Structure-activity relationships of the antimalarial agent artemisinin. 7. Direct modification of (+)-artemisinin and in vivo antimalarial screening of new, potential preclinical antimalarial candidates

Article abstract of DOI:10.1021/jm020142z

On the basis of earlier reported quantitative structure-activity relationship studies, a series of 9β-16-(arylalkyl)-10-deoxoartemisinins were proposed for synthesis. Several of the new compounds 7 and 10-14 were synthesized employing the key synthetic intermediate 23. In a second approach, the natural product (+)-artemisinic acid was utilized as an acceptor for conjugate addition, and the resultant homologated acids were subjected to singlet oxygenation and acid treatment to provide artemisinin analogues. Under a new approach, we developed a one step reaction for the interconversion of artemisinin 1 into artemisitene 22 that did not employ selenium-based reagents and found that 2-arylethyliodides would undergo facile radical-induced conjugate addition to the exomethylene lactone of 22 in good yield. The lactone carbonyls were removed sequentially by diisobutylaluminum hydride reduction followed directly by a second reduction (BF₃-etherate/Et₃SiH) to afford the desired corresponding pyrans. Six additional halogen-substituted aromatic side chains were installed via 22 furnishing the bioassay candidates 15-20. The analogues were examined for in vitro antimalarial activity in the W-2 and D-6 clones of Plasmodium falciparum and were additionally tested in vivo in Plasmodium berghei- and/or Plasmodium yoelii-infected mice. Several of the compounds emerged as highly potent orally active candidates without obvious toxicity. Of these, two were chosen for pharmacokinetic evaluation, 14 and 17.

Full text of DOI:10.1021/jm020142z

J . Med. Chem. 2002, 45, 4321-4335
4321
Str u ctu r e-Activity Rela tion sh ip s of th e An tim a la r ia l Agen t Ar tem isin in . 7.
Dir ect Mod ifica tion of (+)-Ar tem isin in a n d In Vivo An tim a la r ia l Scr een in g of
New , P oten tia l P r eclin ica l An tim a la r ia l Ca n d id a tes
Mitchell A. Avery,*,†,‡ Maria Alvim-Gaston,†,§ J effrey A. Vroman,†,| Baogen Wu,†, Arba Ager,∧ Wallace Peters,O
Brian L. Robinson,O and William Charman]
Department of Medicinal Chemistry, School of Pharmacy, National Center for Natural Products Research, and
Department of Chemistry, University of Mississippi, University, Mississippi 38677
Received April 2, 2002
On the basis of earlier reported quantitative structure-activity relationship studies, a series
of 9â-16-(arylalkyl)-10-deoxoartemisinins were proposed for synthesis. Several of the new
compounds 7 and 10-14 were synthesized employing the key synthetic intermediate 23. In a
second approach, the natural product (+)-artemisinic acid was utilized as an acceptor for
conjugate addition, and the resultant homologated acids were subjected to singlet oxygenation
and acid treatment to provide artemisinin analogues. Under a new approach, we developed a
one step reaction for the interconversion of artemisinin 1 into artemisitene 22 that did not
employ selenium-based reagents and found that 2-arylethyliodides would undergo facile radical-
induced conjugate addition to the exomethylene lactone of 22 in good yield. The lactone carbonyls
were removed sequentially by diisobutylaluminum hydride reduction followed directly by a
second reduction (BF3-etherate/Et3SiH) to afford the desired corresponding pyrans. Six
additional halogen-substituted aromatic side chains were installed via 22 furnishing the
bioassay candidates 15-20. The analogues were examined for in vitro antimalarial activity in
the W-2 and D-6 clones of Plasmodium falciparum and were additionally tested in vivo in
Plasmodium berghei- and/or Plasmodium yoelii-infected mice. Several of the compounds
emerged as highly potent orally active candidates without obvious toxicity. Of these, two were
chosen for pharmacokinetic evaluation, 14 and 17.
In tr od u ction
also causes anemia in children and pregnant women
and increases vulnerability to other diseases. It afflicts
the underprivileged most severely, decreasing produc-
tivity and causing chronic poor health.
Malaria parasites have tormented humankind through-
out history. As early as 6000 B.C.E., there were record-
ings of lethal fevers, which were most likely malaria.1
The Greek physician Hippocrates, in 400 B.C.E., was
the first to describe the intermittent and often relapsing
fever attacks followed by shaking chills, sweating, and
dizziness now known to be caused by malaria. Later,
the Romans described malaria as “a horrid disease,
which comes every summer and kills”.1 Today, malaria
is one of the deadliest diseases on the planet and the
leading cause of sickness and death in the developing
world. According to the World Health Organization
(WHO), malaria causes approximately 500 million clini-
cal cases per year and kills 2.7 million people.2 It is
prevalent in the young, causing one million deaths in
children under the age of five each year. The disease
Medicinal use of the Chinese herb qinghao appears
in several standard Chinese Materia Medica texts as a
treatment for febrile illnesses.3 The herb was specifically
recommended for fevers in the Zhou Hou Bei fi Fang
(The Handbook of Prescriptions for Emergencies) writ-
ten by Ge Heng and published in 341 C.E. The most
detailed description appears in the “Compendium of
Materia MedicasBen Cao Gang Mu”, compiled in 1596,
and is still printed in China today.4,5 The antimalarial
activity of qinghao was rediscovered in China in 1972,
and the antimalarial active principal of qinghao was
named “qinghaosu”. The western name for the com-
pound is artemisinin (1).
Artemisinin (1) is a naturally occurring peroxidic
cadinane sesquiterpene. Additional names found in
China for 1 include qinghaosu, huanghuahaosu, arte-
annuin, and artemisinine. The Chemical Abstracts
adopted artemisinin as the official name; however,
* To whom correspondence should be addressed. Tel: 662-915-5880.
Fax: 662-915-5638. E-mail: mavery@olemiss.edu.
† National Center for Natural Products Research.
‡ University of Mississippi.
§ Present address: Eli Lilly and Company, Lilly Research Labora-
tories, Lilly Corporate Center, Indianapolis, IN 46285.
| Present address: Eastman Kodak Company, Arkansas.
Present address: Genomic Institute, Novartis Research Founda-
tion, San Diego, CA.
∧ Department of Microbiology and Immunology, School of Medicine,
P.O. Box 016960, University of Miami, Miami, Florida 33136.
O Centre for Tropical Antiprotozoal Chemotherapy, Northwick Park
Institute for Medical Research, Y Block, Watford Road, Harrow,
Middlesex HA13UJ , United Kingdom.
] Department of Pharmaceutics, Victorian College of Pharmacy,
Monash University, Parkville, Australia.
10.1021/jm020142z CCC: $22.00 © 2002 American Chemical Society
Published on Web 08/13/2002

4322 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 19
Avery et al.
able in China, has been approved for sale by Novartis
in Switzerland, its first Western market. This product,
licensed from the Institute of Microbiology and Epide-
miology in Beijing, combines artemether (3) with a
synthetic substance, lumefantrine.
Riamet is indicated for the treatment of P. falciparum
malaria, including emergency treatment for travelers
to regions where malaria is prevalent. While the devel-
opment of second-generation artemisinin analogues has
been useful for the treatment of severe and complicated
malaria, there are concerns about recent observations
of fatal neurotoxicity of arteether (5) in animals. Studies
have demonstrated that dihydroartemisinin (2), an in
vitro metabolite of artemether (3), artesunate (4), and
arteether (5), is probably the causative neurotoxic agent.
To avoid metabolism to neurotoxic dihydroartemisinin
(2), it has been suggested that future generations should
avoid a hemiacetal type structure.10,11
The objectives we have embraced are to develop a
commercially viable, nontoxic, orally potent antimalarial
agent designed after artemisinin (1) worthy of entering
a costly development process. To achieve these goals,
we relied on quantitative structure-activity relation-
ship (QSAR) studies, using comparative molecular field
analysis (CoMFA) techniques to predict biological ac-
tivities of new compounds and infer possible activity-
enhancing modifications.12 The set of compounds fo-
cused upon for scale-up and in vivo screening was
halogenated aromatic derivatives (11-20) of our previ-
ously reported base structure 10. Additional analogues
7-9 were prepared based upon our prior report of high
potency of 7 in vitro.13
F igu r e 1. 10-Deoxoartemisinin analogues developed for
synthesis and bioassay.
earlier entries as qinghaosu can be found. Systemati-
cally, it is named [3R-(3R,5aâ,6â,9R,12â,1aR*)]-octahy-
dro-3,6,9-trimethyl-3,12-epoxy-12H-pyrano[4,3-j]-1,2-
benzodioxepin-10(3H)-one.
Chinese researchers isolated artemisinin (1) in 1972
from Artemisia annua L. (annual wormwood), and its
structure was elucidated in 1979.6 The plant now grows
in many countries, although it is originally from north-
ern China.7 Artemisinin (1) yields in the plant can vary
considerably, depending on plant material and growth
conditions. It is present in the leaves and the flowers of
the plant in 0.01-0.8% dry weight.2
Artemisinin (1) has the ability to quickly lower
parasite levels, even in severe cases of cerebral malaria.
Additionally, artemisinin (1) has demonstrated activity
against drug resistant strains of Plasmodium falci-
parum such as W-2 Indochina (chloroquine resistant)
and D-6 Sierra Leone (mefloquine resistant) clones.
Because of this outstanding pharmacological profile in
combination with its novel chemical structure, arte-
misinin (1) became a target studied worldwide.
However, problems associated with artemisinin (1),
including a short plasma half-life, limited bioavailabil-
ity, poor solubility in oil and water, and the low yield of
artemisinin (1) from natural sources, prompted scien-
tists to develop new syntheses of artemisinin (1) deriva-
tives for more than a decade.8 Several analogues have
been obtained semisynthetically from artemisinin (1).
For example, artemisinin (1) can be reduced to dihy-
droartemisinin (2), which is approximately 10 times
more potent than the parent compound in vitro. Deriva-
tives such as artemether (3) and sodium artesunate (4),
obtained by simple modification to the lactol (2), are
recognized as clinically useful drugs in Southeast Asia
(Figure 2).8
These analogues were tested in vitro against known
drug resistant strains of P. falciparum: W-2 Indochina
clone (mefloquine sensitive, chloroquine resistant), D-6
Sierra Leone (chloroquine sensitive, mefloquine resis-
tant), K1 (chloroquine resistant), and NF54 (chloroquine
sensitive). In vivo screening was conducted in Plasmo-
dium berghei- or Plasmodium yoelii-infected mice by
available methods.
Meth od s
Three routes were employed to synthesize the homo-
logues substituted at the 16-position of 10-deoxoarte-
misinin shown in Figure 1. The first route, via total
synthesis, was amenable to multigram scale-up and has
been described before for milligram quantities of 7, 10,
and 11 for in vitro screening.13,14 The other two routes
involved the conversion of co-occurring natural products,
artemisitene and artemisinic acid, each utilizing unique
strategies that will be discussed below.
As shown in Scheme 1, I may be built either from the
accessible artemisinin (1) (route A) or from artemisinic
acid (21) (route B), both of which are produced by
Artemisia annua L. Structures I and II differ only in
the lactonic CdO bond. The lactone carbonyl could be
reduced under heterogeneous conditions with sodium
borohydride (NaBH4) in the presence of a Lewis acid to
give the target class I.15 Two independent solution
reactions may also be employed to solve the same
problem with better reproducibility. Thus, reduction of
the lactone II to the respective lactol with diisobutyl-
aluminum hydride (DIBAL) can be followed by treat-
ment with triethylsilane (Et3SiH) and boron trifluoride
Artesunate (4) has been licensed to the pharmaceuti-
cal company Sanofi, which has launched an oral prepa-
ration, Arsumax, in Western Africa, whereas Rhone
Poulenc Rorer has licensed artemether (3) only for use
in severe malaria. In the western world, the U.S. Army
has expended considerable effort toward clearance of
arteether (5), currently available only in Asia, for
human use.9 The Walter Reed Army Institute of Re-
search has also developed a more stable, water soluble
derivative related to artesunate (4), artelinic acid (6)
(Figure 2). A new malaria drug, Riamet, already avail-

Direct Modification of (+)-Artemisinin
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 19 4323
Sch em e 1
Sch em e 2
combination. We found that reduction of these lactones
to lactols with DIBAL, followed by treatment of the
crude lactol with triethylsilane and borontrifluoride
etherate, gave the desired pyrans 7 and 10-14 in 75-
95% yield after purification.14,22
In the attempted optimization of this reduction meth-
odology, numerous conditions were explored attempting
to find a one pot conversion. Several different aluminum-
based Lewis acids were investigated, which did in fact
produce the desired products. However, a high-yielding
reproducible one pot procedure that was an improve-
ment over the current two step methods could not be
found readily. For example, a one pot procedure using
artemisinin as the substrate gave 10-deoxoartemisinin
in 59% as compared to 95% with the two step se-
quence.14
diethyl etherate complex (BF3‚OEt2) to afford the pyran
adduct I. Exposure of III to light and oxygen, followed
by in situ acidification, serves to catalyze the photooxy-
genation reaction to yield the adduct II. Starting with
artemisinic acid (21), a one pot protection, conjugate
addition, and deprotection can be applied. This route
has been developed in early studies, in which in four
chemical reactions the conversion of artemisinic acid
(21) into our lead compound was achieved. Another
approach is to use artemisitene (22), the oxidized form
of the natural product artemisinin (1), for conjugate
addition (route A).
Sem isyn th esis fr om Ar tem isin ic Acid . The meth-
odology for the conversion of artemisinic acid (21) into
the C-16-substituted-10-deoxo analogues was developed
in earlier studies (Scheme 2).23,24 Transient protection
of the acid moiety of 21 by a silicon protecting group
furnishes an acrylate ester capable of conjugate addi-
tion. With this methodology, the conversion can be
accomplished in one pot. Artemisinic acid (21) is depro-
tonated with n-butyllithium, and the silyl ester is
formed after the addition of trimethylsilyl chloride
(TMSCl) at 0 °C. Upon addition of 10 mol % of copper-
(I) iodide, a Grignard reagent could then be added
leading to 1,4-addition products. Unfortunately, we
found that an unexpected reaction occurs, providing
along with the desired conjugate addition product 36,
the reduced artemisinic acid (38). Furthermore, singlet
oxygenation of the phenylpropylartemisinic acid deriva-
tive (36), followed by acid treatment with Amberlyst-
15, gave the 9â-phenylpropyl artemisinin (31) in a
modest yield of 36% (Scheme 3) and suffered from
reproducibility problems on scale-up. To prepare the 9â-
p-chlorophenylpropyl artemisinin analogue (32), the
same procedure was applied for the p-chlorophenyl-
propylartemisinic acid (37). In this case, photolysis did
not produce the desired product 32. Instead, several
compounds were obtained in which the aryl ring ap-
peared to have undergone oxidation. Thus, while this
method worked well for 31 and other robust unsubsti-
Syn th etic Ap p r oa ch es
Tota l Syn th esis. Starting with the synthetic inter-
mediate, ketal-acid 23,16 alkylation was facilitated by
treatment with 2.6 equiv of lithium diisopropylamine
(LDA) and heated to 50 °C. The orange-colored solution
that resulted was indicative of dianion formation. After
it was cooled to 0 °C, the appropriate alkylating agent
was added giving homologated ketal-acids 24-29
(Scheme 2). After silica gel, purification yields ranged
from 80 to 90%. The alkylating agents were either
purchased (7 and 10) from Aldrich or were prepared by
known procedures.13,17-20 Ozonolysis of a solution of the
ketal-acids 24-29, until a faint blue color appeared,
resulted in the primary ozonide. The transformation to
artemisinin analogues was accomplished by subsequent
acid-catalyzed rearrangement/cyclization giving 30-35
in 25-38% yields after purification.13,16,21
We investigated the transformation of artemisinin to
10-deoxoartemisinin for various artemisinin analogues
such as 30-35 following the procedure of J ung using
BF3‚OEt2 and sodium borohydride (NaBH4) in tetra-
hydrofuran (THF)/MeOH.14,15 It was found that the
yields were variable and somewhat lower than those
reported for artemisinin (1) when using this reagent

