In vitro and in vivo evaluation of 6-azido-2',3'-dideoxy-2'-fluoro-β- D-arabinofuranosylpurine and N6-methyl-2',3'-dideoxy-2'-fluoro-β-D- arabinofuranosyladenine as prodrugs of the anti-HIV nucleosides 2'-F-ara-ddA and 2'-F-ara-ddI

Article abstract of DOI:10.1021/jm960140c

In an effort to improve the pharmacokinetic properties and tissue distribution of 2'-F-ara-ddI, two lipophilic prodrugs, 6-azido-2',3'-dideoxy- 2'-fluoro-β-D-arabinofuranosylpurine (FAAddP, 4) and N⁶-methyl-2',3'- dideoxy-2'-fluoro-β-D-arabinof

Full text of DOI:10.1021/jm960140c

4
676
J . Med. Chem. 1996, 39, 4676-4681
Notes
In Vitr o a n d in Vivo Eva lu a tion of
6
-Azid o-2′,3′-d id eoxy-2′-flu or o-â-D-a r a bin ofu r a n osylp u r in e a n d
6
N -Meth yl-2′,3′-d id eoxy-2′-flu or o-â-D-a r a bin ofu r a n osyla d en in e a s P r od r u gs of th e
An ti-HIV Nu cleosid es 2′-F -a r a -d d A a n d 2′-F -a r a -d d I
†
†
†
†
Tanya Koudriakova, Konstantine K. Manouilov, Kirupa Shanmuganathan, Lakshmi P. Kotra,
†
‡
‡
§
F. Douglas Boudinot, Erica Cretton-Scott, J ean-Pierre Sommadossi, 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, University of Alabama,
Birmingham, Alabama 35294, and Veterans Affairs Medical Center and Department of Pediatrics,
Emory University School of Medicine, Atlanta, Georgia 30033
X
Received February 14, 1996
In an effort to improve the pharmacokinetic properties and tissue distribution of 2′-F-ara-ddI,
two lipophilic prodrugs, 6-azido-2′,3′-dideoxy-2′-fluoro-â-D-arabinofuranosylpurine (FAAddP,
6
4
) and N -methyl-2′,3′-dideoxy-2′-fluoro-â-D-arabinofuranosyladenine (FMAddA, 5), were syn-
thesized and their biotransformation was investigated in vitro and in vivo, in mice. Compounds
and 5 were synthesized via the intermediate 2. For the in vitro studies, FAAddP and FMAddA
4
were incubated in mouse serum, liver homogenate, and brain homogenate. FAAddP was
metabolized in liver homogenate by the reduction of the azido to the amino moiety followed by
deamination, yielding 2′-F-ara-ddI. The conversion of FAAddP to 2′-F-ara-ddA was mediated
by microsomal P-450 NADPH reductase system, as shown by the liver microsomal assay.
FAAddP was also converted to 2′-F-ara-ddI at a slower rate in the brain than in the liver.
FMAddA, however, was stable in brain homogenate and was slowly metabolized in the liver
homogenate. Metabolic conversion of FMAddA in vitro was stimulated by the addition of
adenosine deaminase. In the in vivo metabolism study, FAAddP underwent reduction to 2′-
F-ara-ddA followed by deamination to 2′-F-ara-ddI. FMAddA did not result in increased brain
delivery of 2′-F-ara-ddI in vivo, probably due to the slow conversion as observed in the in vitro
studies. However, there was an increase in the half-life of 2′-F-ara-ddI produced from FMAddA.
This report is the first example in the design of prodrugs using the azido group for adenine-
and hypoxanthine-containing nucleosides. This interesting and novel approach can be extended
to other antiviral and anticancer nucleosides.
