Enantioselective synthesis of the carbocyclic nucleoside (-)-abacavir

Article abstract of DOI:10.1039/c2ob06775g

An enantiopure β-lactam with a suitably disposed electron withdrawing group on nitrogen, participated in a π-allylpalladium mediated reaction with 2,6-dichloropurine tetrabutylammonium salt to afford an advanced cis-1,4-substituted cyclopentenoid with both high regio- and stereoselectivity. This advanced intermediate was successfully manipulated to the total synthesis of (-)-Abacavir.

Full text of DOI:10.1039/c2ob06775g

View Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Organic &
Biomolecular
Chemistry
Cite this: Org. Biomol. Chem., 2012, 10, 1870
www.rsc.org/obc
PAPER
Enantioselective synthesis of the carbocyclic nucleoside (−)-abacavir†
Grant A. Boyle,a Christopher D. Edlin,a Yongfeng Li,b Dennis C. Liotta,b Garreth L. Morgans*a and
Chitalu C. Musondaa
Received 21st October 2011, Accepted 5th December 2011
DOI: 10.1039/c2ob06775g
An enantiopure β-lactam with a suitably disposed electron withdrawing group on nitrogen, participated in
a π-allylpalladium mediated reaction with 2,6-dichloropurine tetrabutylammonium salt to afford an
advanced cis-1,4-substituted cyclopentenoid with both high regio- and stereoselectivity. This advanced
intermediate was successfully manipulated to the total synthesis of (−)-Abacavir.
the furanose ring has been replaced by an isoteric methylene or
ethylene group (vida infra).
Introduction
Acquired immune deficiency syndrome (AIDS) has rapidly
become one of the major causes of death in the world.1 It is esti-
mated that over 60 million people have been infected with the
human immunodeficiency virus (HIV), which is the causative
agent of AIDS. Since the discovery of HIV in the 1980s, remark-
able progress has been made in the development of novel anti-
viral drugs.2 Entecavir 1, Carbovir 2a, and Abacavir 2b are some
well known examples of carbocyclic nucleosides that have arisen
from this extraordinary effort (Fig. 1). Carbocyclic nucleosides
have been the subject of extensive investigation into their poten-
tial uses as antiviral agents for example as therapeutic agents to
address hepatitis and herpes virus and HIV.3,4 The antiviral prop-
erties exhibited by these carbocyclic nucleosides are due, in part,
to their metabolic and chemical stability towards phosphorylases
and phosphotransferases.5 These compounds are structural ana-
logues of endogenous nucleosides in which the oxygen atom in
The construction of carbocyclic nucleosides, more specifically
the putting together of a heterocyclic base with a 1,4-substituted
carbocyclic sugar analogue is generally carried out in one of
three ways. Firstly, by direct substitution of the hydroxyl group
by a Mitsunobu coupling reaction on the general carbocyclic
ring 3 with a heterocyclic base (Scheme 1).6 These reactions
result in a net inversion of hydroxyl stereochemistry.
A second approach is to use Trost palladium catalysed allylic
alkylation chemistry of activated esters such as 4. This protocol
follows a double displacement/inversion and a resultant net
retention of configuration. This is a particularly useful method
for the convergent synthesis of carbocyclic nucleosides and can
be achieved directly from a chiral sugar or from desymmetriza-
tion of a meso intermediate, amongst others.7–11
A final general approach to this class of molecules follows the
synthesis of the carbocyclic nucleoside by linear construction of
the heterocyclic fragment from an amino group of the cyclopen-
tyl moiety 5.12–14 Also, Jung reported the use of an activated
amine as a leaving group in a palladium catalysed substitution
reaction.15
The utilization of functionalized cyclopentene as the source
of the sugar fragment is a common approach to the synthesis
of carbocyclic nucleosides, particularly in conjunction with
Fig. 1 Some examples of relevant carbocyclic nucleosides.
aiThemba Pharmaceuticals PTY (Ltd.), PO BOX 21Modderfontein, 1645
Gauteng, South Africa. E-mail: gmorgans@ithembapharma.com;
Fax: +27 (0)11 372-4518; Tel: +27 (0)11 372-4500
bEmory University, Department of Chemistry, 1521 Dickey Dr Emerson
403, Atlanta, GA, USA, 30322. Tel: 404-727-8130
†Electronic supplementary information (ESI) available: Copies of 1H
NMR and 13C NMR spectra of new compounds. See DOI: 10.1039/
c2ob06775g
Scheme 1 General approaches to carbocyclic nucleosides.
1870 | Org. Biomol. Chem., 2012, 10, 1870–1876
This journal is © The Royal Society of Chemistry 2012

View Online
asymmetric synthesis, chiral pool and enzymatic resolution
approaches.16,17
Among these, the enantioselective construction of the cyclo-
pentenoids has involved two discreet stages, an initial [4 + 2]
cycloaddition reaction of cyclopentadiene with either CvN,
NvO and O2 dienophiles, followed by enzymatic hydrolysis.
