Synthesis of conolidine, a potent non-opioid analgesic for tonic and persistent pain

Article abstract of DOI:10.1038/nchem.1050

Management of chronic pain continues to represent an area of great unmet biomedical need. Although opioid analgesics are typically embraced as the mainstay of pharmaceutical interventions in this area, they suffer from substantial liabilities that include

Full text of DOI:10.1038/nchem.1050

ARTICLES
PUBLISHED ONLINE: 23 MAY 2011 | DOI: 10.1038/NCHEM.1050
Synthesis of conolidine, a potent non-opioid
analgesic for tonic and persistent pain
Michael A. Tarselli1, Kirsten M. Raehal2,3, Alex K. Brasher1, John M. Streicher2,3, Chad E. Groer2,3,
Michael D. Cameron2, Laura M. Bohn2,3 and Glenn C. Micalizio1
*
*
Management of chronic pain continues to represent an area of great unmet biomedical need. Although opioid analgesics
are typically embraced as the mainstay of pharmaceutical interventions in this area, they suffer from substantial liabilities
that include addiction and tolerance, as well as depression of breathing, nausea and chronic constipation. Because of their
suboptimal therapeutic profile, the search for non-opioid analgesics to replace these well-established therapeutics is an
important pursuit. Conolidine is a rare C5-nor stemmadenine natural product recently isolated from the stem bark of
Tabernaemontana divaricata (a tropical flowering plant used in traditional Chinese, Ayurvedic and Thai medicine).
Although structurally related alkaloids have been described as opioid analgesics, no therapeutically relevant properties of
conolidine have previously been reported. Here, we describe the first de novo synthetic pathway to this exceptionally rare
C5-nor stemmadenine natural product, the first asymmetric synthesis of any member of this natural product class, and the
discovery that (+)-, (1)- and (2)-conolidine are potent and efficacious non-opioid analgesics in an in vivo model of tonic
and persistent pain.
he search for the next generation of therapeutics for the treat- of this natural product class has been described, and no robust
ment of pain continues to define an active area of scientific source of conolidine has been identified to fuel medicinal evalu-
T
pursuit. Among current pharmaceutical interventions, ation. The lack of a source of 1, the scarcity of chemical approaches
opioid analgesics (i.e. Fig. 1a) represent the most widely embraced1. suitable for the synthesis of the C5-nor stemmadenine skeleton and
Unfortunately, these agents are clinically problematic as a result of the compelling drug-like molecular structure of conolidine motiv-
their well-established adverse properties that include addiction, tol- ated our initial chemical investigations aimed at securing ample
erance, depression of breathing, nausea and chronic constipation2. quantities of this rare alkaloid by synthetic means.
Hence the identification of effective non-opioid analgesics to
replace these well-established therapeutics is widely held as an
important area of investigation. Here, we describe an interdisciplin-
ary study that has led to the discovery that a rare plant-derived
natural product, first isolated in 20043, possesses potent non-
opioid analgesic properties and is effective in alleviating chemically
induced, inflammatory and acute tonic pain. Overall, efforts have
defined the first de novo synthetic pathway to an exceptionally
rare C5-nor stemmadenine natural product, and the first asym-
metric synthesis of any natural product in this class. These advances
have fuelled the discovery that (+)-, (þ)- and (2)-conolidine are
potent non-opioid analgesics in vivo.
Tabernaemonta divaricata is a flowering tropical plant that has
been used historically in traditional Chinese, Ayurvedic and Thai
medicines, with applications spanning treatment of fever, pain,
scabies and dysentery4. The search for the medicinally relevant
components of this plant has resulted in the isolation of a vast
a
b
HO
MeO
MeO
N Me
N
Me
N Me
O
O
O
H
H
OH
Oxycodone
O
O
HO
Morphine
Hydrocodone
N
N
5
N
5
N
H
H
H
R
O
H
HN
HN
Me
Me
Me
HO
1
2
3
Conolidine
Pericine
Stemmadenine
(R=CO2Me)
array of indole alkaloids that possess diverse biological profiles4. Figure 1 | Opioid analgesics and stemmadenine-based alkaloids.
Conolidine (1) is an exceedingly rare component of a Malayan a, Common opioid analgesics morphine, hydrocodone and oxycodone.