In tr od u ction
ddI) and 2′,3′-dideoxy-2′-fluoro-â-D-arabinofuranosyl-
7
adenine (2′-F-ara-ddA) were synthesized. Marquez et
Didanosine (ddI) is a synthetic nucleoside structurally
related to inosine with proven activity against human
immunodeficiency virus (HIV). It is approved for use
7
al. reported that 2′-F-ara-ddA and 2′-F-ara-ddI were
stable in acidic media and were as potent as ddI in
1
+
protecting CD4 ATH8 cells from the cytopathogenic
in patients who are intolerant to zidovudine (AZT) or
who have developed drug resistance to zidovudine
effects of HIV-1. However, 2′-F-ara-ddI, as well as ddI,
is relatively hydrophilic and does not readily penetrate
2
2
therapy. However, its various side effects, chemical
8
the blood-brain barrier (BBB) in mice. Recently, we
3
instability in gastric acid, and low oral bioavailability
synthesized a more lipophilic prodrug, 2′,3′-dideoxy-2′-
fluoro-â-D-arabinofuranosylpurine (2′-F-ara-ddP), which
was converted to 2′-F-ara-ddI by xanthine oxidase in
limit the usefulness of ddI. Furthermore, there is
evidence that ddI enters the central nervous system and
the cerebrospinal fluid (CSF) less readily than AZT. The
extent of ddI uptake in brain tissue and CSF relative
to that in plasma was only 4.7 and 1.5%, respectively.
In an effort to overcome the instability of ddI and 2′,3′-
dideoxyadenosine (ddA) in acidic conditions, 2′,3′-dideoxy-
8
vivo. Pharmacokinetic studies indicated that 2′-F-ara-
4
-6
ddP increased the delivery of 2′-F-ara-ddI to the brain
in mice with an increased AUCbrain/AUCserum ratio of
approximately 36% after oral and intravenous admin-
8
istration of 2′-F-ara-ddP.
2
′-fluoro-â-D-arabinofuranosylhypoxanthine (2′-F-ara-
As a continuation of the development of prodrug
strategies in enhancing the brain delivery of anti-HIV
nucleosides by utilizing the in vivo biotransformation
systems, we have synthesized 6-azido-2′,3′-dideoxy-2′-
*
To whom correspondence should be sent at: Department of
Medicinal Chemistry, College of Pharmacy, The University of Georgia,
Athens, GA 30602-2352. Tel: (706) 542-5379. Fax: (706) 542-5381.
E-mail: dchu@rx.uga.edu.
†
6
The University of Georgia.
The University of Alabama.
Emory University School of Medicine.
Abstract published in Advance ACS Abstracts, October 1, 1996.
fluoro-â-D-arabinofuranosylpurine (FAAddP, 4) and N -
‡
§
methyl-2′,3′-dideoxy-2′-fluoro-â-D-arabinofuranosyl-
X
16
adenine (FMAddA, 5) as prodrugs of 2′-F-ara-ddI. The
S0022-2623(96)00140-9 CCC: $12.00 © 1996 American Chemical Society

Notes
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 23 4677
Sch em e 2. Biotransformation of the Prodrug FAAddP to
Sch em e 1. Synthesis of FAAddP and FMAddP
2
′-F-ara-ddI
Over the initial 45 min, the prodrug was rapidly
metabolized with a t1/2 of 0.48 h. Subsequently, the rate
of conversion was much slower (t1/2 ) 7.74 h) than the
initial rate, probably due to the depletion of cofactors.
The formation of 2′-F-ara-ddI paralleled the decline of
the prodrug with most of the prodrug being converted
to 2′-F-ara-ddI. Only low concentrations of 2′-F-ara-ddA
were detected in the liver homogenate. The biotrans-
formation of FAAddP in brain homogenate was some-
what slower than that in the liver with a half-life of 6.1
h. However, only 10% of the prodrug was converted to
synthesis, in vitro biotransformation, and in vivo dis-
position of the prodrugs are discussed in this report.
Resu lts a n d Discu ssion
2
′-F-ara-ddI and 5% to 2′-F-ara-ddA over a 6 h time
The prodrugs FAAddP (4) and FMAddA (5) were
synthesized from the 6-chloropurine derivative 2 (Scheme
period. Thus, similar to the studies in serum, an
unidentified pathway was responsible for the disap-
pearance of FAAddP in mouse brain.
1
). Compound 2 was synthesized from 5-O-benzoyl-3-
deoxy-1,2-O-isopropylidine-R-D-ribofuranose (1) accord-
ing to published procedures.8 Compound 2 was de-
benzoylated using DIBAL-H in CH2Cl2 at -78 °C to
obtain compound 3. Upon treatment of compound 3
with LiN3 in DMF at room temperature, the 6-azido
derivative 4 was obtained in 73% yield. The treatment
of compound 2 with methylamine in DMF at 80 °C for
The metabolic conversion of FMAddA to 2′-F-ara-ddI
appeared to be a one-step process, facilitated by adeno-
sine deaminase. The prodrug was stable in PBS, mouse
serum, and mouse brain homogenate and was slowly
metabolized to 2′-F-ara-ddI in the liver homogenate (t1/2
) 9.1 h). However, upon addition of adenosine de-
aminase to PBS (t1/2 ) 0.46 h) and brain homogenate
(t1/2 ) 3.7 h), virtually all of the prodrug was converted
to 2′-F-ara-ddI.
