Synthesis of the louisianin alkaloid family via a 1,2,4-triazine inverse-electron-demand Diels-Alder approach

Article abstract of DOI:10.1021/jo901761r

Chemical Equation Presented) Isolated in 1995, the four members of the louisianin family (A, B, C and D) are simple pyridine and 2-pyridone alkaloids that display both antibacterial and anticancer activity. Herein we describe the synthesis of all four members of the louisianin family, from a conveniently prepared 1,2,4-triazine and via a common tetrasubstituted pyridine intermediate. This study includes the synthesis of louisianin B in both racemic form and as the (-)-enantiomer. 2009 American Chemical Society.

Full text of DOI:10.1021/jo901761r

pubs.acs.org/joc
Synthesis of the Louisianin Alkaloid Family via a 1,2,4-Triazine
Inverse-Electron-Demand Diels-Alder Approach
Nicola Catozzi,† Michael G. Edwards,† Steven A. Raw,‡ Pierre Wasnaire,† and
Richard J. K. Taylor*,†
‡
†Department of Chemistry, University of York, Heslington, York YO10 5DD, U.K., and Process Research
and Development, AstraZeneca, Silk Road Business Park, Charter Way, Macclesfield, Cheshire SK10 2NA, U.K.
rjkt1@york.ac.uk
Received August 17, 2009
Isolated in 1995, the four members of the louisianin family (A, B, C and D) are simple pyridine and
2-pyridone alkaloids that display both antibacterial and anticancer activity. Herein we describe the
synthesis of all four members of the louisianin family, from a conveniently prepared 1,2,4-triazine
and via a common tetrasubstituted pyridine intermediate. This study includes the synthesis of
louisianin B in both racemic form and as the (-)-enantiomer.
Introduction
for supramolecular chemistry6 and as ligands for organome-
tallic chemistry has been amply demonstrated.7
Pyridines and pyridones are ubiquitous compounds in
nature and privileged structures in medicinal chemistry and
agrochemistry. For example, the common drugs Atazanavir
(Bristol-Myers-Squibb, antiretroviral HIV)1 and Esomepra-
zole (AstraZeneca, proton pump inhibitor)2 and the pesti-
cide Imidacloprid (Baeyer)3 all feature pyridine cores in
various forms. Pyridine alkaloids also represent a huge class
of natural products encompassing many structures, from the
relatively simple epibatidine4 to the complex cyclothiazamy-
cin B1.5 Moreover, the utility of pyridines as building blocks
The importance of the pyridine core has led to the develop-
ment of many synthetic approaches to pyridines (and dihydro-
pyridines).8 Classically, pyridines have been prepared via
condensation reactions as exemplified by the well-known
Hantzsch synthesis of dihydropyridines from 1,5-dicarbonyl
compounds and ammonia.9 Other important condensation
approaches to pyridines include the Bohlmann-Rahtz reac-
tion10 (recently expanded by Bagley and co-workers11) and
the Krohnke reaction.12 More contemporary approaches to
€
pyridines have also been described. The [2 þ 2 þ 2]-cobalt-
catalyzed cycloaddition of 2 equiv of alkyne with a nitrile
€
*To whom correspondence should be addressed. Tel: þ44-(0)1904-
432606. Fax: þ44-(0)1904-434523.
developed by Bonnenmann and Vollhardt has proved use-
ful,13 especially when used in an intramolecular context.14
(1) (a) Bold, G.; Fassler, A.; Capraro, H. G.; Cozens, R.; Klimkait, T.;
Lazdins, J.; Mestan, J.; Poncioni, B.; Rosel, J.; Stover, D.; Tintelnot-
Blomley, M.; Acemoglu, F.; Beck, W.; Boss, E.; Eschbach, M.; Hurlimann,
T.; Masso, E.; Roussel, S.; Ucci-Stoll, K.; Wyss, D.; Lang, M. J. Med. Chem.
1998, 41, 3387. (b) Xu, Z. M.; Singh, J.; Schwinden, M. D.; Zheng, B.;
Kissick, T. P.; Patel, B.; Humora, M. J.; Quiroz, F.; Dong, L.; Hsieh, D. M.;
Heikes, J. E.; Pudipeddi, M.; Lindrud, M. D.; Srivastava, S. K.; Kronenthal,
D. R.; Mueller, R. H. Org. Process Res. Dev. 2002, 6, 323.
(6) See, for example: (a) Ward, M. D.; McCleverty, J. A.; Jeffery, J. C.
Coord. Chem. Rev. 2001, 222, 251. (b) Mamula, O.; von Zelewsky, A. Coord.
Chem. Rev. 2003, 242, 87.
(7) Kwong, H.-L.; Yeung, H.-L.; Yeung, C.-T.; Lee, W.-S.; Lee, C.-S.;
Wong, W.-L. Coord. Chem. Rev. 2007, 251, 2188.
(2) Cotton, H.; Elebring, T.; Larsson, M.; Li, L.; Sorensen, H.; von Unge,
S. Tetrahedron: Asymmetry 2000, 11, 3819.
(8) For a general review, see: Henry, G. D. Tetrahedron 2004, 60, 6043.
(9) Phillips, A. P. J. Am. Chem. Soc. 1949, 71, 4003.
(3) (a) Shiokawa, K.; Tsuboi, S.; Kagabu, S.; Moriya, K. Heterocyclic
compounds. EP 192060, August 1986. (b) Shiokawa, K.; Tsuboi, S.; Kagabu,
S.; Moriya, K. Chem. Abstr. 1986, 106, 28848.
(4) Spande, T. F.; Garraffo, H. M.; Edwards, M. W.; Yeh, H. J. C.;
Pannell, L.; Daly, J. W. J. Am. Chem. Soc. 1992, 114, 3475.
(5) Hashimoto, M.; Murakami, T.; Funahashi, K.; Tokunaga, T.; Nihei,
K.; Okuno, T.; Kimura, T.; Naoki, H.; Himeno, H. Bioorg. Med. Chem.
2006, 14, 8259.
(10) Bohlmann, F.; Rahtz, D. Chem. Ber. 1957, 90, 2265.
(11) See, for example: (a) Bagley, M. C.; Glover, C.; Merritt, E. A. Synlett
2007, 2459. (b) Bagley, M. C.; Chapaneri, K.; Dale, J. W.; Xiong, X.; Bower,
J. J. Org. Chem. 2005, 70, 1389.
€
(12) Krohnke, F. Synthesis 1976, 1.
(13) Bonnemann, H.; Brijoux, W. In Aspects of Homogeneous Catalysis;
€
Ugo, R., Ed.; D. Reidel: Dordrecht, 1984; Vol. 5, pp 75-196.
(14) Naiman, A.; Vollhardt, K. P. C. Angew. Chem., Int. Ed. 1977, 16, 708.
DOI: 10.1021/jo901761r
r
Published on Web 09/25/2009
J. Org. Chem. 2009, 74, 8343–8354 8343
2009 American Chemical Society

