Synthesis, biotransformation, and pharmacokinetic studies of 9-(β-D- arabinofuranosyl)-6-azidopurine: A prodrug for ara-A designed to utilize the azide reduction pathway

Article abstract of DOI:10.1021/jm960339p

As a part of our efforts to design prodrugs for antiviral nucleosides, 9-(β-D-arabinofuranosyl)-6-azidopurine (6-AAP) was synthesized as a prodrug for ara-A that utilizes the azide reduction biotransformation pathway. 6-AAP was synthesized from ara-A via its 6-chloro analogue 4. The bioconversion of the prodrug was investigated in vitro and in vivo, and the pharmacokinetic parameters were determined. For in vitro studies, 6-AAP was incubated in mouse serum and liver and brain homogenates. The half-lives of 6-AAP in serum and liver and brain homogenates were 3.73, 4.90, and 7.29 h, respectively. 6- AAP was metabolized primarily in the liver homogenate microsomal fraction by the reduction of the azido moiety to the amine, yielding ara-A. However, 6- AAP was found to be stable to adenosine deaminase in a separate in vitro study. The in vivo metabolism and disposition of ara-A and 6-AAP were conducted in mice. When 6-AAP was administered by either oral or intravenous route, the half-life of ara-A was 7-14 times higher than for ara-A administered intravenously. Ara-A could not be found in the brain after the intravenous administration of ara-A. However, after 6-AAP administration (by either oral or intravenous route), significant levels of ara-A were found in the brain. The results of this study demonstrate that 6-AAP is converted to ara-A, potentially increasing the half-life and the brain delivery of ara-A. Further studies to utilize the azide reduction approach on other clinically useful agents containing an amino group are in progress in our laboratories.

Full text of DOI:10.1021/jm960339p

5202
J . Med. Chem. 1996, 39, 5202-5207
Syn th esis, Biotr a n sfor m a tion , a n d P h a r m a cok in etic Stu d ies of
9-(â-D-Ar a bin ofu r a n osyl)-6-a zid op u r in e: A P r od r u g for Ar a -A Design ed To
Utilize th e Azid e Red u ction P a th w a y1
Lakshmi P. Kotra,† Konstantine K. Manouilov,† Erica Cretton-Scott,‡ J ean-Pierre Sommadossi,‡
F. Douglas Boudinot,† Raymond F. Schinazi,§ and Chung K. Chu*,†
Departments of Medicinal Chemistry and Pharmaceutics, College of Pharmacy, The University of Georgia,
Athens, Georgia 30602-2352, Department of Pharmacology, School of Medicine, The University of Alabama,
Birmingham, Alabama 35294, and Georgia Research Center for AIDS and HIV Infections, Veterans Affairs Medical Center,
and Department of Pediatrics, Emory University School of Medicine, Atlanta, Georgia 30033
Received May 6, 1996X
As a part of our efforts to design prodrugs for antiviral nucleosides, 9-(â-D-arabinofuranosyl)-
6-azidopurine (6-AAP) was synthesized as a prodrug for ara-A that utilizes the azide reduction
biotransformation pathway. 6-AAP was synthesized from ara-A via its 6-chloro analogue 4.
The bioconversion of the prodrug was investigated in vitro and in vivo, and the pharmacokinetic
parameters were determined. For in vitro studies, 6-AAP was incubated in mouse serum and
liver and brain homogenates. The half-lives of 6-AAP in serum and liver and brain homogenates
were 3.73, 4.90, and 7.29 h, respectively. 6-AAP was metabolized primarily in the liver
homogenate microsomal fraction by the reduction of the azido moiety to the amine, yielding
ara-A. However, 6-AAP was found to be stable to adenosine deaminase in a separate in vitro
study. The in vivo metabolism and disposition of ara-A and 6-AAP were conducted in mice.
When 6-AAP was administered by either oral or intravenous route, the half-life of ara-A was
7-14 times higher than for ara-A administered intravenously. Ara-A could not be found in
the brain after the intravenous administration of ara-A. However, after 6-AAP administration
(by either oral or intravenous route), significant levels of ara-A were found in the brain. The
results of this study demonstrate that 6-AAP is converted to ara-A, potentially increasing the
half-life and the brain delivery of ara-A. Further studies to utilize the azide reduction approach
on other clinically useful agents containing an amino group are in progress in our laboratories.
In tr od u ction
patients developed toxicities.25 Fludarabine is a 2-fluoro
analogue of ara-A, which is resistant to deamination and
proved to be effective in a number of luekemias.26
However, in some patients reversible neurotoxicity and
other neurological complications were observed.27 Other
approaches comprise the conjugation of ara-AMP to
lactosaminated human serum albumin28 and adminis-
tration of ara-A in nanocapsules to improve the phar-
macokinetic profiles.29 9-(â-D-Arabinofuranosyl)-6-(di-
methylamino)purine (ara-DMAP), after intravenous
administration in rats and monkeys, was rapidly con-
verted to 9-(â-D-arabinofuranosyl)-6-(methylamino)pu-
rine (ara-MAP) and other purine metabolic end prod-
ucts.30 However, less than 4% of the dose of ara-DMAP
was found to be converted to ara-A, and the half-life of
ara-A was 4 times longer.
