An expedient protecting-group-free total synthesis of (±)- dievodiamine

Article abstract of DOI:10.1021/ol4013469

The first total synthesis of the Evodia rutaecarpa derived natural product dievodiamine is described. The convergent synthesis was performed without protecting groups, delivering a route that is short and high yielding and uses limited chromatography. Key

Full text of DOI:10.1021/ol4013469

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
An Expedient Protecting-Group-Free Total
Synthesis of (()-Dievodiamine
William P. Unsworth,* Christiana Kitsiou, and Richard J. K. Taylor*
Department of Chemistry, University of York, Heslington, York, YO10 5DD, U.K.
william.unsworth@york.ac.uk; richard.taylor@york.ac.uk
Received May 14, 2013
ABSTRACT
The first total synthesis of the Evodia rutaecarpa derived natural product dievodiamine is described. The convergent synthesis was performed
without protecting groups, delivering a route that is short and high yielding and uses limited chromatography. Key steps include organometallic
addition into a DHED adduct and the Stille coupling of two advanced intermediates to complete the synthesis.
In recent years, protecting-group-free methods have
received widespread attention as ways to improve the
efficiency of synthesis. Furthermore, such an approach
provides an ‘opportunity for invention’, as the develop-
ment of novel synthetic methodology becomes necessary.1
Herein we report the successful application of these prin-
ciples in the first total synthesis of (()-dievodiamine 1.
Dievodiamine was recently isolated from Evodia
rutaecarpa.2 None of its biological properties have been
reported, although the Evodia fruits are used in numerous
traditional Chinese remedies to treat a wide range of con-
ditions including headaches, abdominal pain, migraine,
chill limbs, postpartum hemorrhage, nausea, inflamma-
tion, and cancer.2 Its structure is closely related to evodia-
mine 2, another Evodia rutaecarpa derived natural product,
which was recently synthesized by our group using direct
imine acylation methodology.3 Evodiamine is a known
thermogenic and stimulant and is included in a number of
dietary supplements, principally used to promote weight
loss. In addition, more recent studies have shown that it
binds to a diverse range of proteins; its therapeutic
potential against a number of diseases, including cancer,
Alzheimer’s disease, and cardiovascular disease, and its
ability to inhibit human DNA topoisomerase I have been
reported and reviewed.4 Furthermore, a recent SAR
study has shown evodiamine analogues to be highly
promising antitumor candidates.5 Dievodiamine 1 is not
a simple dimer of evodiamine, as its name may suggest,
but it does contain the basic framework of two evodi-
amine subunits (following oxidation and ring opening)
suggesting that evodiamine 2 is a likely biosynthetic
precursor.6 Bisindole alkaloids constitute a major class
of natural products, and their interesting biology has been
well documented.7 The isolation, synthesis, and biologi-
cal evaluation of bisindole alkaloids remains a highly
(4) (a) Jiang, J.; Hu, C. Molecules 2009, 14, 1852. (b) Dong, G.;
Sheng, C. S.; Wang, S.; Miao, Z.; Yao, J.; Zhang, W. J. Med. Chem.
2010, 53, 7521. (c) Yu, H.; Jin, H.; Gong, W.; Wang, Z.; Liang, H.
Molecules 2013, 18, 1826 and references therein.
(1) Young, S. N.; Baran, P. S. Nat. Chem. 2009, 1, 193 and references
therein.
(5) Dong, G.; Wang, S.; Miao, Z.; Yao, J.; Zhang, Y.; Guo, Z.;
Zhang, W.; Sheng, C. J. Med. Chem. 2012, 55, 7593.
(2) Wang, Q. Z.; Liang, J. Y.; Feng, X. Chin. Chem. Lett. 2010, 21,
596.
(3) (a) Unsworth, W. P.; Kitsiou, C.; Taylor, R. J. K. Org. Lett. 2013,
15, 258. (b) Unsworth, W. P.; Gallagher, K. A.; Jean, M.; Schmidt, J. P.;
Diorazio, L. J.; Taylor, R. J. K. Org. Lett. 2013, 15, 262.
(6) For the biosynthesis of related Evodia alkaloids, see: (a) Yamazaki,
M.; Ikuta, A. Tetrahedron Lett. 1966, 7, 3221. (b) Yamazaki, M.; Ikuta, A.;
Mori, T.; Kawana, T. Tetrahedron Lett. 1967, 8, 3317.
(7) Ryan, K. S.; Drennan, C. L. Chem. Biol. 2009, 16, 351 and
references therein.
r
10.1021/ol4013469
XXXX American Chemical Society