4324 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 19
Avery et al.
Sch em e 3
Sch em e 4
Sch em e 5
tuted substrates, it failed with halogenated substrates
and was not used further to prepare analogues.
We adopted a new procedure for the synthesis of
artemisitene 22, which avoided the use of selenium, by
concentrating on sulfoxide eliminations instead. In an
unsuccessful attempt to introduce the 9R-thiophenyl
moiety for subsequent oxidation and elimination, we
treated artemisinin with LDA followed by PhSSPh. This
reaction might be slow due to the relative unreactivity
of PhSSPh; therefore, a more reactive reagent capable
of transferring the PhS moiety was investigated. Studies
of LDA/PhSS(O)2Ph were conducted with artemisinin
(1). Upon treatment of artemisinin (1) with LDA in THF
at -78 °C followed by addition of PhSS(O)2Ph, the
desired thioether 39 was obtained in 67% yield. The
intermediate 39 was oxidized by m-CPBA in CH2Cl2 at
-78 °C to the sulfoxide 40, which underwent “Cope
type” elimination at 0 °C, providing artemisitene (22)
as the major product isolated in 78% yield (Scheme 4).
We next focused on radical-induced Michael additions
for the synthesis of the new structures from Scheme 1
(Scheme 5). Reaction of 22 with arylethyl iodides, 2,2′-
azobisisobutyronitrile (AIBN), and tributyltin hydride
resulted in an almost equal mixture of â and R epimers
41 and 42. A more efficient process for the conversion
of 9R-substituted analogues to their 9â congeners was
achieved, by refluxing the 9R epimer with DBU in THF
for 12 h. Previously, this conversion had been performed
by deprotonation of the R epimer at low temperatures
Sem isyn th esis fr om Ar tem isin in . Artemisitene
(22), an oxidized form of artemisinin (1), presents an
opportunity for conjugate addition by a radical or
nucleophilic mechanism. Others have added various
nucleophiles to artemisitene,25,26 but we have had
limited success with cuprate and Grignard chemistry
on artemisitene. On the other hand, a radical addition
could in principle avoid the problematical peroxide
reductions effected by Cu(I) upon attempted Normant
coupling of artemisitene. Early success with additions
of primary radicals to artemisitene prompted the search
for a nontoxic, inexpensive route to artemisitene. Arte-
misinin (1) is commercially available in kilogram quan-
tities from Vietnam and has been converted by others
to artemisitene as follows. The semisynthesis of arte-
misitene (22) from 1 was accomplished by El-Feraly in
1990 in four steps, using a photochemical approach.27
Later, Paitayatat developed a short and convenient
method for the conversion of artemisinin (1) to arte-
misitene (22) using a one pot selenoxide elimination
reaction.26 However, the selenium reagent employed in
this conversion has a high level of toxicity and would
not be suitable for large-scale production. The above
conversion was an important first step, but the develop-
ment of a better method was necessary.

Direct Modification of (+)-Artemisinin
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 19 4325
Sch em e 6
Sch em e 7
with LDA and subsequent kinetic quench in very poor
yields.13 Using the new approach, a combined overall
yield of 60% of the adduct 41 could be obtained.
Iodides required for each synthesis were prepared in
a straightforward manner (Scheme 6) from commercial
available carboxylic acids. Attempted reduction with
lithium aluminum hydride (LiAlH4) led to overreduction
of trifluoromethyl arylacetic acids, giving toluene de-
rivatives by displacement of fluorine atoms with hy-
dride. Borane evolved as a particularly useful method
for the reduction of these carboxyl groups and permitted
selective reduction in the presence of carbon-halogen
bonds.
Sch em e 8
Triphenylphosphine-iodine was used to transform
the alcohols into the iodides, furnishing the radical
precursors. Again, the alcohol 49 was commercially
available and was converted to 50 by the same method.
For the formation of the primary radical and trapping
with artemisitene (22), a very dilute solution of tri-
butyltin hydride (Bu3SnH) was added slowly over 8 h
via a syringe pump, to a refluxing solution of 22, AIBN,
and the corresponding iodide. The toxic byproduct,
tributyltin iodide (Bu3SnI), was precipitated as the
fluoride (Bu3SnF) by treatment of the reaction mixture
with a saturated solution of potassium fluoride (KF).
The resulting solution was filtered, and the product was
chromatographed to provide the readily separable R and
â lactone adducts. The R adducts were isomerized to the
â adducts by heating with DBU and THF and monitor-
ing the reaction for completion by thin-layer chroma-
tography (TLC). A second chromatography provided
another batch of â isomer, and the combined isomers
were then used for the ensuing reactions.
The various lactones from these radical additions and
isomerizations 41a -h were reduced to lactols 51a -h ,
as discussed above, and then reduced with triethylsilane
and borontrifluoride etherate to furnish the targets 8a ,
9, and 15-20 (Scheme 1, Scheme 7).
In the present work, no evidence of the products 41a
and 41b arising from the radical-induced Michael
addition was obtained (Scheme 8). The trifluorometh-
ylbutyl derivative (41b) was designed to restore the oral
activity to the butyl parent compound, which had very
high subcutaneous potency in vivo but poor oral activity.
Nonetheless, drug development requires identifying the
drug candidate’s metabolites. Because metabolites are
frequently hydroxylated analogues, we have continuing
Sch em e 9
interest in obtaining the proposed metabolite 8. In this
context, we noted that it was necessary to develop an
alternate conjugate addition methodology. It should be
possible to prepare the new compound using Grignard
reagents. We know that trifluoromethylphenyl Grignard
behaves as expected, undergoing 1,2-addition to alde-
hydes to provide the secondary alcohols (Scheme 9).
Thus, it may be possible to add the corresponding
Grignard simply using copper(I) iodide (CuI).
Reaction of artemisitene (22) and a catalytic amount
of CuI in THF at -10 °C under inert atmosphere with
the trifluoromethylpropylmagnesium bromide afforded
diastereomeric mixtures of 9-trifluoromethylbutyl ar-
temisinin in 69% yield (Scheme 10). The mixture was
purified by flash chromatography (SiO2) to furnish the
desired 9â-trifluoromethylbutyl artemisinin (41b) as a
white solid in 28% yield. The conversion of the R epimer
was performed using DBU, and an overall yield of 52%
of the adduct 41b was obtained.
Reduction of compound 41b was carried out with
DIBAL producing the lactol 51b in 89% yield. Treat-
ment of the lactol 51b with Et3SiH and BF3‚OEt2 gave
the desired pyran 9 (78% yield after purification).

4326 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 19
Avery et al.
Sch em e 10
The concentrations tested subcutaneously were 128, 32,
and 8 mg/kg/day. The drug was administered on days
3-5 postinfection. The results were compared to mice
treated both at the same concentration with artemisinin
and to a control group in which no treatment for the
disease was administered.
The most potent analogue from the subcutaneous in
vivo studies was 16-propyl-10-deoxoartemisinin 7 hav-
ing an IC100 ) 8 mg/kg/day conducted in P. berghei-
infected mice, with the drug administered on days 3-5
postinfection. However, when 7 was administered orally,
the results were unsatisfactory. Excellent oral potency
was found when the arylpropyl analogues 10-14 were
administered in peanut oil. Moreover, the subcutaneous
activity observed for these compounds could be repro-
duced with oral administration using the oil vehicle. It
is presumed that the oil vehicle aids solubilization of
the drug and hence leads to better absorption in the
stomach. Thus, more of the compound would enter the
blood stream with an intact endoperoxide moiety, which
is essential for antimalarial activity. The improvement
of 16â-propyl-10-deoxoartemisinin 7 over the parent
compound artemisinin can be seen by focusing on the 8
mg/kg/day treatment regiment. After 60 days postin-
fection, 100% of the mice treated with 7 were alive,
whereas 0% of the mice treated with artemisinin 1 were
alive. The parasitemia count after 20 days was an
average of 3.6 in mice treated with 7 as compared to 78
for an average in mice treated with artemisinin 1. The
parasitemia count was undetectable in 80% of the mice
treated with 7 after day 20, whereas 80% of the mice
treated with artemisinin had a detectable level of
parasitemia by day 10. Without treatment, 100% of the
mice had a detectable level of parasitemia by day 6, and
100% of the mice had perished by day 10 of the test.
Sch em e 11
At the higher concentration of 128 mg/kg/day dosage,
both 7 and artemisinin were equally effective at control-
ling the disease with a 100% survival rate. At the 32
mg/kg/day dosage, 7 was again able to completely
eradicate the disease whereas artemisinin 1 was unable
to protect 20% of the mice. This 20% of the mice had a
detectable level of parasitemia by day 20 and death
occurred by day 21 postinfection. At the lower concen-
trations of 2 and 0.5 mg/kg/day dosages, both 7 and
artemisinin were inadequate at protecting the mice
against the disease with a 0% survival rate.
One can readily envisage the synthesis of 41a by
Grignard conjugate addition as described above for 41b.
Protection of the hydroxyl group of the 3-bromo-1-
propanol (57) was necessary for the preparation of the
Grignard reagent. Silylation of 57 using tert-butyldi-
methylsilyl (TBDMS) chloride under standard condi-
tions proved successful (Scheme 11) providing 58.28 The
TBDMS group was our first choice as a protecting group
since it can be removed under mild conditions without
affecting the core structure. While this approach yielded
the desired product 41a , the very poor yield of 3%
inspired the use of different protecting groups. We felt
that it would be worthwhile to explore the same reaction
using the tetrahydropyranyl (THP) ether derivative of
57 and trimethylsilyl (TMS) derivative of 57. However,
compound 41a was not observed. Optimization studies,
which improve the yield of compound 41a , remain a
project for further investigation.
We were encouraged by the improved potency as
compared to artemisinin by subcutaneous administra-
tion; yet, we were disappointed by the marginal im-
provement in potency via oral administration. At all of
the concentration levels of 128, 32, and 8 mg/kg/day,
there was a 0% survival rate with both 7 and artemisi-
nin 1 as the treatment method. At the highest concen-
tration of 128 mg/kg/day, 7 did lower the parasitemia
level at day 10 when compared to artemisinin 1 (3 vs
23.2) and also extended the life of the mice by an
average of 3 days. At the lower concentrations of 32 and
8 mg/kg/day, oral administration of the drugs 7 and
artemisinin 1 was equally ineffective at treating the
disease with a 0% survival rate. Analogue 7 was
retested orally, this time administered in oil; yet, this
failed to give improvement in protection from the
disease. It is postulated that first pass metabolism is
occurring, presumably at the ω position (or ω-1), to an
This work demonstrates that artemisitene can be
used as the precursor to produce artemisinin derivatives
with new functionality at C-16. The routes described
here are reasonably simple pathways to (9â)-16-substi-
tuted artemisinin.
Biologica l Activity
In vivo testing of 7 and 10-14 was conducted in the
Thompson test at various concentrations, both subcu-
taneously and orally, in mice infected with P. berghei.29

Direct Modification of (+)-Artemisinin
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 19 4327
F igu r e 2. Comparison of the mean pharmacokinetic profiles of 17 (circles) and 14 (squares) and the reported profile for DQHS
(triangles, previous study of Bloodworth et al., 1999) following iv administration.
inactive metabolite 8. Attempts were made to synthe-
size this suspected metabolite (vide supra) for subcu-
taneous in vivo testing. If this compound was inactive
subcutaneously, it would lend support to the metabolism
hypothesis.
results demonstrate a lack of correlation between the
in vitro and the in vivo assays.
While these analogues may not be the “cure-all” for
malaria, they are a definite improvement over the
parent compound, artemisinin 1. They are also compa-
rable to other artemisinin analogues that have been
tested in vivo such as artemether 3, arteether 5,
dihydroartemisinin 2, and sodium salt of artelinic acid
6. The above analogues have shown promise as anti-
malarial agents; thus, the in vivo analysis was pursued.
Later, it was discovered that 5 was fatally neuro-
toxic.10,11 The major metabolite of ethers at the 10-
position is dihydroartemisinin 2, which is highly neu-
rotoxic both in vivo and in vitro (IC50 ) 10-6 M).30,31
Unlike the previous analogues, the 10-deoxo analogues
are postulated not to be metabolized to lactol type
compounds, which is presumed to be the culprit, at least
in part, for the fatal neurotoxicity. The analogues 7 and
10-14 have also undergone in vitro neurotoxicity
screening and have not shown substantial neurotoxicity
as compared to dihydroartemisinin 2.32
The WHO/TDR in vitro antimalarial screen was
conducted using modifications of the procedures of
Desjardins et al.,33 to assess the intrinsic activity of the
new analogues 12-19 relative to known controls such
as quinine, chloroquine, mefloquine, and sodium arte-
sunate (4). This method can provide quantitative mea-
surements of the antimalarial activity, based on the
inhibition of uptake of radiolabeled nucleic acid precur-
sor by the parasite during short-term cultures in
microtitration plates.33 Pure samples were tested for
antimalarial activity in vitro in parasitized human red
blood cells (RBC) against four P. falciparum clones:
mefloquine resistant and chloroquine, pyrimethamine,
and sulfadoxine sensitive African Sierra Leone (D6);
chloroquine sensitive NF54; chloroquine resistant K1;
and chloroquine resistant, mefloquine sensitive, and
chloroquine, pyrimethamine, and sulfadoxine resistant
An excellent in vitro activity exhibited by two C-9-
modified analogues [R ) -(CH2)3C6H5 (10) and -(CH2)3-
C6H4(4′-Cl) (11)]14 prompted us to explore the function-
ality at the benzene ring. Initially, we focused upon the
five C-9-modified analogues 10-14. Considering the
significant in vitro activity of 10, R ) -(CH2)3C6H5, and
11, R ) -(CH2)3C6H4(4′-Cl), compounds 12-14 were of
great interest for in vivo study. The fluoro analogue, for
example, would be expected to reduce aromatic metabo-
lism thus improving potency. Other modifications also
intended to provide in vivo SAR and required a large
scale synthesis of 10-14, which was successfully ac-
complished via the synthetic route outlined (Scheme 2).
The 16â-propyl-10-deoxoartemisinin analogue 7 is the
best analogue tested via subcutaneous administration.
All of the analogues containing a phenyl ring 10-14
gave complete protection at 32 mg/kg/day in the sub-
cutaneous test. The p-Cl derivative 11 showed a slight
advantage over the other analogues giving a 40%
survival rate at the 8 mg/kg/day in the subcutaneous
test. The real advantage of the phenyl derivatives
became apparent in the oral test. Analogues 11-14
possess IC100 values less than or equal to 32 mg/kg/day
in the oral test, and the unsubstituted phenyl derivative
10 gave 80% protection at this dosage. The oral test was
also conducted in P. berghei-infected mice, with the drug
administered on days 3-5 postinfection. The p-Cl 11
and the p-F 12 even showed marginal protection (20%)
at the 8 mg/kg/day dosage in the oral test. While the
p-F analogue 12 (relative activity ) 490, where arte-
misinin 1 ) 100) was notably less active than the p-Cl
analogue 11 (relative activity ) 3317, where artemisinin
1 ) 100) in vitro, it performed as well orally. These