Vidarabine, 9-(â-D-arabinofuranosyl)adenine (ara-A),
was originally discovered as an antitumor agent,2 and
in later studies, it was shown to be active against herpes
simplex virus types 1 and 2.3,4 Ara-A is a licensed
compound for the treatment of herpes simplex keratitis5
and encephalitis.6-8 It has also been considered for the
treatment of genital and disseminated herpes infec-
tions,9 cytomegalovirus encephalitis,10 chronic hepatitis
B virus (HBV) infection,11,12 and acute non-lymphoid
leukemia.13 Ara-A may also be an alternative therapy
for acyclovir-resistant herpes simplex virus, cytomega-
lovirus, and varicella-zoster virus infections.14,15 How-
ever, the use of ara-A as a clinically effective agent is
limited due to its rapid deamination to ara-H by
adenosine deaminase (ADA) in vivo16,17 as well as its
poor solubility in water.
As a part of our efforts to design prodrugs of antiviral
agents with improved pharmacokinetic properties, we
have recently synthesized 9-(2,3-dideoxy-2-fluoro-â-D-
arabinofuranosyl)-6-azidopurine (FAAddP) and 9-(2,3-
dideoxy-2-fluoro-â-D-arabinofuranosyl)-N6-methylade-
nine (FMAddA) as prodrugs of 2′-F-ara-ddI.31 We have
demonstrated that the 6-azido analogue FAAddP un-
derwent reduction to 2′-F-ara-ddA by a human microso-
mal P-450 NADPH dependent system as well as in
mice,31 as the first example of the design of prodrugs
taking advantage of the azide reduction pathway. As
an extension of this approach, we have synthesized 9-(â-
D-arabinofuranosyl)-6-azidopurine (6-AAP) as a prodrug
of ara-A and wish to report here in vitro and in vivo
pharmacokinetic studies.
There were several attempts to prevent the rapid
metabolism of ara-A,18 including the coadministration
of adenosine deaminase inhibitors such as deoxy-
coformycin19-23 and N6-benzoyladenosine.24 The effects
of ara-AMP and ara-A in combination with erythro-9-
(2-hydroxy-3-nonyl)adenine (EHNA) were studied in
mouse leukemia L1210/C2 cell culture, and the results
were promising.21 However, in the clinical trials with
the combination of ara-A and deoxycoformycin, some
* Corresponding author. Tel: (706) 542-5379. Fax: (706) 542-5381.
E-mail: dchu@merc.rx.uga.edu.
† The University of Georgia.
‡ The University of Alabama.
§ Emory University School of Medicine.
X Abstract published in Advance ACS Abstracts, December 1, 1996.
S0022-2623(96)00339-1 CCC: $12.00 © 1996 American Chemical Society

6-AAP: A Prodrug for Ara-A
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 26 5203
Ta ble 1. Parameters for the in Vitro Biotransformation of
6-AAP and Ara-A in Mice
Sch em e 1. Synthesis of
9-(â-D-Arabinofuranosyl)-6-azidopurine (6-AAP)a
Kel
T1/2
(h)
compound
6-AAPa
medium
analyte
(h-1
)
liver homogenate
6-AAP
ara-A
ara-H
6-AAP
ara-A
ara-A
ara-H
6-AAP
6-AAP
0.14
0.48
0.26
0.12
0.27
16.87
0.31
0.25
0.01
4.66
1.46
2.45
5.98
2.58
0.04
2.24
3.73
7.29
6-AAP and
coformycin
ara-A
liver homogenate
liver homogenate
6-AAP
6-AAP
serumb
brain homogenate
a
Average of two studies, and the liver homogenate was pre-
b
pared by addition of 1 or 1.5 weight equiv of water. These values
were calculated from 0 to 1 h after the incubation. The decline of
6-AAP concentration in serum was biphasic with a slow decline
in the first 1 h and then with a faster decline (T1/2 ) 1.4 h).
Sch em e 2. Biotransformation of 6-AAP to Ara-A in the
Liver
a
(i) NaNO2, AcOH, 36 h; (ii) adenosine deaminase, water, 16
h; (iii) Ac2O, pyridine, 0 °C, 16 h; (iv) SO2Cl2, CH2Cl2, DMF, reflux,
2 h; (v) NH3, MeOH, 2 h; (vi) lithium azide, DMF, 2 days.
Resu lts a n d Discu ssion
The target compound 5 (6-AAP) was synthesized from
ara-A (Scheme 1). Ara-A was deaminated to 9-(â-D-
arabinofuranosyl)hypoxanthine (1) using adenosine
deaminase in >90% yield. This method was found
superior in comparison to the deamination procedure
with NaNO2/AcOH. Compound 1 was peracetylated
with acetic anhydride in pyridine and then converted
to its 6-chloro derivative 3 under refluxing conditions
with thionyl chloride (26% from 1).32 Compound 3 was
deprotected to compound 4 by treatment with ammonia
in methanol and subsequently treated with LiN3 in
DMF to obtain compound 5 (6-AAP) (38% from 3).
The stability of 6-AAP at pH 2, 7, and 11 and toward
adenosine deaminase hydrolysis was studied by UV
spectroscopy. At pH 2 and 7, 6-AAP did not show any
significant change over a period of 2.75 h at 287.5 nm
(37.1 °C). However, at pH 11, 6-AAP immediately
changed its UV absorption maximum from 287.5 to
222.5 nm. 6-AAP was not hydrolyzed by adenosine
deaminase for up to 3 h in a separate in vitro study
performed in phosphate buffer (pH 7.4) at 25 °C. These
results show that unlike ara-A, 6-AAP is not a substrate
for ADA.
