RESEARCH NEWS & VIEWS
ME DICINAL CHEMISTRY
New lead for
pain treatment
The synthesis of conolidine, a scarce, naturally occurring compound, has enabled
the first studies of its pharmacological properties to be carried out. Excitingly,
conolidine is a painkiller that seems to have an unusual mechanism of action.
et al.1), might pave the way for the development
of new non-opioid analgesics.
The natural product conolidine was origi­
nally isolated2 in extremely small quantities
— just 0.00014% yield — from the stem bark
of the flowering tropical plant Tabernaemontana divaricata. The low natural abundance
of the compound has hindered the study of
its potential therapeutic properties. By com­
pleting the first total chemical synthesis of
conolidine, Micalizio and co-workers provided
Bohn’s team with enough synthetic material to
carry out the first in vivo studies of its analgesic
properties.
A key challenge in the synthesis of cono­
lidine is the construction of its bicyclic ring
system (Fig. 1a), which consists of an eightmembered ring bridged by two carbon atoms
and contains a nitrogen atom at one of the
SARAH E. REISMAN
S
ome of the most powerful painkillers,
such as morphine, hydrocodone and
oxycodone, belong to the opioid family
of analgesics. Unfortunately, prolonged expo­
sure to these analgesics can cause several
adverse side effects, including physical and
psychological addiction. As a result, there
is a continued effort to identify painkillers
that have different biological mechanisms
from opioids and that elicit fewer side effects.
With this in mind, a team of chemists, led
by Glenn Micalizio, has joined forces with a
group of neuroscientists, headed by Laura
Bohn, to synthesize and study the analgesic
properties of a rare, naturally occurring com­
pound called conolidine. Their promising find­
ings, reported in Nature Chemistry (Tarselli
C5
b
a
N
H
N
O
Me
N
H
H
Me
MeO2C
HO
Stemmadenine
Conolidine
c
HN
Bond breaks
OH
N
+
Oxidation
Stemmadenine
H
MeO2C
Me
HO
Loss of
H2C O
N
HN
N
H
H MeO C
2
Me
HO
H
MeO2C
Me
HO
N
Vallesamine
Cyclization precursor
d
NH
H
Me
O
N
+
Acid
N
H
H2C O
H
Me
O
CH2
N
H
Cyclization precursor
N
H
N
H
Me
O
N
H
Conolidine
Figure 1 | Inspiration for the synthesis of conolidine. a, Conolidine is a scarce, plant-derived
compound. Its core structure is shown in blue. Me is a methyl group. b, Stemmadenine is a related
natural product. Its core structure (red) has one more carbon atom than conolidine; the carbon atom is
known as C5. c, The biosyntheses of compounds containing the core structure of conolidine are thought
to involve a process in which the C5 carbon is excised from the stemmadenine core. In this example,
the oxidation of a nitrogen atom in stemmadenine triggers the loss of C5 as formaldehyde (CH2O) and
generates a cyclization precursor, which undergoes an intramolecular reaction to yield vallesamine as a
product. Curly arrows show the electron movement in the intramolecular reaction. d, In their synthesis
of conolidine, Tarselli et al.1 prepared a compound that, on treatment with an acid and formaldehyde,
formed a cyclization precursor similar to that shown in c. The precursor underwent an intramolecular
reaction that formed the desired product.
4 5 8 | NAT U R E | VO L 4 7 3 | 2 6 M AY 2 0 1 1
© 2011 Macmillan Publishers Limited. All rights reserved
bridgehead positions. In devising a strategy
to prepare this system, the authors turned to
nature for inspiration.
Conolidine is a ‘C5-nor stemmadenine’
natural product, which means that it con­
tains the same carbon skeleton as the more
abundant natural product stemmadenine,
except that it lacks one of the carbon atoms
(known as C5; see Fig. 1b). It has been pro­
posed3 that, in the biosynthesis of C5-nor
stemmadenine compounds, the C5 atom of
a stemmadenine framework is excised in a
process that begins with the oxidation of the
bridgehead nitrogen (Fig. 1c). This proposal
has been validated chemically — stemm­
adenine can be converted to its C5-nor
analogue vallesamine in this manner, albeit in
low (25%) yield4.