An example of
a CvN dienophile-enzymatic hydrolysis
approach has made use of 2-azabicyclo[2.2.1]hept-5-en-3-one
( ) 6, which is readily prepared from the [4 + 2] cycloaddition
between cyclopentadiene and tosylcyanide, followed by aqueous
hydrolysis.18 The resultant lactam can be enzymatically resolved
to either enantiomer 6 (Scheme 2).19 Similarly, an NvO dieno-
phile-enzymatic hydrolysis approach has made use of 4-acetami-
docyclopent-2-en-1-yl acetate 8 which is readily synthesized
from the [4 + 2] cycloaddition of cyclopentadiene and nitroso-
carbamate, generated in situ, to afford adduct 7.20 Compound 8
is conveniently resolved via an enzymatic hydrolysis that utilizes
electric eel acetyl cholinesterase to afford enantiomerically pure
9. Finally, di-acetate 11 can formally be regarded as the reduction
and acetylation product of the [4 + 2] cycloaddition product 10
of cyclopentadiene and singlet oxygen.21 The bis-acetate 11 can
be resolved by enantioselective enzymatic hydrolysis with elec-
tric eel acetyl cholinesterase22 or with porcine pancreas lipase to
afford enantiomerically pure 12.23
Scheme 3 Proposed use of π-allylpalladium chemistry for the syn-
thesis of advanced intermediate 15 from lactam 16.
of the π-allylpalladium complex would be facilitated by the
relief of the four membered ring strain25 and the formation of the
requisite cis-1,4-substituted cyclopentenoid moiety would
proceed with a net retention of configuration as a consequence of
the well established double inversion that is operative in these
reactions. Finally, further manipulation of intermediate 15 might
then result in the formation of the carbocyclic nucleoside Abaca-
vir 2b.
While there have been many reported methods for the syn-
thesis of various substituted carbocyclic nucleosides, particularly
Abacavir,26–35 overall yields of the reported synthetic routes are
generally low and the synthetic schemes often require extensive
manipulations of the [4 + 2] cycloadduct being utilized. Because
Abacavir has proven to be an important component of many
HIV combination regimens, particularly paediatrics, there is a
need for improved methods for producing Abacavir that require
fewer synthetic steps and higher overall yields.
Results and discussion
With this in mind, the starting material for the synthesis was the
known β-lactam 13, which was prepared in two steps and in high
enantiomeric purity.24 To facilitate the novel π-allylpalladium
complex formation an anion stabilizing electron withdrawing
group needed to be installed onto the nitrogen of lactam 13, and
cheap readily available sulfonyl chlorides, sulfonyl anhydrides
and carbamates were initially screened (Scheme 4). An initial
screen of alkyl and aromatic sulfonylchlorides (entries 1–3)
proved disappointing as complex reaction mixtures were
obtained. However after further optimization, treatment of the
resolved lactam 13 with n-butyllithium, followed by quenching
with p-toluenesulfonyl chloride gave the desired product 18d in
60% yield (entry 4) after conventional workup. Initially, these
compounds exhibited poor shelf stability. However, subsequent
studies showed that the compounds were indefinitely stable as
long as they were pure, dry and free of residual acid. Sub-
sequently, a more reliable process was developed, i.e., lactam 13
was allowed to react with 4-methylbenzenesulfonic anhydride in
the presence of triethylamine and catalytic DMAP to afford 18d
Scheme 2 Some relevant [4 + 2]–enzymatic hydrolysis approaches to
chiral sugar building blocks.
We hypothesized that carbocyclic nucleosides could be pre-
pared using a complementary approach involving a [2 + 2]
cycloaddition of cyclopentadiene with chlorososulfonyl isocya-
nate (CSI), followed by enzymatic resolution and nucleobase
addition by way of a π-allylpalladium intermediate. Specifically,
racemic lactam 13, readily prepared from cyclopentadiene and
CSI after reductive workup, can be efficiently resolved by
enzyme catalyzed hydrolysis (Scheme 3).24 We envisioned that a
suitably activated derivative 16 might be an attractive precursor
of a π- allylpalladium intermediate 17 that could be captured
directly by an intact nucleobase. Based on the chemistry devel-
oped by Trost, we thought it likely that the reaction of the
nucleobase with a π-allylpalladium intermediate would be both
regioselective and stereoselective (see Scheme 3). The formation
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 1870–1876 | 1871

View Online
Scheme 4 Summary of attempts at installing an electron withdrawing
group onto lactam 13.
in reproducible yields of 75–81% (entry 6). Presumably, this
increase in yield observed when sulfonyl anhydrides are used in
place of the corresponding sulfonyl chlorides is a consequence
of the relatively low reactivity of the conjugate base, p-toluene-
sulfonate, relative to chloride, as well as the absence of a Lewis
acid. Finally, reaction with di-tert-butyl dicarbonate afforded the
known compound 18e36 in 71% yield (entry 7).
With the N-sulfonylamido and Boc derivatives in hand, we
were in a position to attempt the key reaction in our sequence,
i.e., the formation of a nucleoside intermediate through the direct
introduction of 2,6-dichloropurine37 tetrabutylammonium salt by
way of a π-allylpalladium complex. Reaction of 18d with 2,6-
dichloropurine tetrabutylammonium salt using either Pd(OAc)2
or Pd2(dba)3 in the presence of P(i-OPr)3 as the phosphine
ligand in tetrahydrofuran at ambient temperature afforded 19 in
75% yield (Scheme 5). Moreover, analysis of the crude reaction
Scheme 5 Synthesis of Abacavir 2b.