active area of research8 with their potential antimalarial
properties in particular receiving prominent attention
recently.8cÀe An unusual, although not unique,8b struc-
tural feature of dievodiamine is the ethylene bridge link-
ing the two indole-containing portions. This is convenient
from a synthetic viewpoint, as it provides a handle for
a convergent synthesis, via the cross-coupling of two
evodiamine-like fragments (Figure 1). With this in mind,
and in view of the diverse biological profile of evodiamine
and related compounds, we decided to embark on the
total synthesis of (()-dievodiamine, to confirm the re-
ported structure and to enable its therapeutic potential to
be better examined.
This was then converted to DHED HCl 4 by heating with
3
dimethyl anthranilate 10 and POCl3, using a procedure
modified from that of Decker.10 In our hands, a particu-
larly convenient purification was developed; the crude
reaction mixture was poured into water, and the resulting
precipitate was removed by filtration, rinsed with water,
and dried, affording the desired quinazolinium salt 4 as a
yellow solid in high yield. It was then planned to trap this
adduct with an organometallic species. However, sur-
prisingly little is known about the reactivity of DHED
systems11 and, to the best of our knowledge, there are no
reports of CÀC bond formation at the electrophilic carbon
of any DHED. To test this idea, a small excess of
((trimethylsilyl)ethynyl)lithium was added to a suspension
of DHED HCl 4 in THF at À78 °C and allowed to warm
3
to RT before quenching with water. As expected, only a
trace amount of alkyne 11 was isolated, with the bulk
of the starting material 4 recovered by filtration of the
crude reaction mixture. In contrast, when 3 equiv of
((trimethylsilyl)ethynyl)lithium were used all of the start-
ing material 4 was consumed and 11 was isolated cleanly,
suggesting that 1 equiv of the organolithium species must
deprotonate the indole, before the requisite nucleophilic
addition takes place. Conveniently, the progress of the
reaction could be monitored visually, as the mixture
became homogeneous upon completion of the reaction.
Following aqueous workup and treatment of the inter-
mediate alkyne 11 with TBAF, alkynyl dihydroquinazoli-
none 12 was isolated in 90% yield over the two-step
sequence. Of course, the TMS group present during this
sequence necessitates a separate cleavage step and there-
fore does not satisfy the ideals of a protecting group-free
synthesis. Thus, the same transformation was attempted
using an excess of a lithium acetylide ethylenediamine
complex. This was unsucessful, but the use of an excess
of commercially available ethynylmagnesium chloride did
give product 12. Initial results were disappointing, how-
ever, as under the conditions described above the desired
alkyne 12 was only isolated in trace amounts, with the
bulk of the material remaining insoluble as the reaction
progressed and was lost during aqueous workup. The
Figure 1. Retrosynthetic strategy.
Our convergent retrosynthetic strategy hinged upon the
Stille reaction of two indole-containing fragments 3 and 5.
It was thought that the requisite stannane 3 could be
obtained via the novel addition of a metalated alkyne into
poor solubility of DHED HCl 4 was thought to be a
3
limiting factor in this reaction, and pleasingly, when the
solvent was switched to toluene, and lithium chloride12
was included as an additive, alkyne 12 was isolated in a
much improved yield following a single, truly protect-
ing-group-free transformation. Finally, hydrostannyl-
ation with tributyltin hydride and AIBN in refluxing
benzene completed the synthesis of stannane coupling
partner 3, which was isolated as a single regio- and
stereoisomer in reasonable yield. It should be noted
that column chromatography was only used in the final
step of either of these three- or four-step sequences from
lactam 9.
dehydroevodiamine hydrochloride (DHED HCl, 4), itself
3
an alkaloid derived from Evodia rutaecarpa, followed by
hydrostannylation. It was hoped that the 3-iodo-indole
fragment 5 could be synthesized from 2-(methylamino)-
benzamide 6 and commercially available indole-2-
carboxylic acid 7 (Figure 1).
The synthesis began with the conversion of indole 8 into
known lactam 99 via a Curtius rearrangement and sub-
sequent electrophilic aromatic substitution (Scheme 1).
(8) (a) Fernandez, L. S.; Buchanan, M. S.; Carroll, A. R.; Feng, Y. J.;
Quinn, R. J.; Avery, V. M. Org. Lett. 2009, 11, 329. (b) Vougogiannopoulou,
K.; Fokialakis, N.; Aligiannis, N.; Cantrell, C.; Skaltsounis, A.-L. Org. Lett.
2010, 12, 1908. (c) Dethe, D. H.; Erande, R. D.; Ranjan, A. J. Am. Chem.
Soc. 2011, 133, 2864. (d) Zeldin, R. M.; Toste, F. D. Chem. Sci. 2011, 2, 1706.
(e) Vallakati, R.; May, J. A. Synlett 2012, 2577. (f) Vallakati, R.; May, J. A.
J. Am. Chem. Soc. 2012, 134, 6936. (g) Welch, T. R.; Williams, R. M.
Tetrahedron 2013, 69, 770.
(10) Decker, M. Eur. J. Med. Chem. 2005, 40, 305.
(11) Limited examples of reduction or hydrolysis reactions of DHED
adducts can be found in refs 5 and 10.
(12) Krasovskiy, A.; Knochel, P. Angew. Chem., Int. Ed. 2004, 43,
3333.
(9) Judd, K. E.; Mahon, M. F.; Caggiano, L. Synthesis 2009, 2809.
B
Org. Lett., Vol. XX, No. XX, XXXX