Rather than pursuing a strictly biomimetic
sequence in which a compound containing the
stemmadenine framework was converted to
conolidine, Micalizio and the chemistry team1
adopted a more subtle approach. They devel­
oped a short synthesis of a compound that has
all but one of the carbons found in conolidine
(Fig. 1d). They then incorporated the final
carbon by reacting the compound with formal­
dehyde (CH2O), generating a precursor similar
to that produced in the biomimetic synthesis
of vallesamine. This precursor then underwent
an intramolecular cyclization reaction to forge
the critical eight-membered ring of the bicyclic
system.
This approach enabled the team to prepare
conolidine in a more straightforward manner
than would have been possible by first making
the stemmadenine framework and then frag­
menting it. In addition, using this approach,
they were able to independently prepare not
only the naturally occurring isomer of cono­
lidine, but also its unnatural mirror-image
isomer (known as (–)-conolidine), in only ten
synthetic steps from a commercially available
starting material.
With access to synthetic conolidine, Bohn
and her group1 went on to evaluate the com­
pound’s analgesic properties in mice. In
experiments designed to evaluate conolidine’s
effects on both acute and persistent pain,
they found that it was indeed a painkiller of
similar potency to morphine. However, the
authors’ pharmacological studies revealed that
conolidine does not bind to opioid receptors
(the biological targets responsible for the
analgesic effects of morphine and other opi­
oid drugs). Furthermore, the authors found
that conolidine does not seem to adversely
affect the locomotor activity of mice, as opioid
anal­gesics do, suggesting that conolidine might
have fewer side effects than opioids.
Intriguingly, Bohn and colleagues also
observed that (–)-conolidine has compara­
ble in vivo activities to (+)-conolidine, the
naturally occurring isomer. This is unusual,
because the two mirror-image isomers (enan­
tiomers) of a compound commonly elicit
NEWS & VIEWS RESEARCH
different biological responses. For example, the
enantiomer of morphine is a poor painkiller5.
The fact that both enantiomers of conolidine
are analgesic may indicate something about its
biological target. Taken together, the biologi­
cal findings suggest that conolidine may have
a previously undiscovered pharmacological
mechanism for inducing analgesia.
Just what the mechanism of action is has not
been determined, and this is clearly a prior­
ity for future research. Micalizio, Bohn and
colleagues’ preliminary studies1 nevertheless
indicate that conolidine is a promising candi­
date for further study as a non-opioid analge­
sic. Moreover, the authors’ concise, modular
and high-yielding chemical synthesis should
provide ample quantities of conolidine for the
further study and development of this pain­
killer — quantities that would be extremely
difficult to extract from the natural source of
the compound. ■
Sarah E. Reisman is in the Division of
Chemistry and Chemical Engineering,
P RECISIO N MEASUREMEN T
A search for electrons
that do the twist
One might think that physicists know everything about the electron. But the
latest measurement of its shape could alter expectations for results at high-energy
particle accelerators. See Letter p.493
AARON E. LEANHARDT
I
f I were to tell you about an elementary
particle that has mass and charge, but
neither size nor structure, yet still has a
well-defined orientation and can point in a
specific direction in space, you would prob­
ably think I am describing something from a
science-fiction novel. In fact, I am telling you
about the electron. On page 493 of this issue,
Hudson et al.1 describe an experiment aimed
at refining our understanding of this funda­
mental particle and, more broadly, the basic
laws of nature.
Described colloquially, their experiment
searches for evidence of an aspheric distor­
tion to the shape of the electron, or, more
technically, to the shape of its interactions with
electric fields. Hudson et al.1 observe no such
distortion. However, a detailed understanding
of their apparatus allows them to report their
null result as a new limit on the magnitude of
the electric dipole moment of the electron.
This work has important ramifications for
the types of particles that can be discovered
at high-energy accelerators, and may even­
tually help to explain the composition of the
observable Universe.
It is well established that the electron has a
magnetic dipole moment, which means that
it behaves like a tiny bar magnet with north
and south poles. For example, a magnetic field
can rotate the orientation of an electron, just as
it can move the needle of a compass. Hudson
et al.1 are searching for the as-yet undiscovered
electric analogue, the electric dipole moment
of the electron. An electric dipole moment
can be depicted as a battery with positive and
negative terminals, and its orientation can be
rotated by electric fields. Therefore, the experi­
mental effort of Hudson and colleagues can be
viewed as an attempt to answer the question:
does an electric field twist the orientation of
an electron?