T. divaricata, isolated in only 0.00014% yield from the stem bark Although extremely effective, well-known side effects compromise
of this small flowering plant (Fig. 1b)3. Although no biological or the therapeutic profile of these agents. b, Conolidine, pericine and
medicinal properties of conolidine have been described, other stemmadenine are rare alkaloids derived from plants used in traditional
more abundant alkaloids common to this species have been impli- Chinese, Ayurvedic and Thai medicine. Conolidine itself can be isolated in
cated as opioid analgesics4,5.
just 0.00014% yield from the stem bark of T. divaricata. The low natural
The molecular structure of 1 defines it as a member of the C5- abundance, established utility of T. divaricata in traditional medicine and
nor stemmadenine family of natural products, other members of synthetic challenge posed by the structure of conolidine served to motivate
which have represented significant challenges to modern asym- our pursuit of a programme to accomplish the first synthetic pathway to
metric synthesis. In fact, no asymmetric synthesis of any member conolidine and evaluate its biological properties.
1Department of Chemistry, The Scripps Research Institute, Jupiter, Florida, USA, 2Department of Molecular Therapeutics, The Scripps Research Institute,
Jupiter, Florida, USA, 3Department of Neuroscience, The Scripps Research Institute, Jupiter, Florida, USA. e-mail: lbohn@scripps.edu; micalizio@scripps.edu
*
NATURE CHEMISTRY | VOL 3 | JUNE 2011 | www.nature.com/naturechemistry
© 2011 Macmillan Publishers Limited. All rights reserved.
449

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1050
ARTICLES
a
Excision of C5
H
N
N
5
N
(1) H2O2
N
R
(2) (CF3CO)2O, –5 °C
then dilute NaOH
H
R
H
H
R
N
H
HN
HO
HO
R2O
Me
Me
Me
3
4
5
H
N
b
PMB
H
N +
N
N
20
N
15
+
1
19
HO
HO
N
H
H
19
H
R
Me
H
Me
O
H
N
O
Me
Me
Me
H
6
7
(S)-8
9
Figure 2 | Development of a synthesis strategy for conolidine inspired by the biosynthetic proposal for the conversion of stemmadenine to vallesamine.
a, Biomimetic semisynthesis of vallesamine (5) from stemmadenine (3). b, Retrosynthesis of conolidine (1) that features a cyclization process supported by
the minimization of A-1,3 strain (7 ꢀ 1) — a strategy inspired by the conformational constraints likely operative in the cyclization of 4 (a).
PMB
N
PMB
N
PMB
N
(1) 10, Dess-Martin
(1) KH, Bu3SnCH2I
THF
N
PMBCl, DCE, reflux
periodinane, DCM/H2O
20
+
Me
HO
HO
19
(2) 12, THF
(2) n-BuLi, THF
then, NaBH4, MeOH
57%
76%
Me
Me
Me
10
OH
OH
Li
N
9
8
(E:Z =12:1) 11
12
PhO2S
PMB
N
H
N
(1) Na(Hg) amalgam
N
N
CF3CO2H, then
(2) MnO2, DCM
+
N
N
(3) ACE-Cl, DCE,
84°C
(43% over 5 steps)
H
O
H
H
Me
(CH2O)n, CH3CN
80°C
Me
Me
N
H
O
H
N
HO
O
Me
H
PhO2S
95%a
1
( )-Conolidine
9-steps from 9
18% overall yield
13
7
Figure 3 | Execution of a synthesis pathway to conolidine (1). Synthesis of (+)-conolidine (1) is accomplished in just nine steps and 18% overall yield from
the commercially available pyridine 9. A Still’s allyl stannylmethyl ether-based [2,3]-Wittig rearrangement is used as a key stereoselective transformation in
this pathway that culminates in the acid-mediated conversion of 7 to conolidine 1. aYield reported is for the preparation of the 2 H2SO4 salt (the compound
1
used in subsequent in vivo experiments); spectral data reported in the Supplementary Information are of the free base of 1. PMB, p-methoxybenzyl; DCE,
1,2-dichloroethane; DCM, dichloromethane; ACE-Cl, a-chloroethyl chloroformate.
stability to the potentially unstable b,g-unsaturated ketone 7 (ref. 9).
The preparation of substrate 7 was then deemed possible from
allylic alcohol 8, and ultimately from the readily available pyridine 9.