4328 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 19
Avery et al.
Ta ble 1. In Vivo Antimalarial Activity of Substituted
Analogues of Artemisinin against P. bergheia
The ED90 data indicated that doses of 10 mg/kg or
less are effective in the mouse models for these com-
pounds, so in the future, lower oral doses will be
considered. The reason for the use of high oral doses is
that the liquid chromatography (LC)/mass spectrometry
(MS) assay has different intrinsic sensitivities for the
different series of peroxides under PK analysis in the
Charman lab. Also, to compare all compounds from the
different series at the same oral molar dose (dictated
by the compound with the highest LOQ in the assay),
it was necessary to use a consistent dose.
survival (SC)b
survival (PO)b,c
dose (mg/kg/day) dose (mg/kg/day)
structure
R
128 32
8
128 32
8
1
7
(artemisinin)
CH3(CH2)3
C6H5(CH2)3
p-ClC6H4(CH2)3
p-FC6H4(CH2)3
p-MeOC6H4(CH2)3 5/5
5/5
5/5
5/5
5/5
5/5
3/5 0/5 0/5
5/5 5/5
0/5 0/5
2/5 0/5
4/5 0/5
5/5 1/5
5/5 1/5
5/5 0/5
5/5 0/5
Compounds 17 and 14 had a reasonably low plasma
LOQ of about 10 ng/mL and will be tested with lower
oral doses in the future. If it is assumed that it is only
the fraction of drug in solution that is absorbed, then
the oral bioavailability may well increase with lower
doses. However, these two compounds still have very
low aqueous solubilities (pH 6.5, phosphate buffer) of
approximately 0.1 µg/mL; hence, this remains a key
issue.
10
11
12
13
14
5/5 0/5 5/5
5/5 2/5 5/5
5/5 0/5 5/5
5/5 0/5 5/5
5/5 0/5 5/5
3,4-Cl2C6H4(CH2)3 5/5
a
b
Drug administered on days 3-5 postinfection. Parasitemia
levels were 0 after 60 days on survivors. c Administered in peanut
oil by gavage.
Indochina (W2). All test compounds were solubilized in
dimethyl sulfoxide (DMSO) and diluted 400-fold (to rule
out a DMSO effect) in culture medium with plasma for
a starting concentration of at least 12 500 ng/mL. Drugs
were subsequently diluted 5-fold using the Cetus Pro/
Pette system utilizing a range of concentrations from
0.8 to 12 500 ng/mL. Fifty percent inhibitory concentra-
tions were reported in nanograms per milliliter.
The results of the in vivo Peters’ 4 day antimalarial
testing34 of our compounds revealed that the new
analogues have good oral activities (Tables 3-5). Fac-
tors including stability, synthesis cost, and activity led
to the selection of compounds 17 (the most potent) and
14 (the most economical) for synthesis of 5 g scale-up
for pharmacokinetic analysis (PK).
Preliminary data for the PK results of compounds 17
and 14 have been completed with the objective to
evaluate the key pharmacokinetic parameters of the
artemisinin derivatives 17 and 14 in rats after intra-
venous administration and to then compare these data
to dihydroartemisinin, which is being used as a refer-
ence compound. The absolute oral bioavailability of 17
and 14 was also evaluated. Because of their very low
aqueous solubility, the absolute oral bioavailability of
each compound was low. The low value is partially a
function of the high oral dose (∼120 mg/kg), but even
with these high doses, the plasma concentrations are
still relatively low.
In terms of the IV PK data, these two compounds have
clearance values of 20-30 mL/min/kg for 17 and 14. The
half-life for 17 is also reasonably good at 73 min, with
the shorter value for 14 being an issue. Solubility
limitations meant that 14 could only be dosed IV at
approximately 2 mg/kg (as compared with the 12 mg/
kg dose for 17). It is possible that the shorter measured
half-life for 14 was a function of lower plasma concen-
trations, which have reflected part of the distribution
phase rather than a true elimination phase. Future
studies should explore the plasma protein binding of this
series, as this may be an important factor governing the
changes in their pharmacokinetic profile relative to
dihydroartemisinin. Furthermore, it is possible that the
p-substituted phenyl ring is the key in the longer half-
life for 17. As a rule of thumb, high plasma protein
binding of a drug has the potential to decrease the free
fraction present in plasma. This could have a negative
impact on activity and tissue distribution. The assess-
ment of structural changes on the intrinsic clearance
of a compound can only be determined by correcting for
changes in the plasma free fraction. In attempting to
address these issues, the very low aqueous solubility of
17 and 14 becomes a significant limitation that needs
to be addressed as further compounds are explored.
Ta ble 2. In Vitro IC50 Data for Various Artemisinin Analogues and Control Drugs against Four Strains of P. Falciparum
P. falciparum (IC50
)
D6
W2
K1
NF54
compda
MW
(ng/mL)b
(ng/mL)b
(ng/mL)b
(ng/mL)b
chloroquine
mefloquine
quinine
10 (H)
11 (p-Cl)
9 (n-C4H6F3)
15 (m-Cl)
16 (m-F)
17 (m-CF3)
18 (p- CF3)
19 (3,5-CF3)
20 (3,5-F2)
319.88
442.28
342.42
372.50
406.94
364.40
406.94
390.49
440.50
440.50
508.49
408.48
7.58 (4.10)c
45.55 (8.35)c
159.61 (129.6)c
0.90 (1.1)c
1.90 (1.9)c
12.46
118.04 (102.5)c
34.06 (2.78)c
149.88
79.01
20.34
99.60
7.36
35.38
9.08
1.35 (0.8)c
2.60 (1.6)c
8.87
NT
NT
7.55 (0.46)c
8.28
21.60 (0.38)c
26.83
10.07
11.56
12.00
39.55
4.06
18.27 (40.19)c
41.76 (27.33)c
35.60 (30.76)c
101.52 (48.78)c
NT
8.48
30.32
28.13 (0.47)c
0.78
58.01 (0.45)c
7.11 (0.58)c
6.50 (0.45)c
0.82
3.20
NT
a
Compounds 11-20 are all 10-deoxoartemisinins with a 9-(CH2)3PhX (Figure 1), whereas compound 9 has a n-CH2CH2CH2CF3 side
chain. Assay performed by TIBOTEC. c Assay performed by Walter Reed Army Institute of Research (WRAIR).
b

Direct Modification of (+)-Artemisinin
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 19 4329
Ta ble 3. In Vivo ED50 and ED90 Data in a Mouse Malaria
Model (P. berghei N) for Artemisinin Analogues and Control
Drugs
The new compounds were fully characterized and
evaluated for in vitro and in vivo antimalarial activity.
On the basis of the data from the in vivo antimalarial
experiment (Peters’ 4 day) and factors such as stability
and synthesis cost, we chose two compounds, 17 (the
most potent) and 14 (the most economical), for synthesis
of 5 g scale-up for PK studies. The preliminary PK data
led us to believe that 17 has the better PK profile,
although the series as a whole still suffers from limiting
solubility issues. Future chemistry needs to address
these limitations, as the potency seems reasonable,
considering that absorption is limited.
P. berghei N (SC)
P. berghei N (PO)
compda
ED50
ED90
ED50
ED90
chloroquine
mefloquine
quinine
sodium artesunate
10 (H)
11 (p-Cl)
15 (m-Cl)
16 (m-F)
17 (m-CF3)
18 (p-CF3)
19 (3,5-CF3)
20 (3,5-F2)
1.5
1.1
83.0
0.8
0.5
0.7
0.5
0.46
0.4
0.44
0.9
1.3
3
2.3
1050
1.8
3.3
71
3.4
5.1
118
13.0
3.4
2.1
8.3
3.4
4.4
2.5
1.0
1.2
0.9
0.85
0.75
0.8
1.7
2.0
2.4
1.4
1.0
2.0
0.9
1.25
0.63
2.6
1.8
2.1
5.0
5.7
Exp er im en ta l Section
All solvents were purchased as reagent grade, dried ap-
propriately, and stored over dry 4 Å molecular sieves. Solvent
and reagent transfers were accomplished either via dried
syringe or via cannula, and all reactions were routinely
conducted under argon atmosphere unless otherwise indicated.
Flash chromatography was accomplished using silica gel
(Whatman 60, 230-400 mesh). Unless otherwise noted, all
NMR analyses were conducted in CDCl3 on a Bruker Avance
DPX-400 and were referenced to chloroform at δ 7.27. IR
spectra were obtained using a Digilab FTS-40 FT-IR or a ATI
Mattson Genesis Series FT-IR and run neat. Melting points
were obtained on a MelTemp. Mass spectral data was obtained
on a high-resolution ESI FT-ICR or VG 7070E-HF mass
spectrometer. Elemental analyses were within 0.35 as deter-
mined by QTI, Whitehouse, NJ , or by Desert Analytics, Tucson,
AZ.
a
Compounds 10, 11-20 are all 10-deoxoartemisinins with a
9-(CH2)3PhX (Figure 1).
Ta ble 4. In Vivo ED50 and ED90 Data in Two Mouse Malaria
Models (P. yoelli ssp. NS and P. berghei N) for Artemisinin
Analogue 9 and Control Drugsa
mg/kg × 4
%
compd
strain route ED50
ED90 suppression
9 (n-C4H6F3)
9 (n-C4H6F3)
sodium artesunate
sodium artesunate
N
NS
N
PO >10
>10
6.5
0.3
77.2
56.5
PO NA 10 NA 10
PO >10
PO >10
>10
>10
NS
a
Compound 9 is a 10-deoxoartemisinins with a 9-(CH2)3CF3
(Figure 1).
Gen er a l P r oced u r e for P r ep a r a tion Keta l-Acid s 24-
29. To a solution of ketal-acid 23 (2 mmol) in THF (5 mL) at
-78 °C was added, via cannula, a freshly prepared solution of
LDA (2.6 mmol). After 10 min, the resultant solution was
allowed to warm to room temperature over a 30 min period.
The solution was then heated at 50 °C for 2 h, an orange-
colored solution being indicative of dianion formation. The
solution was then cooled to 0 °C, and 2.6 equiv of the
appropriate alkylating agent was added via syringe. After 2.5
h, TLC showed that no starting material remained. The
reaction was quenched with saturated NH4Cl solution (15 mL)
and then extracted with CHCl3 (3 × 15 mL). The combined
organic layers were then washed with brine (2 × 30 mL), dried
over MgSO4, filtered, and concentrated in vacuo. The crude
material was purified via flash chromatography (200:1; silica
gel: compound) eluting with 50% EtOAc/hex + 1% HOAc. Pure
products 24-29 were obtained as yellow oils in 80-90% yield.
Gen er a l P r oced u r e for P r ep a r a tion of Ta r get Ar te-
m isin in An a logu es. Ozon olysis a n d Acid -Ca ta lyzed Cy-
cliza tion of Keta l-Acid s to Yield An a logu es 30-35. O3/
O2 (7 PSI, 0.5 L/min, 80 V) was bubbled into a solution of ketal-
acid 24-29 (1 mmol) in dry CH2Cl2 (125 mL) at -78 °C, until
the reaction mixture had turned faint blue. After the reaction
mixture was purged with O2 followed by Ar, silica gel (750 mg)
and 15% aqueous H2SO4 (75 µL) were added. The mixture was
allowed to come to ambient temperature and stirred overnight,
and the solids were filtered and washed with CH2Cl2 (2 × 10
mL) and then EtOAc (2 × 10 mL). The resulting filtrate was
washed with saturated aqueous NaHCO3, dried over Na2SO4,
and evaporated in vacuo to give crude products, which were
easily purified by preparative TLC (PTLC) on silica gel. Pure
products 30-35 were recovered as white crystalline solids in
25-39% yield.
Ta ble 5. In Vivo Data (P. yoelli ssp. NS)
P. yoelli
ssp. NS (SC)
P. yoelli
ssp. NS (PO)
compda
chloroquine
ED50
ED90
ED50
ED90
1.6
2.3
28
6.0
mefloquine
quinine
sodium artesunate (SSV)
10 (H)
120
1.8
0.5
56
22.0
0.9
128
1.6
29
100
11 (p-Cl)
0.6
1.0
15 (m-Cl)
16 (m-F)
17 (m-CF3)
18 (p- CF3)
0.6
0.58
0.6
1.00
1.00
1.00
1.25
0.67
19 (3,5-CF3, SSV*)
19 (DMSO*)
20 (3,5-F2, SSV*)
20 (DMSO*)
4.2
1.5
2.2
2.3
12.0
3.6
6.5
7.5
a
Compounds 10-20 are all 10-deoxoartemisinins with
a
9-(CH2)3PhX (Figure 1). *SSV and DMSO refer to formulations
for oral administration; see Experimental Section.
Con clu sion s
In summary, we have developed a short and efficient
methodology from the easily accessible natural product
artemisinin (1) to artemisitene (22). Because the syn-
theses of artemisitene (22) can be carried out in a
straightforward and nontoxic procedure, synthetic ap-
plication of this methodology as a key step in the
synthesis of new artemisinin derivatives was achieved.
Our four step synthesis of the 9â-16-substituted 10-
deoxyartemisinin derivatives compares favorably in
terms of the use of simple reagents and transformations,
as well as better yields with the previously reported
synthesis.
P r oced u r e for Red u ction of Ar tem isin in An a logu es to
10-Deoxoa r t em isin in An a logu es
7 a n d 10-14 u sin g
Et3SiH/BF 3. To a stirred solution of artemisinin analogues
30-35 (0. 320 mmol or 1.0 equiv) in dry CH2Cl2 (3 mL) at -78
°C was added DIBAL (335 µL or 0.335 mmol, 1.0 M in
hexanes). The dry ice/acetone bath was removed long enough
to allow complete dissolution of the substrate (about 1 min)
and then replaced. After 1.25 h, the reaction was quenched
with saturated NaHCO3 (1 mL), diluted with CH2Cl2 (3 mL),