Comparative in vitro studies were performed in mice
liver homogenate by separately incubating 6-AAP and
6-AAP/coformycin to investigate the biotransformation
of 6-AAP alone and in the presence of the ADA inhibitor
coformycin (Table 1). The half-lives of 6-AAP in the
absence and presence of coformycin were 4.90 and 5.98
h, respectively. Ara-A was detected in both cases, and
the corresponding half-lives were 1.46 and 2.58 h,
respectively. When ara-A alone was incubated in the
mouse liver homogenate, its half-life was 0.04 h. Thus,
the apparent half-life of ara-A generated from 6-AAP
alone is 36 times greater than that of ara-A itself. Azido
reduction assay utilizing the microsomal fraction of
human liver homogenate also confirmed that 6-AAP was
converted to ara-A as shown in mice liver homogenate
studies. The biotransformation of 6-AAP to ara-A
involves the cytochrome P-450 NADPH dependent
system (Scheme 2). A similar reduction of the azido
moiety of AZT to an amino function by the cytochrome
P-450 system has been reported.33,34 The stability and
metabolism of 6-AAP were also studied in mice serum
and brain homogenate (Table 1). The half-life of 6-AAP
in mice serum and brain homogenate were 3.73 and 7.29
h, respectively. The decline of 6-AAP in serum was
biphasic with a slow decline in the initial 1 h period
(T1/2 ) 3.73 h) and then with a faster decline rate (T1/2
) 1.41 h). No ara-A was detected in the serum in vitro,
and thus the reason for the biphasic decline is not
known at the present time. More rigorous biochemical
studies are needed to be performed in order to explain
this behavior of 6-AAP in serum.
Following these interesting in vitro results, in vivo
pharmacokinetic studies were performed in mice. Fig-
ure 1 shows the mean serum concentrations of 6-AAP
versus time after intravenous and oral administration
of 100 mg/kg 6-AAP. The corresponding pharmacoki-
netic parameters for 6-AAP are presented in Table 2.
Maximum concentration of 6-AAP in serum was ob-
served after 5 min of intravenous and after 60 min of
oral dosing (Figure 1). Maximum concentrations of
6-AAP in serum after intravenous and oral dosing were
465 ( 167 and 7.8 ( 2.51 µg/mL, respectively. The
terminal mean half-life values (0.55 and 0.58 h for
intravenous and oral, respectively) were similar for both
routes of administration. The area-under-curve (AUC)

5204 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 26
Kotra et al.
curves for ara-A in serum after the intravenous admin-
istration of 6-AAP were different from those after the
intravenous administration of ara-A. This curve re-
vealed a “retard decline” of ara-A in serum with a
significant increase in the half-life (0.89 h). Ara-A level
was 0.28 ( 0.15 µg/mL after 3 h of the injection. The
AUC value (6.84 ( 0.89 mg‚h/mL) is 73% higher than
that after the ara-A administration (3.95 ( 0.20 mg‚h/
mL) (Table 2). When 6-AAP (100 mg/kg) was admin-
istered orally, the serum AUC value of ara-A (1.15 (
0.13 mg‚h/L) was 29% of that after the intravenous
administration of ara-A (3.95 ( 0.20 mg‚h/L), and the
half-life was 0.45 h.
The brain concentrations of ara-A after the adminis-
tration of 6-AAP versus time are illustrated in Figure
3. Ara-A was not found in the brain after its intra-
venous administration in a dose of 100 mg/kg. There
were no toxicities observed in mice at this level of ara-
A. However, ara-A converted from 6-AAP exhibited a
relatively constant mean concentration in the brain of
0.3-0.1 µg/g from 5 to 120 min after intravenous
administration and from 5 to 240 min after oral admin-
istration of 6-AAP. The greater distribution in the brain
was characterized by the increase in the brain AUC, re,
and half-life values for ara-A after oral administration
(1.55 ( 0.57 mg‚h/L, 1.35, and 5.03 h, respectively)
versus intravenous administration (0.35 ( 0.04 mg‚h/
L, 0.05, and 1.47 h, respectively) of 6-AAP (Table 2).
In summary, we have synthesized 9-(â-D-arabinofura-
nosyl)-6-azidopurine (6-AAP) as a prodrug of ara-A. In
the in vitro studies in mice and human liver, 6-AAP was
converted to ara-A by the cytochrome P-450 system. It
was found that 6-AAP was not a substrate for adenosine
deaminase. The in vivo pharmacokinetics in mice
showed a significant improvement in the brain distribu-
tion and an increase in the half-lives of ara-A in the
serum and the brain, after 6-AAP administration. It
was also observed that the distribution of both 6-AAP
and ara-A is independent of the route of administration
of 6-AAP. Additional studies on 6-AAP as well as on
utilizing the azido moiety in designing prodrugs for
anticancer, antiviral, and other clinically useful agents
containing an amino functionality with undesirable
pharmacokinetic and toxicity profiles are in progress in
our laboratories.
F igu r e 1. Mean ( SD concentrations of 6-AAP in serum
(µg/mL) after intravenous (3) and oral (1) administration and
in brain (µg/g) after intravenous (O) and oral (b) administra-
tion of 100 mg/kg 6-AAP to mice.
values for serum concentration versus time for 6-AAP
were 201.1 ( 17.9 and 13.77 ( 1.4 mg‚h/L, respectively,
following intravenous and oral administration of 100
mg/kg 6-AAP. After intravenous administration of 20
mg/kg 6-AAP, the AUC value was 85.6 mg‚h/L (data not
shown), a 5-fold difference in the AUC values after
intravenous dosing of 20 and 100 mg/kg 6-AAP, which
indicates that the disposition of 6-AAP in mice followed
linear kinetics in the dose interval between 20 and 100
mg/kg. There were no toxicities observed in the mice
at either dose level, and the doses were well tolerated.