1
sample of 19 by H NMR indicated ca. 7% of the N7 regio-
isomer, attesting to the high regioselectivity of the reaction proto-
col. Also, reaction of the 2,6-dichloropurine salt with 18e, using
identical conditions to those described above, failed to give the
desired product, suggesting that the Boc group was not suffi-
ciently electron-withdrawing to facilitate the ring-opening
reaction.
Next, the transformation of 19 to 21 was carried out. That is,
N-methylation of 19 which can be achieved using methyl iodide
(74%), dimethylsulfate (70%) or under Mitsunobu conditions
(94%) to afford 20 in good to excellent yields. Smooth reductive
amido bond cleavage of 20 with sodium borohydride followed
by heterocyclic amination with cyclopropylamine in refluxing
ethanol finally afforded 21 in 63% over the two steps.
The synthesis of Abacavir 2b was completed by reaction of 21
with hydrazine hydrate in warm methanol/water mixture fol-
lowed by treatment with sodium nitrite to afford the putative
intermediate 22. The product was not characterized but immedi-
ately subjected to stannous chloride mediated reduction to afford
Abacavir 2b in 70% yield over the three steps.
Finally, a more direct approach to introduce the NH2 function-
ality at the 6-position of the purine base was investigated using
4-methoxybenzylamine as an ammonia surrogate (Scheme 6).
This route appeared attractive as it avoids the use of potentially
explosive and toxic reagents as exhibited in the closing synthetic
sequences in Scheme 5. Thus, reaction of 21 with an excess of
Scheme 6 Use of 4-methoxybenzylamine as an ammonia surrogate.
4-methoxybenzylamine in hot DMSO afforded compound 23 in
90% isolated yield. This compound when warmed in DMSO in
the presence of HCl resulted in full decomposition of starting
material as assessed by TLC analysis. However, when 23 was
exposed to neat TFA at 50 °C for 72 h, (−)-Abacavir 2b was
obtained in 73% yield.
In summary, an efficient enantioselective synthesis of the car-
bocyclic nucleoside, (−)-Abacavir 2b has been accomplished by
exploiting an enzymatic resolution sequence for the rapid asym-
metric construction of the sugar fragment of the nucleoside. In
addition, a direct palladium catalyzed coupling of the sugar frag-
ment to the purine base allowed for the highly convergent
assembly of the nucleoside analogue 19 which was seamlessly
1872 | Org. Biomol. Chem., 2012, 10, 1870–1876
This journal is © The Royal Society of Chemistry 2012

View Online
converted to (−)-Abacavir 2b. Lastly, introduction of the amine
functional group at the 6-position of the purine base was
approached in a more direct manner from 21 using 4-methoxy-
benzylamine as an ammonia surrogate. This allowed for one less
step as well as avoiding potentially explosive intermediates
(azide 22) and toxic reagents (SnCl2) en route to (−)-Abacavir
2b.
CHCl3), lit.,36 [α]D25 −35.0 (c 0.2 in CHCl3); δH (400 MHz,
CDCl3, Me4Si) 2.40–2.50 (m, 1H), 2.65–2.75 (m, 1H),
3.75–3.85 (m, 1H), 4.5 (s, 1H), 5.90–5.95 (m, 1H), 6.00–6.05
(m, 1H), 6.4 (br s, 1H); δC (100 MHz, CDCl3) 172.5, 136.7,
130.6, 59.1, 53.0, 30.6
(1S,5R)-6-Tosyl-6-azabicyclo[3.2.0]hept-3-en-7-one 18d
Method 1. Compound 13 (15.0 g, 137 mmol) was added to
dichloromethane (160 mL) and cooled to 0 °C under nitrogen.
DMAP (1.68 g, 13.8 mmol) and triethylamine (16.7 g,
165 mmol) were added and the mixture stirred for 2 min to com-
pletely dissolve the solids. 4-Methylbenzenesulfonic anhydride
(55.0 g, 169 mmol) was added portion wise to the reaction
mixture over 5 min. The reaction mixture went to a dark brown
red colour and was allowed to warm to room temperature and
stirred at ambient temperature for an additional 24 h. The reac-
tion mixture was washed with brine (2 × 100 mL) and water (2 ×
100 mL). The organic extract was dried over sodium sulfate,
filtered and evaporated to afford a brown residue. The residue
was dissolved into a minimum amount of dichloromethane and
filtered through a silica gel pad with dichloromethane as the
eluent. The washings were monitored by TLC to ensure all the
product had eluted. The residue was recrystallized from ethyl
acetate–hexane mixtures to afford 18d (27.0 g, 75%) as a crystal-
line white solid. mp: 93–95 °C (ethyl acetate–hexane); [α]D20
−125.3 (c 0.50 in CHCl3); νmax/cm−1 3068, 2921, 1781, 1348,
119; δH (400 MHz, CDCl3, Me4Si) 7.86 (d, J = 8.4 Hz, 2H),
7.34 (d, J = 8.0 Hz, 2H), 6.04–6.02 (m, 2H), 5.02–4.99 (m, 1H),
3.85–3.81 (m, 1H), 2.71–2.66 (m, 1H), 2.54–2.45 (m, 1H), 2.44
(s, 3H); δC (100 MHz, CDCl3) 167.0, 145.0, 138.7, 136.4,
129.9, 128.3, 127.3, 66.0, 51.8, 31.0, 21.6; m/z 264.0651 (MH+,
C13H14NO3S requires 264.0694).