4330 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 19
Avery et al.
and allowed to warm to ambient temperature with stirring.
The mixture was diluted with CH2Cl2 (25 mL) and washed
with 10% HCl/saturated NH4Cl (1:15 v/v) (1 × 10 mL). The
CH2Cl2 layer was then dried (MgSO4), filtered, and concen-
trated in vacuo to give 90 mg (100%) of a white solid, which
was pure enough for the next step. The 90 mg was dissolved
in dry CH2Cl2 (3 mL) at -78 °C. Et3SiH (210 µL or 1.31 mmol)
was added. The reaction was stirred for 10 min, and then, BF3-
OEt2 (45 µL or 0.366 mmol) was added. The resultant solution
was allowed to stir 3 h at -78 °C. After 3 h, the reaction was
quenched at -78 °C with pyridine (150 µL or 1.85 mmol) and
was allowed to warm to ambient temperatures. The reaction
mixture was poured into aqueous saturated NH4Cl (10 mL)
and extracted with EtOAc (3 × 30 mL). The combined EtOAc
layers were washed with NH4Cl (2 × 15 mL), dried (MgSO4),
and concentrated in vacuo to give a white solid, which was
purified by PTLC (silica gel) eluting with 20:80 EtOAc/hexanes
to give pure products 7 and 10-14 as white crystalline solids
in 75-90% yield.
In Vivo Testin g. The in vivo antimalarial testing was
performed at the University of Miami, Department of Micro-
biology, Miami, FL, by Dr. Arba Ager and co-workers. The mice
used in this testing were CD-1 mice, 16-18 g in size. The mice
were infected with P. berghei after a 2 week quarantine. The
treatment regiment began on day three postinfection and
lasted for 3 days (days 3-5 postinfection). The subcutaneous
administration route was by single injection, and the oral
administration was accomplished by using the gavage method
with the drug suspended in peanut oil.35,36
solution) and 14 to be administered at a 2 mg/kg dose (i.e., 2
mL dose of a 0.3 mg/mL solution). IV formulations were
prepared on a day of dosing and sterile-filtered through a 0.2
µm syringe filter immediately prior to use. An aliquot of the
formulation was retained and assayed for drug content to
confirm the concentration of the administered dose. For
administration of the dose, the IV solution was administered
as a 5 min infusion (2 mL volume) via the jugular vein cannula,
after which the cannula was flushed with heparinized saline
(2 U/mL) to ensure complete administration of the dose.
Oral Formulation and Administration. The oral formula-
tions of 14 and 17 were prepared as standard aqueous
suspensions in 1 mL of 1% carboxymethylcellulose (medium
viscosity, Sigma), with 50 µL of 1% Tween 80 (Sigma), at a
nominal dose of 120 mg/kg (0.3 mmol/kg) in the expectation
that it would provide quantifiable plasma concentrations. The
exact amount of 14 and 17 to be administered was weighed
into 12 mL polypropylene tubes and then suspended in the
total administration volume. Any drug-related material re-
maining in the tube after dosing was dissolved in a known
volume of acetonitrile and assayed by LC-MS to enable
determination of the administered dose. For administration
of the dose, rats were lightly anaesthetized with isofluorane
to facilitate insertion of a gavage tube. The suspension
formulation was taken up into a syringe, attached to the
gavage tube, and the dose was administered into the stomach.
The polypropylene tube (previously containing the suspension)
was then rinsed with 1 mL of water, which was also then
administered to each rat.
In the in vivo Peters’ 4 day antimalarial testing34 of our
compounds conducted by Dr. Wallace Peters, two batches of
five male random-bred Swiss albino mice weighing 18-22 g
were inoculated intravenously with 2 × 106 RBC parasitized
with P. berghei N. Animals were then given a fixed dose of 30
mg/kg once a day for four consecutive days. Compounds were
first dissolved in DMSO (due to lower water solubility), and
subsequently, an aqueous dilution was prepared for use and
administered subcutaneously to one group and orally to a
second group. The parasitemia was determined on the day
following the last treatment, day 4, to determine qualitatively
the presence and degree of activity at the screening dose.
SSV is a formula used by Roche for biological, toxicological,
and PK studies whose components are Na-carboxymethylcel-
lulose (CMC), medium viscosity (Merck), benzyl alcohol p.a.
(Merck), Tween 80 (Fluka), NaCl 0.9% (Braun) and whose
composition is Na-CMC, 5 g; benzyl alcohol, 5 mL; Tween 80,
4 mL; NaCl, 0.9% at 1000 mL.
P h a r m a cok in etic Stu d ies. All experimental procedures
were approved and performed in accordance with the guide-
lines of the Institutional Animal Experimentation Ethics
Committee. The studies were conducted in fasted male Spra-
gue-Dawley rats, and the treatments included IV administra-
tion of 14 or 17, and oral administration of 14 or 17 in the
SSV suspension, which is a formulation amenable to use in
later developmental and toxicology studies.
Su r gica l P r oced u r es. On the day prior to dosing, poly-
ethylene cannulae were inserted into the right jugular vein
and/or right carotid artery of rats (280-350 g) under isoflurane
anaesthesia. Once they were implanted, the cannulae were
flushed with heparinised saline (2 U/mL). The cannulae were
then exteriorized by tunneling subcutaneously to emerge above
the scapulae. Rats were housed in cages fitted with swivel
attachments to enable free movement about the cages while
protecting the integrity and patency of the cannulae. All rats
returned to normal grooming, drinking, and sleeping behavior
within 1 h of surgery. Animals were fasted but had free access
to water for the duration of the study. At the conclusion of
each experiment, rats were euthanized by a lethal injection
of pentobarbitone.
P la sm a Sa m p le Collect ion a n d P r ocessin g. Blood
samples were taken from the indwelling carotid cannula. An
initial blood volume of 0.2 mL was withdrawn to clear the line
of heparinized saline. A fresh syringe was then used to
withdraw a 0.3 mL blood sample that was placed in a clean
eppendorf tube containing approximately 20 µL of a 1 U/µL
heparin solution to prevent clotting. Blood samples were stored
on ice until centrifuged to separate plasma from red blood cells,
and a 150 µL aliquot of plasma was transferred to a fresh
eppendorf tube and then stored frozen until analysis. The blood
sample that was initially drawn to clear the cannula was then
readministered via the carotid cannula, followed by a small
volume (0.5 mL) of heparinized saline (2 U/mL). This procedure
facilitated patency of the cannula and minimized any possible
changes to blood haematocrit and blood volume that can occur
with serial blood sampling.
Blank blood samples were collected from each rat prior to
dosing. After IV administration of 14 or 17, blood samples were
collected at 0 (end of infusion), 5, 15, 30, 60, 90, 120, 150, 180,
240, 300, and 360 min postdosing. After oral administration
of each compound as the SSV oral suspension formulations,
blood samples were collected at 5, 15, 30, 60, 90, 120, 180, 240,
300, and 360 min postdosing.
An a lysis of P la sm a Con cen t r a t ion s of 14 a n d 17.
Preparation of Plasma Standards. Calibration standards for
14 and 17 were prepared separately by spiking 1 mL of blank
rat plasma with 10 µL of a methanol solution containing 1 mg/
mL of compound. This process provided a plasma standard
with a concentration of 10 000 ng/mL. Further dilutions were
made from this stock with blank plasma to afford plasma
standards in the range of 5-10 000 ng/mL for 17 and 10-
10 000 ng/mL for 14. Aliquots of each plasma standard (150
µL) were then transferred to eppendorf tubes for processing.
Acetonitrile (300 µL) was added to precipitate proteins, and
the tubes were vortexed and centrifuged for 10 min at 3500
rpm (Beckman G5-6R refrigerated centrifuge). A 300 µL
aliquot of the supernatant was then transferred to a glass
autosampler vial, and 15 µL of the internal standard solution
(100 µg/mL TDR 7551/2) was added. An aliquot of the solution
(50 µL) was then injected onto the LC-MS system. For each
conpound, calibration curves were linear over the stated
concentration range, and correlation coefficients typically
exceeded 0.999.
Dr u g F or m u la tion s. Intravenous Formulation and Ad-
ministration. The IV dose of 14 and 17 comprised an aqueous
solution comprising 20% dimethylsulfoxide (DMSO)/5% Cre-
mophore EL. This formulation provided for 17 to be admin-
istered at a 10 mg/kg dose (i.e., 2 mL dose of a 1.75 mg/mL
Plasma Sample Preparation. A simple plasma sample
preparation protocol was developed, which involved acetoni-

Direct Modification of (+)-Artemisinin
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 19 4331
trile precipitation of plasma proteins, addition of an internal
standard, and injection onto the LC-MS system. To a 150 µL
aliquot of rat plasma was added 300 µL of acetonitrile to
precipitate plasma proteins. All samples were then processed
as described above for the preparation of plasma standards.
Sample Analysis. Samples were analyzed on a Waters
Alliance 2690 HPLC interfaced with a Micromass LCZ mass
spectrometer operating under atmospheric pressure chemical
ionization conditions (APCI). The full scan spectrum of 17
resulted in a substantial parent ion, MH+ (441 amu), which
was chosen as the analytical species. The full scan spectrum
of 14 resulted in a substantial parent ion, MH+ (407 amu),
with the expected isotopic ratio resulting from a monochlori-
nated molecule, and this was chosen as the analytical species.
The assay validation criteria were chosen to facilitate rapid
assay development while still providing confidence in data
integrity. The validation of the assay included linearity of
response over the concentration range analyzed, interday
variability, system suitability, and replicate analysis of samples
at the limit of quantitation (LOQ). Calibration curves were
accepted provided that not more than two standards on each
day had a back-calculated value that deviated by more than
20% from the nominal concentration. All calibration curves
run during the course of assay validation and analysis of
plasma samples met this criterion and were linear over the
stated concentration range. Individual calibration curves were
weighted (1/x2) given the range of plasma concentrations. The
LOQ was determined as the lowest standard for which at least
three out of four back-calculated concentrations were within
30% of the nominal value. The LOQ was determined to be 5
ng/mL for 17 and 10 ng/mL for 14.
Suitability of storage and sample processing was investi-
gated by spiking blank plasma to a concentration of 5 µg/mL
and then comparing the measured value with a sample that
had previously been frozen and then thawed. Both 14 and 17
were shown to be unaffected by the freezing process, and the
sample preparation procedure resulted in measured concen-
trations in agreement with nominal concentrations.
Reproducibility and on-system stability after sample pro-
cessing was assessed by reinjecting samples from the calibra-
tion curves following storage in the autosampler (10 °C)
overnight. The consistency of the slope of the calibration curves
from day-to-day was used as a simple measure of interday
variability. The %CV of the slopes of the calibration curves
remained within acceptable limits and did not vary signifi-
cantly from the slope of a solvent standard curve, indicating
that samples are stable if stored overnight in the autosampler
and that there were no matrix effects associated with the
injection of plasma samples.
cycloh exyl]eth yl]-2,5,5-tr im eth yl-1,3-d ioxa n e (26). Yield
74%. 1H NMR (CDCl3): δ 7.23 (d, 2H, J ) 8.3 Hz), 7.10 (d,
2H, J ) 8.4 Hz), 5.26 (s, 1H), 3.54 (dd, 2H, J ) 2.3, 11.6 Hz),
3.38-3.48 (m, 2H), 2.71 (br dd, 2H, J ) 10.5, 10.5 Hz), 2.56
(m, 1H), 2.06 (br dd, 1H, J ) 2.9, 10.7 Hz), 1.72-1.94 (m, 3H),
1.31 (s, 3H), 1.15 (m, 1H), 1.01 (s, 3H) 0.86-0.94 (m, 6H), 0.09
(s, 9H). IR (CH2Cl2): 3400-2500, 1700, 1598, 1491, 1454, 1246,
1091, 861, 748 cm-1. DCIMS-NH3: m/z 566 (M + NH4), 549
(M+). Anal. [C31H49O4ClSi (as hydrate, 1/4H2O)] C, H.
(1′′S,1′′′S,3′′S,6′′R)-(2′′E)-2-[2′-[3′′-1′′′-Ca r boxy-4′′′-p-flu o-
r op h en ylbu tyl)-6′′-m eth yl-2′′-[(tr im eth ylsilyl)m eth ylen e]-
cycloh exyl]eth yl]-2,5,5-tr im eth yl-1,3-d ioxa n e (27). Yield
75%. 1H NMR (CDCl3): δ 7.04 (d, 2H, J ) 5.5 Hz), 6.90 (d,
2H, J ) 8.8 Hz), 5.26 (s, 1H), 3.53 (d, 2H, J ) 11.5 Hz), 3.40
(ddd, 2H, J ) 2.7, 7.6, 10.3 Hz), 2.53 (m, 1H), 2.06 (br dd, 1H,
J ) 2.9, 10.7 Hz), 1.72-1.94 (m, 3H), 1.30 (s, 3H), 1.15 (m,
1H), 1.00 (s, 3H), 0.91 (d, 3H, J ) 6.8 Hz), 0.89 (s, 3H), 0.08
(s, 9H). IR (CH2Cl2): 3400-2700, 1705, 1506, 1456, 1376, 1259,
1093 cm-1. DCIMS-NH3: m/z 550 (M + NH4), 533 (M + H),
464, 357, 110. Anal. (C31H49O4SiF) C, H.
(1′′S,1′′′S,3′′S,6′′R)-(2′′E)-2-[2′-[3′′-1′′′-Car boxy-4′′′-p-m eth -
oxyph en ylbu tyl)-6′′-m eth yl-2′′-[(tr im eth ylsilyl)m eth ylen e]-
cycloh exyl]eth yl]-2,5,5-tr im eth yl-1,3-d ioxa n e (28). Yield
88%. 1H NMR (CDCl3): δ 7.03 (d, 2H, J ) 8.5 Hz), 6.78 (d,
2H, J ) 8.6 Hz), 5.28 (s, 1H), 3.76 (s, 3H), 3.55 (dd, 2H, J )
2.3, 11.5 Hz), 3.38-3.44 (m, 2H), 2.52 (dd, 2H, J ) 6.1, 6.1
Hz), 2.07 (br dd, 1H, J ) 2.9, 10.5 Hz), 1.70-1.95 (m, 3H),
1.31 (s, 3H), 1.15 (m, 1H), 1.02 (s, 3H), 0.92 (dd, 3H, J ) 3.3,
3.3 Hz), 0.88 (s, 3H), 0.09 (s, 9H). IR (CH2Cl2): 3400-2500,
1736, 1704, 1611, 1513, 1462, 1376, 1300, 1247, 1092, 1041,
848 cm-1. FABMS: m/z 545 (M + H), 441, 383, 369, 323, 157,
121. Anal. (C32H50O5Si) C, H.
(1′′S,1′′′S,3′′S,6′′R)-(2′′E)-2-[2′-[3′′-1′′′-Ca r b oxy-4′′′-3,4-
d ich lor op h en ylbu tyl)-6′′-m eth yl-2′′-[(tr im eth ylsilyl)m e-
th ylen e]cycloh exyl]eth yl]-2,5,5-tr im eth yl-1,3-dioxan e (29).
Yield 90%. 1H NMR (CDCl3): δ 7.23 (dd, 2H, J ) 2.0, 20.1
Hz), 6.94 (dd, 1H, J ) 2.0, 8.2 Hz), 5.27 (s, 1H), 3.55 (dd, 2H,
J ) 1.8, 11.5 Hz), 3.36-3.43 (m, 2H), 2.53 (dd, 2H, J ) 6.8,
6.8 Hz), 2.07 (br dd, 1H, J ) 2.9, 10.5 Hz), 1.70-1.95 (m, 3H),
1.32 (s, 3H), 1.18 (m, 1H), 1.01 (s, 3H), 0.92 (d, 3H, J ) 6.8
Hz), 0.87 (s, 3H), 0.09 (s, 9H). IR (CH2Cl2): 3440-2600, 1732,
1705, 1599, 1469, 1396, 1373, 1248, 1093, 1032, 848 cm-1
Anal. (C31H48O4SiCl2) C, H.
.
P r ep a r a tion of 39 fr om Ar tem isin in (1). In a flame-
dried, 10 mL round-bottom flask equipped with an argon line,
1 (282 mg, 1 mmol, 1 equiv) was placed and was dissolved in
THF (2 mL). This solution was added via cannula to a solution
of LDA (1.2 equiv, 1.2 mmol) in dry THF (2 mL) at -78 °C.
After 30 min, a solution of S-phenylbenzenethiosulfonate (325
mg, 1.3 mmol, 1.3 equiv) in THF was added. The reaction
mixture was left stirring for another 30 min at a temperature
of -78 °C. The reaction was quenched with distilled water (2
mL) and then allowed to warm to room temperature. The
reaction mixture was diluted with H2O (5 mL) and extracted
with EtOAc (3 × 25 mL). The combined organic layers were
then washed with brine (1 × 25 mL), dried over MgSO4,
filtered, and concentrated in vacuo. The crude oil was purified
via flash column chromatography eluting with 20% ethyl
acetate/hexanes. Compound 39 (255 mg) was isolated in 67%
yield.
(1′′S,1′′′S,3′′S,6′′R)-(2′′E)-2-[2′-[3′′-1′′′-Ca r boxyp en tyl)-6′′-
m eth yl-2′′-[(tr im eth ylsilyl)m eth ylen e]cycloh exyl]eth yl]-
2,5,5-t r im et h yl-1,3-d ioxa n e (24). Yield 88%. 1H NMR
(CDCl3): δ 5.32 (s, 1H), 3.55 (dd, 2H, J ) 1.4, 11.5 Hz), 3.44
(dt, 2H, J ) 1.4, 11.5 Hz), 2.72 (br dd, 1H, J ) 11.2, 11.2 Hz),
2.40 (br dd, 1H, J ) 5.0, 11.9 Hz), 2.11 (br dd, 1H, J ) 1.9,
10.5 Hz), 1.36 (s, 3H), 1.03 (s, 3H), 0.94 (d, 3H, J ) 7.1 Hz),
0.89 (s, 3H), 0.88 (dd, 3H, J ) 7.2, 7.2 Hz), 0.13 (s, 9H). IR
(CH2CL2): 3600-2500, 1734, 1703, 1600, 1456, 1395, 1374,
1246, 1212, 1164, 1091, 845 cm-1. EIMS: m/z 452 (M+), 437
(M - Me), 348, 338, 296. Anal. (C25H46O4Si) C, H.
(+)-Oct a h yd r o-3,6r-d im et h yl-3,12-ep oxy-9â-(p h en yl-
th io)-12H-pyr an o[4,3j]-1,2-ben zodioxepin -10(3H)-on e (39).
Yield 67%; mp 171-174 °C. 1H NMR (CDCl3): δ 7.47 (d, 2H J
) 8 Hz), 7.38 (d, 1H J ) 1.2 Hz), 7.32 (t, 2H, J ) 7.28 Hz),
6.02 (s, 1H), 2.42 (d, 1H J ) 2.92 Hz), 2.34 (m, 1H), 2.06 (d,
1H J ) 3.6 Hz), 2.03 (d, 1H J ) 3.8 Hz), 1.92 (d, 1H J ) 3.9
Hz), 1.81-1.74 (m, 2H), 1.53 (s, 3H), 1.43 (s, 3H), 1.00 (m, 1H),
(1′′S,1′′′S,3′′S,6′′R)-(2′′E)-2-[2′-[3′′-1′′′-Ca r boxy-4′′′-p h en -
y lb u t y l)-6′′-m e t h y l-2′′-[(t r im e t h y ls ily l)m e t h y le n e ]-
cycloh exyl]eth yl]-2,5,5-tr im eth yl-1,3-d ioxa n e (25). Yield
91%. 1H NMR (CDCl3): δ 7.10-7.30 (m, 5H), 5.29 (s, 1H), 3.55
(dd, 2H, J ) 3.1, 11.5 Hz), 3.41 (dt, 2H, J ) 1.4, 11.0 Hz), 2.74
(br dd, 1H, J ) 11.5, 11.5 Hz), 2.59 (br dd, 2H, J ) 6.8, 6.8
Hz), 2.39 (br dd, 1H, J ) 4.8, 12.0 Hz), 2.08 (br dd, 1H, J )
2.8, 11.5 Hz), 1.40-2.00 (m, 12H), 1.37 (m, 2H), 1.02 (s, 3H),
0.93 (d, 3H, J ) 7.1 Hz), 0.87 (s, 3H), 0.10 (s, 9H). IR (Nujol):
3600-2400, 1700, 1600, 1450, 1380, 1250, 1220, 1130, 1100,
850, 750, 710 cm-1. DCIMS-NH3: m/z 587 (M + TMS), 515
(M + H), 446, 428, 411, 359, 339. Anal. (C31H50O4Si) C, H.
(1′′S,1′′′S,3′′S,6′′R)-(2′′E)-2-[2′-[3′′-1′′′-Ca r boxy-4′′′-p-ch lo-
r op h en ylbu tyl)-6′′-m eth yl-2′′-[(tr im eth ylsilyl)m eth ylen e]-
13
0.99 (d, 3H, J ) 5.0 Hz). CNMR (CDCl3): δ 171.74, 138.55,
130.35, 128.99, 105.44, 94.60, 82.71, 54.28, 51.51, 50.30, 37.67,
36.38, 34.75, 29.67, 26.19, 26.02, 25.05, 20.31. FABMS: m/z
391.2 (M + H). Anal. (C21H26O5S) C, H.
P r ep a r a tion of Ar tem isiten e (22) fr om 39. In a flame-
dried 10 mL round-bottom flask equipped with an argon line,
39 (30 mg, 0.1 mmol, 1 equiv) was placed and was dissolved