Scheme 1. Synthesis of Stannane 3
Scheme 2. Synthesis of Indole 5
and led only to the partial reduction of iodide 5 (entry iii)
and its eventual decomposition (entry iv) at elevated
temperatures. The additive Et4NCl, which is more com-
monly used as an additive in Heck reactions,16 has found
limited use in related Stille reactions,17 but under the
conditions trialled (Et4NCl, PdCl2(PPh3)2 at 80 °C in
DMF, entry v) no reaction wasobserved. Itwas considered
that the reason for the poor reactivity of 5 may be due to its
steric bulk inhibiting the transmetalation step. Copper
salts are known to accelerate sluggish Stille couplings by
promoting an initial transmetalation of the organostan-
nane to generate a more reactive organocopper intermedi-
ate.18 Pleasingly when the additives CuI and Et4NCl were
The synthesis of iodide coupling partner 5 was achieved
extremely efficiently from commercially available indole-
2-carboxylic acid 7. Acid chloride formation was followed
by reaction with aniline 613 to form amide 14, which was
then heated at reflux in aqueous KOH.14 The insoluble
material was then collected by filtration, affording quina-
zolinone 15 in excellent yield over the three-step sequence.
The synthesis of indole 5 was completed by reaction with
N-iodosuccinimide in acetone, affording the desired pro-
duct 5 in 77% overall yield from 7 (Scheme 2) following
column chromatography. Note that this was the only
chromatography required throughout the four-step syn-
thesis, which could be performed on a multigram scale.
The limited use of chromatography is an important feature
in the syntheses of both coupling partners 3 and 5, espe-
cially during scale-up. Unprotected indoles are often diffi-
cult to handle due to their relatively low solubility in many
organic solvents, but pleasingly, we were instead able to
exploit this property to our advantage by developing
efficient workup conditions that may not have been pos-
sible had protecting groups been employed on the indole
nitrogen atoms.
Table 1. Optimization of Stille Conditionsa
additives
(equiv)
temp (°C)/
time (h)
outcome
(yield)c
entry
cat.
ib
ii
A
A
A
A
B
B
none
70/20
45/1
no reaction
no reaction
3:2 15:5
CsF (2), CuI (0.1)
CsF (2), CuI (0.1)
CsF (2), CuI (0.1)
Et4NCl (1.0)
Et4NCl (1.0),
CuI (0.1)
iii
iv
v
80/1
100/1
80/20
80/20
decomp.
no reaction
17 (10%)
vi
Conditions for the final Stille coupling were established
using vinyl tributylstannane 16 with iodide 5 (Table 1).
First, no reaction was observed following their treatment
with Pd(PPh3)4 in refluxing THF (entry i). Baldwin’s
conditions,15 which exploit the synergistic effect of CuI
and CsF along with Pd(PPh3)4 in DMF, were also ineffec-
tive on this system, affording no product at 45 °C (entry ii)
vii
B
Et4NCl (1.0),
CuI (1.5)
80/2
17 (81%)
a Reactions were performed on a 0.2À0.5 mmol scale using iodide 5
(1.0 equiv), stannane 16 (1.5 equiv), and a Pd catalyst [A = Pd(PPh3)4 or
B = PdCl2(PPh3)2, 0.05 equiv], with the additives and conditions shown,
in DMF unless stated. b Reaction performed in THF. c Isolated yield
following column chromatography.
(13) Coyne, W. E.; Cusic, J. W. J. Med. Chem. 1968, 46, 4179.
(14) Gardner, B.; Kanagasooriam, A. J. S.; Smyth, R. M.; Williams,
A. J. Org. Chem. 1994, 59, 6245.
(15) Mee, S. P. H.; Lee, V.; Baldwin, J. E. Angew. Chem., Int. Ed.
2004, 43, 1132.
(16) Jeffrey, T. Tetrahedron 1996, 52, 10113.
(17) Kanekiyo, N.; Kuwada, T.; Choshi, T.; Nobuhiro, J.; Hibino, S.
J. Org. Chem. 2001, 66, 8793.
Org. Lett., Vol. XX, No. XX, XXXX
C

