A total synthesis of millingtonine a

Article abstract of DOI:10.1021/ol203158p

A total synthesis of millingtonine A, a diglycosylated alkaloid, has been accomplished. Millingtonine A possesses a unique racemic tricyclic core structure not known from any other natural or synthetic source until now. The synthesis features a key bond-f

Full text of DOI:10.1021/ol203158p

ORGANIC
LETTERS
2012
Vol. 14, No. 3
696–699
A Total Synthesis of Millingtonine A
Jens Wegner,†,‡ Steven V. Ley,† Andreas Kirschning,‡ Anne-Lene Hansen,†
Javier Montenegro Garcia,† and Ian R. Baxendale†
Department of Chemistry, University of Cambridge, Lensfield Road, CB2 1EW
Cambridge, U.K., and Institute of Organic Chemistry, and Center of Biomolecular Drug
Research (BMWZ), Leibniz University of Hannover, Schneiderberg 1B, 30167
Hannover, Germany
theitc@ch.cam.ac.uk
Received November 25, 2011
ABSTRACT
A total synthesis of millingtonine A, a diglycosylated alkaloid, has been accomplished. Millingtonine A possesses a unique racemic tricyclic core
structure not known from any other natural or synthetic source until now. The synthesis features a key bond-forming radical UenoꢀStork
cyclization to form the heterocyclic core.
Millingtonine A (1), a glycosidal alkaloid, was first
isolated in Thailand from the methanolic extract of the
flower buds of Millingtonia hortensis.1 The structural
elucidation wassubsequently achievedthrough acombina-
tion of chemical and spectroscopic methods. Millingtonine
A is uniquely a racemic tricyclic aglycon that has a trans
configuration of the two outermost rings to the central
hydrofuran core. The two pendant hydroxyethane groups
are both glycosylated with D-glucose in its β-anomer form.
To date, little is known about the biological profile of
millingtonine A (1); however, Millingtonia hortensis is an
important source of herbal medicine in Southeast Asia
as the plant is cultivated throughout the area and used
for the treatment of tuberculosis or sinusitis and as an
antiasthmatic agent.2 The aqueous methanolic leaf extracts
of Millingtonia hortensis have also been screened against a
series of bacterial strains and yeast cultures demonstrat-
ing comparable antimicrobial activity to gentamycin and
nystatin.3 In addition, numerous other pharmacologically
active substances have been isolated from different parts of
the plant, such as scutellarein and hispidulin from the
petals of the plant, acetyl oleanolic acid from the fruits,
and β-sitosterol from the heartwood and bark. Isolated
extracts containing novel molecular architectures are of
general interest as potential leads for new pharmaceuticals.
Consequently, millingtonine’spromisingbut yetundefined
biological activity together with the unique core structure
and the fact that no previous syntheses have been reported
makes it an attractive target for total synthesis.
Retrosynthetically, we devised a convergent strategy
allowing the rapid and parallel synthesis of the different
building blocks as outlined below (Scheme 1).
Accordingly, we anticipated that millingtonine A (1)
could be derived from diol 2 via a late-stage glycosidation
employing trichloroacetimidate 3 followed by a final global
deprotection. The corresponding glycosyl donor 3 could in
turn be accessed in two steps from commercially available
D-glucose pentaacetate 9. The main heterocyclic core 2would
be obtained from racemic amine 4 and bromide 5 via a
HartwigꢀBuchwald coupling reaction. Finally, the main
fragment 4 would be prepared from the tertiary alcohol 8
and enamine 7 through a UenoꢀStork cyclization.4
† University of Cambridge.
‡ Leibniz University of Hannover.
(1) Hase, T.; Ohtani, K.; Kasai, R.; Yamasaki, K.; Picheansoonthon,
C. Phytochemistry 1996, 41, 317–321.
(2) Anulakanapakorn, K.; Bunyapraphatsara, N.; Satayavivad, J.
J. Sci. Soc. Thailand 1987, 13, 71–83.
(3) Jetty, A.; Iyengar, D. S. Pharmaceut. Biol. 2000, 38, 157–160.
r
10.1021/ol203158p
Published on Web 01/19/2012
2012 American Chemical Society