Experimental
All the reactions dealing with air or moisture sensitive com-
pounds were carried out in a dry reaction vessel under a positive
pressure of nitrogen. Analytical thin-layer chromatography was
performed on an aluminium backed plates coated with 0.25 mm
230–400 mesh silica gel containing a fluorescent indicator. Thin
layer chromatography plates were visualized by exposure to
ultraviolet light (254 nm) and/or by immersion in basic KMnO4
followed by heating. Organic solutions were concentrated by
rotary evaporation at c.a. 30 mmHg. Flash column chromato-
graphy was performed on Silica gel 60 (230–400 mesh, ASTM).
IR spectra were recorded as neat liquids or KBr pellets and
absorptions are reported in cm−1. NMR spectra were measured
1
on a Bruker 400 MHz Avance instrument at 400 MHz for H
and 100 MHz for 13C NMR, using tetramethylsilane as an
internal reference and CDCl3 or d6-DMSO as a solvent. Chemi-
cal shift values for protons are reported in parts per million
(ppm, δ scale) downfield from tetramethylsilane and are refer-
enced to residual proton of CDCl3 (δ 7.26) or residual protons of
d6-DMSO (δ 2.50). Carbon nuclear magnetic resonance spectra
(13C NMR) were recorded at 100 MHz: chemical shifts for
carbons are reported in parts per million (ppm, δ scale)
downfield from tetramethylsilane and are referenced to the
carbon resonance of CDCl3 (δ 77.0) or the carbon resonance of
d6-DMSO (δ 39.51). Data are presented as follows: chemical
shift, multiplicity (s = singlet, d = doublet, t = triplet, q =
quartet, quint = quintet, sext = sextet, sept = septet, m = multi-
plet and/or multiplet resonances, br = broad), coupling constant
in hertz (Hz), and signal area integration in natural numbers.
High resolution mass spectra were measured using a Waters API
Q-TOF Ultima instrument. Optical rotations were recorded on a
Jasco DIP-370 Polarimeter at 20 °C or 25 °C.
Method 2. A solution of 13 (2.00 g, 18.32 mmol) in dry THF
(33 mL) was added drop wise to a stirred mixture of 1.6 M n-
BuLi in hexane (19.5 mL, 31.21 mmol) in dry THF (33 mL) at
−78 °C under Argon. The mixture was stirred at −78 °C for 1 h
and p- toluenesulfonyl chloride (4.65 g, 24.40 mmol) was
added. The reaction mixture was gradually warmed to ambient
temperature. The solvent was evaporated in vacuo, and the
residue was purified by flash chromatography (SiO2, hexane–
ethyl acetate, 4 : 1) to give 18d (2.90 g, 60%) as a white solid.
Analytical data were identical to those reported above.
All reagents were purchased from commercial sources unless
otherwise noted.
(1S,5R)-6-Azabicyclo[3.2.0]hept-3-en-7-one 13
(1S,5R)-tert-Butyl-7-oxo-6-azabicyclo[3.2.0]hept-3-ene-6-
carboxylate 18e
6-Azabicyclo[3.2.0]hept-3-en-7-one38 (55.0 g, 504 mmol) was
added to diisopropyl ether (1000 mL) and filtered through a por-
osity 3 glass sintered funnel to give a yellow homogeneous sol-
ution. Novozyme® 435 (53.0 g, 53 mg mL−1) and distilled
water (9.10 mL, 1.00 equiv.) was added. The reaction mixture
was heated, without stirring, to 70 °C over 1 h and maintained at
that temperature for 11 h. The reaction mixture was allowed to
cool overnight and the enzyme filtered off. The enzyme was
washed with diisopropyl ether (5 × 75 mL) and the combined
organic fractions were evaporated to give an orange solid. The
solid was recrystallized from diisopropyl ether (75 mL) to afford
13 (23.0 g, 42%) as a crystalline tan solid. [α]2D0 −42.0 (c 0.5 in
To a solution of 13 (1.00 g, 9.16 mmol), DMAP (0.22 g,
1.8 mmol) and Et3N (3.80 mL, 27.3 mmol) in tetrahydrofuran
(20 mL), Boc2O (2.00 g, 9.16 mmol) was added in several por-
tions at 0 °C. The mixture was stirred at ambient temperature
over the weekend. The reaction mixture was diluted with ethyl
acetate (60 mL) and washed with water (3 × 20 mL). The
organic extract was dried over sodium sulfate, filtered and con-
centrated under vacuum. The crude product was purified by flash
chromatography (SiO2, hexane–ethyl acetate, 4 : 1) to afford 18e
(1.44 g, 75%) as an orange oil. The spectral and analytical prop-
erties matched those reported in the literature.36
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 1870–1876 | 1873

View Online
(1S,4R)-4-(2,6-Dichloro-9H-purin-9-yl)-N-tosylcyclopent-2-
enecarboxamide 19
to afford 19 (12.8 g, 75%) as a white solid. The solid was iso-
lated as a tenacious acetone adduct which could not be removed
after extended high vacuum drying at 60 °C. This adduct had no
consequence in the next step though.