4332 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 19
Avery et al.
in CH2Cl2 (2 mL). This solution was added via cannula to a
solution of m-CPBA (1.05 equiv, 0.11 mmol) in dry CH2Cl2 (2
mL) at -78 °C. After 6 h at -78 °C, the reaction was allowed
to warm to room temperature. The reaction mixture was
diluted with H2O (5 mL), washed with a saturated solution of
NaHCO3, and extracted with EtOAc (3 × 25 mL). The
combined organic layers were then washed with brine (1 × 25
mL), dried over MgSO4, filtered, and concentrated in vacuo.
The crude oil was purified via flash column chromatography
eluting with 20% ethyl acetate/hexanes. Artemisitene (22)
(16.8 mg) was isolated in 78% yield.
and was dissolved in ether/acetonitrile (3:1). To this solution,
triphenylphosphine (3 equiv), imidazole (3 equiv), and iodine
(3 equiv) were added in this order. The reaction mixture was
kept at room temperature for 1 h and monitored by TLC. The
reaction mixture was filtered and washed with ether. The
organic layers were then washed with brine, dried over MgSO4,
filtered, and concentrated in vacuo. The residue was purified
via flash chromatography on silica gel eluting with 1% ethyl
acetate/hexanes to give 43b -48b and 50 in 58-88%
yield.
Gen er a l P r oced u r e for P r ep a r a tion of th e Alcoh ols
43a -48a fr om 43-48. In a flame-dried three neck 1000 mL
round-bottom flask equipped with a reflux condenser and an
argon line, the appropriate carboxylic acid (43-48) (1 equiv)
was placed and was dissolved in THF. To this solution, a
solution of 1 M BH3‚THF was added via cannula (2 equiv) at
0 °C. The reaction mixture was kept at 0 °C for 4 h and
monitored by TLC. The reaction mixture was diluted with cold
H2O, washed with a saturated solution of NaHCO3, and
extracted with ether. The combined organic layers were then
washed with brine (1 × 25 mL), dried over MgSO4, filtered,
and concentrated in vacuo. The residue was purified via flash
chromatography on silica gel eluting with 20% ethyl acetate/
hexanes to give 43a -48a in 58-88% yield.
m -Ch lor op h en eth yl Iod in e (43b). 1H NMR (CDCl3): δ
13
7.21-7.19 (q, 3H), 7.12 (t, 1H), 3.52 (t, 2H), 3.28 (t, 2H).
-
CNMR (CDCl3): δ 140.90, 134.14, 129.69, 129.06, 127.02,
126.45, 38.69, 11,27. IR (liquid film): 2948, 2879, 1598, 1574,
1429. FT-ICR MS: m/z 267.52 (M + H).
m -F lu or op h en eth yl Iod in e (44b). 1H NMR (CDCl3): δ
7.19-6.89 (q, 3H), 6.79 (t, 1H), 3.47 (t, 2H), 3.16 (t, 2H).
13CNMR (CDCl3): δ 162.02, 141.90, 130.14, 127.39, 114.70,
112.70, 38.05, 11.29. IR (liquid film): 2942, 2878, 1510, 1436,
1416, 1098. FT-ICR MS: m/z 251 (M + H).
m -(Tr iflu or om eth yl)p h en eth yl Iod in e (45b). 1H NMR
(CDCl3): δ 7.30-7.28 (q, 3H), 7.12 (t, 1H), 3.56 (t, 2H), 3.24
(t, 2H). 13CNMR (CDCl3): δ 139.95, 132.58, 130.59, 129.00,
127.11, 125.75, 123.32, 38.90, 10.89. IR (liquid film): 2948,
2879, 1598, 1574, 1429, 1080, 1047. FT-ICR MS: m/z 301 (M
+ H).
1
m -Ch lor op h en eth yl Alcoh ol (43a ). H NMR (CDCl3): δ
7.21-7.19 (q, 3H), 7.12 (t, 1H), 3.76 (t, 2H), 2.78 (t, 2H).
13CNMR (CDCl3): δ 140.90, 134.14, 129.69, 129.06, 127.02,
126.45, 62.96, 38.69. IR (liquid film): 3464, 2948, 2879, 1598,-
1574, 1429, 1096, 1047, 958. FT-ICR MS: m/z 157 (M + H).
p-(Tr iflu or om eth yl)p h en eth yl Iod in e (46b). 1H NMR
(CDCl3): δ 7.46 (d, 2H J ) 8.6 Hz), 7.14 (d, 2H J ) 7.8 Hz),
3.41 (t, 2H, J ) 7.7 Hz), 3.22 (t, 2H, J ) 8.0 Hz). 13CNMR
(CDCl3): δ 139.95, 132.58, 130.59, 129.00, 127.11, 125.75,
123.30, 38.93, 11.02. IR (liquid film): 2948, 2879, 1598, 1574,
1429, 1080, 1047. FT-ICR MS: m/z 301.02 (M + H).
3,5-Di(tr iflu or om eth yl)ph en eth yl Iodin e (47b). 1H NMR
(CDCl3): δ 7.46 (s, 1H), 7.31 (s, 2H), 3.46 (t, 2H, J ) 7.2 Hz),
3.12 (t, 2H J ) 7.6 Hz). 13CNMR (CDCl3): δ 140.97, 131.37,
128.65, 119.72, 39.31, 11.5. IR (liquid film): 2947, 2878, 1530,
1436, 1416, 1098, 1010, 958. FT-ICR MS: m/z 368.08 (M +
H).
1
m -F lu or op h en eth yl Alcoh ol (44a ). H NMR (CDCl3): δ
7.19-6.89 (q, 3H), 6.83 (t, 1H), 3.76 (t, 2H), 2.74 (t, 2H).
13CNMR (CDCl3): δ 162.52, 141.90, 130.14, 127.59, 114.70,
112.70, 63.46, 38.29. IR (liquid film): 3343, 2942, 2878, 1510,
1436, 1416, 1098, 1016, 958. FT-ICR MS: m/z 141 (M + H).
m -(Tr iflu or om eth yl)p h en eth yl Alcoh ol (45a ). 1H NMR
(CDCl3): δ 7.30-7.28 (q, 3H) 7.12 (t, 1H), 3.86 (t, 2H), 2.74 (t,
2H). 13CNMR (CDCl3): δ 139.95, 132.58, 130.59, 129.00,
127.11, 125.75, 123.30, 63.11, 38.90. IR (liquid film): 3323,
2948, 2879, 1598, 1574, 1429, 1080, 1047, 1000, 968. FT-ICR
MS: m/z 191.02 (M + H).
3,5-Diflu or op h en eth yl Iod in e (48b). 1H NMR (CDCl3):
δ 6.60 (s, 2H), 6.30 (s, 1H), 3.48 (t, 2H, J ) 7.8 Hz), 3.16 (t,
2H J ) 7.9 Hz). 13CNMR (CDCl3): δ 166.62, 143.40, 110.50,
99.70, 39.71, 12.5. IR (liquid film): 2942, 2878, 1510, 1436,
1416, 1098. FT-ICR MS: m/z 269.06 (M + H).
1
p-(Tr iflu or om eth yl)p h en eth yl Alcoh ol (46a ). H NMR
(CDCl3): δ 7.26 (d, 2H J ) 8.6 Hz), 7.04 (d, 2H J ) 7.8 Hz),
3.91 (t, 2H, J ) 7.9 Hz), 2.76 (t, 2H, J ) 8.0 Hz). 13CNMR
(CDCl3): δ 139.95, 132.58, 130.59, 129.00, 127.11, 125.75,
123.30, 63.11, 38.90. IR (liquid film): 3463, 2948, 2879, 1598,
1574, 1429, 1080, 1047, 1000, 968. FT-ICR MS: m/z 191.02
(M + H).
Gen er a l P r oced u r e for th e Ra d ica l-In d u ced Mich a el
Ad d ition Ap p r oa ch . P r ep a r a tion of 9â-Ar tem isin in An a -
logu es 41c-41h fr om Ar tem isiten e (22). To a solution of
artemisitene (22) (1 mmol) in dry benzene (40 mL) was added
the iodo derivative 43b-48b (1.5 mmol) and AIBN (0.1 mmol)
at 25 °C. The reaction mixture was heated to reflux temper-
ature, and tributyltin hydride (1.4 mmol) was added as a
benzene solution (25 mL) over a period of 8 h via syringe pump.
After the mixture was completely added, the reaction mixture
was cooled to room temperature. The solvent was evaporated
to dryness, and diethyl ether (15 mL) followed by a saturated
solution of potassium fluoride (4 mL) was added. The solution
was stirred at 25 °C for 12 h and was then filtered, washed
with water, and evaporated to dryness. The crude product was
purified by flash chromatography on silica gel using a mixture
of 15% ethyl acetate/hexanes as eluent to give a mixture of
â-isomer 41c-41h in 35-40% yield and R-isomer in 60-65%
yield.
3,5-Di(t r iflu or om et h yl)p h en et h yl Alcoh ol (47a ). 1H
NMR (CDCl3): δ 7.38 (s, 1H), 7.23 (s, 2H), 3.82 (t, 2H, J ) 8.2
Hz), 2.75 (t, 2H J ) 8.7 Hz). 13CNMR (CDCl3): δ 140.90, 131.3,
128.05, 119.70, 65.76, 39.31. IR (liquid film): 3468, 2947, 2878,
1530, 1436, 1416, 1098, 1010, 958. FT-ICR MS: m/z 257.18
(M + H).
1
3,5-Diflu or op h en eth yl Alcoh ol (48a ). H NMR (CDCl3):
δ 6.63 (s, 2H), 6.03 (s, 1H), 3.88 (t, 2H, J ) 8.0 Hz), 2.78 (t,
2H J ) 8.8 Hz). 13CNMR (CDCl3): δ 163.52, 142.90, 110.70,
99.70, 65.46, 38.71. IR (liquid film): 3453, 2942, 2878, 1510,
1436, 1416, 1098, 1010, 958. FT-ICR MS: m/z 159.23 (M +
H).
Gen er a l P r oced u r e for th e P r ep a r a tion of th e Iod id es
43b-48b a n d 50 fr om 43a -48a a n d 49. In a flame-dried
250 mL round-bottom flask equipped with an argon line, the
appropriate alcohol (43a -48a and 49) (1 equiv) was placed
Con ju ga te Ad d ition of 1,1,1-Tr iflu or op r op a n ylm a g-
n esiu m Br om id e to Ar tem isiten e (22). A solution of arte-