combined with PdCl2(PPh3)2 and heated at 80 °C in DMF
(entry vi), this led to the isolation of a small quantity of
coupled product 17. Furthermore, the yield was increased
dramatically by using an excess of CuI (81%, entry vii).
Focus then switched to completing the synthesis of (()-
dievodiamine 1 (Scheme 3). The coupling of iodide 5 with
stannane 3 was slower than with test substrate 17, but
nonetheless, the desired coupled product 1 was obtained in
35% yield using the conditions developed above with a
20h reaction time. Furthermore, increasingthe amounts of
PdCl2(PPh3)2 and Et4NCl (0.2 and 2.0 equiv respectively)
led to a reduced reaction time (2 h) and a cleaner reaction
mixture, allowing the target compound to be isolated in a
much improved 65% yield following column chromato-
graphy and recrystallization.19 The spectral properties of
the synthetic material (1H and 13C NMR data, IR, mass
spectrum)20 closely matched those reported for the natural
product, with the 13C NMR data being particularly con-
clusive (see Supporting Information), thus confirming its
assigned structure.2
Scheme 3. Total Synthesis of (()-Dievodiamine
Not only did this help to reduce the total number of steps,
but also imparted suitable solubility properties that en-
abled us to minimize chromatography, thus facilitating
scale-up. Key steps include the first example of an orga-
nometallic addition into a DHED adduct and a Stille
reaction in which two sterically hindered components
coupled using PdCl2(PPh3)2 and the unusual combination
of Et4NCl and CuI as additives. Future work will focus on
testing the biological properties of dievodiamine and, if
these results show promise, performing SAR studies on
related analogues.
The first total synthesis of (()-dievodiamine 1 has there-
fore been completed. The two key coupling partners 3 and
5 were each synthesized in just four steps, in 33% and 77%
yield respectively, and the final Stille coupling completed
the synthesis of this potentially valuable natural product in
65% yield. The brevity and efficiency of the synthesis was
undoubtedly aided by the absence of protecting groups.
Acknowledgment. The authors thank the EPSRC for
postdoctoral support (W.P.U., EP/G068313/1) and the
University of York Wild Fund for a PhD bursary (C.K.)
(18) Han, X.; Stoltz, B. M.; Corey, E. J. J. Am. Chem. Soc. 1999, 121,
7600.
Supporting Information Available. Synthetic proce-
dures and spectral data. This material is available free
of charge via the Internet at http://pubs.acs.org.
(19) A lower yield (22%) was obtained when Et4NCl was omitted
under otherwise identical conditions, and no product was isolated in the
absence of CuI, thus confirming the importance of both additives.
(20) We also obtained a melting point for 1, but no literature melting
point has been reported for comparison.
The authors declare no competing financial interest.
D
Org. Lett., Vol. XX, No. XX, XXXX