Small scale. To a stirred solution of tetrabutylammonium salt
of 2,6-dichloropurine (0.67 g, 1.5 mmol) in anhydrous THF
(20 mL) was added DMF (10 mL). Pd(OAc)2 (34 mg,
0.15 mmol) and triisopropyl phosphite (0.21 mL, 0.91 mmol)
were added and stirred under argon at ambient temperature for
1 h. A solution of 18d (0.40 g, 1.5 mmol) in dry THF (5 mL)
was added drop wise to the resultant mixture, which was then
stirred for a further 2 h. The solvent was evaporated in vacuo,
and the residue was purified by flash chromatography (SiO2,
dichloromethane–methanol, 19 : 1) to give 19 (0.33 g, 49%) as a
yellow solid. mp 214–215 °C (from MeOH); [α]2D5 −81.6 (c 0.50
in MeOH); νmax/cm−1 3256, 2966, 1686, 1683, 1609, 1503; δH
(400 MHz, CDCl3, Me4Si) 12.35 (brs, 1H), 7.80 (d, J = 8.0 Hz,
2H), 7.39 (d, J = 8.0 Hz, 2H) 6.22–6.18 (m, 1H), 6.12–6.07 (m,
1H), 5.7.–5.63 (m, 1H), 3.76–3.68 (m, 1H), 2.80–2.70 (m,1H),
2.38 (s, 3H), 2.00–2.05 (m, 1H); δC (100 MHz, CDCl3) 171.1,
152.9, 150.8, 149.5, 146.1, 144.3, 136.2, 135.1, 130.9, 130.6,
129.5, 127.5, 59.6, 50.7, 32.8, 21.0; m/z 452.0342 (MH+,
C18H16Cl2N5O3S requires 452.0351).
Routes to (1S,4R)-4-(2,6-dichloro-9H-purin-9-yl)-N-methyl-N-
tosylcyclopent-2-enecarboxamide 20
Iodomethane as the electrophile. Compound 19 (5.00 g,
11.05 mmol) in acetone (50 mL) was reacted with methyl iodide
(6.81 g, 48.0 mmol) in the presence of K2CO3 (4.58 g,
33.2 mmol) at reflux temperature for 4 h. The reaction mixture
was filtered through Celite and the Celite pad was washed with
additional acetone (50 mL). The solvent was evaporated and the
residue was taken into ethyl acetate (50 mL), and washed with
water (25 mL) and brine (25 mL). The organic extract was dried
over sodium sulfate, filtered and evaporated. The residue was tri-
turated with diethyl ether to afford 20 (3.80 g, 74%). [α]2D0 −73.3
(c 0.5 in MeOH); νmax/cm−1 3256, 2966, 1686, 1683, 1609,
1503; δH (400 MHz, CDCl3, Me4Si) 8.33 (s, 1H), 7.76 (d, J =
8.0 Hz, 2H), 7.38 (d, J = 8.0 Hz, 2H), 6.21–6.78 (m, 1H),
6.01–5.99 (m, 1H), 5.87–5.83 (m, 1H), 4.60–4.57 (m, 1H), 3.27
(s, 3H), 2.93–2.85 (m, 1H), 2.47 (s, 3H), 2.20–2.14 (m, 1H); δC
(100 MHz, CDCl3) 173.7, 152.7, 152.6, 151.5, 145.5, 145.3,
136.8, 135.8, 130.8, 130.5, 130.2, 127.2, 59.4, 50.9, 35.5, 33.4,
21.6; m/z 466.0524 (MH+, C19H18Cl2N5O3S requires 466.0507).
Large scale preparation. 2,6-Dichloropurine (15.0 g,
79.0 mmol) was partially dissolved into tetrahydrofuran
(50 mL). A freshly prepared solution of tetrabutylammonium
hydroxide hydrate (Sigma-Aldrich, 63.5 g, 79.0 mmol) in deio-
nised water (200 mL) was added to the reaction mixture. The
mixture solubilised and after 2 h the solvents were evaporated.
Toluene (50 mL) was added and evaporated to dryness again.
This process was repeated twice more. The semi-solid residue
was triturated with diethyl ether (300 mL) under rapid stirring
for 3 h. The solids were filtered and dried under high vacuum to
afford tetrabutylammonium 2,6-dichloropurin-9-ide (32.2 g,
94%) as a white free-flowing solid. The solid was stored in a
well sealed desiccator and away from light when not in use.
Pd2(dba)3 (1.74 g, 1.90 mmol) and triisopropylphosphite
(2.00 mL, 7.70 mmol) were added to dry tetrahydrofuran
(100 mL), degassed and purged with nitrogen 5 times, and
allowed to stir for 30 min during which time the solution
changed from purple to dark green. Tetrabutylammonium 2,6-
dichloropurin-9-ide (16.4 g, 38.0 mmol) was added and allowed
to stir until all the solids had dissolved. Finally, 18d (10.0 g,
38.4 mmol) was added and the reaction mixture was degassed
and purged with nitrogen 5 times. The reaction mixture was
allowed to stir for 90 min. The reaction mixture was filtered
through Celite and the Celite pad was washed with additional
tetrahydrofuran (3 × 50 mL). The solvent was evaporated to give
a red oil. Analysis by proton NMR suggested that the N7 vs. N9
substitution ratio as 7% and 93% respectively. The oil was taken
up into ethyl acetate (500 mL) and washed with 10% HCl (4 ×
50 mL) and brine (50 mL). The organic extract was dried over
magnesium sulfate, filtered and evaporated to dryness to afford a
red glass. The glass was rapidly stirred in hexane (50 mL) and
acetone (75 mL) was added until solids started to crash out of
solution. The heterogeneous solution was stirred at 0 °C for
30 min and filtered. The solids were washed with chilled
acetone–hexane mixtures (1 : 1, 3 × 15 mL), and air-dried on the
filter for 5 min. The solids were further dried under high vacuum
Dimethylsulfate as the electrophile. Compound 19 (1.00 g,
2.21 mmol) in acetone (50 mL) was reacted with dimethylsulfate
(0.73 g, 5.7 mmol) in the presence of K2CO3 (0.76 g,
5.53 mmol) at ambient temperature overnight. The reaction
mixture was filtered through Celite and the Celite pad was
washed with additional acetone (50 mL). The solvent was evap-
orated and the residue was taken into ethyl acetate (50 mL), and
washed with water (25 mL) and brine (25 mL). The organic
extract was dried over sodium sulfate, filtered and evaporated.