Scheme 1. Retrosynthetic Analysis
Scheme 2. Synthesis of Fragment 8
Scheme 3. Synthesis of Pyrrole 7
The preparation of protected diol 8 began from p-hydro-
xyanisol (10) employing a one-pot oxidation/protection
protocol as described by Wong and co-workers5 to afford
ketoketal 11. In a subsequent sequence, the freshly prepared
ketone 11 was reacted with a preformed solution of ethyl
acetate lithium enolate to form a β-hydroxy ester followed by
lithium borohydride reduction and selective protection of the
resulting primary alcohol. Three different protecting groups
were selected to assess their influence on the stereochemical
outcome of the UenoꢀStork cyclization, namely, the TBS
ether 8a, the TBDPS ether 8b,andthetritylether8c (Scheme2).
The second partner needed for the UenoꢀStork cycliza-
tion is enamine 7, which can be readily obtained from
commercial 1-Cbz-2,5-dihydro-1H-pyrrole (12) by isomer-
ization (Scheme 3).
was easily isomerized to the conjugated molecule 7. Inter-
estingly, the use of microwave heating was found particu-
larly beneficial in forming enamine 7 (60 min, 1.25 mol %
catalyst). The corresponding classical reflux conditions
using an oil bath gave only incomplete conversion even
when much higher loadings of catalyst were used.
In the next stage of the synthesis, the fragments 8aꢀc
and 7 were coupled in a two-step procedure (Scheme 4).
First, the enamine derivative 7 was brominated yielding a
transient bromonium/iminium ion which was trapped in
situ by the corresponding tertiary alcohol 8aꢀc. Then the
resultant intermediate bromide 14aꢀc was further reacted
without isolation via a tin hydride mediated radical cycli-
zation toform aneasily separable mixture of amines15aꢀc
(mixture of cis/trans relative ring junctures). After inten-
sive optimization studies, a yield of 86% over the two
consecutive steps was achieved with a cis/trans ring ratio of
1:2.7. The structure elucidation of the two configurational
isomers of 15 (cis and trans) was deteremined on the basis
of detailed 1H NMR anaylsis and NOE experiments of the
tricycle 15c. In compound 15c, multiple conclusive NOE
couplings between H10ꢀH20ꢀH7ꢀH2 in the cis product
were observed, whereas only NOE contacts between
We found that the pyrrole 12 on treatment with
Wilkinson’s catalyst6 (13) under microwave irradiation
(4) (a) Ueno, Y.; Chino, K.; Wantanabe, M.; Moyira, O.; Okawara,
M. J. Am. Chem. Soc. 1982, 104, 5564–5566. (b) Stork, G.; Mook, R., Jr.;
Biller, S. A.; Rychnovsky, S. D. J. Am. Chem. Soc. 1983, 105, 3741–3742.
(5) Tran-Huu-Dau, M.-E.; Wartchow, R.; Winterfeldt, E.; Wong,
Y.-S. Chem.;Eur. J. 2001, 7, 2349–2369.
(6) Osborn, J. A.; Jardine, F. H.; Young, J. F.; Wilkinson, G.
J. Chem. Soc. A 1966, 1711–1732.
Org. Lett., Vol. 14, No. 3, 2012
697

H10ꢀH20 and H2ꢀH7 for the corresponding trans product
as P(t-Bu)3, (rac)-BINAP, P(otolyl)3, and (o-biphenyl)-
10
Pcy2 proved effective for this transformation. However,
were recorded.7
the catalytic system which gave the best result was Pd2-
(dba)3, P(tBu)3 and NaO-t-Bu in toluene under microwave
heating, which afforded a 76% yield of the protected diol 16.
Scheme 4. UenoꢀStork Cyclization toward Amines 15aꢀc
Scheme 5. Synthesis of Protected Aglycon 16
By inspection of the data (Scheme 4), a general stereo-
chemical trend for the cyclization reaction can be identi-
fied, namely, that the largerthe protecting group the higher
the selectivity. Although due to cost they were not eval-
uated here potentially larger protecting groups such
as ꢀSi(TMS)3 or ꢀSi(TES)3 could be investigated.
In order to obtain the racemic aglycon 2, both silyl
protecting groups had to be removed. Initially a polymer
supported HF.pyridine complex11 was selected since
HF pyridine is known12 to deprotect TBDPS-silyl ethers.
3
However, in our case, only deprotection of the TBS-silyl
ether occurred to yield 18 (Scheme 6). Thus, we resorted to
the use of TBAF in THF to afford the bis-diol 2 ready for
glycosidation.
Following construction of the tricyclic ring system the
Cbz-group was removed to obtain amine 4 which was then
subjected to a Hartwig-Buchwald coupling reaction to
form the silyl protecteddiol 16(Scheme 5). For this process
only the trans isomer 15b was used consistent with the
stereochemistry of the final natural product.
Scheme 6. Silyl Deprotection toward Diol 2
When palladium on charcoal was used as the reduction
catalyst, this resulted in only loss of the dioxolane ring,
while the Cbz-group remained intact. At elevated tempera-
tures double bond reduction was also apparent. Conse-
quently, we investigated the use of Pearlman’s catalyst8
(Pd(OH)2/C), which gave the desired amine 4 in essentially
quantitative yield at ambient temperature in only 60 min
using a hydrogen pressure of less than 1 bar.
Although palladium-catalyzed allylation of amines is
well-known, arylation of aminals has not been reported.
Nevertheless, we found that aryl bromides with Pd2(dba)3,
(o-biphenyl)P(t-Bu)2, and NaO-t-Bu9 or other ligands such
(7) See the Supporting Information for more details.
Glycosidation of diol 2 was first investigated with tri-
chloracetimidate 3, which in turn was obtained via a known
literature procedure in two steps from pentaacetate 9.13
(8) (a) Carpes, M. J. S.; Cesar, P.; Miranda, M. L.; Correia, C. R. D.
Tetrahedron Lett. 1997, 38, 1869–1872. (b) Takahata, H.; Kubota, M.;
Ihara, K.; Okamoto, N.; Momose, T.; Azer, N.; Eldefrawi, A. T.;
Eldefrawi, M. E. Tetrahedron: Asymmetry 1998, 9, 3289–3301.
(9) (a) Hartwig, J. F.; Kawatsura, M.; Hauck, S. I.; Shaughnessy,
K. H.; Alcasar-Roman, L. M. J. Org. Chem. 1999, 64, 5575–5580.
(b) Wolfe, J. P.; Singer, R. A.; Yang, B. H.; Buchwald, S. L. J. Am.
Chem. Soc. 1999, 121, 9550–9561.
(11) (a) Gregorcic, A.; Zupan, M. J. Fluorine Chem. 1984, 24, 291–
302. (b) Ley, S. V.; Baxendale, I. R.; Bream, R. N.; Jackson, P. S.; Leach,
A. G.; Longbottom, D. A.; Nesi, M.; Scott, J. S.; Storer, R. I.; Taylor,
S. J. J. Chem. Soc., Perkin Trans. 1 2000, 3815–4195.
(12) Takahashi, D.; Hirono, S.; Hayashi, C.; Igarashi, M.; Nishimura,
Y.; Toshima, K. Angew. Chem., Int. Ed. 2010, 49, 10096–10100.
(10) (a) Wagaw, S.; Rennels, R. A.; Buchwald, S. L. J. Am. Chem.
Soc. 1997, 119, 8451–8458. (b) Wolfe, J. P.; Buchwald, S. L. Angew.
Chem., Int. Ed. 1999, 38, 2413–2416.
698
Org. Lett., Vol. 14, No. 3, 2012