Direct Modification of (+)-Artemisinin
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 19 4333
1
misitene (1 equiv) and CuI (catalytic) in THF at -10 °C under
argon was treated dropwise with the appropriate Grignard
reagents (1.10 equiv). The mixture was at -10 °C for a further
1 h after which saturated aqueous NH4Cl solution was added
and the mixture was allowed to warm to room temperature.
After the organic layer was separated, the aqueous solution
was extracted with Et2O. The combined organic layer was
washed with brine, dried with MgSO4, and concentrated in
vacuo. The residue was then purified by flash chromatography
on silica gel using a mixture of 15% ethyl acetate/hexanes as
eluent to afford a mixture of â isomer 41b in 38% yield and R
isomer in 58% yield.
Gen er a l P r oced u r e for th e Con ver sion of r Ep im er s
to th e 9â-Ar tem isin in An a logu es 30-35 a n d 45-52 u sin g
DBU. To a solution of 9R-artemisinin analogues (1 equiv) in
dry THF was added DBU (2 equiv) at 25 °C. The reaction
mixture was refluxed for 24 h. After it was cooled to room
temperature, water was added, and the organic material was
extracted with ethyl acetate. The solution was dried over
MgSO4, filtered, and evaporated to dryness. The crude product
was purified by silica gel column chromatography using 15%
ethyl acetate/hexanes as eluent to give the â isomer 7-20 in
50-70% yield.
ep in -10(3H)-on e (35). Yield 31%; mp 135-136 °C. H NMR
(CDCl3): δ 7.33 (d, 1H, J ) 8.2 Hz), 7.26 (s, 1H), 7.01 (dd, 1H,
J ) 2.0, 8.2 Hz), 5.27 (s, 1H), 3.18-3.24 (m, 1H), 2.52-2.72
(m, 2H), 2.34-2.48 (m, 1H), 1.96-2.09 (m, 3H), 1.67-1.82 (m,
4H), 1.44 (s, 3H), 0.99 (d, 3H, J ) 5.8 Hz). IR (KBr): 2993,
2925, 2868, 1738, 1474, 1394, 1184, 1115, 1032, 993, 885, 815
cm-1. Anal. (C23H28O5Cl2) C, H.
(+)-Octa h yd r o-3,6r-d im eth yl-3,12-ep oxy-9â-(3′-(tr iflu o-
r om et h yl)p r op yl)-12H -p yr a n o[4,3j]-1,2-b en zod ioxep in -
10(3H )-on e (41b ). Yield 58%; mp 93-94 °C. 1H NMR
(CDCl3): δ 5.91 (s, 1H), 2.23 (dd, 2H J ) 1.5 Hz), 2.11-2.03
(m, 6H), 1.74-1.66 (m, 6H), 1.47 (s, 3H), 1.01 (d, 3H, J ) 5.0
Hz). 13CNMR (CDCl3): δ 171.73, 105.76, 94.21, 80.58, 50.87,
45.22, 43.59, 37.93, 36.30, 34.43, 33.74, 31.96, 25.84, 25.09,
20.52, 20.26. IR (KBr): 2930, 1740 cm-1. FABMS: m/z 379.45
(M + H). Anal. (C18H25O5F3) C, H.
(+)-Octa h yd r o-3,6r-d im eth yl-3,12-epoxy-9â-(3′-(m-ch lo-
r op h en yl)p r op yl)-12H -p yr a n o[4,3j]-1,2-b en zod ioxep in -
10(3H)-on e (41c). Yield 62%; mp 116-117 °C. 1H NMR
(CDCl3): δ 7.23 (d, 2H, J ) 8.2 Hz), 7.09 (d, 2H, J ) 8.3 Hz),
5.83 (s, 1H), 3.14-3.24 (m, 1H), 2.52-2.72 (m, 2H), 2.34-2.48
(m, 1H), 1.96-2.09 (m, 3H), 1.67-1.82 (m, 4H), 1.44 (s, 3H),
0.98 (d, 3H, J ) 5.7 Hz). 13CNMR (CDCl3): δ 171.76, 146.47,
129.13, 127.09, 125.71, 122.45, 106.42, 93.92, 79.60, 50.81,
43.50, 38.09, 37.91, 36.30, 35.89, 33.75, 29.17, 26.87, 25.57,
25.44, 25.04. IR (KBr): 2951, 2925, 1740, 1491, 1112, 1031,
1000 cm-1. FABMS: m/z 427(M + Li). Anal. (C23H29O5Cl) C,
H.
(+)-Octa h yd r o-3,6r-d im eth yl-3,12-ep oxy-9â-(3′-(m -flu o-
r op h en yl)p r op yl)-12H -p yr a n o[4,3j]-1,2-b en zod ioxep in -
10(3H)-on e (41d ). Yield 52%; mp 152-153 °C. 1H NMR
(CDCl3): δ 7.15-6.93 (m, 4H), 5,83 (s, 1H), 3.22 (ddd, 1H, J )
5.5, 2.7, 7.0 Hz), 2.65 (ddd, 1H, J ) 6.2, 8.0, 10.4 Hz), 2.58
(ddd, 1H, J ) 6.2, 7.6, 8.4 Hz), 2.42 (ddd, 1H, J ) 4.4, 13.0,
15.1 Hz), 1.60-2.19 (m, 9H), 1.44 (s, 3H), 1.25-1.50 (m, 4H),
1.05 (m, 1H), 0.98 (d, 3H, J ) 5.8 Hz). 13CNMR (CDCl3): δ
172.84, 143.47, 133.08, 128.6, 127.09, 125.71, 122.45, 106.42,
93.92, 79.60, 50.81, 43.50, 38.09, 37.91, 36.30, 35.89, 33.75,
29.17, 26.87, 25.57, 25.44, 23.84, 19.92. IR (KBr): 1736, 1382,
1261, 1184, 1116, 881, 850, 831, 798 cm-1. FABMS m/z: 411
(M + Li). Anal. (C23H29O5F) C, H.
(+)-Octah ydr o-3,6r-dim eth yl-3,12-epoxy-9â-(bu tyl)-12H-
pyr an o[4,3j]-1,2-ben zodioxepin -10(3H)-on e (30). Yield 27%;
1
mp 129-130 °C. H NMR (CDCl3): δ 5.86 (s, 1H), 3.20 (ddd,
1H, J ) 5.7, 5.7, 8.2 Hz), 2.43 (ddd, 1H, J ) 3.9, 13.0 14.4
Hz), 2.04 (m, 3H), 1.80 (m, 3H), 1.20-1.60 (m, 8H), 1.46 (s,
3H), 1.09 (m, 2H), 1.01 (d, 3H, J ) 6.0 Hz), 0.92 (dd, 3H, J )
7.1, 7.1 Hz). IR (CH2Cl2): 3054, 2958, 1736, 1421, 1274, 1254,
1114, 1001, 895, 740 cm-1. DCIMS-NH3: m/z 342 (M + NH4),
325 (M + H), 307, 289, 274, 261, 251.
(+)-Octa h yd r o-3,6r-d im eth yl-3,12-ep oxy-9â-(3′-p h en yl-
p r op yl)-12H -p yr a n o[4,3j]-1,2-b e n zod ioxe p in -10(3H )-
on e (31). Yield 28%; mp 137-138 °C. 1H NMR (CDCl3): δ
7.10-7.35 (m, 5H), 5.85 (s, 1H), 3.25 (ddd, 1H, J ) 5.8, 5.8,
8.8 Hz), 2.71 (ddd, 1H, J ) 5.6, 8.4, 13.8 Hz), 2.61 (ddd, 1H, J
) 7.2, 8.4, 13.8 Hz), 2.44 (ddd, 1H, J ) 4.4, 13.5, 5.1 Hz), 1.50-
2.20 (m, 10H), 1.45 (s, 3H), 1.20-1.50 (m, 4H), 1.05 (m, 1H),
0.99 (d, 3H, J ) 5.9 Hz). IR (CH2Cl2): 1740, 1200, 1040, 1010,
890, 840 cm-1. DCIMS-NH3: m/z 404 (M + NH4), 387 (M +
H), 369, 351, 341, 323, 313.
(+)-Octa h yd r o-3,6r-d im eth yl-3,12-ep oxy-9â-(3′-(p-ch lo-
r op h en yl)p r op yl)-12H -p yr a n o[4,3j]-1,2-b en zod ioxep in -
10(3H)-on e (32). Yield 39%; mp 154-155 °C. 1H NMR
(CDCl3): δ 7.23 (d, 2H, J ) 8.2 Hz), 7.09 (d, 2H, J ) 8.3 Hz),
5.83 (s, 1H), 3.14-3.24 (m, 1H), 2.52-2.72 (m, 2H), 2.34-2.48
(m, 1H), 1.96-2.09 (m, 3H), 1.67-1.82 (m, 4H), 1.44 (s, 3H),
0.98 (d, 3H, J ) 5.7 Hz). IR (KBr): 2951, 2925, 1740, 1491,
1112, 1031, 1000 cm-1. CIMS: m/z 459 (M + K+), 443 (M +
Na+), 409, 347. Anal. (C23H29O5Cl) C, H.
(+)-Octa h yd r o-3,6r-d im eth yl-3,12-ep oxy-9â-(3′-(m -tr i-
flu or om eth ylp h en yl)p r op yl)-12H-p yr a n o[4,3j]-1,2-ben zo-
d ioxep in -10(3H)-on e (41e). Yield 62%; mp 112-114 °C. H
1
NMR (CDCl3): δ 7.51 (d, 2H), 7.27 (d, 2H), 5.83 (s, 1H), 3.21
(ddd, 1H, J ) 5.3, 2.5, 7.0 Hz), 2.65 (ddd, 1H, J ) 6.1, 8.3,
10.2 Hz), 2.61 (ddd, 1H, J ) 6.2, 7.6, 8.4 Hz), 2.42 (ddd, 1H, J
) 4.4, 13.0, 15.1 Hz), 1.60-2.21 (m, 9H), 1.44 (s, 3H), 1.25-
1.40 (m, 4H), 1.05 (m, 1H), 0.96 (d, 3H, J ) 5.6 Hz). 13CNMR
(CDCl3): δ 171.79, 147.47, 131.56, 128.6, 127.09, 125.71,
122.45, 119.6, 104.42, 93.92, 79.60, 50.81, 43.50, 38.09, 37.91,
36.30, 35.93, 33.75, 29.17, 26.77, 25.57, 25.28, 23.81, 20.20.
IR (KBr): 1734, 1383, 1261, 1187, 1116, 881, 856, 831, 796
cm-1. FABMS m/z: 461 (M + Li). Anal. (C24H29O5F3) C, H.
(+)-Oct a h yd r o-3,6r-d im et h yl-3,12-ep oxy-9â-(3′-(p -t r i-
flu or om eth ylph en yl)pr opyl)-12H-pyr an o[4,3j]-1,2-ben zodi-
oxep in -10(3H)-on e (41f). Yield 67%; mp 153-154 °C. 1H
NMR (CDCl3): δ 7.52 (d, 2H), 7.29 (d, 2H), 5,83 (s, 1H), 3.22
(ddd, 1H, J ) 5.5, 2.7, 7.0 Hz), 2.65 (ddd, 1H, J ) 6.2, 8.0,
10.4 Hz), 2.58 (ddd, 1H, J ) 6.2, 7.6, 8.4 Hz), 2.42 (ddd, 1H, J
) 4.4, 13.0, 15.1 Hz), 1.60-2.19 (m, 9H), 1.44 (s, 3H), 1.25-
1.50 (m, 4H), 1.05 (m, 1H), 0.98 (d, 3H, J ) 5.8 Hz). 13CNMR
(CDCl3): δ 171.76, 146.47, 129.09, 125.69, 125.65, 105.81,
93.88, 79.59, 50.48, 43.50, 38.09, 37.91, 36.30, 35.93, 33.95,
29.07, 26.76, 25.55, 25.24, 23.71, 20.18. IR (KBr): 1736, 1382,
1261, 1184, 1120, 886, 850, 831, 798 cm-1. FABMS m/z: 461
(M + Li). Anal. (C24H29O5F3) C, H.
(+)-Octa h yd r o-3,6r-d im eth yl-3,12-ep oxy-9â-(3′-(3,5-bis-
tr iflu or om eth ylp h en yl)pr opyl)-12H-p yr a n o[4,3j]-1,2-ben -
zod ioxep in -10(3H)-on e (41g). Yield 58%; mp 109-110 °C.
1H NMR (CDCl3): δ 7.46 (d, 1H), 7.31 (d, 2H), 5.83 (s, 1H),
3.22 (ddd, 1H, J ) 5.5, 2.7, 7.0 Hz), 2.65 (ddd, 1H, J ) 6.2,
8.0, 10.4 Hz), 2.58 (ddd, 1H, J ) 6.2, 7.6, 8.4 Hz), 2.42 (ddd,
1H, J ) 4.4, 13.0, 15.1 Hz), 1.60-2.19 (m, 9H), 1.44 (s, 3H),
(+)-Octa h yd r o-3,6r-d im eth yl-3,12-ep oxy-9â-(3′-(p-flu o-
r op h en yl)p r op yl)-12H -p yr a n o[4,3j]-1,2-b en zod ioxep in -
10(3H)-on e (33). Yield 25%; mp 152-153 °C. 1H NMR
(CDCl3): δ 7.15-7.10 (m, 2H), 6.99-6.93 (m, 2H), 5.83 (s, 1H),
3.22 (ddd, 1H, J ) 5.5, 2.7, 7.0 Hz), 2.65 (ddd, 1H, J ) 6.2,
8.0, 10.4 Hz), 2.58 (ddd, 1H, J ) 6.2, 7.6, 8.4 Hz), 2.42 (ddd,
1H, J ) 4.4, 13.0, 15.1 Hz), 1.60-2.19 (m, 9H), 1.44 (s, 3H),
1.25-1.50 (m, 4H), 1.05 (m, 1H), 0.98 (d, 3H, J ) 5.8 Hz). IR
(KBr): 1736, 1382, 1261, 1184, 1116, 881, 850, 831, 798 cm-1
.
LRMS m/z: 422 (M + NH4), 405 (M + H). Anal. (C23H29O5F)
C, H.
(+)-Octah ydr o-3,6r-dim eth yl-3,12-epoxy-9â-(3′-(p-m eth -
oxyp h en yl)p r op yl)-12H-p yr a n o[4,3j]-1,2-ben zod ioxep in -
10(3H)-on e (34). Yield 34%; mp 161-162 °C. 1H NMR
(CDCl3): δ 7.08 (d, 2H, J ) 8.6 Hz), 6.82 (d, 2H, J ) 8.7 Hz),
5.84 (s, 1H), 3.79 (s, 3H), 3.19-3.25 (m, 1H), 2.51-2.75 (m,
2H), 2.34-2.45 (m, 1H), 2.01-2.09 (m, 3H), 1.67-1.82 (m, 4H),
1.44 (s, 3H), 0.99 (d, 3H, J ) 5.8 Hz). IR (KBr): 2958, 2920,
2859, 1741, 1511, 1247, 1114, 1034, 997, 810 cm-1. FABMS:
m/z 417 (M + H), 399, 371, 343, 201, 147, 121. Anal. (C24H32O6)
C, H.
(+)-Oct a h yd r o-3,6r-d im e t h yl-3,12-e p oxy-9â-(3′-(3,4-
d ich lor op h en yl)p r op yl)-12H-p yr a n o[4,3j]-1,2-ben zod iox-