The concentrations of 6-AAP in brain versus time
after intravenous and oral administrations are shown
in Figure 1. After intravenous and oral administrations
of 6-AAP, the maximum concentrations of 5.92 ( 0.7
and 1.87 ( 0.60 µg/g in the brain were observed at 5
and 30 min, respectively. The brain AUC values for
6-AAP were almost the same for both (intravenous and
oral) routes of administration (4.41 ( 0.37 and 4.12 (
0.37 mg‚h/L, respectively) (Table 2). The serum AUC
levels of 6-AAP were 201 ( 17.9 and 13.77 ( 1.40
mg‚h/L after intravenous and oral administration of
6-AAP. In comparison, the brain AUC levels of 6-AAP
are 2% and 33% of serum AUC levels after intravenous
and oral administration of 6-AAP, respectively. These
data suggest that there may be a saturable transport
process of 6-AAP into the brain. Half-life of 6-AAP in
the brain was approximately 2-fold greater after oral
administration (1.29 h) than after intravenous admin-
istration (0.77 h). The relative brain exposure (re) value
of 6-AAP was also greater after oral dosing (0.3) than
after intravenous administration (0.02) of 6-AAP.
The serum concentrations of ara-A versus time after
the administration of ara-A (intravenous) and 6-AAP
(oral and intravenous) are shown in Figure 2. After the
intravenous administration of ara-A, a significant frac-
tion of the compound was rapidly metabolized, and its
level declined from 18.5 ( 3.5 to 0.33 ( 0.25 µg/mL in
25 min. The AUC value was 3.95 ( 0.2 mg‚h/L, and
the half-life was 0.07 h. However, the pharmacokinetic
Exp er im en ta l Section
Syn th esis. Melting points were determined on a Mel-Temp
II laboratory apparatus and are uncorrected. The 1H NMR
spectra were recorded on a J EOL FX 90 Q FT spectrometer
with TMS as the internal standard; chemical shifts are
reported in parts per million (δ), and the signals are quoted
as s (singlet), d (doublet), t (triplet), or m (multiplet). UV
spectra were recorded on a Beckman DU-650 spectrophotom-
eter. TLC was performed on Uniplates (silica gel) purchased
from Analtech Co. Elemental analysis was performed by
Atlantic Microlab, Inc., Norcross, GA. Ara-A and adenosine
deaminase (type II crude powder from calf intestinal mucosa,
1-5 units/mg of activity) were purchased from Sigma Chemical
Co., St. Louis, MO. All other chemicals were of reagent grade.
9-(â-D-Ar a bin ofu r a n osyl)h yp oxa n th in e (1). Meth od A:
A solution of ara-A (500 mg, 1.87 mmol) in glacial acetic acid
(8 mL) was treated with a solution of NaNO2 (258 mg, 3.73
mmol) in 1 mL of water and stirred for 6 h. Three equal,
additional portions of NaNO2 (600 mg total, 8.4 mmol) were
added over a 6 h period, and the stirring was continued. After
36 h, the solvent was evaporated in vacuo, and the residue
was recrystallized from hot water (25 mL) to obtain pure 1
(339 mg, 67.6%): UV(MeOH) λmax 249.0, 207.0 nm.

6-AAP: A Prodrug for Ara-A
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 26 5205
Ta ble 2. Pharmacokinetic Parameters of Ara-A, 6-AAP, and Ara-A Released from 6-AAP after Dosing of 100 mg/kg Ara-A or 6-AAP
compound
administered
route of
administration
compound
measured
AUC
(mg‚h/L)
tissue
T1/2 (h)
re (brain)
ara-A
iv
iv
ara-A
6-AAP
ara-A
6-AAP
ara-A
serum
brain
serum
brain
serum
brain
serum
brain
serum
brain
3.95 ( 0.20
NDa
0.07
6-AAP
201.1 ( 17.9
4.41 ( 0.37
6.84 ( 0.89
0.35 ( 0.04
13.77 ( 1.40
4.12 ( 0.37
1.15 ( 0.13
1.55 ( 0.57
0.55
0.77
0.89
1.47
0.58
1.29
0.45
5.03
0.02
0.05
0.30
1.35
6-AAP
oral
a
ND, not detected.
6-Ch lor o-9-(2,3,5-tr i-O-a cetyl-â-D-a r a bin ofu r a n osyl)p u -
r in e (3). Crude compound 2 (130 mg) was dissolved in dry
CH2Cl2 (10 mL) and heated to 55 °C. Dry DMF (1 mL) followed
by a 2 M solution of SOCl2 in CH2Cl2 (2.43 mL, 0.57 mmol)
were added dropwise over a period of 45 min. The reaction
mixture was gently refluxed for an additional 75 min. The
reaction mixture was cooled to room temperature and diluted
with CH2Cl2. The organic layer was washed with saturated
NaHCO3 solution (2 × 50 mL) and brine (50 mL) and dried
(anhydrous sodium sulfate). The organic phase was concen-
trated in vacuo and purified by preparative TLC (5% MeOH/
CHCl3) to obtain pure 3 (50 mg, 26% from 1): UV (MeOH)
λmax 263.0, 212.5 nm.
