Coupling the petasis condensation to an iron(III) chloride-promoted cascade provides a short synthesis of Relenza congeners

Article abstract of DOI:10.1021/ol102326b

Iron(III) chloride hexahydrate promotes a cascade of transformations on a Petasis condensation product that sets up the right dihydropyran precursors of valuable Relenza congeners.

Full text of DOI:10.1021/ol102326b

ORGANIC
LETTERS
2010
Vol. 12, No. 22
5322-5325
Coupling the Petasis Condensation to
an Iron(III) Chloride-Promoted Cascade
Provides a Short Synthesis of Relenza
Congeners
Jean-Franc¸ois Soule´,† Aure´lie Mathieu,† Ste´phanie Norsikian,*,† and
Jean-Marie Beau*,†,‡
Centre de Recherche de Gif, Institut de Chimie des Substances Naturelles, CNRS,
AVenue de la Terrasse, F-91198 Gif-sur-YVette, France, and UniVersite´ Paris-Sud,
Institut de Chimie Mole´culaire et des Mate´riaux, F-91405 Orsay, France
jean-marie.beau@u-psud.fr; stephanie.norsikian@icsn.cnrs-gif.fr
Received September 27, 2010
ABSTRACT
Iron(III) chloride hexahydrate promotes a cascade of transformations on a Petasis condensation product that sets up the right dihydropyran
precursors of valuable Relenza congeners.
Influenza A viruses cause a severe infection in the respiratory
system and are responsible for seasonal epidemics and
sporadic pandemics. Although the primary method for
prevention is through vaccination, effective antiviral agents
are required to prepare for a possible pandemic.1
oseltamivir,4 but an alternative would be to use zanamivir
analogues that resemble the natural substrate more closely.5
The present work addresses this issue.
We report here a novel synthetic pathway combining the
three-component borono-Mannich condensation (Petasis re-
action)6 with an efficient and selective iron(III)-promoted
(FeCl3·6H2O) deprotection-cyclization one-pot process. Iron
salts have recently attracted considerable attention as inex-
pensive and environmentally friendly agents in a wide range
of selective processes in organic synthesis.7 The result is a
Among anti-influenza drugs, selective inhibitors of the
surface glycoprotein neuraminidase have been developed, and
two of them, oseltamivir phosphate 1 (Tamiflu)2 and zan-
amivir 2 (Relenza),3 which mimick the high-energy oxycar-
benium intermediate 3, have been approved for human use.
Most “anti-neuraminidase” strategies involve the use of
(4) Recent reviews on Tamiflu synthesis: (a) Farina, V.; Brown, J. D.
Angew. Chem., Int. Ed. 2006, 45, 7330–7334. (b) Shibasaki, M.; Kanai,
M. Eur. J. Org. Chem. 2008, 2008, 1827. (c) Magano, J. Chem. ReV. 2009,
109, 4398–4438.
† Institut de Chimie des Substances Naturelles.
‡ Institut de Chimie Mole´culaire et des Mate´riaux.
(1) (a) De Clercq, E. Nat. ReV. Drug DiscoVery 2006, 5, 1015–1025.
(b) von Itzstein, M. Nat. ReV. Drug DiscoVery 2007, 6, 967–974.
(2) Kim, C. U.; Lew, W.; Williams, M. A.; Liu, H.; Zhang, L.;
Swaminathan, S.; Bischofberger, N.; Chen, M. S.; Mendel, D. B.; Tai, C. Y.;
Laver, W. G.; Stevens, R. C. J. Am. Chem. Soc. 1997, 119, 681–690.
(3) von Itzstein, M.; Wu, W.-Y.; Kok, G. B.; Pegg, M. S.; Dyason, J. C.;
Jin, B.; Phan, T. V.; Smythe, M. L.; White, H. F.; Oliver, S. W.; Colman,
P. M.; Varghese, J. N.; Ryan, D. M.; Woods, J. M.; Bethell, R. C.; Hotham,
V. J.; Cameron, J. M.; Penn, C. R. Nature 1993, 363, 418–423.
(5) Collins, P. J.; Haire, L. F.; Lin, Y. P.; Liu, J.; Russell, R. J.; Walker,
P. A.; Skehel, J. J.; Martin, S. R.; Hay, A. J.; Gamblin, S. J. Nature 2008,
453, 1258–1261.
(6) (a) Petasis, N. A.; Akritopoulou, I. Tetrahedron Lett. 1993, 34, 583–
586. (b) Petasis, N. A. In Multicomponent Reactions; Zhu, J., Bienayme´,
H., Eds.; Wiley-VCH: New York, 2005; pp 1999-1223.
(7) For selected reviews, see: (a) Bolm, C.; Legros, J.; Le Paih, J.; Zani,
L. Chem. ReV. 2004, 104, 6217–6254. (b) Diaz, D.; Miranda, P.; Padron,
J.; Martin, V. Curr. Org. Chem. 2006, 10, 457–476.
10.1021/ol102326b  2010 American Chemical Society
Published on Web 10/14/2010