5
h followed by the deprotection with saturated NH3/
1
6
MeOH for 15 h gave compound 5 quantitatively.
The biotransformation of the prodrug FAAddP to 2′-
F-ara-ddI probably involves a two-step metabolic process
The pharmacokinetics of the active nucleoside, 2′-F-
ara-ddI, were investigated in mice. Concentrations of
2′-F-ara-ddI in serum and brain after intravenous
administration of 20 mg/kg of the compound are il-
lustrated in Figure 1. Serum concentrations of 2′-F-
ara-ddI declined rapidly with a half-life of 0.41 h (Table
1). Brain concentrations of the nucleoside peaked at
approximately 20 min, remained relatively constant for
30 min, and subsequently declined in parallel with
serum concentrations. Relative brain exposure (re) of
the 2′-fluoronucleoside was 16.5%. Total clearance of
2′-F-ara-ddI was 2.18 (L/h)/kg and was moderate rela-
tive to the hepatic blood flow (5 (L/h)/kg) and renal blood
(Scheme 2). The prodrug was first metabolized to 2′-
F-ara-ddA by the P-450 reductase system. A similar
reduction of the azido moiety of AZT to an amino
function by the cytochrome P-450 system has been
recently demonstrated.1
0,11
2′-F-ara-ddA was then me-
tabolized to 2′-F-ara-ddI by adenosine deaminase. The
azido reduction assay confirmed that FAAddP was
metabolized to 2′-F-ara-ddA and 2′-F-ara-ddI by the
microsomal fraction of the human liver homogenate.
Furthermore, in vitro biotransformation studies showed
that direct conversion of FAAddP to 2′-F-ara-ddI by
adenosine deaminase occurred at a negligible rate.
FAAddP was stable in phosphate buffer saline (PBS,
pH 7.4) at 37 °C, indicating that the compound is not
susceptible to chemical hydrolysis. The in vitro biotrans-
formation of this prodrug in mouse serum, however, was
relatively rapid with a degradation half-life of 2.41 h.
Although FAAddP was metabolized in serum, no me-
tabolites were identified. In the liver homogenate,
FAAddP concentrations declined in a biphasic fashion.
1
2
flow (3.6 (L/h)/kg) in mice.
Steady-state volume of
distribution was 0.78 L/kg, hence indicating that the
compound was distributed intracellularly to a moderate
extent.
Concentrations of FAAddP, 2′-F-ara-ddA, and 2′-F-
ara-ddI in the serum and the brain after intravenous
administration of 55 mg/kg of FAAddP are shown in
Figure 2A,B. Serum concentrations of the prodrug

4
678 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 23
Notes
2
′-F-ara-ddA is the rate-limiting step in the formation
of 2′-F-ara-ddI.
As with FAAddP, a higher dose of FMAddA had to
be administered to measure the levels of 2′-F-ara-ddI.
Concentrations of FMAddA and 2′-F-ara-ddI in serum
and brain after intravenous administration of 112 mg/
kg of FMAddA are depicted in Figure 3. Serum con-
centrations of FMAddA declined rapidly with a half-life
of 0.46 h (Table 1). Total clearance (1.93 (L/h)/kg) and
steady-state volume of distribution of FMAddA (0.79
L/kg) were similar to those of 2′-F-ara-ddI. Only 5.6%
of the administered prodrug was converted to 2′-F-ara-
ddI due the low levels of ADA in mice. The relative
brain exposure of the prodrug was 7.5%; however,
inconsistent with in vitro studies, no 2′-F-ara-ddI was
detected in brain samples even at the relatively high
dose administered.
F igu r e 1. Mean ( SD serum (b) and brain (O) concentrations
of 2′-F-ara-ddI after intravenous administration of 20 mg/kg
2
′-F-ara-ddI to mice.