As depicted in Fig. 3a, efforts commenced with commercially
available pyridine 9. Initial N-alkylation, followed by hydride
reduction, provided 8 in 57% yield. With precedent firmly estab-
lished for stereoselective Claisen rearrangement of related substrates
providing the undesired alkene isomer11,12, an unusual stereoselective
transformation of 8 was required to establish the desired alkene
geometry at C19–C20. Although our initial experiments aimed at
accomplishing such a stereoselective transformation focused on the
application of reductive cross-coupling chemistry between 8 and a
suitably functionalized alkyne13, these studies encountered unantici-
pated failure as a result of low levels of reactivity associated with the
allylic alcohol component in this class of metallacycle-mediated con-
vergent coupling reaction. We sought to identify a stereoselective
transformation suitable for the conversion of allylic alcohol 8 to a
functionalized (E)-exo-ethylidine-containing product.
Results and discussion
At the outset of our studies, two anticipated challenges guided the
conception of a chemical pathway to this class of natural products:
(1) formation of a strained 1-azabicyclo[4.2.2]-decane and (2)
stereochemical control in the generation of the exocyclic
trisubstituted alkene.
Our retrosynthetic planning began with an examination of the
biosynthetic pathway for conversion of stemmadenine (3) to valles-
amine (5), a proposal later realized in vitro by Scott and colleagues
(Fig. 2a)6,7. Here, excision of the C5 carbon occurs by a sequence of
(1) N-oxidation, (2) fragmentation, (3) hydrolysis and (4) cycliza-
tion (through 4)8.
The ease with which cyclization takes place in this system is prob-
ably influenced by a conformational preference imparted by the
stereodefined C19–C20 (E)-ethylidene unit, thereby providing a
bias to support cyclization through 4 (Fig. 2a). With this as a
guiding principle, we speculated that conolidine (1) could derive
from iminium ion 6 through a related ring closure by means of a
conformation that would be similarly stabilized by minimization
of allylic strain (Fig. 2b)9. While providing a conformational bias
to facilitate the establishment of the 1-azabicyclo[4.2.2]-core, the
stereochemistry of this alkene was expected to impart kinetic
With this goal in mind, we turned our attention to Still’s allyl
stannylmethyl ether-based [2,3]-Wittig rearrangement14. Whereas
few reports describe the application of this reaction to the establish-
ment of stereodefined exo-alkylidene cycloalkanes15–17, the unique
450
NATURE CHEMISTRY | VOL 3 | JUNE 2011 | www.nature.com/naturechemistry
© 2011 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1050
ARTICLES
PMB
PMB
N
( )-8
N
N
N
8-stepsb
8-stepsb
HO
HO
N
N
H
H
O
H
O
Amano lipase PS
vinyl acetate/hexanes
4 Å mol. sieves, 18h
H
Me
Me
Me
40%a
Me
(–)-1
39%
(+)-1
≥ 90% e.e.
(≥ 92% e.e.)
(≥ 92% e.e.)
≥ 90% e.e.
Figure 4 | Asymmetric synthesis of (1)- and (2)-conolidine. Resolution of racemic 8, by lipase-catalysed acylation, defines a robust means to access
optically active samples of either antipode of conolidine (1). aAfter hydrolysis. bApart from the conditions for oxidation of an intermediate primary alcohol,
this eight-step sequence is identical to that described in Fig. 3.
stereochemical course of this transformation in acyclic settings pad produces a biphasic pain response that is demonstrated by the
suggested its potential utility in solving the problem at hand.
animals attending to and licking the injected paw (nocifensive beha-
As illustrated in Fig. 3, allylic alcohol 8 was converted to its corre- viours)21. Responses to the initial phase (Phase 1) occur within
sponding tributylstannylmethyl ether, and subsequently exposed to the first 10 min of formalin injection and reflect the direct chemical
n-BuLi. The products from the [2,3]-Wittig rearrangement were gen- action at nociceptive primary afferent nerve fibres22. Following a
erated in 76% yield (over two steps), favouring the formation of the brief intermediary phase (10–20 min after formalin injection) in
desired isomer 10 (10:11 ¼ 12:1). Oxidation to the corresponding which no pain response occurs, a second pain response (Phase 2)
b,g-unsaturated aldehydewith Dess–Martin periodinane, and nucleo- becomes evident (20–40 min after formalin injection) during
philic addition of the organolithium reagent 12, then gave a mixture of which the animal once again attends to the injected paw. This
stereoisomeric alcohols 13 that was subsequently advanced to the second phase models a persistent pain state that is believed to be
cyclization substrate 7 by a simple three-step sequence.
due in part to injury-provoked sensitization of central nervous
Conversion of amino ketone 7 to its corresponding trifluoroace- system neurons and to inflammatory factors18,23,24. The application
tate salt, followed by treatment with paraformaldehyde in aceto- of local anaesthetics, such as lidocaine, suppresses the first phase of
nitrile at 80 8C, delivered (+)-conolidine 1 in 95% yield. The the response, whereas non-steroidal anti-inflammatory drugs are
success of this process validated our proposed strategy to access effective in blocking the second phase. Few drugs, such as the
the strained azabicyclo[4.2.2] core of the C5-nor stemmadenine opioid narcotics, are effective in suppressing both phases of the
family of natural products, and delivered the first synthetic sample pain response18.
of conolidine by a synthesis pathway that proceeds in just nine
steps and 18% overall yield from a readily available pyridine.