very concise and high-yielding synthesis of constructs of the
zanamivir family useful to explore the catalytic sites of the
neuraminidases. Our retrosynthetic approach to zanamivir
congeners 6 modified at the C6 position is outlined in
Scheme 1. We envisioned acyclic derivative 7 with appropri-
Scheme 2
.
Construction of Key Building Blocks Boronic Acid
9a and R-Hydroxyaldehyde 8a
Scheme 1. Structure of the Potent Neuraminidase Inhibitors
Oseltamivir Phosphate 1 and Zanamivir 2 and Retrosynthetic
Analysis of Zanamivir Congeners 6 from an
R-Hydroxyaldehyde 8
opening with Grignard reagent 11a in the presence of a
solution of dilithium tetrachlorocuprate,12 the resulting
adduct was deprotected with TBAF to diol 14 in a good
84% yield over two steps. Diol 14 was then submitted to
selective oxidation of the primary alcohol under biphasic
conditions using a sodium hypochlorite solution in the
presence of KBr and a catalytic amount of TEMPO,
buffering with NaHCO3.13 Since this R-hydroxyaldehyde
is difficult to characterize and purify, it was used directly
after the oxidation workup in the next Petasis condensa-
tion.
In initial experiments of the Petasis reaction, using 1 equiv
of each partner at room temperature with diallylamine in
various solvents, the CH2Cl2/HFIP (9:1) system proved to
be the most appropriate, providing aminoalcohol 7a as a
single diastereoisomer in a 41% yield (entries 1 and 2, Table
1). This modest result is explained by the formation of
ketoamine 16 (21%) issued from an intermediate enamine.14
A high yield of 7a (95%) was produced with the use of an
excess of aldehyde and amine (1.5 equiv of each) and heating
ate stereodirecting groups at either N5 or O6 as a key
precursor to cyclic dihydropyrans 4-6 with the proper
functionality and stereochemistry at C4. In turn, 7 could be
derived from the Petasis condensation8 using organoboronic
acid 9 as the nucleophilic partner. This reaction performed
with chiral R-hydroxyaldehydes 8 proceeds with a remark-
ably high stereocontrol, producing 1,2-aminoalcohols with
an anti configuration.9
We first investigated the synthesis of a dihydropyran with
an iso-pentyl side chain at C6 4a-6a (R1 ) CHEt2), a
substituent analogous to the Tamiflu hydrophobic side chain.
The required building blocks were prepared from starting
materials 10 and 12,10 respectively, as summarized in Scheme
2. The synthesis of boronic acid 9a commenced by conden-
sation of lithiated trimethylsilyl-acetylene on Weinreb amide
10. The desired propargyl ketone 11 was quantitatively
protected as a dimethylketal which was detrimethylsilylated
by potassium carbonate in i-PrOH and efficiently hydrobo-
rated using di(isopropylprenyl)borane.11 After oxidation,
boronic acid 9a was obtained in 82% yield.
Table 1. Optimization of the Petasis Coupling Reaction of
R-Hydroxyaldehyde 8a with Boronic Acid 9a
The required R-hydroxyaldehyde (S)-8a was prepared
from the protected (R)-glycidol 12.13 After epoxide ring
(8) For a recent use of the Petasis reaction in a de novo synthesis of the
sialic acids, see: Hong, Z.; Liu, L.; Hsu, C. C.; Wong, C. H. Angew. Chem.,
Int. Ed. 2006, 45, 7417–7421.
8aa
yield
%
b
amine (equiv) solvent (temp °C)
t (h)
160-48c
product
(9) Petasis, N. A.; Zavialov, I. A. J. Am. Chem. Soc. 1998, 120, 11798–
11799.
1
2
3
4
5
NHAll2 (1)
NHAll2 (1)
solventc (22)
7a
7a
7a
7a
17
14-39c
41
CH2Cl2/HFIPd (22) 24
(10) Laine´, D.; Fujita, M.; Ley, S. V. J. Chem. Soc., Perkin Trans. 1
1999, 1639–1646.
NHAll2 (1.5) CH2Cl2/HFIPd (80) 12
NHAll2 (1.5) CH2Cl2, (120, MW) 0.25
95
95
(11) Kalinin, A. V.; Scherer, S.; Snieckus, V. Angew. Chem., Int. Ed.
2003, 42, 3399–3404.
NHBn2
(1.5) CH2Cl2, (120, MW)
2
71
a Equiv of 8a based on the starting diol 14. b Isolated yield following
(12) Dixon, D. J.; Ley, S. V.; Tate, E. W. J. Chem. Soc., Perkin Trans.
1 1998, 3125–3126.
chromatography, based on boronic acid. c Dioxane, 160 h, 14% yield; EtOH,
160 h, 19% yield; CH2Cl2, 48 h, 29% yield; toluene, 48 h, 39% yield. d As
a 9/1 mixture. HFIP ) hexafluoro-iso-propanol.
(13) (a) Lucio Anelli, P.; Biffi, C.; Montanari, F.; Quici, S. J. Org. Chem.
1987, 52, 2559–2562. (b) Siedlecka, R.; Skarzewski, J.; Mlochowski, J.
Tetrahedron Lett. 1990, 31, 2177–2180.
Org. Lett., Vol. 12, No. 22, 2010
5323