In summary, FAAddP underwent reduction to 2′-F-
ara-ddA followed by deamination to the active com-
pound 2′-F-ara-ddI. FMAddA did not result in increased
brain delivery of the prodrug and was slowly converted
to 2′-F-ara-ddI to prove to be effective. In this study
we have demonstrated a new approach in the design of
azido prodrugs, by utilizing the P-450 NADPH reduc-
tase system. This has a great potential in increasing
the lipophilicity of the drugs and, thus, achieving the
effective concentrations in the brain even though it was
unsuccessful in the present study. The application of
this biotransformation system utilizing azido moiety to
design prodrugs for other potentially active antiviral
and anticancer agents with the possibility of improving
the pharmacokinetics is in progress.
declined rapidly with a half-life of 0.22 h (Table 1).
Total clearance and steady-state volume of distribution
of the prodrug were 1.12 (L/h)/kg and 0.58 L/kg,
respectively. Thus, clearance of FAAddP was 2-fold
slower than that of 2′-F-ara-ddI, and the distribution
was slightly less extensive. In serum samples, low
levels of 2′-F-ara-ddA and higher concentrations of 2′-
F-ara-ddI were observed. The elimination half-life of
2
′-F-ara-ddI after FAAddP administration was longer
than that after the administration of 2′-F-ara-ddI. The
elimination half-life of 2′-F-ara-ddA was 2.9 h. The
higher concentration of 2′-F-ara-ddI compared to 2′-F-
ara-ddA is in agreement with the results of the in vitro
studies. Approximately 30% of the intravenously ad-
ministered dose of the FAAddP was converted to 2′-F-
ara-ddI.
FAAddP distributed rapidly into the brain with peak
brain levels observed at the first sampling time (Figure
Exp er im en ta l Section
Melting points were determined on a Mel-Temp II labora-
tory device and are uncorrected. The H NMR spectra were
2
6
B). The relative brain exposure of the prodrug was
.3%, while exposures of 2′-F-ara-ddA and 2′-F-ara-ddI
1
recorded on a J EOL FX 90 Q FT spectrometer, with tetra-
methylsilane 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), m (multiplet), dm (double
of multiplet), or brt (broad triplet). UV spectra were recorded
on a Beckman DU-7 spectrophotometer. TLC were performed
on Uniplates (silica gel) purchased from Analtech Co. Ele-
mental analyses were performed by Atlantic Microlab, Inc.,
Norcross, GA.
were 55.8% and 19.7%, respectively. Thus, brain ex-
posure to FAAddP was relatively low, and the relative
brain exposure to 2′-F-ara-ddI after intravenous admin-
istration of FAAddP was similar to that after adminis-
tration of 2′-F-ara-ddI. Although the relative brain
exposure for 2′-F-ara-ddA was relatively high, brain
concentrations were low.
To compare the disposition of 2′-F-ara-ddI after the
administration of FAAddP to that after the administra-
tion of 2′-F-ara-ddI, AUC (area under curve) values were
normalized for dose. As shown in Table 1, the dose-
normalized AUC values for 2′-F-ara-ddI in the serum
and the brain after intravenous administration of 55 mg/
kg of FAAddP were 3-4-fold lower than those after the
administration of 20 mg/kg of 2′-F-ara-ddI.
Concentrations of FAAddP, 2′-F-ara-ddA, and 2′-F-
ara-ddI in the serum after oral administration of 55 mg/
kg FAAddP are depicted in Figure 2C. Absorption of
FAAddP was rapid with peak serum concentrations of
the compounds achieved 0.5 h after dosing. Oral
bioavailability of FAAddP was 19%, indicating incom-
plete absorption, owing in part to its poor solubility.
Brain concentrations of FAAddP, 2′-F-ara-ddA, and 2′-
F-ara-ddI were below the limit of quantitation because
of the low oral bioavailability of FAAddP. Similar to
intravenous study of FAAddP, higher concentrations of
6-Ch lor o-9-(5-O-ben zoyl-2,3-d id eoxy-2-flu or o-â-D-a r a -
bin ofu r a n osyl)p u r in e (2). Compound 2 was prepared from
8
compound 1 according to previously published procedures: UV
8
(
MeOH) λmax 263.5 nm (lit. UV (MeOH) λmax 263.5 nm).
6
-Ch lor o-9-(2,3-dideoxy-2-flu or o-â-D-ar abin ofu r an osyl)-
p u r in e (3). A solution of compound 2 (1.34 g, 3.56 mmol) in
CH Cl (40 mL) was cooled to -78 °C under nitrogen, and
DIBAL-H (10.5 mL, 1 M solution in CH Cl ) was added slowly.