The residue was triturated with diethyl ether to afford 20 (0.72 g,
70%). The analytical data matched those reported above.
Mitsunobu reaction. To a stirring solution of 19 (0.65 g,
1.44 mmol) in THF–CH2Cl2 (1 : 1, 20 mL) was added methanol
(0.2 mL), PPh3 (1.33 g, 5.76 mmol), and diisopropyl azodicar-
boxylate (0.98 mL, 5.74 mmol). After stirring for 30 min under
argon, the solvent was evaporated in vacuo to give a residue,
which was purified by flash chromatography (SiO2 hexane–ethyl
acetate, 1 : 2) to afford 20 (0.63 g, 1.35 mmol, 94%) as a solid.
The analytical data matched those reported above.
((1S,4R)-4-(2-Chloro-6-(cyclopropylamino)-9H-purin-9-yl)
cyclopent-2-en-1-yl)methanol 21
A solution of 20 (2.80 g, 6.00 mmol) in isopropyl alcohol
(8 mL) and tetrahydrofuran (32 mL) was cooled to 0 °C. Sodium
borohydride (0.227 g, 6.00 mmol) was added portion wise over
1 min, and stirring was continued until TLC analysis indicated a
complete reaction. This took 60 min. The reaction was quenched
with water (20 mL) at 0 °C and allowed to warm to ambient
temperature for 2 h. The reaction mixture was taken into ethyl
acetate (100 mL) and washed with water (50 mL) and brine
1874 | Org. Biomol. Chem., 2012, 10, 1870–1876
This journal is © The Royal Society of Chemistry 2012

View Online
(2 × 50 mL). The aqueous layer was re-extracted with ethyl
acetate (2 × 50 mL). The pooled organic extract was dried over
magnesium sulfate, filtered and evaporated to give a brown
residue. Flash chromatography (SiO2, hexane–ethyl acetate,
1 : 2) afforded ((1S,4R)-4-(2,6-dichloro-9H-purin-9-yl)cyclopent-
2-en-1-yl)methanol (1.48 g, 86%) as a viscous oil. δH
(400 MHz, CDCl3, Me4Si) 8.52 (s, 1H), 6.25–6.23 (m, 1H),
5.88–5.85 (m, 1H), 5.80–5.77 (m, 1H), 3.90 (dd, J = 10.5, 3.7,
1H), 3.75 (dd, J = 10.5, 3.7, 1H), 3.50 (brs, 1H), 3.12–3.10 (m,
1H), 2.93–2.84 (m, 1H), 1.92–1.86 (m, 1H); δC (100 MHz,
CDCl3) 152.7, 152.6, 151.2, 145.7, 140.3, 130.6, 128.8, 64.0,
60.6, 47.5, 34.0. The data matched those in the literature.35 Dry
ethanol (10 mL) was added to the above oil (1.48 g, 5.19 mmol)
and stirred until the oil dissolved completely. Cyclopropylamine
(2.00 mL, 28. 4 mmol) was added and the reaction mixture was
warmed to 50 °C and maintained at that temperature for 5 h. The
solvent was evaporated and the residue taken up into acetone
(25 mL) and stirred with sodium bicarbonate (500 mg) for
30 min. The organic layer was filtered and evaporated to give a
residue that was purified by flash chromatography (SiO2, ethyl
acetate) to afford 21 (1.15 g, 3.76 mmol, 73%) as a white foam
that collapsed into a gum over time. [α]2D0 = 23.3 (c 0.50 in
CHCl3) νmax/cm−1 3322, 3256, 1686, 1683, 1503; δH
(400 MHz, CDCl3, Me4Si) 7.81 (s, 1H), 6.80 (brs, 1H), 6.14 (m,
1H), 5.81 (m, 1H), 5.62 (m, 1H), 3.77 (m, 1H), 3.66 (m, 1H),
2.97 (m, 3H), 2.81 (m, 1H), 1.81 (m, 1H), 0.87 (m, 2H), 0.6 (m,
2H); δC (100 MHz, CDCl3) 156.2, 154.1, 138.9, 138.8, 129.6,
118.6, 64.5, 60.0, 47.5, 34.0, 7.2; m/z 306.1122 (MH+,
C14H17ClN5O requires 306.1122).
overnight. The reaction mixture was diluted with ethyl acetate
(50 mL) and washed with water (3 × 25 mL) and brine (20 mL).