9-(â-D-Ar a b in ofu r a n osyl)-6-ch lor op u r in e (4). Com-
pound 3 (200 mg, 0.5 mmol) was dissolved in saturated. NH3/
MeOH (5 mL) and stirred at room temperature for 2 h. The
solvent was evaporated in vacuo to obtain crude 4 (170 mg)
which was used for the subsequent reaction without any
further purification: UV (MeOH) λmax 263.0 nm.
F igu r e 2. Mean ( SD serum concentrations of ara-A after
intravenous administration of ara-A (1) and after oral (b) and
intravenous (3) administration of 100 mg/kg 6-AAP to mice.
9-(â-D-Ar a bin ofu r a n osyl)-6-a zid op u r in e (5). A solution
of 4 (170 mg, 0.63 mmol) in DMF (5 mL) was treated with
lithium azide (270 mg, 5.52 mmol) and stirred for 2 days at
room temperature. The solvent was evaporated under reduced
pressure at 40 °C, and the crude oil was recrystallized from
MeOH to obtain pure 5 (67 mg, 38.4%): mp 185-190 °C dec;
UV λmax (water) pH 2, 205.0 (15 506), 287.0 nm (6496); pH 7,
1
208.5 (12 943), 287.5 nm (6033); pH 11, 222.5 nm (6730); H
NMR (DMSO-d6) δ 3.66-3.90 (m, 3H, H-5′, H-4′), 4.16-4.33
(m, 2H, H-2′, H-3′), 5.14 (t, 1H, 5′-OH, exchangeable with D2O),
6.50 (d, 1H, H-1′), 8.75 (s, 1H, H-8), 10.12 (s, 1H, H-2); IR (KBr)
2037, 1649. Anal. (C10H11N7O4‚0.65CH3OH) C, H, N.
Sta bility Stu d ies of 6-AAP . A kinetic study at varying
pHs (pH 2, 7, and 11 at 37.1 °C) was performed on a UV
spectrophotometer to investigate the stability of 6-AAP. At
pH 11, the UV absorption maximum for compound 5 shifted
from 287.5 to 222.5 nm immediately. At pH 7, the compound
did not show any significant change in UV absorption maxima
over a period of 2.75 h at 287.5 nm indicating that it is stable
at the neutral pH. At pH 2, the compound was stable.
Ad en osin e Dea m in a se Stu d ies. 6-AAP (0.22 µΜ/mL)
was incubated with adenosine deaminase (0.05 mg/mL) in
phosphate buffer (pH 7.4) at 25.1 °C, and the change in the
concentration was observed at 278.5 nm for 3 h.
An a lysis. Concentrations of ara-A and 6-AAP in serum,
brain (whole or homogenate), and liver homogenate were
measured by high-performance liquid chromatography (HPLC).
Chromatographic separations were carried out on a Millipore
gradient system, which was equipped with a Model 486
tunable UV detector, two Model 510 pumps, a Model 717 plus
autosampler, and Millennium 2010 Chromatography Manager
software (Millipore Corp., Milford, MA). All the solvents were
of HPLC grade. Chromatography was performed on an Altech
Hypersil ODS C18 column (5 µm particle size, 4.6 × 150 mm;
Altech Associates, Deerfield, IL). The mobile phase A was
0.5% acetonitrile in 2.5 mM KH2PO4 (pH 6.8), mobile phase B
was 12% acetonitrile in 2.5 mM KH2PO4 (pH 6.8), mobile phase
C was 2.5 mM KH2PO4 (pH 5.2), and mobile phase D was 23%
methanol in 2.5 mM KH2PO4.
F igu r e 3. Mean ( SD brain concentrations of ara-A after oral
(b) and intravenous (3) administration of 100 mg/kg 6-AAP
to mice.
Meth od B: A suspension of ara-A (500 g, 1.87 mmol) in
distilled water (30 mL) was treated with adenosine deaminase
(4 mg) and stirred for 16 h. The water was evaporated under
reduced pressure, and the white residue was recrystallized
from hot water (20 mL) to obtain compound 1 as a soft white
solid (462 mg, 92%): UV (MeOH) λmax 248.5, 205.5 nm.
9-(2,3,5-Tr i-O-a cet yl-â-D-a r a b in ofu r a n osyl)h yp oxa n -
th in e (2). Acetic anhydride (1 mL, 10.5 mmol) was added to
a suspension of 1 (335 mg, 1.25 mmol) in dry pyridine (5 mL)
at 0 °C and stirred for 16 h. The solvent was evaporated in
vacuo, the residue was dissolved in 50 mL of methylene
chloride, and the organic layer was washed with water (2 ×
50 mL), saturated NaHCO3 solution, and brine and dried
(anhydrous sodium sulfate). The organic layer was concen-
trated in vacuo to obtain 2 as a brownish yellow solid (339
mg, crude yield 68%) which was used in the subsequent
reaction without any further purification: UV (MeOH) λmax
250.0, 206.0 nm.
In vitr o Meta bolism Stu d y. Female NIH Swiss mice
(Harland Sprague-Dawley, Indianapolis, IN) were sacrificed,

5206 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 26
Kotra et al.
and serum and liver and brain tissues were collected before
each experiment. Serum was collected from several animals.