2
2
2
2
The reaction mixture was stirred at -78 °C for 45 min and
quenched by the slow addition of MeOH. The reaction mixture
was warmed to room temperature, and the solvent was
evaporated in vacuo. The residue was dissolved in hot MeOH
and filtered through a pad of Celite. Upon concentration of
the filtrate, the residue was purified on a silica gel column
3
(5% MeOH in CHCl ) to obtain pure compound 3 (0.74 g,
9
75%): mp 157-158 °C; UV (MeOH) λmax 263.5 nm (lit. UV
(
+
MeOH) λmax 260 nmm); [R]25
D
+52.3 (c 0.5, MeOH) (lit. [R]
9
25
D
55.7 (c 1.4, MeOH)).
-Azid o-9-(2,3-d id eoxy-2-flu or o-â-D-a r a bin ofu r a n osyl)-
p u r in e (4). A solution of 3 (1.0 g, 3.67 mmol) in DMF (25
mL) and LiN (0.90 g, 18.3 mmol) was stirred at room
6
3
temperature for 24 h. The DMF was evaporated under high
vacuum to yield a white solid which was boiled in MeOH and
filtered three times to yield pure 4 (0.75 g, 73.5%) as a white
2
′-F-ara-ddI when compared to 2′-F-ara-ddA were ob-
served, suggesting that the metabolism of FAAddP to

Notes
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 23 4679
Ta ble 1. Pharmacokinetics Parameters for 2′-F-ara-ddI, FAAddP, 2′-F-ara-ddA, and FMAddA following Administration of 20 mg/kg
′-F-ara-ddI, 55 mg/kg FAAddP, or 112 mg/kg FMAddA to Mice
2
dose
route of
AUC/dose
compd measd
(mg/kg)
administration
substrate
AUC0-τ (µM‚h)
AUC (µM‚h)
t1/2 (h)
((µM‚h/µmol)/kg)
2
′-F-ara-ddI Administered
2
′-F-ara-ddI
20
55
iv
iv
serum
brain
36.2
5.6
38.0
6.3
0.41
0.47
0.48
0.08
FAAddP Administered
FAAddP
serum
brain
serum
brain
serum
brain
199.3
12.0
3.0
3.9
26.5
4.8
199.8
12.6
8.6
4.8
26.9
5.3
0.22
0.22
2.92
0.91
0.68
0.96
1.01
0.06
0.04
0.02
0.13
0.03
2
2
′-F-ara-ddA
′-F-ara-ddI
FAAddP Administered
FAAddP
55
oral
serum
brain
serum
brain
serum
brain
34.2
NDa
0.8
ND
6.7
37.5
ND
1.0
ND
9.2
1.53
ND
1.74
ND
2.92
ND
0.19
ND
0.005
ND
0.046
ND
2
2
′-F-ara-ddA
′-F-ara-ddI
ND
ND
FMAddA Administered
FMAddA
′-F-ara-ddI
112
iv
serum
brain
serum
brain
296.8
39.0
10.1
ND
298.8
39.5
11.2
ND
0.46
0.98
0.52
ND
0.68
0.09
0.025
ND
2
a
ND, not detected.
F igu r e 2. Mean ( SD concentrations of FAAddP (O), 2′-F-ara-ddA (3), and 2′-F-ara-ddI (b) in serum (A) and brain (B) after
intravenous administration and in serum (C) after oral administration of 55 mg/kg FAAddP to mice.
F igu r e 3. Mean ( SD serum (A) and brain (B) concentrations of FMAddA (b) and 2′-F-ara-ddI (3) after intravenous administration
of 55 mg/kg FMAddA to mice.
solid: mp 209 °C dec; UV (H
2
O) λmax 287.5 (ꢀ 7471, pH 7), 287.5
mixture was stirred overnight. Evaporation of the solvent
yielded crude product, which was purified by silica gel column
chromatography to yield pure 59 (1.13 g, quantitative yield)
1
(ꢀ 7327, pH 2), 234 nm (ꢀ 9478, pH 11); H NMR (DMSO-d
6
) δ
,16
2
.03-2.95 (m, 2H), 3.67 (brt, 1H), 5.09 (t, 1H, D
2
O exchange-
9
9
,16
able), 5.57 (m, 1H), 6.61 (dd, 1H), 8.84 (d, 1H), 10.15 (s, 1H).