The organic extract was dried over sodium sulfate, filtered and
evaporated to give an oil that was chromatographed (SiO2, ethyl
acetate–methanol, 98 : 2) twice by way of preparative TLC to
afford ((1S,4R)-4-(6-(cyclopropylamino)-2-((4-methoxybenzyl)
amino)-9H-purin-9-yl)cyclopent-2-en-1-yl)methanol 23 (0.21 g,
90%) as an orange gum. [α]2D0 −2.7 (c 0.50 in CHCl3); δH
(400 MHz, CDCl3, Me4Si) 7.41 (s, 1H), 7.30 (d, J = 8.7, 2H),
6.81 (d, J = 8.7, 2H), 6.09–6.03 (m, 1H), 5.90 (brs, 1H),
5.79–5.73 (m, 1H), 5.46–5.37 (m, 1H), 5.24–5.16 (m, 1H), 4.56
(brs, 1H), 4.54 (brs, 1H), 3.83–3.77 (m, 1H), 3.76 (s, 3H),
3.74–6.37 (m, 1H), 3.05 (brs, 1H), 2.94 (brs, 1H), 2.82–2.64 (m,
1H), 2.21 (brs, 1H), 2.08–1.97 (m, 1H), 0.83–0.72 (m, 2H),
0.62–0.49 (m, 2H); δC (100 MHz, CDCl3) 159.0, 158.6, 156.0,
137.7, 136.2, 132.5, 130.6, 129.1, 114.9, 113.7, 65.2, 60.8, 55.2,
47.7, 45.4, 32.7, 23.6, 7.2; m/z 407.2208 (MH+, C22H27N6O2
requires 407.2195).
Compound 23 (0.16 g, 0.38 mmol) was dissolved into chloro-
form (5 mL) and stirred under nitrogen. TFA (1.5 mL) was
added and the reaction mixture was heated at 50 °C for 18 h.
More TFA (2.5 mL) was added and stirring was continued an
additional 18 h at 50 °C. Finally, an additional aliquot of TFA
(2.5 mL) was added and stirring was continued for 18 h at
50 °C. The solvent was evaporated and the residue was neutral-
ized with saturated sodium carbonate solution (20 mL) and
extracted with chloroform (4 × 10 mL) and dried over mag-
nesium sulfate. Preparative TLC (SiO2, chloroform–methanol,
97 : 3) afforded 2b (0.080 g, 73%) as a solid. The analytical data
matched those given earlier for 2b above.
Routes to Abacavir 2b
Azide route. Compound 21 (0.20 g, 0.66 mmol) was dis-
solved in hydrazine monohydrate (10 mL) and methanol (5 mL)
and heated at 50 °C overnight. The reaction mixture was concen-
trated to dryness and co-evaporated with 2-propanol (2 × 30 mL)
until a white gum was obtained. The residue was dissolved in a
10% aqueous acetic acid solution (10 mL) and cooled in an ice
bath. Sodium nitrite (75 mg, 1.10 mmol) was added, and the
mixture was stirred for 1 h. After evaporating the solvent, the
crude product was dissolved in ethanol (20 mL) and tin(II) chlor-
ide dihydrate (315 mg, 1.41 mmol) was added and the reaction
mixture was refluxed for 2 h. The solvent was evaporated and
the residue was purified by column chromatography (SiO2,
dichloromethane–methanol, 19 : 1) to afford 2b (0.13 g, 70%
yield) as a solid. [α]2D5 −37.9 (c 0.29 in MeOH), lit.35 [α]D24
−37.5 (c 0.51 in MeOH), νmax/cm−1 3221, 3207, 1589, 1474; δH
(400 MHz, CDCl3, Me4Si) 7.48 (s, 1H), 6.37 (brs, 1H),
6.09–6.08 (m, 1H), 5.76–5.74 (m, 1H), 5.42–5.40 (m, 1H), 5.12
(brs, 2H), 3.81–3.78 (m, 1H), 3.72–3.69 (m, 1H), 3.4 (brs, 1H),
3.05–2.92 (m, 2H), 2.76–2.70 (m, 2H), 2.00–1.94 (m, 1H),
0.82–0.77 (m, 2H), 0.58–0.54 (m, 2H); δC (100 MHz, CDCl3)
159.4, 156.3, 149.9, 138.1, 136.6, 130.3, 115.1, 65.0, 61.1, 47.6,
32.6, 23.5, 7.3
References
1 UNAIDS, WHO (December 2009). “The Global Aids Epidemic”.http://
data.unaids.org/pub/FactSheet/2009/20091124_FS_global_en.pdf.
Retrieved 18 December 2010.
2 E. De Clercq, “The design of drugs for HIV and HCV”, Nat. Rev. Drug
Discovery, 2007, 6, 1001.
3 D. M. Huryn and M. Okabe, Chem. Rev., 1992, 92, 1745.
4 C. Simons, Nucleoside Mimetics: Their Chemistry and Biological Prop-
erties, Gordon and Breach Science Publishers, Amsterdam, 2001.
5 G. N. Jenkins and N. J. Turner, Chem. Soc. Rev., 1995, 24, 169.
6 For an example see M. Asami, J. Takahashi and S. Inoue, Tetrahedron:
Asymmetry, 1994, 5, 1649.
