conversions.8 In addition, previous reports using P2O5 on
model oxamate 7 were also moderately successful;6 how-
ever the mixture formed an intractable solid mass making
handling and isolation on scale difficult. Interestingly,
T3P and Et3N in either DMF or EtOAc did provide
moderate conversion (60ꢀ80%) but only after extended
heating (>24 h).
However to our delight not only did the cyclodehydra-
tion to 2 occur efficiently upon treatment with triflic
anhydride/pyridine at ambient temperatures but the ensu-
ing [4 þ 2] cycloaddition and concomitant dehydration/
aromatization to the desired dihydropyrano[2,3-c]pyridine
1 also occurred in a single flask reaction. To our knowledge
this is the first example of such a cascade cycloaddition/
aromatization sequence occurring at ambient temperatures,
under mild conditions, with a completely unactivated alkene
and an oxazole. Ultimately the new route delivered the
desired pyridine 1 in a two-pot process from relatively
inexpensive materials in 36ꢀ56% overall yields.
With confirmation that Tf2O/pyridine was optimal for
cyclodehydration, the substrate scope was then explored
(Figure 2). Employing amino acids other than glycine
allowed for substitution at R2 of the pyridine ring in high
yields (10bꢀd).
Surprisingly, the rate of the reaction was markedly
increased with the 4-alkyl substituted oxazoles (10b and
Having successfully demonstrated the viability of the
key intramolecular cycloaddition we were keen to explore
the substrate scope of the [4 þ 2] reaction sequence. As a
(7) See refs 2 and 4 as well as: (a) Maeda, I.; Takehara, M.; Togo, K.;
Asai, S.; Yoshida, R. Bull. Chem. Soc. Jpn. 1969, 42, 1435. (b) Han, W.;
Egberton, M.; Wai, J.; Zhuang, L.; Ruzek, R. D.; Perlow, D.; Isaacs,
R. C.; Cameron, M.; Foster, B.; Dolling, U. H.; Hoerrner, R.; Obligado,
V. E.; Neilson, L. A.; Kim, B.; Payne, L. S.; Morrisette, M. M.; Williams,
P. D.; Pye, P. J.; Angelaud, R.; Mancheno, D. E.; Askin, D.
WO2005087768, 2005.
(5) Formation of 3 with methyl chlorooxalate in general gave higher
yields and purity, whereas the dimethyloxalate invariably led to the
formation of 10ꢀ20% of the undesired oxalamide via bis-addition.
However, from a cost standpoint the use of dimethyl oxalate is much
preferred.
(6) Thalhammer, A.; Mecinovic, J.; Schofield, C. J. Tetrahedron Lett.
2009, 50, 1045.
(8) Unsuccessful reagents included: triphosgene/Et3N; COCl2/Et3N;
Ms2O/Et3N; BF3 OEt2; SOCl2; POCl3/pyr; ClSO3H/pyr; TFAA/TFA;
H2SO4/Ac2O; Ph3P/I2/Et3N.
3
Org. Lett., Vol. 14, No. 6, 2012
1605

Scheme 2. Cycloaddition To Give Oxepine-Fused Pyridines
c]pyridine (12a) in a 41% combined yield. Whereas the
4-iPr-substituted hexenyloxy oxazole arising from 11b
(R = iPr) provided a reversed selectivity giving the
tetrahydrooxepino[2,3-c]pyridine (13b) in a 2:1 ratio with
the dihydropyrano[2,3-c]pyridine (13a) in a 65% combined
yield.11 Lastly, heteroatoms can also be incorporated into
the fused ring, for example the dioxepino[5,6-c]pyridine
(14b). Interestingly, this substrate (11c, X = O) gave
exclusively the dioxepine-fused pyridine with none of the
dioxane-fused product observed. If the constitutional iso-
mers formed (12a and 13a) arise from acid-catalyzed olefin
isomerization, then one can explain the lack of formation of
14a through β-alkoxy destabilization of the carbocation
due to the inductive effects of the oxygen.12
An unexpected product was observed when the isolated
oxazole 2 was submitted to thermal conditions. It was
found that brief treatment of oxamide 3 with triflic anhy-
dride followed by basic workup allowed for isolation of
the oxazole 2. When 2 was heated to 120 °C in toluene
for 24 h, the ring-opened 3-hydroxypyridine 17a was
obtained in excellent yields. The benzamide (15b) and the
p-nitrobenazamide (15c) oxazole precursors were also
subjected to the cascade reaction conditions. Following
oxazole formation with triflic anhydride no cycloaddition
was found to occur. However, after the isolated oxazoles
(16b/c) were heated at 120 °C for 24 h the ring-opened
3-hydroxypyridines (17b/c) were isolated as the only cy-
cloadducts (Scheme 3).13 Thus dependent upon the condi-
tions chosen to perform the cycloaddition, complete regio-
selective control over which CꢀO bond is broken can be
obtained to provide the fused dihydropyranopyridine
(bond b breaks under acidic conditions) or the 3-hydro-
xypyridine (bond a breaks under thermal conditions)
(Scheme 4).14 It should be noted that under thermal
Figure 2. Substrate scope for the [4 þ 2].