Using a well-characterized mouse line in this particular assay,
the sulfate salts of (+)-, (þ)- and (2)-conolidine were administered
Building on the successful synthesis of racemic 1, we explored the (10 mg kg21, i.p.) 15 min before formalin (5%, 25 ml, intraplantar)
potential of this route to accomplish the first asymmetric synthesis challenge20. Compared with vehicle, all three forms of conolidine
of a C5-nor stemmadenine alkaloid. As illustrated in Fig. 4, resol- sulfate reduce the display of nocifensive behaviours in both Phase
ution of (+)-8 with a commercially available lipase provided opti- 1 (P, 0.001) and Phase 2 (P, 0.001, two-way analysis of variance
cally enriched substrates (≥92% e.e.) that were advanced to (þ)- (ANOVA)) (Fig. 5b). Further, the unnatural antipode (2)-conoli-
or (2)-conolidine (≥90% e.e.) by a simple eight-step sequence. dine is effective in a dose-dependent manner (Fig. 5c), displaying
Here, the only preparative modification required was a substitute potency similar to that shown for morphine sulfate (see Fig. 5d
for the Dess–Martin oxidation used to generate the b,g-unsaturated and Table 1)20.
aldehyde en route to enantio-enriched carbinol 13. Whereas
To investigate the duration of drug action, (2)-conolidine sulfate
Dess–Martin oxidation of optically active samples of 10 resulted was administered at increasing time intervals before formalin chal-
in inconsistent partial racemization, a modified Parikh–Doering lenge. Whereas one-way ANOVA reveals no time effect for either
oxidation ensured a reproducible pathway to enantio-defined phase (P . 0.05), subsequent Student’s t-test analysis reveals that
samples of either enantiomer of conolidine.
although conolidine significantly attenuates the pain response com-
With a concise asymmetric chemical pathway established, and pared with vehicle at 15 and 30 min before injection in Phase 1, and
the first synthetic samples of (þ)-, (2)- and (+)-conolidine in at 15 min before that in Phase 2, it was less efficacious when admi-
hand, hundreds of milligrams of synthetic material was prepared nistered 1 h before formalin challenge in Phase 1 or 30–60 min
and our attention shifted to exploring the potential analgesic prop- before that in Phase 2 (Fig. 5e).
erties of these substances in vivo.
Taken together, these data demonstrate that, like opioid analge-
Because related members of this natural product class have been sics, conolidine is effective in suppressing responses to both
postulated to have opioid properties4,5, we assessed the efficacy of chemically induced and inflammation-derived pain20. Further,
conolidine in several nociceptive tests in C57BL/6J mice. Unlike pharmacokinetic measures demonstrate that conolidine is present
opioids, (þ)- and (2)-conolidine sulfate (1) do not produce antino- in the brain and plasma at relatively high concentrations during
ciception in the hot plate (51 8C) or in the warm water tail immer- the period of its analgesic efficacy and remains present up to 4 h
sion assay (49 8C) following systemic injection of 10 mg kg21
intraperitoneally (i.p.) (Supplementary Fig. S1). However, in a
,
after injection (Fig. 5f)25.
To further demonstrate that conolidine differs from opioid
visceral pain model, the acetic acid abdominal constriction analgesics, pharmacological studies were performed. On the basis
assay18, conolidine (10, 20 mg kg21) proved to be efficacious in of b-arrestin translocation assays, radioligand binding displacement
preventing the acetic acid-induced writhing response (Fig. 5a). and G-protein coupling assays, conolidine has no affinity
Therefore, conolidine differs from morphine in that it does not for or efficacy at the m-opioid receptor, the primary target
promote antinociception to acute thermal stimulation; however, of morphine (Supplementary Fig. S2 and Table S1). Moreover,
like morphine, and a host of known analgesics, it does suppress radioligand displacement and b-arrestin translocation studies also
the acetic acid-induced writhing response19,20
A pain model designed to assess both acute tonic and persistent tors (Supplementary Fig. S2a and Table S1).
