A catalytic route to ampakines and their derivatives

Article abstract of DOI:10.1021/ol200111a

A catalytic domino reaction that efficiently provides access to an important class of heterocycles, the ampakines, is reported. Our approach is based on the cobalt-catalyzed hydroformylation of dihydrooxazines and allows for the facile synthesis of the pharmaceutically interesting compound CX-614 and related substances.

Full text of DOI:10.1021/ol200111a

ORGANIC
LETTERS
2011
Vol. 13, No. 6
1426–1428
A Catalytic Route to Ampakines and
Their Derivatives
Michael Mulzer and Geoffrey W. Coates*
Department of Chemistry and Chemical Biology, Baker Laboratory, Cornell University,
Ithaca, New York 14853-1301, United States
gc39@cornell.edu
Received January 13, 2011
ABSTRACT
A catalytic domino reaction that efficiently provides access to an important class of heterocycles, the ampakines, is reported. Our approach is
based on the cobalt-catalyzed hydroformylation of dihydrooxazines and allows for the facile synthesis of the pharmaceutically interesting
compound CX-614 and related substances.
Neurodegenerative diseases such as Alzheimer’s or Par-
kinson’s are on the rise globally and will lead to substantial
financial and societal costs if effective treatments are not
found in the near future.1 One promising therapy currently
under investigation is the administration of ampakines.2
These compounds allow positive allosteric modulation
of R-amino-3-hydroxy-5-methyl-4-isoxazole propionate
(AMPA) receptors and could help alleviate the symptoms
of these diseases by restoring diminished glutamatergic
neurotransmission.2b,d Important examples of ampakines
are CX-5543 and CX-6144 (Scheme 1), and their effects on
cell receptors and potential medicinal applications are well
documented. Structurally related to CX-614, 3-sub-
stituted-[1,2,3]-benzotriazinones, derived from simple am-
pakines (Scheme 1), also show potential for the treatment
of neurodegenerative diseases by modulating AMPA re-
ceptors at even nanomolar concentrations.5a
Scheme 1. Examples of Ampakines
(1) Forman, M. S.; Trojanowski, J. Q.; Lee, V. M.-Y. Nat. Med. 2004,
10, 1055.
(2) (a) Arai, A. C.; Kessler, M. Curr. Drug Targets 2007, 8, 583. (b)
O’Neill, M. J.; Bleackman, D.; Zimmerman, D. M.; Nisenbaum, E. S.
CNS Neurol. Disord.: Drug Targets 2004, 3, 181. (c) O’Neill, M. J.;
Witkin, J. M. Curr. Drug Targets 2007, 8, 603. (d) Yamada, K. A.
Neurobiol. Disease 1998, 5, 67. (e) Lynch, G. Curr. Opin. Pharmacol.
2006, 6, 82.
As part of our research efforts to use carbon monoxide
as a feedstock for the preparation of synthetically useful
(3) Arai, A.; Kessler, M.; Ambros-Ingerson, J.; Quan, A.; Yigiter, E.;
Rogers, G.; Lynch, G. Neuroscience 1996, 75, 573.
(4) (a) Jourdi, H.; Hsu, Y.-T.; Zhou, M.; Qin, Q.; Bi, X.; Baudry, M.
J. Neurosci. 2009, 29, 8688. (b) Jin, R.; Clark, S.; Weeks, A. M.; Dudman,
J. T.; Gouaux, E.; Partin, K. M. J. Neurosci. 2005, 25, 9027. (c) Arai, A.;
Kessler, M.; Rogers, G.; Lynch, G. Mol. Pharmacol. 2000, 58, 802.
(d) Dicoua, E.; Rangona, C.-M.; Guimiota, F.; Spedding, M.; Gressens,
P. Brain Res. 2003, 970, 221. (e) Jourdi, H.; Hamo, L.; Oka, T.; Seegan, A.;
Baudry, M. Neuropharmacology 2009, 56, 876.
(5) (a) Mueller, R.; Lee, S.; O’Hare, S.; Rogers, G.; Rachwal, S.;
Street, L. US Pat. Appl. 2010/0041647 A1. (b) Rogers, G.; Lynch, G. Int.
Pat. Appl. WO 97/36907.
(6) Selected articles: (a) Rowley, J. M.; Lobkovsky, E. B.; Coates,
G. W. J. Am. Chem. Soc. 2007, 129, 4948. (b) Schmidt, J. A. R.;
Lobkovsky, E. B.; Coates, G. W. J. Am. Chem. Soc. 2005, 127, 11426.
r
10.1021/ol200111a
Published on Web 02/16/2011
2011 American Chemical Society