Catozzi et al.
JOCArticle
Recently, [4 þ 2]-cycloadditions have also been popularized
for the synthesis of pyridines. For example, Ghosez and co-
workers15 and Moody and co-workers16 have demonstrated
the use of an inverse-electron-demand hetero-Diels-Alder
sequence involving aza-dienes and alkynes to afford either
pyridines or pyridones depending upon the constitution of
the aza-diene. Furthermore, the use of triazines as masked
aza-dienes in [4 þ 2]-inverse-electron-demand Diels-Alder
(IED-DA) reaction sequences has been successfully ex-
ploited (vide infra).
To date, there have been five reports on the synthesis of
members of the louisianin family, including our own pre-
vious communication on the synthesis of louisianins C and
D.19 Prior to any total syntheses, as part of the structural
elucidation studies in 1997, Omura and co-workers20 de-
scribed the chemical interconversion of natural louisianin A
to louisianin C (and then subsequently into louisianin D). A
reaction sequence involving formation of the chloropyridine
from the pyridone (with phosphorus oxychloride) and chlor-
ide reduction (with zinc/hydrochloric acid) afforded louisia-
nin C, and then isomerization of the allyl group in louisianin
C to the 1-propenyl group of louisianin D was accomplished
with DBU. However, it was not until 2003 that Kelly and co-
workers described the first total synthesis of one of the family
members, louisianin C (3).21 Their nine-step (11% overall
yield) synthesis featured a key intramolecular, fluoride-
induced silylpyridine to aldehyde cyclization to form the
cyclopentane ring. Subsequently, in 2006, Chang and co-
workers reported the first total synthesis of louisianin A (1)
in seven steps (24% overall yield).22 In a manner similar to
that for the Kelly route, annulation of the cyclopentane ring
was achieved at a late stage of the synthesis via a Dieck-
mann-Thorpe cyclization. A recent publication from Chang
and co-workers extended this Dieckmann-Thorpe cycliza-
tion approach to the synthesis of louisianins C and D23 and
utilized an orthogonal cross-coupling strategy to install
either the C-5 allyl or 1-propenyl group of louisianins C
and D, respectively, without recourse to an isomerization
step; this enabled both louisianins C and D to be prepared in
22% and 20% yield, respectively, for the seven overall steps.
These first three approaches, from different commercially
available pyridines, are characterized by a stepwise synthesis
featuring a late-stage annulation. An alternative approach
was conceived by Chen and co-workers in which an early
stage [3 þ 3]-annulation of ethyl cyclopentene-1-carboxylate
was used to construct an aza-bicyclic system which ulti-
mately gave louisianin D in 10 steps (20% overall yield).24
It is noteworthy that, to date, louisianins A, C, and D have all
been prepared, but there are no literature reports concerning
the synthesis of louisianin B.
CHART 1
Isolated in 1995 by Omura and co-workers17,18 from the
cultured broth of Streptomyces sp. WK-4028 obtained
from a soil sample collected in Louisiana, the louisianins
A-D (Chart 1, 1-4) are a family of simple pyridine
and pyridone alkaloids. A common core structure with
a cyclopentane ring fused to a pyridine in the C3-C4
position is present in all of the family members. The
most notable difference between the family members is
the oxidation level at the C-2 position. Whereas louisia-
nins C and D are pyridines, louisianins A and B are
2-pyridones. Other differences can be found in either
the C-5 substituent [louisianin D (4) contains a 1-propenyl
group rather than the allyl group present in the other
three members] or the oxidation level of the C-4 benzylic
position [louisianin B (2) contains a secondary alcohol
whereas the other three members exist as ketones]. It is
noteworthy that the stereochemistry of the secondary
alcohol center in louisianin B was not reported by Omura
and co-workers and thus remains undefined. Though
the biological properties of the louisianins have been
explored only in an initial screening, some interesting
biological properties have been identified. For instance,
louisianin A proved to be a potent inhibitor of testoster-
one-responsive carcinoma SC115, while louisianins C and
D were potent suppressors of cultured vascular endo-
thelial cells.17
In contrast to the other reported approaches, our synthetic
route to the louisianins is based upon the use of 1,2,4-
triazines as suitable aza-dienes for construction of the pyr-
idine core via [4 þ 2]-cycloaddition chemistry. 1,2,4-Tria-
zines25,26 have been known since the initial report of Bischler
in 1889,27 and although some natural products featuring this
template are known,25 such compounds generally repre-
sented chemical curiosities until the realization that they
were capable electron-deficient dienes in Diels-Alder chem-
istry (with electron-rich olefins). Though the first report of
(19) Catozzi, N.; Wasnaire, P.; Taylor, R. J. K. Tetrahedron Lett. 2008,
49, 2865.
(20) Sunazuka, T.; Zhi-Ming, T.; Harigaya, Y.; Takamatsu, S.; Hayashi,
M.; Komiyama, K.; Omura, S. J. Antibiot. 1997, 50, 274.
(21) Beierle, J. M.; Osimboni, E. B.; Metallinos, C.; Zhao, Y.; Kelly,
T. R. J. Org. Chem. 2003, 68, 4970.
(15) Ghosez, L.; Jnoff, E.; Bayard, P.; Sainte, F.; Beaudegnies, R.
Tetrahedron 1999, 55, 3387.
(16) (a) Fletcher, M. D.; Hurst, T. E.; Miles, T. J.; Moody, C. J.
Tetrahedron 2006, 62, 5454. (b) Hurst, T. E.; Miles, T. J.; Moody, C. J.
Tetrahedron 2008, 64, 874.
(17) Isolation and biological properties: Komiyama, K.; Takamatsu, S.;
Kim, Y. P.; Matsumoto, A.; Takahashi, Y.; Hayashi, M.; Woodruff, H. B.;
Omura, S. J. Antibiot. 1995, 48, 1086.
(18) Physiochemical properties and characterization data: Takamatsu,
S.; Kim, Y. P.; Hayashi, M.; Furuhata, K.; Takayanagi, H.; Komiyama, K.;
Woodruff, H. B.; Omura, S. J. Antibiot. 1995, 48, 1090.
(22) Chang, C. Y.; Liu, H. M.; Chow, T. J. J. Org. Chem. 2006, 71, 6302.
(23) Chang, C. Y.; Liu, H. M.; Hsu, R. T. Tetrahedron 2009, 65, 748.
(24) Chen, H. W.; Hsu, R. T.; Chang, M. Y.; Chang, N. C. Org. Lett.
2006, 8, 3033.
(25) Lindsey, C. W.; Layton, M. E. In Science of Synthesis; Weinreb, S.
M., Ed.; Thieme: Stuttgart, 2004; Vol. 17.2.2, pp 357-447.
€
(26) Neunhoffer, H.; Wiley, P. F. Chemistry of 1,2,3-Triazines and 1,2,4-
Triazines, Tetrazines and Pentazines; Wiley: New York, 1978.
(27) Bischler, A. Chem. Ber. 1889, 22, 2801.
8344 J. Org. Chem. Vol. 74, No. 21, 2009

Catozzi et al.
JOCArticle
an IED-DA with a 1,2,4-triazine, to afford pyridines and
SCHEME 2
€
€
dihydropyridines, was described by Neunhoffer and Fruhauf
in 1969,28 this chemistry was plagued by dubious regioselec-
tivity and modest yields. However, the rigorous study of
Boger and Panek established that the reaction of pyrrolidine
enamines with triazines provides an efficient and regioselec-
tive route to pyridines and dihydropyridines via reaction
across the C3-C6 diene (Scheme 1).29 Boger and co-workers
also showed that the enamines required can be conveniently
prepared in situ from the desired amine and ketone substrate.
Further work established robust procedures for the synthesis
of 1,2,4-triazines and utilized these IED-DA reactions with
electron-rich olefins in natural product synthesis.30,31
SCHEME 1
interested in the preparation of the louisianin family of
alkaloids. The synthesis of louisianins C and D was reported
in a prior communication.19 Herein we present our full
findings on the preparation of all four members of the
louisianin family by a unified synthetic approach.
Results and Discussion
Retrosynthesis. We envisaged that the preparation of all
four of the louisianin family members could be achieved
from a single suitably functionalized intermediate; this led us
to the retrosynthetic analysis illustrated in Scheme 3. The key
intermediate in such a plan would be the tetrasubstituted
pyridine ester 9. Ester 9 has the potential to undergo both
reductive transformation (toward louisianins C and D) and
oxidative transformation (toward louisianins A and B).
Thus, it was postulated that the ester group of pyridine 9
could undergo hydrolysis and subsequent decarboxylation
to afford the reduced pyridine 8, the intermediate necessary
to synthesize louisianins A and B. Conversely, it seemed
likely that oxidative transformation of the ester group via
one of several potential methods (Dakin oxidation,36
Baeyer-Villiger oxidation,36 or carboxy inversion37) would
give rise to the pyridone oxidation level necessary to prepare
louisianins A and B via pyridone 7. A series of common steps
would then lead to the natural products. First, the key
oxygenation in the C-4 benzylic position was to be achieved
via the anionic oxidation method of Chen,24 and then
the allyl group was to be revealed via sulfide oxidation/
elimination. The final steps would then be standard oxida-
tion and/or isomerization steps. We anticipated that the
required tetrasubstituted pyridine 9 could be prepared
via the IED-DA/retro-Diels-Alder/elimination reaction
between 3,6-substituted-1,2,4-triazine 10 and the enamine
derived from cyclopentanone and pyrrolidine; this required
the preparation of 1,2,4-triazine 10, which we anticipated
should arise via the oxidative cyclocondensation of the novel
R-hydroxyketone 12 and amidrazone 11.31
We investigated the use of the triazine IED-DA chemistry
methodology to prepare highly substituted pyridines but
noticed a reluctance of the elimination/aromatization to
proceed, and aminodihydropyridines 5 were obtained. A
tethered imine-enamine approach was therefore developed
in which a secondary amine linked to a primary amine via a
two-carbon spacer provides an imine-enamine 6 when
treated with the requisite ketone. Upon completion of the
IED-DA/retro-DA sequence, the resulting dihydropyran
contains a second basic center which facilitates the required
elimination/aromatization (Scheme 2, eq a).32 Further stu-
dies into the development of a one-pot pyridine synthesis
resulted in the identification of a rapid, solvent-free, micro-
wave-mediated variant of the Boger reaction (Scheme 2,
eq b).33 More recently, we have also disclosed a more practical
procedure which utilizes a silica-mediated aromatization/
elimination protocol (Scheme 2, eq c).34 These studies
prompted our interest in the application of this triazine
methodology to natural product targets. First, it was realized
that the dihydropyridine 5 produced from the initial IED-DA
reaction could function as an aza-diene for a second IED-DA
reaction producing complex tetracycles.35 Second, we became
€
€
(28) Neunhoffer, H.; Fruhauf, H.-W. Tetrahedron Lett. 1969, 10, 3151.
(29) Boger, D. L.; Panek, J. S. J. Org. Chem. 1981, 46, 2179.
(30) (a) Boger, D. L.; Panek, J. S.; Meier, M. M.; Duff, S. R. J. Org. Chem.
1982, 47, 895. (b) Boger, D. L.; Panek, J. S. J. Am. Chem. Soc. 1985, 107, 5745.
(31) Boger, D. L.; Panek, J. S.; Yasuda, M. Org. Synth. 1988, 66, 142.
(32) Raw, S. A.; Taylor, R. J. K. Chem. Commun. 2004, 508.
(33) Sainz, Y. F.; Raw, S. A.; Taylor, R. J. K. J. Org. Chem. 2005, 70, 10086.
(34) Catozzi, N.; Bromley, W. J.; Wasnaire, P.; Gibson, M.; Taylor, R. J.
K. Synlett 2007, 2217.
Preparation of 1,2,4-Triazine 10 and Tetrasubstituted Pyri-
dine 9. In order to construct triazine 10, the R-hydroxy ketone
12 was required; the two-step preparation of this fragment
is illustrated in Scheme 4. Ultimately, the preparation of 12
(35) (a) Raw, S. A.; Taylor, R. J. K. J. Am. Chem. Soc. 2004, 126, 12260.
(b) Bromley, W. J.; Gibson, M.; Lang, S.; Raw, S. A.; Whitwood, A. C.;
Taylor, R. J. K. Tetrahedron 2007, 63, 6004.
(36) Krow, G. R. Org. React. 1993, 43, 251.
(37) Denney, D. B.; Sherman, N. J. Org. Chem. 1965, 30, 3760.
J. Org. Chem. Vol. 74, No. 21, 2009 8345