Anal. (C10 ) C, H, N.
N -Meth yl-9-(2,3-d id eoxy-2-flu or o-â-D-a r a bin ofu r a n o-
as a hygroscopic foam: UV (MeOH) λmax 264 nm (lit.
UV
(MeOH) λmax 265 nm); [R]25
+56.57 (c 1.9, MeOH)).
+56.1 (c 0.58, MeOH) (lit. [R]
25
H
10FN
7
O
2
D
D
6
9
,16
syl)a d en in e
(5). A solution of 2 (1.60 g, 4.25 mmol) in
In Vitr o Sta bility in Ser u m , Br a in , a n d Liver Hom o-
gen a te. Liver and brain homogenate were prepared in a 1:1
(g:mL) ratio with isotonic 0.05 M phosphate buffer, pH 7.4.
FAAddP (70 µM) or FMAddA (50 µM) was added to the mouse
DMF (50 mL) and methylamine (3 mL) was sealed in a steel
bomb and heated at 80 °C for 5 h. After cooling, the solvent
was evaporated, NH in MeOH (150 mL) was added, and the
3

4
680 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 23
Notes
serum, the brain homogenate, or the liver homogenate, and
the mixture was incubated in a shaker water bath at 37 °C.
Aliquots of 100 µL were removed at time zero and at selected
times for up to 6 h. Concentrations of the compounds were
determined by HPLC.
and AZddU of 7.8 and 4.3 min, respectively. Eluants were
monitored at a UV wavelength of 260 nm.
Nucleoside standards ranging from 0.01 to 100 µg/mL,
prepared in blank PBS, serum, brain homogenate, and liver
homogenate were treated the same as unknown samples.
Samples with nucleoside concentrations greater than 100 µg/
mL were diluted with the appropriate blank matrix. The limit
of quantitation (signal-to-noise ratio of 3:1) for the 2′-fluoro-
nucleosides in all biological media was 0.1 µg/mL. Extraction
recovery was greater than 80% for all compounds. The intra-
and interassay relative standard deviations (RSDs) for each
compound were less than 10% in all media.
Da ta An a lysis in Vitr o Stu d ies. Linear regression of the
natural logarithm of nucleoside analogue concentrations as a
function of time were used to determine first-order degradation
rate constants (k) and associated half-lives (t1/2 ) 0.693/k) in
PBS, serum, liver homogenate, and brain homogenate.
Da ta An a lysis in Vivo Stu d ies. Nucleoside concentration
as a function of time data were analyzed by a noncompart-
mental technique. The area under the serum or brain nucleo-
side mean (n ) 3) concentration versus time curve and the
first moment (AUMC) were determined by Lagrange poly-
nomial interpolation and integration from time zero to the last
Azid o Red u ction Assa y. The analysis of azido-reducing
1
1
activity was described previously.
Dea m in a tion of 2′-F -a r a -d d A a n d F MAd d A by Ad en o-
sin e Dea m in a se. Samples (1.5 mL) of 2′-F-ara-ddA (80 µM)
or FMAddA (50 µM) were prepared in 0.05 M isotonic
phosphate buffer, pH 7.4, and placed into a shaking water bath
at 37 °C. The reaction was initiated by the addition of 15 µL
of adenosine deaminase (type VII from calf intestinal mucosa,
Sigma Chemical Co., St. Louis, MO). The final activity in the
incubation media was 0.05 unit/mL for 2′-F-ara-ddA and 1.0
unit/mL for FMAddA. At specified time intervals, aliquots of
1
00 µL were withdrawn for the determination of 2′-F-ara-ddA
and 2′-F-ara-ddI or FMAddA and 2′-F-ara-ddI concentrations.
An im a l Stu d ies. Animal studies were approved by the
University of Georgia Animal Care and Use Committee and
conducted in accordance with guidelines established by the
Animal Welfare Act and the National Institutes of Health
Guide for the Care and Use of Laboratory Animals. Female
NIH-Swiss mice (Harland Sprague-Dawley, Indianapolis, IN)
weighing 24-28 g were housed in 12 h light/12 h dark,
constant-temperature (22 °C) environment and had free access
to standard laboratory chow and water. Animals were ac-
climatized to this environment for 1 week prior to the experi-
ments.
sample time (AUC0-τ) with extrapolation to time infinity using
14
the least-squares terminal slope (λ
time points were used to obtain λ
from 0.693/λ For intravenously administered compounds,
total clearance (CL ) was calculated from dose/AUC and
z
). The last three to five
z
. Half-life was calculated
z
.