.
demonstrate a lack of affinity or efficacy for k- and d-opioid recep-
pain responses was also used to ascertain conolidine’s analgesic effi-
In an attempt to determine the biological target of conolidine, we
cacy. The injection of a dilute formalin solution into a rodent’s paw subjected (+)-conolidine sulfate to screening by the Psychoactive
NATURE CHEMISTRY | VOL 3 | JUNE 2011 | www.nature.com/naturechemistry
© 2011 Macmillan Publishers Limited. All rights reserved.
451

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1050
ARTICLES
(–) Conolidine sulfate
Vehicle
(+)Conolidine
(–)Conolidine
Vehicle
(+)Conolidine
(–)Conolidine
Vehicle
10mgkg–1
(
)Conolidine
(
)Conolidine
5mg kg–1
20mgkg–1
a
c
b
250
200
150
100
50
250
200
150
100
50
60
40
20
0
60
40
20
0
40
30
20
10
0
*
**
*
**
**
**
***
***
***
**
**
**
**
***
0
0
Veh 10 10 10 20
Phase 1
Phase 2
Phase 1
Phase 2
Conolidine sulfate
(mgkg–1, i.p.)
Morphine sulfate
Vehicle
(–)Conolidine
Vehicle
4.5mgkg–1
3mg kg–1
6mgkg–1
e
f
80
60
40
20
0
d
–5
Phase 1
Phase 2
250
200
150
100
50
250
200
150
100
50
Brain
Plasma
60
40
20
0
–5.5
–6
#
**
*
*
**
–6.5
–7
**
***
***
0
0
15 30 60
15 30 60
Phase 1
Phase 2
0 30 60 120
Time (min)
240
Time before formalin challenge (min)
Figure 5 | Conolidine is antinociceptive in visceral, tonic and persistent pain models and is present at micromolar levels in the brain after systemic
injection. Male adult (12–16 weeks old) C57BL/6J mice from Jackson Labs were treated with conolidine sulfate or vehicle (10% dimethyl sulfoxide in water,
10
in H2O)19. Abdominal constrictions were recorded during the 30 min immediately following acetic acid injection. Drug effect versus vehicle: **P , 0.01,
l, 5% in saline) was injected into the hind-paw pad (intraplantar) 15 min after
m
l g21, i.p.) at the times indicated. a, Conolidine sulfate (10 or 20 mg kg21, i.p. of the indicated enantiomer) was injected 15 min before acetic acid (0.75%
***P , 0.001, Student’s t-test, n ¼ 4–8 mice per treatment. b, Formalin (25
m
conolidine sulfate (10 mg kg21, i.p.); paw licking was recorded for 40 min following formalin injection. Phase 1 refers to the sum of time spent licking in the
first 10 min of response and Phase 2 represents the 20–40-min response period. Drug effect versus vehicle: *P , 0.05, **P , 0.01, ***P , 0.001, Student’s
t-test, n ¼ 6–12 mice per treatment. c, Dose effect of (2)-conolidine sulfate in each phase of the formalin test: drug effect by one-way analysis of variance
(ANOVA): P ¼ 0.0004 for Phase 1 and P , 0.0001 for Phase 2; Bonferroni post-hoc analysis: vehicle vs. drug within each phase: **P , 0.01; ***P , 0.001,
n ¼ 7–12 mice per treatment. d, Dose effect of morphine sulfate in 10% dimethyl sulfoxide (i.p.) compared to vehicle. Drug effect by one-way ANOVA:
P ¼ 0.03 for Phase 1 and P , 0.0001 for Phase 2; Bonferronni post-hoc analysis: vehicle vs. drug within each phase: *P , 0.05; **P , 0.01; ***P , 0.001,
n ¼ 4–12 mice per treatment. e, Time effect of (2)-conolidine sulfate suppression of both phases of the formalin response. Conolidine (10 mg kg21, i.p.) or
vehicle was administered at the times indicated before formalin challenge. Conolidine vs. vehicle: *P , 0.05; **P , 0.01; 60 vs. 15 min: #P , 0.05, Student’s
t-test, n ¼ 6–12 mice per treatment. f, Pharmacokinetic profile of (2)-conolidine sulfate in the plasma and brain after a 10 mg kg21 i.p. injection in male
C57BL/6J mice at the times indicated20. In all cases, the mean+standard error of the mean is given; statistical analyses were performed with GraphPad
Prism 5.0 (GraphPad Software, San Diego, CA).