Catozzi et al.
JOCArticle
SCHEME 3
was achieved via an alkyne oxidation and subsequent sub-
stitution reaction; however, the preparation of the desired
R-hydroxy-γ0-haloketone proved to be nontrivial. Although
many methods are known for the preparation of R-hydro-
xyketones, including the bis(trifluoroacetoxy)iodobenzene
(PIFA) oxidation of methyl ketones38 and the Rubottom
oxidation of silyl enol ethers,39 the ready availability of
5-bromopentene and 5-chloropentyne suggested the oxida-
tion of either of these precursors. Both the RuO4-40 and
KMnO4-promoted41 oxidations of alkenes have been re-
ported in the literature for the convenient preparation of
R-hydroxy ketones. Unfortunately, application of either
method to the oxidation of 5-bromopentene proved unsuc-
cessful, and the rapid formation of considerable amounts
of polymeric material was observed. In contrast, the hyper-
valent iodine (PIFA)-promoted oxidation of 5-chloropen-
tyne 13 under the conditions reported by Tamura and
co-workers42 afforded R-hydroxy-γ0-chloroketone 14. It is
noteworthy that an aqueous bicarbonate quench was re-
quired to hydrolyze the large proportion of the trifluoroa-
cetate ester formed during the reaction; the method
described by the authors of passing the material through
silica gel had no effect. Chloride 14 proved somewhat
unstable, and any attempts at purification resulted in a
substantial loss of material. Given the presence of an
electrophile and nucleophile in the same molecule, this is,
perhaps, unsurprising! However, chloride 14, without pur-
ification, could be readily subjected to an SN2 displacement
with potassium thiophenolate. This reaction readily pro-
ceeded in a straightforward fashion in aqueous solution to
afford the desired R-hydroxyketone 12 in a 60% yield over
the two steps.
With R-hydroxyketone 12 in hand, we were then able to
examine the preparation of the 3,6-disubstituted 1,2,4-tria-
zine 10. In order to obtain the required triazine regioisomer
(i.e., the desired 3,6-regioisomer vs the alternative 3,5-
regioisomer), the triazine construction had to be carried
out sequentially with the initial formation of amidrazone
hydrazone 15 followed by oxidative cyclocondensation
(Scheme 5).43 Thus, amidrazone 11 (prepared according to
the procedure of Boger and Panek)31 and R-hydroxyketone
12 were condensed in ethanolic solution to afford the desired
hydrazone 15. Compound 15 could be isolated and purified,
in 62% yield, as a 3:2 mixture of isomers (presumably
isomeric around the N-N bond), which rapidly interconvert
in solution. However, the polarity of this compound made it
more convenient to telescope the condensation with the
subsequent oxidative cyclocondensation. Toward this goal,
hydrazone 15 was heated in toluene solution in the presence
of 10 equiv of MnO2 to afford the desired triazine 10 in
75% optimal yield over the two steps. The regioselective
nature of this two-step process is noteworthy; none of the
SCHEME 4
(38) Moriarty, R. M.; Berglund, B. A.; Penmasta, R. Tetrahedron Lett.
1992, 33, 6065.
(39) Rubottom, G. M.; Vazquez, M. A.; Pelegrin, D. R. Tetrahedron Lett.
1974, 15, 4319.
(40) Plietker, B. J. Org. Chem. 2004, 69, 8287.
(41) Bonini, C.; Chiummiento, L.; Funicello, M.; Lupattelli, P.; Pullez,
M. Eur. J. Org. Chem. 2006, 80.
(42) Tamura, Y.; Yakura, T.; Haruta, J.; Kita, Y. Tetrahedron Lett. 1985,
26, 3837.
(43) Laphookhieo, S.; Jones, S.; Raw, S. A.; Sainz, Y. F.; Taylor, R. J. K.
Tetrahedron Lett. 2006, 47, 3865.
8346 J. Org. Chem. Vol. 74, No. 21, 2009