T
steady-state volume of distribution (Vss) from dose × AUMC/
2
AUC . The fraction of the prodrug converted to parent
2
′-F-ara-ddI, dissolved in physiological saline (15 mg/mL),
compound (f
where AUCprpd is the AUC of the parent compound after
administration of the prodrug (dosepd) and CL is the clearance
of the parent compound. Relative brain exposure (r ) was
calculated from AUCbrain/AUCserum
c
) was calculated from AUCprpd × CL
T
/dosepd,
was administered intravenously via tail vein injection at a dose
of 20 mg/kg (79 µmol/kg). FAAddP (55 mg/kg; 197 µmol/kg)
was administered intravenously as a solution in DMSO (15
mg/mL) or orally by a gavage as a suspension in physiological
saline. FMAddA, dissolved in saline (15 mg/mL), was admin-
istered intravenously at a dose of 112 mg/kg (437 µmol/kg).
At selected time intervals, mice (three animals per each time
point) were anesthetized with diethyl ether and sacrificed by
exsanguination via left ventricular heart puncture. Serum
was harvested from blood collected. The brain was excised,
rinsed with normal saline, blotted dry, and weighed. Serum
and brain samples were frozen at -20 °C until analysis.
T
15
e
.
Ack n ow led gm en t. This work was supported by
NIH Grants AI 25899, AI 32351, and HL-42125. J .P.S.
was supported by Faculty Research Award from the
American Cancer Society.
Refer en ces
An a lytica l Meth od ology. Concentrations of FAAddP,
FMAddA, 2′-F-ara-ddA, and 2′-F-ara-ddI in PBS, serum, brain,
and liver homogenate were determined by high-performance
liquid chromatography (HPLC). The brain or liver tissue were
homogenized in a 1:1 (g:mL) ratio with ice-cold isotonic 0.05
M phosphate buffer, pH 7.4. Buffer, serum or tissue homo-
genate (100 µL) was mixed with 10 µL of internal standard
(1) Faulds, D.; Brogden, R. N. Didanosine: a review of its antiviral
activity, pharmacokinetics properties and therapeutic potential
in human immunodeficiency virus infection. Drugs 1992, 44, 94-
1
16.
(
(
(
2) Tartaglione, T. A.; Fletcher, C. V.; Collier, A. C. Principles and
management of the acquired immunodeficiency syndrome. In
Pharmacotherapy. A pathophysiologic approach; DiPiro, J . T.,
et al., Eds.; Appleton and Lange: Norwalk, CO, 1993; pp 1837-
(
25 µg/mL of 3′-azido-2′,3′-dideoxyuridine, AZddU). Aceto-
1
867.
nitrile (600 µL) containing 0.1% acetic acid was added while
the mixture was vortexed to precipitate proteins. The tubes
were centrifuged at 3000 rpm for 5 min, and the supernatant
was transferred to a clean tube. Supernatant was evaporated
to dryness under a stream of nitrogen gas at room tempera-
ture. The residual film was reconstituted in 110 µL of mobile
phase, and 50 µL was injected onto the HPLC.
Chromatographic separations were performed using a Hy-
persil ODS column 150 × 4.6 mm, 5 µm particle size (Alltech
Associates, Deerfield, IL), preceded by a guard column packed
with 30-40 µm pellicular Perisorb RP-18. Mobile phase flow
rate was 2 mL/min. For the analysis of FAAddP and 2′-F-
ara-ddA, the mobile phase consisted of 7% (v/v) acetonitrile
in 80 mM sodium acetate, pH 5.0. The retention times for
3) Drusano, G. L.; Yuen, G. J .; Morse, G.; Cooley, T. P.; Seidlin,
M.; Lambert, J . S.; Liebman, H. A.; Valentine, F. T. ; Dolin, R.
Impact of bioavailability on determination of the maximal
tolerated dose of 2′,3′-dideoxyinosine in phase I trials. Antimi-
crob. Agents Chemother. 1992, 36, 1280-1283.