Drug Screening Program (PDSP) sponsored by the National Institute
To gain further insight into conolidine’s pharmacological profile
of Mental Health (Bethesda, Maryland). Conolidine was able to dis- in vivo, we performed open-field activity monitoring in C57BL/6J
place more than 50% of radioligand binding to the serotonin-3 recep- mice. Although conolidine (10 mg kg21, i.p.) readily enters the
tor ion channel, the norepinephrine transporter as well as the a2B brain, it does not alter locomotor activity—specifically horizontal
adrenergic, a2C adrenergic and histamine-2 receptors, with low affi- activity, total distance travelled, vertical rearing or stereotypic move-
nity (Ki . 1 mM, Supplementary Table S2). Functional assays, ments—in C57BL/6J mice (Supplementary Fig. S3). This is in direct
however, show marginal efficacy and potency for conolidine as an contrast to morphine, which is classically known to stimulate loco-
agonist or antagonist at these potential protein targets motor activity in mice33. Although it is clear that conolidine is not
(Supplementary Table S2), suggesting that these candidates may not an opioid analgesic, its mechanism of action remains to be deter-
represent the primary mechanism of action for this compound.
mined. Regardless, its lack of adverse effects in the open-field test
There may be some significance in the fact that conolidine does is encouraging because this may be indicative of fewer side effects,
not produce antinociception in the hot plate and tail flick assays, yet such as sedation, or elevations of dopamine—a characteristic of all
is efficacious in other pain tests because it may give some clues as to addictive substances34.
its mechanism of action. For example, if conolidine were acting at
Conolidine’s analgesic efficacy is promising, and the enantio-
M2/M4 muscarinic receptors26, a2C receptors27 and opioid recep- convergence in analgesic efficacy is interesting (Fig. 5a,b). If both anti-
tors or by blocking serotonin or norepinephrine transporters28, podes of 1 operate by a similar pharmacological mechanism, then the
some degree of hot plate and/or tail flick antinociception would local symmetry about the azabicyclo[4.2.2]-decane indicates a toler-
be expected. Interestingly, other analgesics, such as gabapentin, do ance for ethylidene substitution at either site of the system. Future
not suppress responses in the acute thermal antinociception tests; studies will explore the impact that diverse substitution at these sites
moreover, gabapentin, which also lacks a definitive mechanism of has on analgesic properties of unnatural C5-nor stemmadenines.
action, is primarily effective against the second phase of the forma-
Overall, we describe the first total synthesis of the rare C5-nor
stemmadenine alkaloid, conolidine, and the first asymmetric
lin test29–32 when administered systemically.
452
NATURE CHEMISTRY | VOL 3 | JUNE 2011 | www.nature.com/naturechemistry
© 2011 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1050
ARTICLES
16. Wang, X. J., Hart, S. A., Xu, B., Mason, M. D., Goodell, F. R. & Etzkorn, F. A.
Serine-cis-proline and serine-trans-proline isosteres: stereoselective synthesis of
(Z)- and (E)-alkene mimics by Still–Wittig and Ireland–Claisen rearrangements.
J. Org. Chem. 68, 2343–2349 (2003).
Table 1 | Relative drug potencies in both phases of the
formalin test (i.p.; 95% confidence interval) from data
presented in Fig. 5c,d.
17. Fristad, W. E. & Paquette, L. A. Oxidation chemistry of a cis-2-(3-
flurylidene)ethanol. Heterocycles 31, 2219–2224 (1990).
Relative potencies (ED50, mg kg21
)
18. Le Bars, D., Gozariu, M. & Cadden, S. W. Animal models of nociception.
Pharmacol. Rev. 53, 597–652 (2001).
19. Collier, H. O. J., Dinneedn L. C., Johnson, C. A. & Schneider, C. The abdominal
constriction response and its suppression by analgesics in the mouse. Br. J.
Pharmac. Chemother. 32, 295–310 (1968).
Drug
Phase 1
Phase 2
(2)-Conolidine sulfate
Morphine sulfate
5.6 (4.0–7.8)
4.6 (3.3–6.4)
6.0 (3.7–9.7)
2.4 (1.7–3.2)
20. Mogil, J. S., Kest, B., Sadowski, B. & Belknap, J. K. Differential genetic mediation
of sensitivity to morphine in genetic models of opiate antinociception: influence
of nociceptive assay. J. Pharmacol. Exp. Ther. 276, 532–544 (1996).