intermediates,6 we recently reported the carbonylative ring
expansion of 2-substituted oxazolines to oxazinones7,8 and
hydroformylation of 2-oxazolines to N-acylated amino-
aldehydes.9 Based on these findings we proposed that the
in situ generation of an aminoaldehyde moiety followed by
cyclization could produce the desired ampakine targets or
synthetic precursors shown in Scheme 1. This methodol-
ogy would be very atom economical, as opposed to current
synthetic routes which use hard-to-obtain precursors,5
suffer from low yields,10 or require stoichiometric,11
toxic,12 or expensive13 reagents. In this report we describe
a domino reaction14 that readily yields the desired ampa-
kine framework while using simple starting materials and
Co2(CO)8 as an inexpensive precatalyst.
isolated yield. Highly polar and nonpolar solvents both
completely impeded the reaction (entries 1 and 2). Low-
ering the reaction temperature or the overall pressure
decreased the yield (entries 6 and 7). Furthermore, adding
˚
molecular sieves (3 A) to the reaction mixture led to no
improvement, suggesting that the water produced in the
reaction has no deleterious effect. Lastly, the transforma-
tion scales up well with virtually no loss in isolated yield
(entry 5).
Table 2. Hydroformylation of Substituted 2-Aryl-dihydrooxa-
zines
Table 1. Influence of Solvent, Pressure, and Temperature
entry
R1
R2
R3
product
yield (%)a
65
70, 73b
68, 87b
72b
72b
43, 73,b 74c
79
76b
40, 71b
64, 80b
1
2
3
4
5
6
7
8
9
10
H
H
H
H
H
H
Br
H
H
H
H
H
F
Me
OMe
Ph
H
2b
2c
2d
2e
2f
H
pressure
(psi)
temp
H
entry
solvent
MeCN
(°C)
yield (%)a
Me
OMe
H
1
2
3
4
5
6
7
1000
1000
1000
1000
1000
1000
800
80
80
80
80
80
60
80
<1b
<1b
50
H
2g
2h
2i
H
n-Hexane
THF
(-OCH2O-)
(-OCH2CH2O-)
(-CHdCH-)2
2j
1,4-Dioxane
PhMe
70
81, 81c
23b
61b
2k
PhMe
a Isolated yield for reactions carried out on a 0.5 mmol scale.
b [Substrate] = 0.12 M, 0.4 mmol scale. c 8 mol % Co2(CO)8,
[substrate] = 0.12 M, 0.4 mmol scale.
PhMe
a Isolated yield for reactions carried out on a 0.5 mmol scale.
b Determined by 1H NMR spectroscopy. c 7.5 mmol scale.
Next, we subjected a variety of substituted dihydroox-
azines to the optimized conditions (Table 2). The intro-
duction of either electron-withdrawing (entry 1) or
electron-donating (entries 3, 6, 9, and 10) substituents
onto the aryl moiety of the dihydrooxazine decreased the
isolated yield despite complete consumption of the corre-
sponding starting materials.16 The decrease in yield in
entry 1 is consistent with observations made by Jia and co-
workers for the carbonylative ring expansion of aryl-
substituted 2-oxazolines.8 In entries 3, 6, 9, and 10 the
electron-donating substituents should increase the nucleo-
philicity of the dihydrooxazine which could facilitate
initiation of unproductive ring-opening polymerization,
a well-known reaction of oxazolines and oxazines.17 Initia-
tion of polymerization requires the attack of a substrate
molecule on an already activated substrate molecule
(Scheme 2, path A: SN2 attack by 1a in the place of Co
(CO)4-; and path B: attack by 1a at Co acyl).17a Conse-
quently, we decided to decrease the overall substrate con-
centration to favor the desired hydroformylation pathway.
We hypothesized that a dihydrooxazine bearing an
ortho-substituted phenol would give the desired ampakine
structure following stepwise hydroformylation and cycli-
zation. To test this theory, a model substrate was synthe-
sized and subjected to hydroformylation conditions in the
presence of catalytic amounts of Co2(CO)8 (Table 1).
Solvents of different polarities and Lewis basicities were
investigated as these parameters are crucial in related
carbonylation reactions.15 Toluene proved to be the best
solvent (entry 5), yielding the desired product in 81%
(7) Byrne, C. M.; Church, T. L.; Kramer, J. W.; Coates, G. W.
Angew. Chem., Int. Ed. 2008, 47, 3979.
(8) For a prior example of carbonylative ring expansion of 2-oxazo-
lines, see: Xu, H.; Jia, L. Org. Lett. 2003, 5, 1575.
(9) Laitar, D. S.; Kramer, J. W.; Whiting, B. T.; Lobkovsky, E. B.;
Coates, G. W. Chem. Commun. 2009, 5704.
€
€
(10) Bohme, H.; Boing, H. Arch. Pharm. 1961, 294, 556.
ꢀ
(11) Cayley, A. N.; Cox, R. J.; Menard-Moyon, C.; Schmidt, J. P.;
Taylor, R. J. K. Tetrahedron Lett. 2007, 48, 6556.
(12) Takacs, J. M.; Helle, M. A. Tetrahedron Lett. 1989, 30, 7321.
(13) Chiou, W.-H.; Mizutani, N.; Ojima, I. J. Org. Chem. 2007, 72,
1871.
(14) Tietze, L. F. Chem. Rev. 1996, 96, 115.
(15) Church, T. L.; Getzler, Y. D. Y. L.; Coates, G. W. J. Am. Chem.
Soc. 2006, 128, 10125.
(16) Based on 1H NMR spectroscopy or TLC analysis of the crude
reaction mixture.
(17) (a) Saegusa, T.; Ikeda, H.; Fujii, H. Macromolecules 1973, 6, 315.
(b) Culbertson, B. M. Prog. Polym. Sci. 2002, 27, 579.
Org. Lett., Vol. 13, No. 6, 2011
1427