7 B. M. Trost, G.-H. Kuo and T. Benneche, J. Am. Chem. Soc., 1988, 110,
621.
8 B. M. Trost, L. Li and S. D. Guile, J. Am. Chem. Soc., 1992, 114, 8754.
9 B. M. Trost, R. Madsen, S. D. Guile and A. E. H. Elia, Angew. Chem.,
Int. Ed. Engl., 1996, 35, 1569.
10 B. M. Trost, R. Madsen, S. D. Guile and B. Brown, J. Am. Chem. Soc.,
2000, 122, 5947.
11 B. M. Trost, M. Osipov, P. S. J. Kaib and M. T. Sorum, Org. Lett., 2011,
13, 3222.
12 J. R Huff, Bioorg. Med. Chem., 1999, 7, 2667.
13 M. Tanaka, Y. Norimine, T. Fujita, H. Suemune and K. Sukai, J. Org.
Chem., 1996, 61, 6952.
14 M. Tanaka, M. Yoshioka and K. Sakai, Tetrahedron: Asymmetry, 1993, 4,
981.
15 M. E. Jung and H. Rhee, J. Org. Chem., 1994, 59, 4719.
16 M. T. Crimmins, Tetrahedron, 1998, 54, 9229.
17 C. A. M. Afonso and V. B. Kurteva, Chem. Rev., 2000, 109, 6809.
18 R. Vince, M. Hua, J. Brownell, S. M. Daluge, F. C. Lee, W. M. Shannon,
G. C. Lavelle, J. Quails, O. S Weislow, R. Kiser, P. G. Canonico,
R. H. Schultz, V. L. Narayanan, J. G. Mayo, R. H. Shoemaker and
M. R. Boyd, Biochem. Biophys. Res. Commun., 1988, 156, 1046.
Ammonia surrogate route. Compound 21 (0.17 g,
0.57 mmol) was added to DMSO (2.00 mL). 4-Methoxybenzyl-
amine (0.40 mL, 3.1 mmol) was added and the reaction mixture
was heated to 150 °C and maintained at that temperature
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 1870–1876 | 1875

View Online
19 S. J. C. Taylor, R. McCague, R. Wisdom, C. Lee, K. Dickson,
G. Ruecroft, F. O’Brien, J. Littlechild, J. Bevan, S. M Roberts and C.
T. Evans, Tetrahedron: Asymmetry, 1993, 4, 1117.
20 M. J. Mulvihill, J. A. Gage and M. J. Miller, J. Org. Chem., 1998, 63,
3357.
29 G.-I. An and H. Rhee, Nucleosides, Nucleotides Nucleic Acids, 2002, 21, 65.
30 M. T. Crimmins, B. W King, W. J. Zuercher and A. L. Choy, J. Org.
Chem., 2000, 65, 8499.
31 S. M. Daluge, M. T. Martin, B. R. Sickles and D. A. Livingston, Nucleo-
sides, Nucleotides Nucleic Acids, 2000, 19, 297.
21 C. Kaneko, A. Sugimoto and S. Tanaka, Synthesis, 1974, 876.
22 D. R. Deardorff, C. Q. Windham and C. L. Craney1, Org. Synth., 1996,
73, 25.
23 K. Laumen and M. P. Schneider, J. Chem. Soc., Chem. Commun., 1986,
1298.
24 E. Forró and F. Fülöp, Tetrahedron: Asymmetry, 2004, 15, 573.
25 For an analogous reaction on 2-azabicyclo[2.2.1]hept-5-en-3-one, see:
N. Katagiri, M. Takebayashi, H. Kokufuda, C. Kaneko, K. Kanehira and
M. Torihara, J. Org. Chem., 1997, 62, 1580.
32 M. T. Crimmins and W. J. Zuercher, Org. Lett., 2000, 2, 1065.
33 H. F. Olivo and J. Yu, J. Chem. Soc. Perkin Trans. 1, 1998, 3, 391.
34 S. M. Daluge, S. S. Good, M. B. Faletto, W. H. Miller, M. H St. Clair,
L. R. Boone, M. Tisdale, N. R. Parry, J. E. Reardon, R. E. Dornsife,
D. R. Averett and T. A Krenitsky, Antimicrob. Agents and Chem., 1997,
41, 1082.
35 M. T. Crimmins and B. W. King, J. Org. Chem., 1996, 61, 4192.
36 L. Kiss, E. Forró, R. Sillanpaa and F. Fülöp, J. Org. Chem., 2007, 72,
8786.
26 A. Vazquez-Romero, J. Rodriguez, A. Lledo, X. Verdaguer and A. Riera,
Org. Lett., 2008, 10, 4509.
27 A. Kim and J. H. Hong, Bull. Korean Chem. Soc., 2007, 28, 1545.
28 M. Freiria, A. J. Whitehead and W. B. Motherwell, Synthesis, 2005, 18,
3079.
37 Utilizing 2,6-dichloro-9H-purine has two advantages over 2,6-diamino-
9H-purine and 2-amino-6-chloro-9H-purine as the nucleophile: it has
good solubility, and the resulting product does not require protection for
subsequent alkylation chemistry. For example see ref. 25.
38 T. Durst and M. J. O’sullivan, J. Org. Chem., 1970, 56, 2043.
1876 | Org. Biomol. Chem., 2012, 10, 1870–1876
This journal is © The Royal Society of Chemistry 2012