Scheme 7. Glycosidation of Diol 2
Scheme 8. Final Deprotection Steps toward Millingtonine A (1)
high β/R ratio (Scheme 7). Indeed, in the case of the
pivaloyl-protected system the β-anomer was formed ex-
clusively. Therefore, considering the high anomeric ratio
and since we anticipated selective deprotection of the
benzyl groups of compounds 22 could be problematic
(due to the presence of the internal double bond), it was
decided to attempt the synthesis with compound 23
(Scheme 8). First, the pivaloyl groups were removed by
treatment with aqueous LiOH in THF/MeOH at 60 °C,
and then cleavage of the dioxolane was conducted with
PPTS in a mixture of acetone and water to finally afford
millingtonine A (1). The spectral analysis of this material
matched that of the previously isolated natural product.7
In conclusion, the route described constituted a longest
linear synthesis of 12 steps giving millingtonine A (1) in
6.2% overall yield. Of particular note, although not dis-
cussed in this paper, was the considerable lability of the
intermediate structures which were exceedingly prone to
rearrangement at each stage of the synthetic scheme.
Consequently realization of this total synthesis is a con-
siderable synthetic achievement. The analysis of the ac-
companying rearrangements and the elucidation of the
resulting chemical architectures will be discussed in detail
at a later date.
Unfortunately, attempted glycosidation using AgOTf as the
activator14 at ꢀ20 °C furnished only the bis-ortho ester 19
(Scheme 7). All successive attempts to rearrange this ortho-
ester to the alternative di-β-glycoside were unsucessful.
Many alternative methods for the glycosylation of diol 2
via the use of glycosyl sulfides, glycosyl sulfoxides, glycosyl
bromides, and glycosyl trichloroacetimidates in each
case with the pivaloyl, benzoate, or acetate protecting
groups and various activators led to low yields and/or
poor selectivity.
Finally, diglycosidation was achieved to give the benzyl-
or pivaloyl-protected structure in reasonable yields and
Acknowledgment. We thank the Royal Society (I.R.B.),
the BP Endowment and Alexander von Humboldt award
(S.V.L.), and the Fonds der Chemischen Industrie (J.W.).
(13) (a) Cai, T. B.; Lu, D.; Tang, X.; Zhang, Y.; Landerholm, M.;
Wang, P. G. J. Org. Chem. 2005, 70, 3518–3524. (b) Schmidt, R. R.;
Michel, J. Tetrahedron Lett. 1984, 25, 821–824. (c) Cook, B. N.; Bhakta,
S.; Biegel, T.; Bowman, K. G.; Armstrong, J. I.; Hemmerich, S.;
Bertozzi, C. R. J. Am. Chem. Soc. 2000, 122, 8612–8622.
(14) Maruyama, M.; Takeda, T.; Shimizu, N.; Hada, N.; Yamada, H.
Carbohydr. Res. 2000, 325, 83–92.
Supporting Information Available. Experimental proce-
dures and compound characterization. This material is avail-
able free of charge via the Internet at http://pubs.acs.org.
Org. Lett., Vol. 14, No. 3, 2012
699