Table 3. Hydroformylation of Alkyl-Substituted 2-Dihydroox-
azines
Scheme 2. Proposed Catalytic Cycle
entry
R1
R2
R3
product
yield (%)a
1
2
3
4
Me
Ph
Me
H
H
H
2l
66, 90b (1.6:1)c
80 (2.9:1)c
93
H
H
2m
2n
2o
Me
H
H
Me
80 (1.3:1)c
a Isolated yield for reactions carried out on a 0.5 mmol scale.
b [Substrate] = 0.12 M, 0.4 mmol scale. c Diastereomeric ratio (cis/trans).
dihydrooxazine through protonation and subsequent ring
opening via an SN2-type pathway should give rise to a
transient cobalt-alkyl species. Insertion of CO19 followed
by hydrogenolysis of the resulting cobalt-acyl intermedi-
ate20,21 produces an N-acylaminoaldehyde. The presence
of the Brønsted acid HCo(CO)4 should then facilitate
isomerization to the corresponding hemiaminal and sub-
sequent formation of an iminium ion by loss of water. A
fast second cyclization to the final product is likely given
the close vicinity of the phenol group to the iminium ion.
In conclusion, we have developed an atom economical
route to an important ampakine framework via hydro-
formylation of dihydrooxazines. Our methodology provides
quick access to the pharmaceutically interesting compound
CX-614 (Table 2, entry 9) and to a central building block
for the synthesis of 3-substituted-[1,2,3]-benzotriazinones
(Table 2, entry 5). In both cases, the modular nature of the
starting materials allows for the synthesis of a wide range of
derivatives.
This modification of the reaction conditions led to a pro-
nounced increase in yield for substrates with electron-
donating substituents (entries 3, 6, and 10) and furnished
CX-554 (entry 8), CX-614 (entry 9), and the intermediate
needed for the synthesis of benzotriazinones (entry 5) in
good yield.
Modifications to the dihydrooxazine backbone were
well tolerated (Table 3, entries 1-4).18 The observed
diastereomeric ratios were relatively small, possibly due
to the flexibility of the five-membered ring and the dis-
tance between the existing and the newly formed stereo-
center. The position of the substituent had little effect on
the ratio (entries 1 and 4), whereas changing the size of the
substituent from methyl to phenyl roughly doubled the
stereoselectivity (entries 1 and 2). Furthermore, the dia-
stereomeric ratios seem to be primarily kinetically con-
trolled. For example, in entry 1 the cis-isomer forms in
23% excess; a pure sample of this diastereomer is recov-
ered unchanged when resubjected to our hydroformyla-
tion conditions but epimerizes to approximately a 1:1
mixture when exposed to p-toluenesulfonic acid at 40
°C. Consequently, the two diastereomers of a given sub-
strate seemingly do not interconvert under our reaction
conditions.
Acknowledgment. We thank the Department of Energy
for financial support (DE-FG02-05ER15687).
Supporting Information Available. Crystallographic
data, experimental procedures, and characterization
data. This material is available free of charge via the
Internet at http://pubs.acs.org.
Based on previous literature reports,7,8,13 we expect the
first part of the domino reaction to proceed by the cata-
lytic cycle depicted in Scheme 2. First, activation of the
ꢀ
ꢀ
ꢀ
(20) Ungvary, F.; Laszlo, M. Organometallics 1983, 2, 1608.
(21) Using in situ IR spectroscopy, we observed the buildup and
subsequent persistence of an absorption peak centered around 1740
cm-1. Attributing this peak to the cobalt-acyl species (cf. ref 15 and
references therein), we suspect that the hydrogenolysis reaction might be
the rate-determining step in our proposed catalytic cycle.
(18) Introduction of a methyl or phenyl group onto the oxygen-
bearing methylene group led to low yields or a mixture of products,
respectively.
(19) Roe, C. D. Organometallics 1987, 6, 942.
1428
Org. Lett., Vol. 13, No. 6, 2011