The brain and liver were washed in normal saline at 4 °C,
wiped, and then weighed. Either 1 or 1.5 weight equiv of water
was added to the tissue and homogenized using a homogenizer.
The homogenate was divided into two halves: One portion was
used as the blank and the other for incubation with either ara-
A, 6-AAP or 6-AAP/coformycin in a water bath shaker at 37
°C. Initial concentrations of 6-AAP and ara-A were 100 µg/
mL. Samples and blanks of volume 400 µL were collected at
0, 5, 15, and 30 min and at 1, 2, 3, 4, and 5 h.
To measure the analyte concentrations in serum or liver or
brain homogenate, 400 µL of sample was mixed with 50 µL of
internal standard (AzdU, 5 µg/mL) and 0.7 mL of acetonitrile.
After centrifugation, the supernatant was decanted to another
tube, treated with anhydrous Na2SO4, and then vortexed for
1 min and centrifuged again. The organic layer was separated
and evaporated under nitrogen stream at room temperature.
The residue was reconstituted in mobile phase A and filtered
through a MPS-I micropartition system (3kDa membranes)
(Amicon Inc., Beverly, MA) by centrifugation for 50 min at
2000 rpm to further clean up the samples; 100 µL of filtrate
was injected for analysis.
fuged at 9000 rpm for 10 min. The resulting supernatant from
the serum or the brain was transferred to a clean tube and
dried under a stream of nitrogen gas at 22 °C. The residue
was reconstituted in 220 µL of mobile phase D, and after
centrifugation at 12 000 rpm for 40 min, 100-150 µL was
injected for the HPLC analysis. During the first 28 min, a
linear gradient from 5% C and 95% D to 20% C and 80% D
was run, and then during the next 20 min, a linear gradient
was run to reach 65% C and 35% D at a flow rate of 1.5 mL/
min. After each assay, the column was equilibrated to initial
conditions for 7 min. The λmax was set at 249 nm for the first
15 min to observe ara-H, from 15 to 30 min at 261 nm to
observe ara-A, from 30 to 40 min at 285 nm to observe
6-azidoarabinosylpurine, and then changed to 261 nm to
observe AzdU. The retention times for ara-H, ara-A, 6-AAP,
and AzdU were 13.5, 27.8, 36.7, and 43.5 min, respectively.
Sta n d a r d Cu r ves. Standard curves were prepared for
each type of sample by adding known amounts of ara-A and
6-AAP to the serum, brain, or liver and subjecting them to
the extraction procedure as described above. The limits of
quantitation of the ara-A and 6-AAP were 0.1 and 0.3 µg/mL,
respectively. The percent recoveries of the compounds were
63% for ara-A and 55% for 6-AAP.
Da ta An a lysis. Serum and tissue concentrations versus
time data for 6-AAP and ara-A were analyzed by noncompart-
mental methods. The AUC versus time profiles from time zero
to the last measured concentration were determined by the
linear trapezoidal rule, and the AUC from the time of the last
measured concentration to infinity was determined by dividing
the last determined concentration by the least-squares elimi-
nation rate constant (λz). Half-life was calculated from 0.693/
λz. The relative tissue exposure (re) of the compounds was
calculated from AUCtissue/AUCserum. A variation of AUC was
calculated according to previous procedures.35
For the HPLC analysis, during the first 10 min, the flow
rate was changed linearly from 1.5 to 1.0 mL/min and
continued at 1.0 mL/min until the end of assay (67 min). In
the first 20 min, the mobile phase consisted of 95% A and 5%
B; from 20 min, a linear gradient was run for 55 min to reach
5% A and 95% B. After each analysis, the column was
equilibrated for 10 min to initial conditions. The λmax was set
at 249 nm for the first 15 min to observe ara-H, from 15 to 35
min at 261 nm to observe ara-A, from 35 to 45 min at 285 nm
to observe 6-AAP, and subsequently changed to 261 nm to
observe AzdU. The retention times for ara-H, ara-A, 6-AAP,
and AzdU were 14.1, 33.8, 40.9, and 49.3 min, respectively.
Azid o Red u ction Stu d y. The procedure for the analysis
of azido-reducing activity was described previously.33 Briefly,
incubation mixtures contained either 1.5 mg of human liver
fraction protein (homogenate or supernatant fractions follow-
ing centrifugation) or 1.5 mg of microsomal protein, 5.0 mM
MgCl2, 6.0 mM NADPH, and 0.4 mg/mL 6-AAP in 0.1 M
phosphate buffer saline at pH 7.4 (final volume of 0.2 mL).
The reaction was initiated by adding NADPH and conducted
at 37 °C for 60 min under nitrogen. Reactions were terminated
by heating at 100 °C for 30 s, and the proteins were removed
by centrifugation at 14000g for 6 min. Aliquots (100 µL) were
then analyzed for nucleosides by HPLC. Control incubations
were performed in the absence of protein.
Ack n ow led gm en t. This work was supported by
NIH Grant AI 25899, AI 32351, and HL 42125 and the
Department of Veterans Affairs.
Refer en ces
(1) Preliminary account on this paper was presented at the 211th
American Chemical Society National Meeting, New Orleans, LA,
March 24-28, 1996, Medicinal Chemistry General Session,
Paper 16.
(2) Reist, E. J .; Benitez, A.; Goodman, L.; Baker, B. L.; Lee, W. W.