the reaction mixture at 80 °C for 12 h (entry 3). The reaction
also proceeded smoothly in dichloromethane under micro-
wave radiation (120 °C, 300 W, 15 min) (entry 4). Under
these conditions, N,N-dibenzylamine also provided the
expected adduct 17 in good yield (entry 5).15 The stereo-
chemistry of the trans-amino alcohol 7a16 was confirmed
after removal of the allyl protecting group (N,N-dimethyl-
barbituric acid, cat. Pd(PPh3)4, CH2Cl2, 35 °C)17 and conver-
sion to the corresponding oxazolidinone 15 [CO(OCCl3)2,
NEt3, CH2Cl2, 87% from 7a]. The lH NMR spectrum of 15
exhibits a coupling constant 3J5,6 value of 8.0 Hz, indicating
a cis relationship between H-5 and H-6.
protecting group and selective acetylation of the amino group
to acetamide 21a in 85% for the two steps (Scheme 4). With
Scheme 4
.
Iron Trichloride-Promoted Cyclization of Acetamide
21a to Oxazoline 22a and Alcohol 23a
Removal of the dimethyl acetal and cyclization proved
extremely difficult. Treatment of 7 or 17 with a number
of acidic promoters (H2SO4, TFA, CeCl3·7H2O/NaI,18
Bi(NO3)·5H2O,19 TiCl4/LiI20) in different conditions resulted
in either no reaction or decomposition.
We ultimately found that treatment of amine 17 with iron
trichloride hexahydrate in CH2Cl2 provided directly cyclic
adducts 1821 and 19 (79% yield, ratio of 94:6) resulting from
the deprotection of the acetal, double bond isomerization,
and cyclization (Scheme 3). This expedient transformation
this change, we were pleased to find that FeCl3·6H2O now
promoted cyclization smoothly (45 °C for 2 h) and unexpect-
edly provided directly allylic oxazoline 22a as the only
product (87% yield).22 Moreover, adding water to the
reaction medium led to the exclusive formation (82%) of
alcohol 23a, with the natural configuration of the sialic acids
at C4. This may result from the in situ opening of oxazoline
22a by water as shown by its transformation to 23a under
identical conditions.23 A possible sequence could be an initial
Lewis acid-assisted formation of oxazolinium intermediate
25 from complex 24 that rotates around the C3-C4 bond to
26 (Scheme 5). Cyclization to 27, analogous to 19 (Scheme
Scheme 3
.
Iron Trichloride-Promoted Cyclization of Tertiary
Allylic Amine 17
was however extremely sluggish (5 days with 6 equiv of
FeCl3·6H2O), and all attempts to further transform this tertiary
amine never provided acceptable results. We believe that the
double bond isomerization resulted from the formation of
transient aziridinium intermediate 20 followed by rotation
around the C3-C4 bond for cyclization.
Scheme 5. Iron Trichloride-Promoted Cyclization of Acetamide
21a: Probable Sequence Leading to Oxazoline 22a
We then attributed the lack of reactivity to the absence of
a good internal nucleophile under these conditions to
participate in the cyclization process. For this, exchange of
the N-dialkyl substituents to the natural acetamide was
achieved on 7a by palladium-catalyzed removal of the allyl
(14) This side reaction reveals a poor reactivity of the boronic acid.
(15) With these optimized microwave conditions, the reaction with
aminodiphenylmethane, a primary hindered amine that usually performs
well in the Petasis reaction, did not afford the desired compound in a
satisfying yield.
(16) Preservation of the enantiomeric purity was confirmed by chiral
HPLC.
(17) Garro-Helion, F.; Merzouk, A.; Guibe, F. J. Org. Chem. 1993, 58,
6109–6113.
(18) Marcantoni, E.; Nobili, F.; Bartoli, G.; Bosco, M.; Sambri, L. J.
Org. Chem. 1997, 62, 4183–4184.
3) and ironIII-promoted oxazoline formation, then leads to
product 22a.
(19) Eash, K. J.; Pulia, M. S.; Wieland, L. C.; Mohan, R. S. J. Org.
Chem. 2000, 65, 8399–8401.
(20) Balme, G.; Gore, J. J. Org. Chem. 1983, 48, 3336–3338.
(21) As a 65:35 mixture of anomer in the favor of the ꢀ one, which
was confirmed by a NOE effect between H-6 and the hydrogen of the
hydroxyl proton at the C2.
(22) A similar oxazoline is in fact the key intermediate in the synthesis
of zanamivir from Neu5Ac. See ref 5.
5324
Org. Lett., Vol. 12, No. 22, 2010