4) Collins, J . M.; Klecker, R. W., J r.; Kelley, J . A.; Roth, J . S.;
McCully, C. L.; Balis, F. M.; Poplack, D. G. Pyrimidine dideoxy-
ribonucleosides: selectivity of penetration into cerebrospinal
fluid. J . Pharmacol. Exp. Ther. 1988, 245, 466-470.
(5) Anderson, B. D.; Hoesterey, B. L.; Baker, D. C.; Galinsky, R. E.
Uptake kinetics of 2′,3′-dideoxyinosine into brain and cerebrospi-
nal fluid of rats: intravenous infusion studies. J . Pharmacol.
Exp. Ther. 1990, 253, 113-118.
(
6) Tuntland, T.; Ravasco, R. J .; Al-Habet, S.; Unadkat, J . D. Afflux
of Zidovudine and 2′,3′-dideoxyinosine out of the cerebrospinal
fluid when administered alone and in combination to Macaca
nemestina. Pharm. Res. 1994, 11, 312-317.
2
′-F-ara-ddI, FAAddP, and AZddU were 4.5, 7.9, and 5.1 min,
(
7) Marquez, V. E.; Tseng, C. K.-H.; Mitsuya, H.; Aoki, S.; Kelley,
J . A.; Ford, J r., H.; Roth, J . S.; Broder, S.; J ohns, D. G.; Driscoll,
J . S. Acid-stable 2′-fluoro purine dideoxynucleosides as active
agents against HIV. J . Med. Chem. 1990, 33, 978-985.
respectively. The mobile phase for the analysis of 2′-F-ara-
ddI consisted of 4.2% (v/v) acetonitrile in 40 mM sodium
acetate, pH 4.1, yielding retention times of 3.78 and 7.6 min,
for 2′-F-ara-ddI and AZddU, respectively. For the analysis of
FMAddA in serum and liver homogenate, a mobile phase of
(8) Shanmuganathan, K.; Koudriakova, T.; S. Nampalli, S.; Du, J .;
Gallo, J . M.; Schinazi, R. F.; Chu, C. K. Enhanced brain delivery
of an anti-HIV nucleoside 2′-F-ara-ddI by xanthine oxidase
mediated biotransformation. J . Med. Chem. 1994, 37, 821-827.
7
.5% acetonitrile in 40 mM sodium acetate, pH 6.0, was used.
The retention time for FMAddA was 8.8 min and that for
AZddU was 4.9 min. For FMAddA analysis in brain homo-
genate, the mobile phase consisted of 7.5% acetonitrile in 10
(
9) Barchi, J . J ., J r.; Marquez, V. E.; Driscoll, J . S.; Ford, H., J r.;
Mittsuya, H.; Shirasaka, T.; Aoki, S.; Kelley, J . A. Potential anti-
AIDS drugs. Lipophilic, adenosine deaminase-activated pro-
drugs. J . Med. Chem. 1991, 34, 1647-1655.
mM K
2 2
HPO (pH 7.2), yielding retention times for FMAddA

Notes
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 23 4681
(
10) Placidi, L.; Cretton, E. M.; Placidi, M.; Sommadossi, J .-P.
(14) Rocci, M. L., J r.; J usko, W. J . LAGRAN program for area and
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.
11) Cretton, E. M.; Sommadossi, J -P. Reduction of 2′-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-
moments in pharmacokinetic analysis. Comput. Programs Biomed.
1
983, 16, 203-216.
15) Gibaldi, M.; Perrier, D. Clearance concepts. In Pharmacokinetics,
nd ed.; Marcel Dekker Inc.: New York, 1982; pp 319-353.
16) Chu, C. K.; Ullas, G. V.; J eong, L. S.; Ahn, S. K.; Doboszewski,
B.; Lin, Z. X.; Beach, J . W.; Schinazi, R. F. Synthesis and
structure-activity relationships of 6-substituted 2′,3′-dideoxy-
purine nucleosides as potential anti-human immunodeficiency
virus. J . Med. Chem. 1990, 33, 1553-1561.
(
(
(
2
9
50.
(
12) Gerlowski, L. E.; J ain, R. K. Physiologically based pharmaco-
kinetic modeling: principles and applications. J . Pharm. Sci.
1
983, 72, 1103-1126.
(
13) Camierer, G. W.; Smith, C. G. Studies of the enzymatic de-
amination of cytosine arabinoside. I. Enzyme distribution and
species specificity. Biochem. Pharmacol. 1965, 14, 1405-1416.
J M960140C