Potential anticancer agents. LXXVI. Synthesis of purine nucleo-
sides of â-D-arabinofuranose. J . Org. Chem. 1962, 27, 3274-
3279.
(3) Drach, J . C.; Shipman, C., J r. The selective inhibition of viral
DNA synthesis by chemotherapeutic agents: an indicator of
clinical usefulness? Ann. N.Y. Acad. Sci. 1977, 284, 396-409.
(4) Andrei, G.; Snoeck, R.; Goubou, P.; Desmyter, J .; DeClercq, E.
Comparative activity of various compounds against clinical
strains of herpes simples virus. Eur. J . Clin. Microbiol. Infect.
Diseases 1992, 11, 143-151.
In h ibition of Azid o Red u ction Assa y. Assays were
performed using 1.5 mg of microsomal protein as described
above following either a 45 s exposure to carbon monoxide or
a 5 min preincubation with 1 mM metyrapone prior to the
addition of NADPH.
In Vivo P h a r m a cok in etics. Female NIH Swiss mice
(Harland Sprague-Dawley, Indianapolis, IN) weighing 24-
28 g were used for the pharmacokinetic experiments. Mice
were acclimatized in a 12 h light/12 h dark, constant temper-
ature (20 °C) environment for 1 week before the experiments.
In a randomized study, animals were administered either
20 or 100 mg/kg 6-AAP or 100 mg/kg ara-A (intravenous).
6-AAP was also dosed orally (100 mg/kg po). At least three
animals each were sacrificed at 0.08, 0.025, 0.5, 0.75, 1.0, 2.0,
3.0, 4.0, 5.0, and 6.0 h after drug administration. Blood
(serum) from the heart and whole-brain samples were col-
lected. Serum samples were treated immediately, and brain
samples were frozen at -20 °C until analysis. To determine
the nucleoside concentrations in the serum, a known amount
of serum sample, 50 µL of the internal standard (AzdU, 5 µg/
mL), and 1.0 mL of acetonitrile as a protein precipitant were
added to polypropylene microcentrifuge tubes (1.7 mL). Tubes
were mixed and centrifuged at 9000 rpm for 10 min.
(5) Denis, J .; Chouraqui, P.; Karpouzas, I.; Xuang, T. H.; Pouliquen,
Y. Treatment of superficial herpes simplex keratitis with Vi-
darabine (Vira A): A multicenter study of 100 cases. J . Fr.
Ophthalmol. 1990, 13, 143-150.
(6) Whitley, R. J . Herpes simplex virus infections of the central
nervous system. Encephalitis and neonatal herpes. Drugs 1991,
42, 406-427.
(7) Stula, D.; Lyrer, P. Severe herpes simplex encephalitis: Course
15 years following decompressive craniotomy. Schweiz. Med.
Wochenschr. 1992, 122, 1137-1140.
(8) Whitley, R. J . Neonatal herpes simplex virus infections. J . Med.
Virol. 1993, Suppl. 1, 13-21.
(9) DRUGDEX (R) Information System; Gelman, C. R., Rumack, B.
H., Eds.; MicroMedex, Inc.: Englewood, CO, expired May 31,
1995.
(10) Suzuki, Y.; Kamei, S.; Tamura, M.; Sawai, S.; Takasu, T.
Cytomegalovirus encephalitis in immunologically normal adults.
Rinsho Shinkeigaku 1990, 30, 168-173.
(11) Chien, R. N.; Liaw, Y. F. Drug therapy in patients with chronic
type B hepatitis. J . Formosan Med. Assoc. 1995, 94 (Suppl. 1),
s1-s9.
(12) Fu, X. X. Therapeutic effect of combined treatment with ara-A,
dauricine and Chinese herbs in chronic hepatitis B infection.
Zhonghua Neike Zazhi (Beijing) 1991, 30, 498-501.
(13) Resegotti, L. Treatment of acute non-lymphoid leukemia (ANLL)
in elderly patients. The GIMEMA experience. Leukemia 1992,
6 (Suppl. 2), 72-75.
To measure the ara-A and 6-AAP in the whole brain, 50 µL
of internal standard (AzdU, 10 µg/mL) and 300 µL of water
were added to the weighed tissue samples (approximately 300
mg). After homogenization, 1.8 mL of acetonitrile was added
to tissue homogenates, and samples were mixed and centri-

6-AAP: A Prodrug for Ara-A
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 26 5207
(14) Chatis, P. A.; Crumpacker, C. S. Resistance of herpes viruses
to antiviral drugs. Antimicrob. Agents Chemother. 1992, 36,
1589-1595.
(15) Nitta, K.; Shiota, H.; Naito, J .; Mimura, Y. Sensitivities to other
antiviral drugs and thymidine kinase activity of acyclovir-
resistant herpes simplex virus type 1. Nippon Ganka Gakkai
Zasshi 1994, 98, 513-519.
phase II trial from children’s cancer study group. Am. J . Clin.
Oncol. 1992, 15, 490-493.
(26) Hood, M. A.; Finley, R. S. Fludarabine: A review. DICP Ann.
Pharmacother. 1991, 25, 518-524.
(27) Cheson, B. D.; Vena, D. A.; Foss, F. M.; Sorensen, J . M.
Neurotoxicity of purine analogs: A review. J . Clin. Oncol. 1994,
12, 2216-2228.
(28) J ansen, R. W.; Kruijt, J . K.; Van Berkel, T. J .; Meijer, D. K.