It should be expected that stronger electric
fields and longer measurement times would
enhance the probability of observing the elec­
tron ‘doing the twist’. Herein lies the difficulty.
A free electron will accelerate under the influ­
ence of an electric field and crash into the walls
of the apparatus. This effect is extremely useful
for generating X-rays in medical devices and
security scanners, but in the present experi­
ment it has only the detrimental effect of
limiting the measurement time. This obstacle
can be overcome by binding several electrons
to a heavy nucleus to form a neutral atom com­
prising a central core and some outer valence
electrons. An electric field will not accelerate
this neutral atom, but will polarize it — that
is, it will separate opposite charges within the
atom. Furthermore, the effective electric field
‘seen’ by the valence electrons in a suitably
chosen and properly polarized neutral atom
can be quite large2. The previous best attempt
to detect the electric dipole moment of the
electron was made by probing the valence
electrons in a beam of neutral thallium atoms3.
Even before the thallium-based experiment3
was completed, techniques to improve on it
were being devised. Molecules are typically
easier to polarize than atoms, which trans­
lates into the molecular valence electrons
experiencing even larger effective electric
fields4,5. This benefit was crucial in enabling
Hudson et al.1, who worked with ytterbium
© 2011 Macmillan Publishers Limited. All rights reserved
California Institute of Technology, Pasadena,
California 91125, USA.
e-mail: [email protected]
1. Tarselli, M. A. et al. Nature Chem. 3, 449–453
(2011).
2. Kam, T.-S., Pang, H.-S., Choo, Y.-M. & Komiyama, K.
Chem. Biodiver. 1, 646–656 (2004).
3. Potier, P. & Janot, M. M. C.R. Acad. Sci. 276C, 1727
(1973).
4. Scott, A. I., Yeh, C.-L. & Greenslade, D. J. Chem. Soc.
Chem. Commun. 947–948 (1978).
5. Rice, K. C. in The Chemistry and Biology of
Isoquinoline Alkaloids (eds Phillipson, J. D., Roberts,
M. F. & Zenk, M. H.) 191–203 (Springer, 1985).
monofluoride (YbF), to surpass the measure­
ment sensitivity achieved in the thallium-based
experiment3 — albeit, at present, by a mod­
est factor of 1.5. Specifically, the authors limit
the magnitude of the electron’s electric dipole
moment to less than 10.5 × 10−28 e centi­metres,
where e is the charge of the electron. In electro­
static units, this value is more than 16 orders
of magnitude weaker than the known mag­
netic dipole moment of the electron. Hudson
and colleagues have pioneered the use of
cold polar molecules to push the search for
an electric dipole moment of the electron to
new levels, and their work serves as a gateway
to multiple next-generation molecule-based
experiments. These experiments6–10, as well
as a continued effort by Hudson et al.1, are aim­
ing to improve on the above-mentioned limit
by a factor of 10–100.
How can studying a sizeless and struc­
tureless particle be so interesting? The inter­
est arises from its interaction with another
seemingly featureless entity — empty space,
casually called the vacuum. In reality, empty
space is not always so empty. The vacuum
comprises a sea of particles that are hopping
into and out of existence like waves crash­
ing onto a shore and then receding back to
the ocean. These whimsical particles do not
stick around for long enough to be observed
directly. However, they make their presence
felt through their interaction with common­
place matter, such as the electrons studied by
Hudson and colleagues1.
Physicists contend that it is these particles
that give the electron its electric dipole
moment, almost as if they are the band playing
just the right music required for the electrons
to do the twist. Without these particles, no
electric field would be strong enough and
no measurement time long enough for us to
see the electrons dance. Furthermore, they
are a subset of the new particles that physicists
working at high-energy accelerators are hoping
to create and observe directly. Hence, searches
for the electric dipole moment of the electron
provide crucial information about phenom­
ena that naturally occur at energies 1030 times
greater than those directly measured in the
precision tabletop work of Hudson et al.1.
In 1950, common theoretical arguments
asserted that fundamental particles could not
2 6 M AY 2 0 1 1 | VO L 4 7 3 | NAT U R E | 4 5 9