Catozzi et al.
JOCArticle
SCHEME 5
TABLE 1. Oxidative Cyclocondensation
entry
conditions
yield, %a
1
2
3
4
5
6
7
MnO2, PhMe, Δ
PDC, CH2Cl2, rt
TPAP, NMO, CH2Cl2, rt
RuCl2(Ph3P)3, TEMPO, 4 A ms, MeCN, O2, Δ
DMP, CH2Cl2, rt
20-75
0b
0b
0c
35
60d
0b
˚
SO3 Py, Et3N, DMSO/CH2Cl2, rt
3
DCC, TFA, DMSO/CH2Cl2, rt
aYield of chromatographically homogeneous material. bStarting
material decomposed. cNo reaction occurred. d4% of aldehyde 16 was
also obtained.
triazine 10 (entry 7); unfortunately, only decomposition was
observed.
With a reliable route to triazine 10 established, the IED-
DA/retro-DA/elimination sequence between the enamine
derived from cyclopentanone and pyrrolidine was investi-
gated (Scheme 6). We were delighted to obtain the required
pyridine 9 in excellent yield (80-89%) using both the micro-
wave and silica variants of the Boger procedure. The “one-
pot” microwave variant of this reaction33 proceeded smoothly
and efficiently on a 250 mg scale (80% yield, 1 h), although the
yield diminished with larger scale reactions. In contrast, the
thermal SiO2 procedure34 could be conducted on a multigram
scale with no loss in yield (89%). It should be noted, however,
that optimal yields were obtained by adding the silica gel after
the successful formation of the dihydropyridine and then
subsequently reheating the reaction mixture. As expected,
pyridine 9 was obtained as a single regioisomer as evidenced
by the observation of only one singlet (Ha, 1 H, 8.34 ppm) in
the aromatic region. With pyridine 9 in hand, the common
intermediate necessary for the preparation of all four louisia-
nin family members was available.
alternative 3,5-disubstituted 1,2,4-triazine regioisomer was
obtained. To our disappointment, however, the MnO2 reac-
tion proved to be capricious, and considerable variability
was observed in the yields obtained (20-75% yield). One
possible explanation for this is that MnO2 may oxidatively
decompose hydrazone 15 (or amidrazone 11 present in
equilibrium).44
Thus, we turned our attention to the development of a
more robust tandem oxidation-cyclization procedure
and examined the use of other oxidants (Table 1). Some-
what surprisingly, common oxidants such as PDC and
TPAP led only to decomposition of the starting material
(entries 2 and 3). Alternatively, the metal-catalyzed aerobic
oxidations, e.g., RuCl2(PPh3)3, that we have previously used
in tandem oxidation/heterocycle formation chemistry45
returned only starting material (entry 4). Although both
the Dess-Martin periodinane and Parikh-Doering oxi-
dation cleanly afforded triazine 10 (entries 5 and 6),
Parikh-Doering oxidation proved to be the method of
choice because of the superior yields obtained (repro-
ducibly >50%). Of interest was the identification of some
of the intermediate aldehyde 16 in the reaction mixture; from
the unpurified reaction mixture, a ratio of 4:1 triazine/
aldehyde was observed. However, although a small amount
of the aldehyde (4%) could be isolated from the reaction
mixture, the majority of the product was the cyclized triazine
10. This presumably indicates that a mild source of acid is
sufficient to catalyze cyclocondensation; indeed standing
aldehyde 16 in CDCl3 led to the formation of triazine 10.
With this in mind, the Pfitzner-Moffat oxidation46 of
hydrazone 15 was examined, with the view that the acidic
reaction conditions would favor the complete formation of
SCHEME 6
Louisianins C and D. In order to complete the synthesis of
louisianins C and D it was necessary to achieve the decarbo-
xyethylation of pyridine ester 9. The literature contains few
examples of the direct ethoxydecarbonylation of quinolines
or indoles bearing an ester adjacent to the nitrogen hetero-
atom and only a single example for pyridines.47 Normally,
(44) We have previously observed that this is a problem in tandem
triazine-forming reactions under microwave heating (ref 43).
(45) Robinson, R. S.; Taylor, R. J. K. Synlett 2005, 1003.
(46) Pfitzner, K. E.; Moffatt, J. G. J. Am. Chem. Soc. 1963, 85, 3027.
(47) Messinger, P.; Meyer-Barrientos, H. Liebigs. Ann. Chem. 1981, 2087.
J. Org. Chem. Vol. 74, No. 21, 2009 8347

Catozzi et al.
JOCArticle
SCHEME 7
decarboxylation is accomplished via a two-step saponifica-
tion/decarboxylation protocol. To this end, saponification
of ethyl ester 9 was readily accomplished by treatment with
potassium trimethylsilanolate giving acid 17 in 90% yield
(vide infra). The decarboxylation step is classically accom-
plished by strong heating either neat48 or in the presence of
copper powder49 or copper/quinoline.50 Unfortunately,
none of these decarboxylation methods proved successful
and returned only the carboxylic acid 17. In contrast, we
were delighted to discover that dilute aqueous hydrochloric
acid was able to effect the desired decarboxyethylation of
ester 9 directly in 83% yield (Scheme 7). The conditions for
this transformation were crucial to ensure that the decarbox-
yalkylation proceeded efficiently. In this regard, microwave
irradiation of ester 9 in 0.74% hydrochloric acid (1:50 37%
HCl/H2O) at 185 °C/50 W for 1 h were identified as the
optimum reaction conditions; in the presence of more con-
centrated acid the reaction product was 2-pyridinecarboxylic
acid 17. To the best of our knowledge, this conversion also
represents the first example of the microwave acceleration of
a direct pyridine decarboxylation.51
Benzylic oxidation was attempted next, and this was
potentially one of the most difficult steps because of the
paucity of available methods. However, Chen and co-work-
ers in their synthesis of louisianin D24 demonstrated that
oxidation of the cyclopentapyridine proceeded exclusively in
the C-4 position (para position) upon exposure to 6 equiv of
LHMDS followed by oxygenation. Despite the close simi-
larity of pyridine 8 to Chen’s substrate, we were disappointed
to find that, although oxidation did occur to afford a 2:1
mixture of isomeric alcohols, we could neither drive the
reaction to completion nor identify the nature of the isomeric
byproduct. However, application of Corey and Ensley’s52 in
situ hydroperoxide quench procedure, that is carrying out
the same reaction in the presence of triethyl phosphite, was
entirely successful. Under these conditions, a completely
regioselective oxidation was observed affording the desired
benzylic alcohol 18 in 80% yield (Scheme 7). The finely
balanced nature of this oxidation reaction was further illu-
strated in the completion of the louisianin A and B synthesis
(vide infra).
Having successfully achieved decarboxylation and oxida-
tion, the only functionality awaiting installation was the allyl
group in the C-5 position. This was readily revealed by
chemoselective oxidation of sulfide 18 to the corresponding
sulfoxide (m-CPBA, -78 °C, 1 h) and then subjection of the
“crude” sulfoxide to thermolysis (xylene, 150 °C) in the
presence of calcium carbonate to trap the extruded benze-
nesulfenic acid. This reaction cleanly afforded the desired
alkene 19 in a 76% yield over the two steps (Scheme 7).
Furthermore, no evidence for any isomerization of the allyl
group was observed under these conditions.
With the allyl group in place, we then undertook the final
oxidation necessary to obtain louisianin C (3). Omura’s work
on the interconversion of the louisianin family successfully
utilized the Jones oxidation for this transformation.20 Un-
fortunately, in our hands, this reaction proved to be low
yielding. Equally surprising was the observation that man-
ganese dioxide oxidation led only to decomposition of the
substrate. Ultimately, we were able to successfully use PCC
supported on alumina to achieve this oxidation in 80% yield
(Scheme 7). The product obtained displayed data identical to
those reported for louisianin C (3) [e.g., δC (CDCl3) 208.0; δH
(CDCl3) 8.76 (1 H, s, Ha) [lit.18 δC (CDCl3) 207.3; δH
(CDCl3) 8.75 (1 H, s, Ha)]].
Next, louisianin D (4) was prepared from louisianin C (3)
via isomerization of the allyl to the required 1-propenyl
substituent. Omura’s group had shown that this could be
accomplished by treatment of louisianin C with DBU in
benzene at reflux.20 In order to expedite the workup proce-
dure we attempted the use of polymer-supported methyl-
1,5,7-triazabicyclo[4.4.0]dec-5-ene (MTBD) but were sur-
prised to observe extensive decomposition and only obtained
a 10% yield of louisianin C (4). However, when heated with
MTBD in toluene (80 °C, 6 h), lousianin D (4) was obtained
(48) Stanforth, S. P.; Tarbit, B.; Watson, M. D. Tetrahedron 2004, 60,
8893.
(49) Kenyon, J.; Thaker, K. J. Chem. Soc. 1957, 2531.
(50) Schultz, C.; Link, A.; Kunick, C. Arch. Pharm. (Weinheim) 2001, 334,
163.
(51) Subsequent to our initial communication on the synthesis of lousia-
nins C and D (ref 19), a report on the copper-catalyzed microwave-promoted
decarboxylation of arylcarboxylic acids was disclosed: Goossen, L. J.;
Manjolinho, F.; Khan, B. A.; Rodriguez, N. J. Org. Chem. 2009, 74, 2620.
(52) Corey, E. J.; Ensley, H. E. J. Am. Chem. Soc. 1975, 97, 6908.
8348 J. Org. Chem. Vol. 74, No. 21, 2009