Coupling of antiviral drug ara-AMP to lactosaminated albumin
leads to specific uptake in rat and human hepatocytes. Hepa-
tology 1993, 18, 146-152.
(29) Yang,T. Y.; Hu, J .; Li, H. X. Studies on pharmacokinetics of 9-â-
D-arabinosyladenine nanocapsules. Yaoxue Xuebao 1992, 27,
923-927.
(30) Lambe, C. U.; Resetar, A.; Specter, T.; Koszalka, G. W.; Nelson,
D. J . Metabolism and pharmacokinetics of the anti-varicella-
zoster virus agent 6-dimethylaminopurine arabinoside. Antimi-
crob. Agents Chemother. 1992, 36, 353-360.
(31) Koudriakova, T.; Manouilov, K. K.; Shanmuganathan, K.; Kotra,
L. P.; Boudinot, F. D.; Sommadossi, J .-P.; Chu, C. K. In vitro
and in vivo evaluation of 6-azido-2′,3′-dideoxy-2′-fluoro-â-D-
arabinofuranosylpurine (FAAddP) and 6-methyl-2′,3′-dideoxy-
2′-fluoro-â-D-arabinofuranosyladenine (FMAddA) as prodrugs of
the anti-HIV nucleosides, 2′-F-ara-ddA and 2′-F-ara-ddI. J . Med.
Chem. 1996, 39, 4676-4681.
(16) Cass, C. E. 9-â-D-Arabinofuranosyladenine (Ara-A). In Antibiot-
ics. Mechanism of Action of Anti-eukaryotic and Antiviral
Compounds; Hahn, F. E., Ed.; Springer-Verlag: New York, 1979;
Vol. V, pp 87-109.
(17) Whitley, R.; Alford, C.; Hess, F.; Buchanan, R. Vidarabine:
a
preliminary review of its pharmacological properties and thera-
peutic use. Drugs 1980, 20, 267-282.
(18) Plunkett, W.; Cohen, S. S. Two approaches that increase the
activity of analogs of adenine nucleosides in animal cells. Cancer
Res. 1975, 35, 1547-1554.
(19) Cass, C. E.; Ah-Yeung, T. H. Enhancement of 9-â-D-arabino-
furanosyladenine cytotoxicity to mice leukemia L1210 in vitro
by 2′-deoxycoformycin. Cancer Res. 1976, 36, 1486-1491.
(20) LePage, G. A.; Worth, L. S.; Kimball, A. P. Enhancement of
antitumor activity of arabinofuranosyl adenine by 2′-deoxyco-
formycin. Cancer Res. 1976, 36 (4), 1481-1485.
(21) Cass, C. E.; Tan, T. H.; Selner, M. Antiproliferative effects of
9-â-D-arabinofuranosyladenine-5′-monophosphate and related
compounds in combination with adenosine deaminase inhibitors
against a mouse leukemia L1210/C2 cells in culture. Cancer Res.
1979, 39 (5), 1563-1569.
(32) Robins, M. J .; Bason, G. L. 6-Chloro-9-(2-deoxy-â-D-erythro-
pentofuranosyl)purine from the chlorination of 2′-deoxyinosine.
In Nucleic Acid Chemistry; Townsend, L. B., Tipson, R. S., Eds.;
J ohn-Wiley & Sons: New York, 1978; Part II, pp 601-606.
(33) Cretton, E. M.; Sommadossi, J .-P. Reduction of 3′-azido-2′,3′-
dideoxynucleosides to their 3′-amino metabolite is mediated by
cytochrome P-450 and NADPH-cytochrome P-450 reductase in
rat liver microsomes. Drug Metab. Dispos. 1993, 21, 946-950.
(34) Placidi, L.; Cretton, E. M.; Placidi, M.; Sommadossi, J .-P.
Reduction of 3′-azido-3′-deoxythymidine to 3′-amino-3′-deoxythy-
midine in human liver microsomes and its relationship to
cytochrome P-450. Clin. Pharmacol. Ther. 1993, 54, 168-176.
(35) Yuan, J . Estimation of variance for AUC in animal studies. J .
Pharm. Sci. 1993, 82, 761-763.
(22) Plunkett, W.; Benzamin, R. S.; Keating, M. J .; Freireich, E. J .
Modulation of 9-â-D-arabinofuranosyladenine-5′-triphosphate
and deoxyadenosine-triphosphate in leukemic cells by 2′-deoxy-
coformycin during therapy with 9-â-D-arabinofuranosyladenine.
Cancer Res. 1982, 42 (5), 2092-2096.
(23) Agarwal, R. P.; Blatt, J .; Miser, J .; Sallan, S.; Lipton, J . M.;
Reaman, G. H.; Holcenberg, J .; Poplack, D. G. Clinical pharma-
cology of 9-â-D-arabinofuranosyladenine in combination with 2′-
deoxycoformycin. Cancer Res. 1982, 42 (9), 3884-3886.
(24) Tritach, G. L.; Nechaev, A.; Mittelman, A. Synergism between
the antiproliferative activities of arabinofuranosyladenine and
N6-benzoyladenosine. Cancer Biochem. Biophys. 1977, 2 (2), 87-
90.
(25) Miser, J . S.; Roloff, J .; Blatt, J .; Reaman, G. H.; Krailo, M. D.;
Hammond, G. D. Lack of significant activity of 2′-deoxycofor-
mycin alone or in combination with adenine arabinoside in
relapsed childhood acute lymphoblastic leukemia. A randomized
J M960339P