21. Dubuisson, D. & Dennis, S. G. The formalin test: a quantitative study of the
analgesic effects of morphine, meperidine, and brain stem stimulation in rats
and cats. Pain 4, 161–174 (1977).
22. McNamara, C. R. et al. TRPA1 mediates formalin-induced pain. Proc. Natl Acad.
Sci. USA 104, 13525–13530 (2007).
23. Woolf, C. J. Evidence for a central component of post-injury pain
hypersensitivity. Nature 306, 686–688 (1983).
synthetic pathway to any C5-nor stemmadenine. With an efficient
source of this alkaloid secured by synthetic means (nine steps,
18% overall yield), the production of sufficient quantities of the
natural product (and related antipode) was followed by the first
evaluation of this alkaloid in vivo. These combined studies have
resulted in the discovery that conolidine (1) is a potent non-
opioid analgesic that is effective at alleviating chemically induced,
acute and persistent tonic pain. Further studies in neuropathic
pain models will be undertaken to determine more widespread
therapeutic promise for the treatment of chronic pain. Finally,
although determination of the pharmacological mechanism of
action associated with the potent analgesic properties of this
alkaloid remains an area of intense current investigation, the
results of these studies mark the establishment of a chemical
foundation suitable for investigating the therapeutic potential of
this unique alkaloid as a potent non-opioid analgesic.
24. Tjølsen, A., Berge, O. G., Hunskaar, S., Rosland, J. H. & Hole, K. The formalin
test: an evaluation of the method. Pain 51, 5–17 (1992).
25. Li, X., Kamenecka, T. M. & Cameron, M. D. Bioactivation of the epidermal
growth factor receptor inhibitor gefitinib: implications for pulmonary and
hepatic toxicities. Chem. Res. Toxicol. 22, 1736–1742 (2009).
26. Wess, J. et al. Muscarinic receptor subtypes mediating central and peripheral
antinociception studied with muscarinic receptor knockout mice: a review. Life
Sci. 72, 2047–2054 (2003).
27. Stone, L. S., Fairbanks, C. A. & Wilcox, G. L. Moxonidine, a mixed alpha(2)-
adrenergic and imidazoline receptor agonist, identifies a novel adrenergic target
for spinal analgesia. Ann. NY Acad. Sci. 1009, 378–385 (2003).
28. Jann, M. W. & Slade, J. H. Antidepressant agents for the treatment of chronic
pain and depression. Pharmacotherapy 27, 1571–1587 (2007).
29. Shimoyama, N., Shimoyama, M., Davis, A. M., Inturrisi, C. E. & Elliot, K. J.
Spinal gabapentin is antinociceptive in the rat formalin test. Neurosci. Lett. 222,
65–67 (1997).
30. Hunter, J. C. et al. The effect of novel anti-epileptic drugs in rat
experimental models of acute and chronic pain. Eur. J. Pharmacol. 324,
153–160 (1997).
31. Munro, G. Pharmacological assessment of the rat formalin test utilizing the
clinically used analgesic drugs gabapentin, lamotrigine, morphine, duloxetine,
tramadol and ibuprofen: influence of low and high formalin concentrations.
Eur. J. Pharmacol. 605, 95–102 (2009).
Received 16 December 2010; accepted 4 April 2011;
published online 23 May 2011
References
1. Melnikova, I. The pain market. Nat. Rev. Drug Discov. 9, 589–590 (2010).
2. Crofford, L. J. Adverse effects of chronic opioid therapy for chronic
musculoskeletal pain. Nat. Rev. Rheum. 6, 191–197 (2010).
3. Kam, T.-S., Pand, H.-S., Choo, Y.-M. & Komiyama, K. Biologically active ibogan
and vallesamine derivatives from Tabernaemontana divaricata. Chem. Biodivers.
1, 646–656 (2004).
4. Pratchayasakul, W., Pongchaidecha, A., Chattipakorn, N. & Chattipakorn, S.
Ethnobotany & ethnopharmacology of Tabernaemontana divaricata. Indian J.
Med. Res. 127, 317–335 (2008).
5. Ingkaninan, K., Ijzerman, A. P., Taesotikul, T. & Verpoorte, R. Isolation of
opioid-active compounds from Tabernaemopntana pachysiphon leaves.
J. Pharm. Pharmacol. 51, 1441 (1999).
6. Scott, A. I., Yeh, C.-L. & Greenslade, D. Laboratory model for the biosynthesis of
vallesamine, apparicine, and related alkaloids. J. Chem. Soc. Chem. Commun.