Introduction of nitrogen at C4 followed the previous
reports in this area.8 Opening of oxazoline 22a with
trimethylsilyl azide24 gave azido compound 28a selectively
reduced with indium metal,25 followed by saponification of
the iso-propylester to furnish the corresponding amine 5a
(Scheme 6). Treatment with amino-iminomethanesulfonic
Scheme 7. Extension of the Synthetic Sequence for the
Preparation of Zanamivir Analogues from Aldehydes 8b and 8ca
Scheme 6. Completion of the Synthesis of Zanamivir and
2,3-Unsaturated Sialic Acid (Neu5Ac2en) Analogues
a Experimental details are given in the Supporting Information.
in the presence of limited amounts of water also provided
the C4-alcohol 23b in the cyclo-pentyl series.
acid26 afforded the guanidine derivative 6a, a zanamivir
congener with a hydrophobic side chain at C6. An analogue
of the 2,3-unsaturated sialic acid (Neu5Ac2en) was also
obtained in the form of lithium salt 4a by saponification of
the iso-propylester 23a.
We also briefly explored the efficiency of this newly
developed synthetic sequence for the preparation of other
Neu5Ac2en/zanamivir analogues. Thus, the above-mentioned
synthetic route was successfully applied to aldehydes 8b (R1
) cyclo-pentyl) and 8c [R1 ) CH(CH2OBz)2] (Scheme 7).
The sequence involved the Petasis coupling (7b and 7c),
exchange of the protecting groups at nitrogen (21b and 21c),
iron(III)-promoted cyclization (22b and 22c), azide formation
(28b and 28c), and reduction to amine/hydrolysis (5b and
5c). The additional route by this iron(III)-mediated cascade
In summary, we have developed a short synthetic route
to the highly valuable dihydropyran framework of zanamivir
congeners by coupling the Petasis three-component, boronic
acid Mannich reaction to an iron(III)-promoted one-pot
cascade of deprotection-C-C double bond isomerization-
cyclization-oxazoline formation. Moreover, adding a limited
amount of water to the iron promoter induced direct opening
of the bicyclic oxazoline to Neu5Ac2en congeners in high
yields. Additional flexibility of this synthetic approach is
expected, especially in the development of novel preparations
of structurally diverse ulosonic acids. This is currently being
developed in our laboratory together with biological testing
of the new constructs as antineuraminidases and anti-
influenza agents.
Acknowledgment. This study was financially supported
by the ICSN-CNRS.
(23) Treament of 21a with HCl (18 equiv) replacing FeCl3·6H2O and
water (4 equiv) under otherwise identical conditions (rt for 24 h) provided
a complex reaction mixture from which oxazoline 22a (14%) and a
hemiacetal analogous to 18 (13%) could be isolated.
(24) von Itzstein, M.; Jin, B.; Wu, W.-Y.; Chandler, M. Carbohydr. Res.
1993, 244, 181–185.
Supporting Information Available: Experimental pro-
cedures, characterization data, and 1H and 13C NMR spectra
for all new compounds. This material is available free of
charge via the Internet at http://pubs.acs.org.
(25) Reddy, G. V.; Rao, G. V.; Iyengar, D. S. Tetrahedron Lett. 1999,
40, 3937–3938.
(26) von Itzstein, M.; Wu, W.-Y.; Jin, B. Carbohydr. Res. 1994, 259,
301–305.
OL102326B
Org. Lett., Vol. 12, No. 22, 2010
5325