The retrosynthetic analysis of iminosugar 3 is shown in
Scheme 1. 1,2-Diol and amine protection leads to the
Displacement of the mesylate with N-benzyl-allylamine
gave the diene 18 in a moderate yield (55%). The precursor
of the RCM reaction was then reacted with ruthenium
complex 1, but no reaction was observed even at elevated
temperatures. The diene was then subjected to RCM using
the more reactive ruthenium complex 2 in dichloromethane,
and a single product was isolated, which was identified as
the pyrrole 19. Not only did this indicate that RCM had taken
place, it also suggested that the catalyst had brought about
alkene isomerization and dehydrogenation. Neither a change
of the reaction solvent to toluene nor an adjustment of the
reaction temperature led to a change in the reaction outcome;
in every instance the pyrrole was the only product ever
isolated.
Scheme 1
protected pyrrolidine 4. Retrosynthetic dehydration of the
alcohol affords the alkene 5, which can then be disconnected
to the diene 6. Cleavage of the allylamine and further
functional group interconversion gives the ketone 8 via the
alkene 7. The key reaction in our route would be formation
of the pyrroline 5 via the RCM reaction of diene 6, which
would afford a trisubstituted double bond with the stereo-
chemistry of the C-1′ carbon atom, R to the double bond
already established.
A synthesis of chiral synthon 8 from L-(S)-erythrulose has
been described by Vandewalle.6 However, L-(S)-erythrulose
is no longer commercially available. An alternative route to
ketone 8 is the oxidative cleavage of L-ascorbic acid, a
method first reported by Jung and Shaw.7 Using this
approach, L-ascorbic acid 9 was converted into the diol 12
following the procedure devised by Abushanab et al.8
Selective protection of the primary hydroxyl group followed
by oxidation of the secondary hydroxyl group gave the
ketone 14 in good overall yield (41%).
Scheme 3
There are only a few reports of pyrrole formation during
RCM.9 Most of them mention pyrrole formation as a side
reaction of the RCM reaction. However, Verpoort and co-
workers recently developed a general pyrrole synthesis using
a tandem Grubbs’ carbene-RuCl3 catalytic system; they
assumed that RuCl3 was responsible for the pyrrole
formation.9d They also noticed that amines bearing an
electron-withdrawing group on the nitrogen atom did not
dehydrogenate to the pyrrole, but gave the corresponding
pyrrolines. This suggests that the basicity of the nitrogen
atom plays an important role in the dehydrogenation.
Nevertheless, there is one example in the literature where
the RCM reaction (with 1 in dichloromethane at reflux) of
an N-Boc-protected amine gave, after purification, a mixture
of the expected pyrroline and the corresponding pyrrole.9b
In our case, we observed total conversion of the diene 18
into the pyrrole 19 in the presence of 2 (1 mol % in
dichloromethane or toluene).
Scheme 2
In light of these results, we next decided to synthesize
the N-Boc-protected diene 21 and subject it to the RCM
reaction (Scheme 4). Treatment of the mesylate 17 with
allylamine and subsequent protection afforded the RCM
precursor 21 in good yield (90%). This precursor was then
subjected to RCM using ruthenium complex 2 in dichlo-
romethane; the reaction was complete in 36 h (microwave
(6) De Wilde, H.; De Clercq, P.; Vandewalle, M. Tetrahedron Lett. 1987,
28, 4757.
(7) Jung, M. E.; Shaw, T. J. J. Am. Chem. Soc. 1980, 102, 6304.
(8) Abushanab, E.; Bessodes, M.; Antonakis, K. Tetrahedron Lett. 1984,
25, 3841.
(9) (a) Evans, P.; Grigg, R.; Monteith, M. Tetrahedron Lett. 1999, 40,
5247. (b) Bujard, M.; Briot, A.; Gouverneur, V.; Mioskowski, C. Tetra-
hedron Lett. 1999, 40, 8785. (c) Yang, C.; Murray, W. V.; Wilson, L. J.
Tetrahedron Lett. 2003, 44, 1783. (d) Dieltiens, N.; Stevens, C. V.; De
Vos, D.; Allaert, B.; Drozdzak, R.; Verpoort, F. Tetrahedron Lett. 2004,
45, 8995. (e) Yang, Q.; Xiao, W.-J.; Yu, Z. Org. Lett. 2005, 7, 871.
Ketone 14 was then converted into the corresponding
alkene 15 using Wittig chemistry, and cleavage of the silyl
ether afforded the allylic alcohol 16. Activation of the alcohol
then gave the mesylate 17 in quantitative yield.
3522
Org. Lett., Vol. 7, No. 16, 2005