947–948 (1978).
7. Lim, D-H., Low, T-Y. & Kam, T-S. Biomimetic oxidative transformations of
pericine: partial synthesis of apparicine and valparicine, a new pentacyclic indole
alkaloid from Kopsia. Tetrahedron Lett. 47, 5037–5039 (2006).
32. Baillie, J. K. & Power, I. The mechanism of action of gabapentin in neuropathic
pain. Curr. Opin. Investig. Drugs 7, 33–39 (2006).
33. Goldstein, A. & Sheehan, P. Tolerance to opioid narcotics. I. Tolerance to the
“running fit” caused by levorphanol in the mouse. J. Pharmacol. Exp. Ther. 169,
175–184 (1969).
34. Wise, R. A. & Bozarth, M. A. A psychomotor stimulant theory of addiction.
Psychol. Rev. 94, 469–492 (1987).
Acknowledgements
We gratefully acknowledge T.-S. Kam (University of Malaya, Kuala Lumpur, Malaysia) for
providing authentic spectra of natural (þ)-conolidine for comparison with our synthetic
samples (see Supplementary Information).
´
8. Ahond, A., Cave, A., Kan-Fan, C., Langlois, Y. & Potier, P. The fragmentation of
N,N-dimethyltryptamine oxide and related compounds: a possible implication
in indole alkaloid biosynthesis. J. Chem. Soc. Chem. Commun. 517 (1970).
9. Hoffmann, R. W. Allylic 1,3-strain as a controlling factor in stereoselective
transformations. Chem. Rev. 89, 1841–1860 (1989).
Author contributions
´
´
´
10. Bennasar, M.-L., Zulaica, E., Sole, D., Roca, T., Garcıa-Dıaz, D. & Alonso, S.
Total synthesis of the bridged indole alkaloid apparicine. J. Org. Chem. 74,
8359–8368 (2009).
G.C.M. conceived, initiated and directed the project. M.A.T. and A.K.B. conducted all
chemical experiments. L.M.B. initiated and directed the in vivo and in vitro
pharmacological evaluation. L.M.B. and M.D.C. directed the pharmacokinetic
experiments, and K.M.R. and C.G. conducted all biochemical and in vivo experiments.
Receptor binding profiles were generously provided by the National Institute of Mental
Health’s Psychoactive Drug Screening Program, Contract no. HHSN-271-2008-00025-C
(NIMH PDSP). The NIMH PDSP is directed by Bryan L. Roth (MD, PhD) at the University
of North Carolina at Chapel Hill and Project Officer Jamie Driscol at NIMH, Bethesda,
Maryland, USA. G.C.M. and L.M.B. wrote the manuscript.
11. Amat, M., Dolors Coll, M., Passarella, D. & Bosch, J. An enantioselective
synthesis of the Strychnos alkaloid (2)-tubifoline. Tetrahedron: Asymmetry 7,
2775–2778 (1996).
12. Amat, M., Dolors Coll, M., Bosch, J., Espinosa, E. & Molins, E., Total syntheses
of the Strychnos indole alkaloids (2)-tubifoline, (2)-tubifolidine, and (2)-
19,20-dihydroakuammicine. Tetrahedron: Asymmetry 8, 935–948 (1997).
13. Kolundzic, F. & Micalizio, G. C. Synthesis of substituted 1,4-dienes by direct
alkylation of allylic alcohols. J. Am. Chem. Soc. 129, 15112–15113 (2007).
14. Still, W. C. & Mitra, A. A highly stereoselective synthesis of Z-trisubstituted
olefins via [2,3]-sigmatropic rearrangement. Preference for a pseudoaxially
substituted transition state. J. Am. Chem. Soc. 100, 1927–1928 (1978).
15. Hart, S. A., Trindle, C. O. & Etzkorn, F. A. Solvent-dependent stereoselectivity in
a Still–Wittig rearrangement: an experimental and ab initio study. Org. Lett. 3,
1789–1791 (2001).
Additional information
The authors declare no competing financial interests. Supplementary information and
chemical compound information accompany this paper at www.nature.com/
naturechemistry. Reprints and permission information is available online at http://www.
nature.com/reprints/. Correspondence and requests for materials should be addressed to L.M.B.
and G.C.M.
NATURE CHEMISTRY | VOL 3 | JUNE 2011 | www.nature.com/naturechemistry
453
© 2011 Macmillan Publishers Limited. All rights reserved.