4334 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 19
Avery et al.
13
1.25-1.50 (m, 4H), 1.05 (m, 1H), 0.98 (d, 3H, J ) 5.8 Hz).
-
390 (M + NH4), 373 (M + H), 355, 337, 327, 269. Anal.
CNMR (CDCl3): δ 176.00, 140.07, 131.00, 112.45, 105.14,
92.46, 78.89, 50.93, 44.89, 39.62, 39.45, 35.86, 31.17, 29.24,
28.18, 27.97, 26.53, 23.97, 21.12. IR (KBr): 1740, 1382, 1261,
1184, 1116, 881, 850, 831, 798 cm-1. FT-ICR MS m/z: 529.5
(M + H). Anal. (C23H28O5F2) C, H.
(C23H32O4) C, H.
(+)-Octa h yd r o-3,6r-d im eth yl-3,12-ep oxy-9â-(3′-(p-ch lo-
r oph en yl)pr opyl)-12H-pyr an o[4,3j]-1,2-ben zodioxepin (11).
1
Yield 88%; mp 84-86 °C. H NMR (CDCl3): δ 7.12-7.30 (m,
4H), 5.19 (s, 1H), 3.78 (ddd, 1H, J ) 1.3, 4.3, 7.2 Hz), 3.44 (dd,
1H, J ) 11.7, 11.7 Hz), 2.60 (ddd, 2H, J ) 5.1, 7.4, 7.4 Hz),
2.43-2.53 (m, 1H), 2.37 (ddd, 1H, J ) 4.1, 13.4, 14.7 Hz), 2.01
(ddd, 1H, J ) 3.0, 4.8, 14.7 Hz), 1.83-1.93 (m, 1H), 1.62 (dd,
2H, J ) 7.8, 7.8 Hz), 1.42 (s, 3H), 1.25 (dd, 2H, J ) 6.5, 11.1
Hz), 0.95 (d, 3H, J ) 6.2 Hz). IR (KBr): 2926, 2913, 2866, 1494,
1454, 1091, 1067 cm-1. DCIMS-NH3: m/z 424 (M + NH4),
407 (M + H). Anal. (C23H31O4Cl) C, H.
(+)-Octa h yd r o-3,6r-d im eth yl-3,12-ep oxy-9â-(3′-(3,5-d i-
flu or oph en yl)pr opyl)-12H-pyr an o[4,3j]-1,2-ben zodioxepin -
10(3H)-on e (41h ). Yield 52%; mp 100-104 °C. 1H NMR
(CDCl3): δ 7.15-6.93 (m, 3H), 5,83 (s, 1H), 3.22 (ddd, 1H, J )
5.5, 2.7, 7.0 Hz), 2.65 (ddd, 1H, J ) 6.2, 8.0, 10.4 Hz), 2.58
(ddd, 1H, J ) 6.2, 7.6, 8.4 Hz), 2.42 (ddd, 1H, J ) 4.4, 13.0,
15.1 Hz), 1.60-2.19 (m, 9H), 1.44 (s, 3H), 1.25-1.50 (m, 4H),
1.05 (m, 1H), 0.98 (d, 3H, J ) 5.8 Hz). 13CNMR (CDCl3): δ
173.90, 150.07, 142.00, 112.45, 105.14, 100.02, 90.46, 79.60,
50.93, 44.60, 39.62, 39.47, 35.86, 31.17, 29.24, 28.18, 27.91,
26.53, 23.97, 20.13. IR (KBr): 1736, 1382, 1261, 1184, 1116,
881, 850, 831, 798 cm-1. FT-ICR MS m/z: 429 (M + H). Anal.
(C23H28O5F2) C, H.
Gen er a l P r oced u r e for Red u ction of th e La cton e 41b-
41h Der iva tive to La ctol 51b-h u sin g DIBAL. P r ep a r a -
tion of th e 10-Dih yd r o 9â-Ar tem isin in An a logu es. To a
stirred solution of 41b-h (1.0 equiv) in dry CH2Cl2 at -78 °C
was added 1 M DIBAL in CH2Cl2 (1.1 equiv). After 1 h, the
reaction was quenched with saturated NaHCO3, diluted with
CH2Cl2, and allowed to warm to room temperature. The
mixture was diluted with CH2Cl2 and washed with 10% HCl/
saturated NH4Cl (1:15 v/v). The CH2Cl2 layer was then dried
over MgSO4, filtered, and concentrated in vacuo to give 51b-h
as a white solid in 89-96% yield.
Gen er a l P r oced u r e for Red u ction of th e La ctol 51b-h
Der iva t ive t o t h e P yr a n 13-19 u sin g E t 3SiH/BF 3OE t 2.
P r ep a r a tion of 10-Deoxo 9â-Ar tem isin in An a logu es. To
a stirred solution of 51b-h (1 equiv) in dry CH2Cl2 at -78
°C, Et3SiH (4 equiv) was added. The reaction was stirred for
10 min, and BF3‚OEt2 (1.5 equiv) was added. The resultant
solution was allowed to stir for 3 h at -78 °C. After 5 h, the
reaction was quenched at -78 °C with pyridine (8 equiv) and
was allowed to warm to room temperature. The reaction
mixture was poured into aqueous saturated NH4Cl and
extracted with EtOAc. The combined organic layers were
washed with NH4Cl, dried (MgSO4), and concentrated in vacuo
to give a white solid. The crude products were purified by flash
chromatography on silica gel (80:20 EtOAc/hexanes) to give
13-19 as a pure white crystalline compound in 55-85%
yield.
(+)-Octa h yd r o-3,6r-d im eth yl-3,12-ep oxy-9â-(3′-(p-flu o-
r oph en yl)pr opyl)-12H-pyr an o[4,3j]-1,2-ben zodioxepin (12).
1
Yield 85%; mp 76-77 °C. H NMR (CDCl3): δ 7.12-7.07 (m,
2H), 6.98-6.92 (m, 2H), 5.19 (s, 1H), 3.78 (dd, 1H, J ) 3.2,
10.9 Hz), 3.43 (dd, 1H, J ) 11.7, 11.7 Hz), 2.61-2.32 (m, 4H),
2.15-1.81 (m, 2H), 1.71-1.45 (m, 7H), 1.43 (s, 3H), 1.41-1.04
(m, 5H), 0.97 (d, 3H, J ) 6.3 Hz). IR (KBr): 1633, 1511, 1454,
1216, 1193, 1159, 1126, 910, 877, 829, 761, 736 cm-1. Anal.
(C23H31O4F) C, H.
(+)-Octah ydr o-3,6r-dim eth yl-3,12-epoxy-9â-(3′-(p-m eth -
o x y p h e n y l)p r o p y l)-12H -p y r a n o [4,3j]-1,2-b e n zo d io x -
ep in (13). Yield 94%; mp 95-96 °C. 1H NMR (CDCl3): δ 7.07
(d, 2H, J ) 8.7 Hz), 6.82 (d, 2H, J ) 8.7 Hz), 5.19 (s, 1H), 3.79
(s, 3H), 3.43 (dd, 1H, J ) 11.7, 11.7 Hz), 2.60 (ddd, 2H, J )
5.1, 7.4, 7.4 Hz), 2.43-2.53 (m, 1H), 2.37 (ddd, 1H, J ) 4.1,
13.4, 14.7 Hz), 2.01 (ddd, 1H, J ) 3.0, 4.8, 14.7 Hz), 1.83-
1.93 (m, 1H), 1.62 (dd, 2H, J ) 7.8, 7.8 Hz), 1.43 (s, 3H), 0.95
(d, 3H, J ) 6.2 Hz). IR (KBr): 2958, 2920, 2859, 1612, 1514,
1462, 1377, 1242, 1099, 1090, 1063, 877, 810 cm-1. FABMS:
m/z 403 (M + H), 217, 201, 185, 121. Anal. (C24H34O5) C, H.
HRFAB calcd for C24H34O5 + H, 403.2484; found, 403.2486.
(+)-Oct a h yd r o-3,6r-d im e t h yl-3,12-e p oxy-9â-(3′-(3,4-
d ich lor op h en yl)p r op yl)-12H-p yr a n o[4,3j]-1,2-ben zod iox-
ep in (14). Yield 93%; mp 128-129 °C. 1H NMR (CDCl3): δ
7.33 (d, 1H, J ) 8.2 Hz), 7.24 (d, 1H, J ) 2.0 Hz), 6.98 (dd,
1H, J ) 2.0, 8.2 Hz), 5.20 (s, 1H), 3.78 (dd, 1H, J ) 5.1, 13.0
Hz), 3.44 (dd, 1H, J ) 11.8, 11.8 Hz), 2.57 (ddd, 2H, J ) 5.1,
7.4, 7.4 Hz), 2.41-2.50 (m, 1H), 2.36 (ddd, 1H, J ) 4.1, 13.4,
14.7 Hz), 1.83-1.93 (m, 1H), 1.59 (dd, 2H, J ) 7.7, 7.7 Hz),
1.41 (s, 3H), 1.25 (dd, 2H, J ) 9.8, 9.8 Hz), 0.95 (d, 3H, J )
6.2 Hz). IR (KBr): 2952, 2927, 2864, 1473, 1377, 1192, 1126,
1099, 1061, 935, 875, 831 cm-1. Anal. (C23H30O4Cl2) C, H.
(+)-Octa h ydr o-3,6r-d im eth yl-3,12-ep oxy-9â-(3′-(m-ch lo-
r oph en yl)pr opyl)-12H-pyr an o[4,3j]-1,2-ben zodioxepin (15).
Yield 85%; mp 92-93 °C. 1H NMR (CDCl3): δ 7.00-7.7.15 (m,
4H), 5.19 (s, 1H), 3.56 (ddd, 1H, J ) 1.3, 4.3, 7.2 Hz), 3.44 (dd,
1H, J ) 11.7, 11.7 Hz), 2.60 (ddd, 2H, J ) 5.1, 7.4, 7.4 Hz),
2.43-2.53 (m, 1H), 2.37 (ddd, 1H, J ) 4.1, 13.4, 14.7 Hz), 2.01
(ddd, 1H, J ) 3.0, 4.8, 14.7 Hz), 1.83-1.93 (m, 1H), 1.62 (dd,
2H, J ) 7.8, 7.8 Hz), 1.42 (s, 3H), 1.25 (dd, 2H, J ) 6.5, 11.1
Hz), 0.95 (d, 3H, J ) 6.2 Hz). IR (KBr): 2920, 2915, 2860, 1490,
1452, 1090, 1067 cm-1. FABMS: m/z 413 (M + Li). Anal.
(C23H31O4Cl) C, H.
(+)-Octa h yd r o-3,6r-d im eth yl-3,12-ep oxy-9â-(3′-(m -flu o-
r oph en yl)pr opyl)-12H-pyr an o[4,3j]-1,2-ben zodioxepin (16).
Yield 82%; mp 76-77 °C. 1H NMR (CDCl3): δ 7.12-7-6.92
(m, 4H), 5.19 (s, 1H), 3.78 (dd, 1H, J ) 3.2, 10.9 Hz), 3.43 (dd,
1H, J ) 11.7, 11.7 Hz), 2.61-2.32 (m, 4H), 2.15-1.81 (m, 2H),
1.71-1.45 (m, 7H), 1.43 (s, 3H), 1.41-1.04 (m, 5H), 0.97 (d,
3H, J ) 6.3 Hz). IR (KBr): 1633, 1511, 1454, 1216, 1193, 1159,
1126, 910, 877, 829, 761, 736 cm-1. FABMS m/z: 397 (M +
Li). Anal. (C23H31O4F) C, H.
(+)-Octa h yd r o-3,6r-d im eth yl-3,12-ep oxy-9â-(3′-(m -tr i-
flu or om eth ylp h en yl)p r op yl)-12H-p yr a n o[4,3j]-1,2-ben zo-
d ioxep in (17). Yield 80%; mp 93-94 °C. 1H NMR (CDCl3): δ
7.12-7-6.92 (m, 4H), 5.19 (s, 1H), 3.76 (dd, 1H, J ) 3.2, 10.9
Hz), 3.47 (dd, 1H, J ) 11.6, 11.5 Hz), 2.61-2.36 (m, 4H), 2.13-
1.82 (m, 2H), 1.69-1.48 (m, 7H), 1.43 (s, 3H), 1.41-1.04 (m,
5H), 0.97 (d, 3H, J ) 6.3 Hz). IR (KBr): 1636, 1516, 1457,
1212, 1197, 1158, 1126, 910, 877, 829, 767, 739 cm-1. FABMS
m/z: 447 (M + Li). Anal. (C24H31O4F3) C, H.
(+)-Octah ydr o-3,6r-dim eth yl-3,12-epoxy-9â-(bu tyl)-12H-
p yr a n o[4,3j]-1,2-ben zod ioxep in (7). Yield 80%; mp 91-92
°C. H NMR (CDCl3): δ 5.85 (s, 1H), 3.80 (ddd, 1H, J ) 1.2,
1
4.1, 11.5 Hz), 3.46 (dd, 1H, J ) 11.8, 11.8 Hz), 2.40-2.50 (m,
1H), 2.40 (ddd, 1H, J ) 4.0, 13.5, 14.6 Hz), 2.03 (ddd, 1H, J )
3.1, 4.9, 14.4 Hz), 1.85-19.4 (m, 1H), 1.71 (dq, 1H, J ) 3.3
12.9 Hz), 1.50-1.62 (m, 4H), 1.45 (s, 3H), 1.21-1.38 (m, 5H),
0.98 (d, 3H, J ) 6.2 Hz), 0.90 (dd, 3H, J ) 7.0, 7.0 Hz). IR
(CH2Cl2): 2945, 2925, 2858, 1461, 1373, 1098, 1064, 877 cm-1
.
CIMS-NH3: m/z 328 (M + NH4), 311 (M + H), 293, 275, 265,
207. Anal. (C18H30O4) C, H.
(+)-Octa h yd r o-3,6r-d im eth yl-3,12-ep oxy-9â-(3′-(tr iflu o-
r om eth yl)pr opyl)-12H-pyr an o[4,3j]-1,2-ben zodioxepin (9).
Yield 85%; mp 125-128 °C. H NMR (CDCl3): δ 5.53 (s, 1H),
1
4.02 (s, 1H), 3.82 (d, 2H, J ) 18 Hz), 2.31-2.25 (m, 2H), 2.12-
2.03 (m, 5H), 1.93-189 (m, 2H), 1.71-1.69 (m, 2H), 1.62-1.49
(m, 8H), 1.41 (s, 3H), 1.29-125 (m, 2H), 0.95 (d, 3H, J ) 5.0
Hz). FT-ICR MS: m/z 365 (M + H). Anal. (C18H27O4F3) C, H.
(+)-Octa h yd r o-3,6r-d im eth yl-3,12-ep oxy-9â-(3′-p h en yl-
p r op yl)-12H-p yr a n o[4,3j]-1,2-ben zod ioxep in (10). Yield
92%; mp 80-81 °C. 1H NMR (CDCl3): δ 7.15-7.34 (m, 5H),
5.21 (s, 1H), 3.80 (ddd, 1H, J ) 1.1, 4.2, 11.5 Hz), 3.45 (dd,
1H, J ) 11.7, 11.7 Hz), 2.57-2.70 (m, 2H), 2.50 (dddd, 1H, J
) 4.0, 7.6, 11.6, 11.6 Hz), 2.39 (ddd, 1H, J ) 4.0, 13.4, 14.5
Hz), 2.03 (ddd, 1H, J ) 3.0, 4.8, 14.7 Hz), 1.89 (dddd, 1H, J )
2.9, 3.8, 6.7, 13.6 Hz), 1.63 (dd, 2H, J ) 7.7, 7.7 Hz), 1.44 (s,
3H), 1.27 (dd, 2H, J ) 6.4, 11.3 Hz), 0.96 (d, 3H, J ) 6.2 Hz).
IR: 2927, 2867, 1452, 1097, 1067, 877 cm-1. CIMS-NH3: m/z