Catozzi et al.
JOCArticle
SCHEME 8
in 78% yield (Scheme 7). Once again, the characterization
data obtained showed that the synthetic product was identical
to the natural product [mp 83-85 °C (lit.20 mp 83-86 °C)].
Louisianins A and B. In contrast to the route employed to
access louisianins C and D, the pyridone core structure of
louisianins A and B requires an oxidative transformation of
the ethyl ester group present in pyridine 9. The most direct route
to achieve this aim appeared to be either a Dakin or Baeyer-
Villiger oxidation of the corresponding aldehyde or methyl
ketone, respectively (Scheme 8). In this regard, pyridine ester
9 was successfully transformed into aldehyde 20 via a standard
DIBAL-H reduction in 89% yield. Furthermore, conversion of
the ester 9 into Weinreb amide 21 via the aluminum amide
method53 followed by treatment with methyllithium gave the
methyl ketone 22 in 62% yield over the two steps.
With the required substrates in hand, attention was turned
to the oxidation of aldehyde 20 (Scheme 8). Acknowledging
the absence of precedent in the literature for the Dakin
oxidation of 2-formylpyridines, and the potential for over-
oxidation, a range of oxidants were examined for this
transformation. Unfortunately, none of the conditions ex-
amined produced any of the required formate. The use of 2
equiv of oxidant led to a 3:1 sulfoxide/sulfone mixture (m-
CPBA) or sulfoxide alone (monoperoxymaleic acid) with no
reaction at the formyl group. The strongest oxidant known
for the Dakin oxidation, trifluoroperacetic acid (TFPAA),36
promoted oxidation of the aldehyde to the corresponding
carboxylic acid and concomitant oxidation of the sulfide; the
resulting product, obtained in essentially quantitative but
unpurified yield, was identified as the carboxylate salt 23
(Scheme 8).54 More nucleophilic peroxide sources such as
NaOOH or PhSeO3H55 were also investigated, but unfortu-
nately, only decomposition was observed.
Attention was therefore switched to the attempted
Baeyer-Villiger oxidation of methyl ketone 22, but these
studies were equally frustrating. The acetyl group proved
remarkably inert under any conditions, even with TFPAA.
Consequently, only oxidation to the sulfoxide was observed.
Perhaps this is not surprising given that the only literature
precedent for the Baeyer-Villiger oxidation of 2-acetylpyri-
dine56 required forcing conditions and pyridine 22 is appreci-
ably more electron-rich. These results discouraged examination
of the carboxy inversion reaction37,57 and rapidly curtailed the
direct oxidation approach to the required pyridones.
Next, an alternative approach to the introduction of the
required oxidation at the C-2 position via a hydroxy-Sand-
meyer reaction on the 2-pyridyldiazonium salt was ex-
plored.58 Thus, pyridine ester 9 was converted into the
2-aminopyridine 24, the required diazonium precursor, via
a series of straightforward reaction steps; saponification to
the potassium carboxylate with potassium trimethylsilano-
late,59 acyl azide formation with diphenylphosphoryl azide,
and subsequent Curtius rearrangement60 followed by acidic
deprotection and decarboxylation afforded the desired
2-aminopyridine 24 in a 63% overall yield for this telescoped
three-step sequence (Scheme 9).
The literature scope for the hydroxy-Sandmeyer reaction
with 2-aminopyridine substrates is limited, and moderate
yields are observed.61 Such reactions are usually conducted
(56) Llamas, R.; Jimenez-Sanchidrian, C.; Ruiz, J. R. Tetrahedron 2007,
63, 1435.
(57) For elegant examples of this reaction in synthesis, see, for example:
(a) Danishefsky, S.; Tsuzuki, K. J. Am. Chem. Soc. 1980, 102, 6889. (b)
Breder, A.; Chinigo, G. M.; Waltman, A. W.; Carreira, E. M. Angew. Chem.,
Int. Ed. 2008, 47, 8514.
(58) (a) Cohen, T.; Dietz, A. G.; Miser, J. R. J. Org. Chem. 1977, 42, 2053.
(b) Hanson, P.; Rowell, S. C.; Walton, P. H.; Timms, A. W. Org. Biomol.
Chem. 2004, 2, 1838.
(53) Guanti, G.; Riva, R. Tetrahedron: Asymmetry 2001, 12, 1185.
(54) Compound 23 was characterized by 1H NMR, 13C NMR, and IR and
the structure assigned on the basis of these data. HRMS could not be obtained
because MS analysis of the crude salt provided no sensible information.
Furthermore, the free base of the salt could not be isolated because of the
extremely polar nature of the corresponding acid sulfoxide that evaded
aqueous extraction and could not be purified by chromatography.
(55) Syper, L. Synthesis 1989, 167.
(59) Laganis, E. D.; Chenard, B. L. Tetrahedron Lett. 1984, 25, 5831.
(60) Donald, J. R.; Edwards, M. G.; Taylor, R. J. K. Tetrahedron Lett.
2007, 48, 5201.
(61) See, for example: (a) Van de Poel, H.; Guillaumet, G.; Viaud-
Massuard, M. C. Heterocycles 2002, 57, 55. (b) Manera, C.; Benetti, V.;
Castelli, M. P.; Cavallini, T.; Lazzarotti, S.; Pibiri, F.; Saccomanni, G.;
Tuccinardi, T.; Vannacci, A.; Martinelli, A.; Ferrarini, P. L. J. Med. Chem.
2006, 49, 5947.
J. Org. Chem. Vol. 74, No. 21, 2009 8349

Catozzi et al.
JOCArticle
TABLE 2. Hydroxy-Sandmeyer Reactions of Aminopyridine 24
entry
conditions
product
yield, %
1
2
3
4
5
6
NaNO2, HCl/H2O (1:10), 0 °C; then NaOH, 100 °C
NaNO2, HCl/AcOH (1:10), 0 °C; then CuSO4, 100 °C
NaNO2, HCl/H2O/CH2Cl2 (1:10:11), 0 °C; then CuSO4, 100 °C
NaNO2, HCl/H2O (1:10), 0 °C; then CuSO4, 100 °C
0
0
0
32
0
22
26
25
t-BuONO, BF3 OEt2, THF, 0 °C; then CuCl, aq NaOH
3
NaNO2, H3PO2, H2O, 0 °C to rt
SCHEME 9
Success was eventually achieved using the halodecarbox-
ylation of carboxylic acid 17 and a subsequent SNAr reaction
with a suitable “hydroxide” source (Scheme 10). Classically,
halodecarboxylation via the Hunsdiecker reaction was pro-
moted by either silver or mercury salts.64 Although there
have been many successful advances in Hunsdiecker tech-
nology, particularly the Kochi variants,65 success has never
been demonstrated with 2-pyridinecarboxylic acids.66 In
contrast, the report of Inouye and co-workers67 on the
double bromodecarboxylation of 4-n-butoxypyridine-2,6-
dicarboxylic acid demonstrated that Barton halodecarbox-
ylation68 was a viable method for 2-pyridinecarboxylic acids.
Accordingly, when carboxylic acid 17 was combined with
Barton’s alcohol, DMAP, and DCC and thermolysis in
carbon tetrachloride, the corresponding 2-chloropyridine
27 was obtained in 56% yield. The Barton ester formed in
situ proved to be highly unstable and could be neither
isolated nor detected in the reaction mixture. Despite this,
the reaction, although somewhat slow (24 h reflux in CCl4
was required), was robust, reliable, and scalable. Since the
original disclosure of Barton halodecarboxylation, im-
proved conditions have been reported to provide higher
yields using either photolytic conditions69 or xanthate es-
ters;70 we found that these methods offered no improvement
in yield over the original protocol.
with either CuSO4 or NaOH as the “hydroxide” source.
Several conditions for the hydroxy substitution reaction
using amine 24 were examined, and the results are compiled
in Table 2. Generation of the arene diazonium salt under
standard reaction conditions with NaNO2 and HCl followed
by application of either aqueous CuSO4 or NaOH and
heating led only to decomposition and the formation of a
multitude of products (entries 1-3). These reactions were
hindered by the poor solubility of the diazonium salt in the
reaction media, and attempts to use a variety of solvent
combinations to alleviate this problem met with no success.
Nevertheless, when the reaction was conducted in solely
aqueous solution with hydrochloric acid/water (entry 4),
chloride substitution of the diazonium salt and concomitant
sulfide oxidation were observed producing adduct 26 in 32%
yield. Alternative methods for the generation of the diazo-
nium salt were therefore explored. Though tert-butyl nitrite
has been successfully used for 2-aminopyridines,62 again
only decomposition was observed (entry 5). Hypophos-
phorus acid (H3PO2) is generally used for reduction of
diazonium salts to the corresponding pyridine. However,
as Porter and co-workers63 reported that hypophosphorus
acid can effect the hydroxy-Sandmeyer reaction, the same
reaction with pyridine 24 was studied. Initially, an encoura-
ging 22% yield of the desired pyridone 25 was obtained, but
further investigation failed to give higher yields. In search of
a more robust and successful method of pyridone formation,
other approaches were therefore explored.
With chloropyridine 27 in hand, treatment with a range of
alcohols was explored. Initial results confirmed the literature
view that SNAr reaction on 2-chloropyridines are difficult to
effect in the absence of electron-withdrawing substituents.71
(64) Crich, D. In Comprehensive Organic Synthesis; Trost, B. M., Flem-
ing, I., Eds.; Pergamon Press: Oxford, 1991; Vol. 7, pp 717-734.
(65) Kochi, J. K.; Sheldon, R. A. Org. React. 1972, 19, 279.
(66) Uemura, S.; Tanaka, S.; Okano, M.; Hamana, M. J. Org. Chem.
1983, 48, 3297.
(67) Inouye, M.; Miyake, T.; Furusyo, M.; Nakazumi, H. J. Am. Chem.
Soc. 1995, 117, 12416.
(68) Barton, D. H. R.; Crich, D.; Motherwell, W. B. Tetrahedron 1985, 41,
3901.
(69) Dauben, W. G.; Kowalczyk, B. A.; Bridon, D. P. Tetrahedron Lett.
1989, 30, 2461.
(70) Nyfeler, E.; Renaud, P. Org. Lett. 2008, 10, 985.
(71) There are very few examples of such reactions with 2-chloropyr-
idines. Typically, those reported contain one or more electron-withdrawing
groups to facilitate the reaction. See, for example: (a) Uray, G.; Kriessmann,
I. Synthesis 1984, 679. (b) Sycheva, T. V.; Rubtsov, N. M.; Sheinker, Y. N.;
Yakhontov, L. N. Chem. Heterocycl. Compd. 1987, 23, 82.
(62) Carroll, F. I.; Lee, J. R.; Navarro, H. A.; Ma, W.; Brieaddy, L. E.;
Abraham, P.; Damaj, M. I.; Martin, B. R. J. Med. Chem. 2002, 45, 4755.
(63) Wijtmans, M.; Pratt, D. A.; Brinkhorst, J.; Serwa, R.; Valgimigli, L.;
Pedulli, G. F.; Porter, N. A. J. Org. Chem. 2004, 69, 9215.
8350 J. Org. Chem. Vol. 74, No. 21, 2009