10c) compared to simple glycine substrates, most likely due
to a further increase in the electron density of the oxazole
moiety.9 Variation at the R1 moiety was less successful.
The morpholine amide was a competent substrate to pro-
vide 10e although the yields were only modest and not
optimized further. Other R1 substituents investigated
either gave conversion to the oxazole but failed to undergo
cycloaddition (R1 = Me-, MeO-, (MeO)2CH-, Ph-, Me2N-)
or failed to form oxazole (R1 = CF3, CCl3) altogether.
Internal alkenes (R3 = Me) likewise were successful sub-
strates providing tetrasubstituted pyridines 10f (R3
=
Me).10 A final example demonstrates the utility of the
method as the pentasubstituted pyridine 10g (R2 = R3 =
Me) was accessed in excellent yields in only three steps with
no intermediate isolations. The corresponding dihydrofur-
an-fused pyridines are readily obtained utilizing a butenyl
tether (10h and 10i).
Interestingly the seven-membered-ring homologue from
the hexenyl tether 11a was found to give two cycloaddition
products whose distribution was dependent on the oxazole
1
substitution (Scheme 2). H NMR analysis showed that
the desired tetrahydrooxepino[2,3-c]pyridine (12b, R = H)
was formed as a 1:3 mixture with the dihydropyrano[2,3-
(9) The 4-alkyl substituted oxazoles proceeded to form pyridine
products in 1ꢀ2 h vs 6ꢀ>24 h for the unsubstituted oxazoles.
(10) Use of either cis or trans-4-hexen-1-ol provided similar yields
and similar rates of reaction as well.
(11) At present our rationale for the 6,6- and 6,7-ring system dis-
tribution is due to simple olefin isomerization in the presence of TfOH. It
is interesting to note the accelerating effect the 4-iPr substituted oxazole
has on the rate of the cycloaddition in which olefin isomerization is now
suppressed. A similar rate enhancement was also observed in the
dihydrofuran series arising from the relative rates of ring formation
(5 > 6 . 7).
(12) Another explanation is that olefin migration to the internal enol
ether gives an alkene substrate with improper electronics to undergo the
[4 þ 2] cycloaddition.
1606
Org. Lett., Vol. 14, No. 6, 2012

Scheme 3. Thermal Cycloaddition of Oxazoles
Scheme 4. Thermal vs Acidic Ring Opening
access highly substituted fused pyridines in an extremely
efficient, modular, and regiospecific manner. In addition,
for the majority of substrates the reaction conditions
required were found to be quite moderate needing only
Tf2O/pyridine at ambient temperature to induce oxazole
formation and cycloaddition with completely unactivated
alkenes. At this time sufficient evidence to rule out either a
concerted or stepwise mechanism is not available. Future
work will address this knowledge gap. In addition, it has
been found that complete control of the mechanism re-
garding oxabicycle ring opening is obtained by switching
from acidic to thermal conditions providing the fused
bicyclic pyridine products or the ring-opened 3-hydroxy-
pyridine products.
conditions the products (17aꢀc) possess the substitution
pattern reported in the literature for 3-hydroxypyridines
arising from intermolecular oxazole DielsꢀAlder reactions.15
In summary a novel cascade cyclodehydration/cycload-
dition/aromatization sequence has been developed to
(13) Interestingly both the one carbon shorter and longer alkenyl
tether homologues of 16b reported by Padwa, A. (see ref 4) were found to
be completely unreactive towards cycloaddition under a variety of
conditions, returning only unreacted oxazole.
(14) An alternative mechanism in which fused pyridine product 8
simply undergoes ring opening via nucleophilic attack by adventitious
water under the thermal conditions was ruled out by heating purified
pyridine 1 to 120 °C in toluene for 24 h. No change in the HPLC purity or
NMR profile was observed. Similarly, treating isolated oxazole 2 with
Tf2O/pyr led to exclusive formation of pyridine 1, albeit with a slow
reaction rate compared to the in situ method. Interestingly, treatment of
pure oxazole 2 with 1 equiv of triflic acid gave an 80:20 mixture of
pyridine 1 and 3-hydroxy pyridine 17a after 3 h. It is hypothesized that
adventitiouswater in the triflic acid lead to hydrolysis of pyridine 1, as no
attempts were made to rigorously exclude water.
Acknowledgment. The authors thank the analytical
sciences department at GlaxoSmithKline for support and
analysis in structure determination.
Supporting Information Available. A general procedure
along with 1H, 13C NMR and HRMS for all compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(15) (a) Kondrat’eva, G. Y Khim. Nauka Prom. 1957, 2, 666. (b)
Kondrat’eva, G. Y. Chem. Abstr. 1958, 52, 6345. (c) Lakhan, R.; Ternai,
B. Advances in Heterocyclic Chemistry; Katritzky, A. R., Boulton, A. J.,
Eds.; Academic Press: New York, NY, 1974; Vol. 17, Chapter 4, p 100.
The authors declare no competing financial interest.
Org. Lett., Vol. 14, No. 6, 2012
1607