Direct Modification of (+)-Artemisinin
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 19 4335
(14) Avery, M. A.; Mehrotra, S.; J ohnson, T. L.; Bonk, J . D.; Vroman,
J . A.; Miller, R. Structure-Activity Relationships of the Anti-
malarial Agent Artemisinin. 5. Analogues of 10-Deoxoartemisi-
nin Substituted at C-3 and C-9. J . Med. Chem. 1996, 39, 4149-
4155.
(15) J ung, M.; Li, X.; Bustos, D. A.; ElSohly, H. N.; McChesney, J .
D.; Milhous, W. K. Synthesis and antimalarial activity of (+)-
deoxoartemisinin. J . Med. Chem. 1990, 33, 1516-1518.
(16) Avery, M. A.; Chong, W. K. M.; J ennings-White, C. Stereoselec-
tive total synthesis of (+)-artemisinin, the antimalarial constitu-
ent of Artemisia annua L. J . Am. Chem. Soc. 1992, 114, 974-
979.
(17) J i, Y. Synthesis of 1-(3-bromopropyl)-4-chloro-benzene. Huagong
Shikan 2001, 15, 31-34.
(18) Sakamuri, S.; Enyedy, I. J .; Kozikowski, A. P.; Zaman, W. A.;
J ohnson, K. M.; Wang, S. Pharmacophore-based discovery,
synthesis, and biological evaluation of 4-phenyl-1-arylalkyl
piperidines as dopamine transporter inhibitors. Bioorg. Med.
Chem. Lett. 2001, 11, 495-500.
(19) Botero Cid, M. H.; Holzgrabe, U.; Kostenis, E.; Mohr, K.;
Traenkle, C. Search for the Pharmacophore of Bispyridinium-
Type Allosteric Modulators of Muscarinic Receptors. J . Med.
Chem. 1994, 37, 1439-1445.
(20) Terada, I.; Matsuzaki, K.; Nonoshita, K.; Fujita, F. Preparation
of pyridine derivatives as insecticides and miticides. WO 9002736;
Idemitsu Kosan Co., Ltd.: J apan, 1990; 129 pp.
(21) Avery, M. A.; J ennings-White, C.; Chong, W. K. M. Total
synthesis of (+)-artemisinin and (+)-9-demethylartemisinin.
Tetrahedron Lett. 1987, 28, 4629-4632.
(22) Kraus, G. A.; Frazier, K. A.; Roth, B. D.; Taschner, M. J .;
Neuenschwander, K. Conversion of Lactones into Ethers. J . Org.
Chem. 1981, 46, 2417-2419.
(+)-Oct a h yd r o-3,6r-d im et h yl-3,12-ep oxy-9â-(3′-(p -t r i-
flu or om eth ylph en yl)pr opyl)-12H-pyr an o[4,3j]-1,2-ben zodi-
oxep in (18). Yield 80%; mp 79-80 °C. 1H NMR (CDCl3): δ
7.51 (d, 2H), 7.28 (d, 2H), 5.19 (s, 1H), 3.76 (dd, 1H, J ) 3.2,
10.9 Hz), 3.47 (dd, 1H, J ) 11.6, 11.5 Hz), 2.61-2.36 (m, 4H),
2.13-1.82 (m, 2H), 1.69-1.48 (m, 7H), 1.43 (s, 3H), 1.41-1.04
(m, 5H), 0.97 (d, 3H, J ) 6.3 Hz). IR (KBr): 1636, 1516, 1457,
1212, 1197, 1158, 1126, 910, 877, 829, 767, 739 cm-1. FABMS
m/z: 447 (M + Li). Anal. (C24H31O4F3) C, H.
(+)-Octa h yd r o-3,6r-d im eth yl-3,12-ep oxy-9â-(3′-(3,5-bis-
tr iflu or om eth ylph en yl)p r opyl)-12H-p yr a n o[4,3j]-1,2-ben -
zod ioxep in (19). Yield 76%; mp 109-110 °C. 1H NMR
(CDCl3): δ 7.69 (s, 1H), 7.59 (s, 2H), 5.19 (s, 1H), 3.76 (dd,
1H, J ) 3.3, 11.52 Hz), 3.44 (dd, 1H, J ) 11.67, 11.7 Hz), 2.72
(m, 2H), 2.5 (m, 1H), 2.35 (m, 1H), 2.1 (m, 1H), 1.8 (m, 1H),
1.46-1.7 (m, 6H), 1.42 (s, 3H), 1.0-1.28 (m, 6H), 0.96 (d, 3H,
J
) 6.09 Hz). FT-ICR MS m/z: 509.50 (M + H). Anal.
(C25H30O4F6) C, H.
(+)-Octa h yd r o-3,6r-d im eth yl-3,12-ep oxy-9â-(3′-(3,5-d i-
flu or op h en yl)p r op yl)-12H -p yr a n o[4,3j]-1,2-b en zod iox-
ep in (20). Yield 70%; mp 119-121 °C. 1H NMR (CDCl3): δ
6.6 (m, 3H), 5.18 (s, 1H), 3.77 (dd, 1H, J ) 3.66, 3.63 Hz), 3.46
(dd, 1H, J ) 11.67, 11.67 Hz), 2.3-2.6 (m, 4H), 2.0 (m, 1H),
1.8 (m, 1H), 1.40-1.75 (m, 6H), 1.41 (s, 3H), 1.0-1.4 (m, 6H),
0.94 (d, 3H, J ) 6.03 Hz). FT-ICR MS m/z: 409.50 (M + H).
Anal. (C23H30O4F2) C, H.
Ack n ow led gm en t. This work was variously sup-
ported by the NIH (Allergy and Infectious Diseases),
WHO (Special Program for Tropical Diseases Research),
CDC (Cooperative Agreement), and Biomedical Re-
search Internship Public Health Service Grant No.
NIGMS-GM55379.
(23) Vroman, J . A.; Khan, I.; Avery, M. A. Conjugate addition of
trimethylsilylmethyl cuprates to artemisinic acid. Homologation
and subsequent conversion to 9.beta.-modified artemisinin
analogues. Synlett 1997, 1438-1440.
(24) Vroman, J . A.; Elsohly, H. N.; Avery, M. A. Conjugate addition
of a cyano-Gilman cuprate to an acrylic acid: homologation of
artemisinic acid and subsequent conversion to 16-butylartemisi-
nin. Synth. Commun. 1998, 28, 1555-1562.
(25) Acton, N.; Karle, J . M.; Miller, R. E. Synthesis and antimalarial
activity of some 9-substituted artemisinin derivatives. J . Med.
Chem. 1993, 36, 2552-2557.
(26) Paitayatat, S.; Tarnchompoo, B.; Thebtaranonth, Y.; Yuthavong,
Y. Correlation of Antimalarial Activity of Artemisinin Deriva-
tives with Binding Affinity with Ferroprotoporphyrin IX. J . Med.
Chem. 1997, 40, 633-638.
(27) El-Feraly, F. S.; Ayalp, A.; Al-Yahya, M. A.; McPhail, D. R.;
McPhail, A. T. Conversion of artemisinin to artemisitene. J . Nat.
Prod. 1990, 53, 66-71.
(28) Corey, E. J .; Venkateswarlu, A. Protection of hydroxyl groups
as tert-butyldimethylsilyl derivatives. J . Am. Chem. Soc. 1972,
94, 6190-6191.
Refer en ces
(1) Desowitz, R. S. The Malaria Capers (More Tales of People,
Research and Reality); W. W. Norton & Company: New York,
1991.
(2) Van Agtmael, M. A.; Eggelte, T. A.; Van Boxtel, C. J . Artemisinin
drugs in the treatment of malaria: from medicinal herb to
registered medication. Trends Pharmacol. Sci. 1999, 20, 199-
205.
(3) Hien, T. T.; White, N. J .; Qinghaosu. The Lancet 1993, 341, 603-
608.
(4) Luo, X. D.; Shen, C. C. The chemistry, pharmacology, and clinical
applications of Qinghaosu (artemisinin) and its derivatives. Med.
Res. Rev. 1987, 7, 29-52.
(5) People’s Health Publishing: Lanzhou, China, 1991; pp 944-946.
(6) Klayman, D. L. Qinghaosu (artemisinin): an antimalarial drug
from China. Science 1985, 228, 1049-1055.
(29) Vennerstrom, J . L.; Fu, H. N.; Ellis, W. Y.; Ager, A. L., J r.; Wood,
J . K.; Andersen, S. L.; Gerena, L.; Milhous, W. K. Dispiro-1,2,4,5-
tetraoxanes: a new class of antimalarial peroxides. J . Med.
Chem. 1992, 35, 3023-3027.
(7) Van Geldre, E.; Vergauwe, A.; Van den Eeckhout, E. State of
the art of the production of the antimalarial compound arte-
misinin in plants. Plant Mol. Biol. 1997, 33, 199-209.
(8) Vroman, J . A.; Alvim-Gaston, M.; Avery, M. A. Current progress
in the chemistry, medicinal chemistry and drug design of
artemisinin based antimalarials. Curr. Pharm. Des. 1999, 5,
101-138.
(9) Brossi, A.; Venugopalan, B.; Dominguez Gerpe, L.; Yeh, H. J .
C.; Flippen-Anderson, J . L.; Buchs, P.; Luo, X. D.; Milhous, W.;
Peters, W. Arteether, a new antimalarial drug: synthesis and
antimalarial properties. J . Med. Chem. 1988, 31, 645-650.
(10) Brewer, T. G.; Grate, S. J .; Peggins, J . O.; Weina, P. J .; Petras,
J . M.; Levine, B. S.; Heiffer, M. H.; Schuster, B. G. Fatal
neurotoxicity of arteether and artemether. Am. J . Trop. Med.
Hyg. 1994, 51, 251-259.
(11) Brewer, T. G.; Peggins, J . O.; Grate, S. J .; Petras, J . M.; Levine,
B. S.; Weina, P. J .; Swearengen, J .; Heiffer, M. H.; Schuster, B.
G. Neurotoxicity in animals due to arteether and artemether.
Trans. R. Soc. Trop. Med. Hyg. 1994, 88, 33-36.
(12) Avery, M. A.; Alvim-Gaston, M.; Rodrigues, C. R.; Barreiro, E.
J .; Cohen, F. E.; Sabnis, Y. A.; Woolfrey, J . R. Structure-Activity
Relationships of the Antimalarial Agent Artemisinin. 6. The
Development of Predictive In Vitro Potency Models Using
CoMFA and HQSAR Methodologies. J . Med. Chem. 2002, 45,
292-303.
(30) Wesche, D. L.; DeCoster, M. A.; Tortella, F. C.; Brewer, T. G.
Neurotoxicity of artemisinin analogues in vitro. Antimicrob.
Agents Chemother. 1994, 38, 1813-1819.
(31) Fishwick, J .; McLean, W. G.; Edwards, G.; Ward, S. A. The
toxicity of artemisinin and related compounds on neuronal and
glial cells in culture. Chem.-Biol. Interact. 1995, 96, 263-271.
(32) Avery, M. A.; McLean, G.; Edwards, G.; Ager, A. Structure-
activity relationships of peroxide-based artemisinin antimalari-
als. In Biologically Active Natural Products; Cutler, S. J ., Cutler,
H. G., Eds.; CRC Press: Boca Raton, 2000; pp 121-132.
(33) Desjardins, R. E. C. C. J .; Haynes, J . D.; Chulay, J . D.
Quantitative assessment of antimalarial activity in vitro by a
semiautomated microdilution technique. Antimicrob. Agents
Chemother. 1979, 16, 710-718.
(34) Peters, W.; Robinson, B. L. Malaria. In Handbook of Animal
Models for Chemotherapy; Zak, O., Sande, M., Eds.; Academic
Press: London, 1999; Chapter 92, pp 757-773.
(35) Shmuklarsky, M. J .; Klayman, D. L.; Milhous, W. K.; Kyle, D.
E.; Rossan, R. N.; Ager, A. L., J r.; Tang, D. B.; Heiffer, M. H.;
Canfield, C. J .; Schuster, B. G. Comparison of â-artemether and
â-arteether against malaria parasites in vitro and in vivo. Am.
J . Trop. Med. Hyg. 1993, 48, 377-384.
(36) Andersen, S. L.; Ager, A.; McGreevy, P.; Schuster, B. G.; Wesche,
D.; Kuschner, R.; Ohrt, C.; Ellis, W.; Rossan, R.; Berman, J .
Activity of azithromycin as a blood schizonticide against rodent
and human plasmodia in vivo. Am. J . Trop. Med. Hyg. 1995,
52, 159-161.
(13) Avery, M. A.; Gao, F.; Chong, W. K. M.; Mehrotra, S.; Milhous,
W. K. Structure-activity relationships of the antimalarial agent
artemisinin. 1. Synthesis and comparative molecular field
analysis of C-9 analogues of artemisinin and 10-deoxoartemisi-
nin. J . Med. Chem. 1993, 36, 4264-4275.
J M020142Z