Catozzi et al.
JOCArticle
SCHEME 10
Thus, the use of 2-trimethylsilylethanol,72 potassium tert-
butoxide73 and potassium trimethylsilanolate gave no reac-
tion under any of the reaction conditions applied. In con-
trast, reactive alcohol nucleophiles such as allyl alcohol and
benzyl alcohols successfully underwent the desired SNAr
reaction. For example, when 5 equiv of allyl alcohol, chlor-
opyridine 27, potassium hydroxide, and 18-crown-6 were
combined in toluene under Dean-Stark conditions,74 the
desired allyloxypyridine 28 was obtained in 84% yield
(Scheme 10). Given the plethora of methods available to
remove allyl protecting groups75 we anticipated that the
pyridone could be revealed by deallylation at a later stage.
We next planned to carry out the C-4 oxidation via the
procedure successfully used in the route to louisianins C and
D. However, it was soon established that allyloxypyridine 28
was inert to any of the oxidation conditions examined. A
range of bases (including NaH,76 KO-t-Bu77 and LHMDS)
and conditions failed to return anything other than starting
material. Analysis of the oxidation results suggests that the
reaction is finely balanced by the requirement for deproto-
nation to occur at the benzylic position. The regioselective
nature of the reaction with pyridine 8 occurs with deproto-
nation only in the most acidic position (i.e., 4-picoline vs 3-
picoline). Given the relative pKa of the benzylic position
(∼35, γ-picoline, DMSO)78 and LHMDS (30, DMSO) and
the excess of base required (10 equiv) for the reaction to
occur, it is clear that the other benzylic positions are not
sufficiently acidic for oxygenation to occur. Moreover, the
increased electron density of in the pyridine ring in pyridine
28 would appear to render the substrate insufficiently acidic
to deprotonate.
This revelation required a revision of the planned strategy.
A reversal of the steps so that oxidation preceded SNAr
appeared to be a possible solution. Thus, we were delighted
to find that application of the same oxidation conditions
used on pyridine 8 proceeded smoothly on chloropyridine 27
to afford the desired benzylic alcohol 30 in 86% yield
(Scheme 10). Apparently, the electron-withdrawing chlorine
substituent permits the desired deprotonation to occur with-
out acidifying the other benzylic positions sufficiently to
promote alternative oxidation products. Unfortunately, any
attempts to perform SNAr on pyridine 30 to generate the
corresponding pyridone led only to destruction of material
whether the conditions were basic, neutral or acidic; pre-
sumably, via elimination pathways. We rationalized that
oxidation of the benzylic alcohol to the corresponding
ketone 31 would render the pyridine more activated to any
subsequent SNAr reaction. To this end, Parikh-Doering
oxidation of alcohol 30 proceeded under standard conditions
to provide ketone 31 in 75% yield. Although attempts to
conduct SNAr reactions on compound 31 under basic con-
ditions were unsuccessful, the use of “non-anhydrous”79
acetic acid gave a slow, yet clean, hydrolysis to the corre-
sponding pyridone 7 (92%). Once more we were pleased to
discover that this reaction could be scaled up without any
lowering of the yield. This key transformation was supported
by the characteristic shift of the C-6 proton (31, 8.14 ppm; 7,
7.15 ppm) and the appearance of the pyridone signal in the
IR spectrum (υ 1660 cm-1).
(72) Lluis, J.; Matheu, M. I.; Castillon, S. Tetrahedron Lett. 1998, 39,
1807.
(73) Birkholz, M. N.; Dubrovina, N. V.; Jiao, H. J.; Michalik, D.; Holz,
J.; Paciello, R.; Breit, B.; Borner, A. Chem. Eur. J. 2007, 13, 5896.
(74) Poon, K. W. C.; Dudley, G. B. J. Org. Chem. 2006, 71, 3923.
(75) Greene, T. W.; Wuyts, P. G. M. Greene’s Protective Groups in
Organic Synthesis, 4th ed.; Wiley-Interscience: New York, 2006.
(76) Stork, G.; Niu, D.; Fujimoto, A.; Koft, E. R.; Balkovec, J. M.; Tata,
J. R.; Dake, G. R. J. Am. Chem. Soc. 2001, 123, 3239.
(77) Sanz, R.; Castroviejo, M. P.; Guilarte, V.; Perez, A.; Fananas, F. J. J.
Org. Chem. 2007, 72, 5113.
With the pyridone oxidation level established, all that
remained was to reveal the allyl group from its masked
precursor (Scheme 11). The low solubility of pyridone 7 at
(79) Lavecchia, G.; Berteina-Raboin, S.; Guillaumet, G. Tetrahedron
Lett. 2004, 45, 6633.
(78) Taft, R. W.; Bordwell, F. G. Acc. Chem. Res. 1988, 21, 463.
J. Org. Chem. Vol. 74, No. 21, 2009 8351

Catozzi et al.
JOCArticle
SCHEME 11
-78 °C and the high-water solubility of sulfoxide 32 predicated
the use of slightly unusual conditions for the oxidation-
elimination sequence. Thus, oxidation of sulfide 7 at -15 °C
was followed by the addition of excess triethylamine and
filtration through silica gel to afforded unpurified sulfoxide 32
that was then immediately thermolysis in the presence of CaCO3
to afford louisianin A (1) in 71% yield over the two steps.
Pleasingly, there was again no evidence for isomerization of the
allyl group. The characterization data obtained compares fa-
vorably with the isolation data18 and also with that reported by
Chang and co-workers22 [e.g., mp 185-188 °C (lit.18 mp
189-191 °C, lit.22 mp 187-188 °C); δC (CDCl3) 207.5; δH
(CDCl3) 7.16(1H, s, Ha) [lit. δC (CDCl3) 208.6,18 207.7;22 lit. δH
(CDCl3) 7.18 (1 H, s, Ha),22 7.20 (1 H, s, Ha)18]].
All that remained was to complete the synthesis of loui-
sianin B (2) via reduction of the ketone; treatment of
louisianin A (1) with NaBH4 in methanol afforded louisianin
B (2), in 82% yield. The 13C NMR data (Table 3) and mp
obtained were in good agreement with the isolation data
reported by Omura and co-workers [e.g., mp 163-165 °C,
lit.18 mp 168-170 °C], but the 1H NMR data were found to
vary consistently from those reported. We believe that this is
due to misassignment of the residual solvent (CD2HOH)
peak in the original spectra.80 Thus, when reported proton
data are recalibrated (þ0.36 ppm), the data are comparable
(Table 3). All attempts to prepare crystals of louisianin A (1),
or of the derived p-bromobenzoate derivative 33, that were
suitable for X-ray analysis were unsuccessful. However, in
TABLE 3. NMR Comparison for Natural and Synthetic Louisianin B (2)
recorded 13C NMR
100 MHza
reported18 13C NMR
68 MHzb
recorded 1H NMR
400 MHzc
reported18 1H NMR
270 MHzd
adjusted 1H NMR
270 MHze
1
2
3
4
5
164.1
133.8
158.6
119.6
133.2
28.0
163.9
133.8
158.5
119.7
133.2
28.5
7.09
6.73
7.09
7
2.65, 2.92
1.95, 2.40
5.24
2.30, 2.56
1.59, 2.05
4.89
2.66, 2.92
1.95, 2.41
5.25
8
9
34.7
76.8
35.2
77.1
11
12
13
33.6
138.0
117.2
34.0
137.9
117.3
3.33, 3.46
5.97
5.10, 5.10
2.95, 3.10
5.62
4.73, 4.73
3.31, 3.46
5.98
5.09, 5.09
a 13C NMR (100 MHz) for compound 2 in CD3OD. b 13C NMR data (68 MHz) for reported by Omura and co-workers for louisianin B in CD3OD.
c 1H NMR data (400 MHz) for compound 2 in CD3OD. d 1H NMR data (270 MHz) reported by Omura and co-workers for louisianin B in CD3OD.
e 1H NMR isolation data adjusted by applying a correction factor of þ 0.36 ppm (see text).
8352 J. Org. Chem. Vol. 74, No. 21, 2009

Catozzi et al.
JOCArticle
order to provide further confirmation for the structural
assignment, louisianin B (2) was reoxidized using the
pyridine intermediate was itself assembled from triazine
10 via the inverse-electron-demand Diels-Alder/retro-
Diels-Alder chemistry we have previously developed for
the preparation of highly functionalized pyridines. Louisia-
nins A and B were prepared via transformation of the
2-ethoxycarbonylpyridine into a 2-chloropyridine and sub-
sequent hydrolysis to the pyridone with an acid-catalyzed
SNAr reaction. In contrast, louisianins C and D were pre-
pared through a key decarboxyethylation of the 2-ethoxy-
carbonylpyridine intermediate. In both series, oxidation in
the C-4 position was accomplished via an oxidation reac-
tion and the allyl group obtained from a masked sulfide
precursor. This appears to be the first reported total syn-
thesis of louisianin B. We have also reported the prepara-
tion of (-)-louisianin B in enantio-enriched form via a
Noyori asymmetric transfer hydrogenation of the ketone in
louisianin B.
Parikh-Doering procedure (SO3 Py, Et3N, DMSO) to give
3
louisianin A (1) in 91% yield, and this sample displayed
spectral data entirely consistent with those reported.18
Finally we examined the asymmetric reduction of louisia-
nin A (1) in order to prepare enantioenriched louisianin B
(2). It should be noted that an optical rotation was not
recorded in the original isolation papers.17,18,80 Both Cor-
ey-Baskshi-Shibata (CBS) reduction81 and Noyori trans-
fer hydrogenation82 have proven to be excellent methods for
the enantioselective reduction of ketones. Although neither
method is precedented for the reduction of pyridone ketones,
the Noyori protocol appeared more suited to our purposes
because of its compatibility with alkenes. Accordingly, when
louisianin A (1) was subjected to Noyori reduction with
5 mol % of the [Ru(S,S)-TsDPEN)] complex in the presence of
formic acid/triethylamine (5:2),83 enantioenriched louisianin
B (2) was obtained in 92% yield and >99:<1 er as deter-
mined by chiral HPLC (3 μM AD-3 column, 90:5:5 isohex-
ane/EtOH/MeOH, flow rate 1.00 mL/min, λ=254 nm). We
sought to use Mosher’s ester analysis84 to determine the
absolute configuration of the carbinol center but found that
the MTPA derivative could not be prepared, even under the
forcing conditions used to prepare p-bromobenzoate deri-
vative 33. The reluctance of Mosher’s esters to form with
hindered alcohols has been noted previously (and new
reagents developed to attenuate the problems);85 presum-
ably, this is the reason for the failure of alcohol 2 to react.
There are also limited examples of using sterically more
accessible enantiopure isocyanates to form diastereomeric
carbamate derivatives to assign the absolute configuration;86
unfortunately, this technique also proved to be unsuccessful.
Therefore, we are currently unable to assign the absolute
configuration of the enantioenriched (-)-louisianin B (2).
According to the rationale for enantioselectivity reported by
Noyori and co-workers,87 the product obtained (using the
(S,S)-TsDPEN ligand) might be expected to have the
(S)-configuration, but in the absence of reduction of alterna-
tive pyridone ketones, there is no evidence to corroborate this.
Experimental Section
General Experimental Procedures. See the Supporting Infor-
mation for complete experimental details including full char-
acterization data and general experimental detail.
4-Allyl-6,7-dihydro-1H-cyclopenta[c]pyridine-1,5(2H)-dione,
Louisianin A (1). A solution of sulfide 7 (205 mg, 0.685 mmol, 1.0
equiv) in anhydrous CH2Cl2 (12 mL) under argon was cooled to
-15 °C, and then purified m-CPBA2 (124 mg, 0.718 mmol, 1.05
equiv) was added in a single portion. The reaction mixture was
stirred at -15 °C until disappearance of starting material (TLC,
1 h). Triethylamine (83 mg, 0.822 mmol, 115 μL, 1.20 equiv) was
added via microsyringe and the reaction warmed to rt and
stirred for 1 h. The mixture was then preadsorbed onto SiO2
and passed through a pad of SiO2 with 4:1 EtOAc/EtOH to
afford the intermediate sulfoxide pyridone 32. This material was
sufficiently pure for the following step. The sulfoxide was
dispersed onto the side of a large pressure tube (20 cm ꢀ
2.5 cm) by evaporation from CH2Cl2, and then xylene (14 mL)
and CaCO3 (257 mg, 2.569 mmol, 3.75 equiv) were charged to
the tube. The tube was sealed, placed in a 160 °C oil bath, and
heated for 18 h. The sulfoxide gradually dissolved upon heating,
and the resultant brown suspension was cooled to rt, transferred
to a round-bottom flask, and preadsorbed onto SiO2. Column
chromatography (SiO2, 19:1 EtOAc/EtOH) afforded the title
compound, louisianin A (1), as a cream solid (91.5 mg, 0.484
mmol, 71% for the two steps). Recrystallization from acetoni-
trile afforded 1 as fine, colorless needles. Data for 1: Rf 0.30 (19:1
EtOAc/EtOH, det: KMnO4); mp 185-188 °C (MeCN) (lit.18
mp 189-191 °C); 1H NMR (400 MHz, CDCl3) 12.73 (br s, 1 H,
NH(6)), 7.16 (s, 1 H, CH(5)), 5.93 (ddt, J=16.9, 10.3, 6.6 Hz, 1
H, CH(11)), 5.11-5.03 (m, 2 H, CH2(12a and 12b)), 3.52 (dd, J=
6.6, 1.1 Hz, 2 H, CH2(10)), 3.01-2.97 (m, 2 H, CH2(7)),
2.72-2.68 (m, 2 H, CH2(8)); 13C NMR (100 MHz, CDCl3)
207.5 (C(9)), 163.9 (C(1)), 149.3 (C(2)), 144.6 (C(3)), 135.6
(C(11)), 132.1 (C(5)), 116.9 (C(12)), 116.0 (C(4)), 36.3 (C(8)),
31.4 (C(10)), 22.8 (C(7)); IR νmax (CH2Cl2)/cm-1 3143, 2918,
2848, 1716, 1663, 1612, 1471, 1448, 1424, 1292, 1240, 1198, 1079,
942, 906, 878, 772, 737; MS (ESI, m/z) 212 (MNaþ); HMRS
(ESI) found 212.0689, calcd for C11H11NNaO2 [MNaþ]
212.0682 (3.3 ppm error).
Conclusion
In conclusion, we have described the preparation of all
four members of the louisianin family (A-D) from the
common tetrasubstituted pyridine intermediate 9. The
(80) Professor Omura confirmed that an R-D value had not been recorded
and that no sample remained. Private communication (email dated Aug 12, 2008).
(81) Corey, E. J.; Helal, C. J. Angew. Chem., Int. Ed. 1998, 37, 1986.
(82) Noyori, R.; Hashiguchi, S. Acc. Chem. Res. 1997, 30, 97.
(83) Noyori’s initial report (ref 83a) uses the isolated complex [RuCl(S,
S)-TsDPEN] to catalyze the reaction. We found it more convenient to use the
procedure reported by Tanis and co-workers (ref 83b) to generate the catalyst
in situ: (a) Fujii, A.; Hashiguchi, S.; Uematsu, N.; Ikariya, T.; Noyori, R. J.
Am. Chem. Soc. 1996, 118, 2521. (b) Tanis, S. P.; Evans, B. R.; Nieman, J. A.;
Parker, T. T.; Taylor, W. D.; Heasley, S. E.; Herrinton, P. M.; Perrault, W.
R.; Hohler, R. A.; Dolak, L. A.; Hester, M. R.; Seest, E. P. Tetrahedron:
Asymmetry 2006, 17, 2154.
(84) Hoye, T. R.; Jeffrey, C. S.; Shao, F. Nature Protocols 2007, 2, 2451.
(85) See for example: (a) Takeuchi, Y.; Itoh, N.; Koizumi, T. J. Chem.
Soc., Chem. Commun. 1992, 1514. (b) Williamson, R. T.; Sosa, A. C. B.;
Mitra, A.; Seaton, P. J.; Weibel, D. B.; Schroeder, F. C.; Meinwald, J.;
Koehn, F. E. Org. Lett. 2003, 5, 1745 and references cited therein.
(86) Vodicka, P.; Streinz, L.; Koutek, B.; Budesinsky, M.; Ondracek, J.;
Cisarova, I. Chirality 2003, 15, 472.
(()-4-Allyl-5-hydroxy-2,5,6,7-tetrahydro-1H-cyclopenta[c]-
pyridin-1-one, Louisianin B (2). Sodium borohydride (8 mg,
0.203 mmol, 1.0 equiv) was added in a single portion to a 0 °C
solution of louisianin A (1) (38.5 mg, 0.203 mmol, 1.0 equiv) in
MeOH (5 mL). The reaction was then stirred for 1 h at 0 °C.
Upon completion of the reduction, the reaction was quenched
by the addition of one drop of glacial acetic acid. The reaction
mixture was then preadsorbed onto SiO2 and purified by column
(87) Yamakawa, M.; Yamada, I.; Noyori, R. Angew. Chem., Int. Ed.
2001, 40, 2818.
J. Org. Chem. Vol. 74, No. 21, 2009 8353

Catozzi et al.
JOCArticle
chromatography (SiO2, 9:1 EtOAc/EtOH) to give louisianin
B (2) (32 mg, 0.167 mmol, 82%) as a colorless solid. Recrystalliza-
tion from acetonitrile afforded 2 as a colorless microcrystalline
solid. Data for 2: Rf 0.25 (9:1 EtOAc/EtOH, det: KMnO4); mp
i-PrOH (0.65 mL) and triethylamine (33 mg, 0.327 mmol,
46 μL) were added. The mixture was heated to reflux for 1 h
and the resultant red-brown solution cooled to rt. The vola-
tiles were removed, and the residual solid was dissolved in
2 mL of anhydrous DMF to afford a 0.0813 M catalyst
solution. A flame-dried round-bottom flask under argon
was charged with louisianin A (1) (10 mg, 0.0523 mmol, 1.0
equiv). Anhydrous DMF (0.8 mL) and then argon-sparged
Et3N/HCO2H (5:2, 0.2 mL) were added via syringe followed
by the catalyst stock solution (0.00262 mmol, 5 mol %, 32 μL
of a 0.0813 M stock solution in DMF). The resultant
orange solution was then stirred under argon for 24 h
at 40 °C. The orange-brown solution was then transferred to
a round-bottom flask and concentrated in vacuo (high vacuum
pump, 40 °C). The crude brown oil was preadsorbed on SiO2
and purified by column chromatography (SiO2, 6:1 EtOAc/
EtOH) to give louisianin B (2) (9.2 mg, 0.048 mmol, 92%) as a
pale yellow amorphous solid. Data for (-)-2: [R]2D5 -33.6
(c 0.42, MeOH); HPLC analysis (3 μM AD-3 column, 90:5:5
isohexane/EtOH/MeOH, flow rate 1.00 mL/min, λ=254 nm),
1
163-165 °C (MeCN) (lit.18 mp 168-170 °C); H NMR (400
MHz, CDCl3, limitedsolubility) 12.64 (brs, 1 H, NH(6)), 7.09 (s, 1
H, CH(5)), 5.93 (dddd, J = 16.6, 10.2, 6.4, 6.4 Hz, CH(12)),
5.26 (dd, J=7.2, 3.6 Hz, CH(9)), 5.14-5.06 (m, 2 H, CH2(13a
and 13b)), 3.37 (dddd, J=16.3, 6.4, 2.3, 1.4 Hz, 1 H, CH2(11a)),
3.27 (dddd, J=16.3, 6.4, 2.2, 1.5 Hz, 1 H, CH2(11b)), 2.97 (ddd,
J=16.9, 8.5, 5.8 Hz, 1 H, CH2(7a)), 2.72 (ddd, J=16.9, 9.0, 4.2 Hz,
1 H, CH2(7b)), 2.42 (dddd, J = 13.3, 9.0, 7.4, 5.8 Hz, 1 H,
CH2(8a)), 1.98 (dddd, J=13.3, 8.5, 4.2, 3.6 Hz, 1 H, CH2(8b)),
1
1.67 (br s, 1 H, OH(10)); H NMR (400 MHz, CD3OD) 7.09
(s, 1 H, CH(5)), 5.97 (dddd, J=17.5, 9.6, 7.0, 5.9 Hz, 1 H, CH(12)),
5.24 (dd, J=7.4, 3.8Hz, 1 H, CH(9)), 5.13-5.07 (m, 2 H, CH2(13a
and 13b)), 3.46 (dd, J=15.9, 7.1 Hz, 1 H, CH2(11a)), 3.33 (m, 1 H,
CH2(11b, obscured by CD3 quintet), 2.92 (ddd, J = 16.6, 8.6,
5.8 Hz, 1 H, CH2(7a)), 2.65 (ddd, J = 16.6, 9.0, 4.6 Hz, 1 H,
CH2(7b)), 2.40 (dddd, J=13.9, 9.0, 7.4, 5.8 Hz, 1 H, CH2(8a)),
1.95 (dddd, J=13.9, 8.6, 4.6, 3.8 Hz, 1 H, CH2(8b)); 13C NMR
(100 MHz, CD3OD) 164.1 (C(1)), 158.6 (C(3)), 138.0 (C(12)),
133.8 (C(2)), 133.2 (C(5)), 119.6 (C(4)), 117.2 (C(13)), 76.8 (C(9)),
34.7 (C(8)), 33.6 (C(11)), 28.0 (C(7)); IR νmax (CH2Cl2)/cm-1
3386, 2927, 2884, 1658, 1618, 1561, 1461, 1434, 1322, 1281,
1112, 992, 915; MS (ESI, m/z) 214 (MNaþ); HMRS (ESI)
found 214.0830, calcd for C11H13NNaO2 [MNaþ] 214.0838
(4.1 ppm error).
Asymmetric Noyori Transfer Hydrogenation of Louisianin A:
Preparation of Enantioenriched Louisianin B (2). A stock solu-
tion of Noyori catalyst was prepared according to the gene-
ral precedent of Tanis and co-workers:5 [RuCl2(p-cymene)]2
(50 mg, 0.0818 mmol) and (S,S)-TsDPEN (60 mg, 0.163 mmol)
were charged to a 5 mL round-bottom flask fitted with a reflux
condenser and three-way tap connected to argon and vacuum.
The flask was purged under argon, and then argon-sparged
tR = 5.10 min (major, 99.4%), 6.60 min (minor, 0.6%), er
>99:<1.
Acknowledgment. We thank the ESPRC (N.C.), Elsevier
Science (M.G.E.), and the Fond Special de Recherche,
ꢀ
Universite Catholique de Louvain (P.W.), for postdoctoral
support. We are also grateful to Dr. T. A. Dransfield
(University of York) for assistance with mass spectrometry
and Mr. A. Hard (AstraZeneca) for assistance with chiral
HPLC analysis.
Supporting Information Available: Detailed procedures for
the preparation of all new compounds and complete character-
ization data. This material is available free of charge via the
Internet at http://pubs.acs.org.
8354 J. Org. Chem. Vol. 74, No. 21, 2009