Effect of change in nucleoside structure on the activation and antiviral activity of phosphoramidate derivatives

Article abstract of DOI:10.1016/j.bmc.2005.04.083

Changing the nucleoside group of a series of phosphoramidate derivatives affects the enzyme mediated hydrolysis rate of the compounds. d4T and AZT-substituted analogs were activated by enzymes such as lipases, esterases, and proteases. On the other hand, 3dT-substituted derivatives were comparatively less prone to hydrolysis under similar experimental conditions. From the experimental results, we propose that the most preferable nucleoside group for enzyme activation is d4T rather than AZT or 3dT. Additionally, we also observed that depending on the enzymes used the chiral selectivity of the enzymes for the phosphorus center of these phosphoramidate derivatives differed, demonstrating the importance of the nucleoside structure for this class of compounds.

Full text of DOI:10.1016/j.bmc.2005.04.083

Bioorganic & Medicinal Chemistry 13 (2005) 5408–5423
Effect of change in nucleoside structure on the activation
and antiviral activity of phosphoramidate derivatives
T. K. Venkatachalam, P. Samuel, S. Qazi and F. M. Uckun*
Departments of Chemistry, Pharmaceutical Sciences, Bioinformatics, Immunology and Virology,
Parker Hughes Institute, 2699 Patton Road, Roseville, MN 55113, USA
Paradigm Pharmaceuticals, 2685 Patton Road, Roseville, MN 55113, USA
Received 21 January 2005; accepted 25 April 2005
Available online 1 July 2005
Abstract—Changing the nucleoside group of a series of phosphoramidate derivatives affects the enzyme mediated hydrolysis rate of
the compounds. d4T and AZT-substituted analogs were activated by enzymes such as lipases, esterases, and proteases. On the other
hand, 3dT-substituted derivatives were comparatively less prone to hydrolysis under similar experimental conditions. From the
experimental results, we propose that the most preferable nucleoside group for enzyme activation is d4T rather than AZT or
3dT. Additionally, we also observed that depending on the enzymes used the chiral selectivity of the enzymes for the phosphorus
center of these phosphoramidate derivatives differed, demonstrating the importance of the nucleoside structure for this class of
compounds.
ꢀ 2005 Elsevier Ltd. All rights reserved.
1. Introduction
found to act as poor substrates for nucleoside ki-
nases.9,12–15 Consequently, the development of prodrug
strategies has been sought to bypass the initial nucleo-
side kinase mediated activation step. In an attempt to
overcome the dependence of ddN analogs on intracellu-
lar nucleoside kinase activation, we and others have pre-
pared a number of aryl phosphate derivatives of
zidovudine and stavudine, some of which were found
to be potent anti-HIV agents with subnanomolar IC50
values.10,11,16–25
The HIV/AIDS pandemic continues its spread at a rate
of over 15,000 new infections every day. Three catego-
ries of antiretroviral agents in clinical use are nucleoside
reverse transcriptase inhibitors (NRTI) such as zidovu-
dine and stavudine (d4T),1,2 protease inhibitors, and
non-nucleoside
reverse
transcriptase
inhibitors
(NNRTI).3–6 The 50-triphosphates of 20,30-dideoxynu-
cleoside analogs (ddN), which are generated by nucleo-
side and nucleotide kinases, are potent inhibitors of
human immunodeficiency virus (HIV) reverse transcrip-
tase.7,8 The rate limiting step for the conversion of 30-az-
ido-30-deoxy thymidine (AZT) to its bioactive
metabolite, AZT-triphosphate, is proposed to be the
conversion of the monophosphate derivative to the
diphosphate derivative. By contrast, the rate limiting
step for the intracellular generation of the bioactive
stavudine metabolite, stavudine-triphosphate, was
reported to be the conversion of stavudine to its mono-
phosphate derivative.9–11 Anti-HIV ddN derivatives pri-
marily rely on nucleoside and nucleotide kinases to
convert them into the corresponding 50-triphosphates
as discussed before. However, such compounds were
Our lead anti-HIV compound, stampidine, is a phos-
phoramidate derivative of stavudine.23 Stampidine was
100 times more active than stavudine and twice as active
as zidovudine against nine clinical HIV-1 isolates of
non-B envelope subtypes (A, C, F, and G) originating
from South America, Asia, and sub-Saharan Africa.20
Stampidine was effective against 20 genotypically and
phenotypically nucleoside analog reverse transcriptase
inhibitor (NRTI)-resistant and 6 non-nucleoside inhibi-
tor (NNRTI)-resistant HIV-1 isolates at subnanomolar
to low nanomolar concentrations.20 Orally or intraperi-
toneally administered stampidine exhibited significant
and dose-dependent in vivo anti-HIV activity against
an NRTI-resistant clinical HIV-1 isolate in severe com-
bined immunodeficient (SCID) mice reconstituted with
peripheral blood (PBL) mononuclear cells from seroneg-
ative human donors.21 Orally administered stampidine
*
Corresponding author. Tel.: +1 651 796 5400; fax: +1 651 796
5493; e-mail: fatih_uckun@ih.org
0968-0896/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2005.04.083

T. K. Venkatachalam et al. / Bioorg. Med. Chem. 13 (2005) 5408–5423
5409
showed a dose-dependent antiretroviral effect in chroni-
cally FIV-infected cats.22 Stampidine therapy was not
associated with any clinical or laboratory evidence of
toxicity at dose levels as high as 500 mg/kg or at cumu-
lative dose levels as high as 8.4 g/kg. Stampidine exhib-
ited favorable pharmacokinetic behavior in mice, rats,
dogs, and cats following oral administration.22,24 The
documented in vitro potency of stampidine against pri-
mary clinical HIV-1 isolates with genotypic and/or phe-
notypic NRTI- or NNRTI-resistance as well as non-B
envelope subtypes together with its in vivo antiretroviral
activity in HIV-infected Hu-PBL SCID mice and FIV-
infected cats warrants its further development as a new
anti-HIV drug.
that the lipase-mediated hydrolysis, as well as
protease-mediated hydrolysis, of d4T derivatives was
more efficient than esterase-mediated hydrolysis.29
We extended our investigations to examine the relation-
ship between the nucleoside moiety of these phospho-
ramidate derivatives and their hydrolysis rate and
antiviral activity. For this purpose, a series of com-
pounds were prepared where only the nucleoside group
was altered. The synthesis and characterization of the
nucleoside-substituted phosphoramidate derivatives
have been reported.18–20 We have focused on three
nucleoside analogs for this study, AZT, d4T, and 3dT.
Table 1 shows the first order rate constants for the enzy-
matic hydrolysis observed for the phosphoramidate
derivatives of 3dT with electron donating and withdraw-
ing substituents. Compounds with an electron donating
group, such as the methoxy group-substituted analog,
showed a 2- to 3-fold slower rate for lipase-mediated
hydrolysis. In the case of the lipase-mediated hydrolysis,
compounds with an electorn-withdrawing group, such
as the bromo-substituted analog, showed a faster rate
of hydrolysis. Unexpectedly, the nitro-substituted ana-
log did not show significant hydrolysis, although one
would have expected such a trend. This result suggested
that lipase activation is not efficient in the case of nitro-
substituted analog either due to the highly polar charac-
teristic of the nitro group or to the lack of fit in the
enzyme pocket. For the esterase-mediated hydrolysis,
the trend was almost reversed. The compounds with
electron donating groups on the aryl moiety such as
OMe group hydrolyzed much faster than those with
electron withdrawing groups.
The generation of the active metabolites for the phos-
phoramidate derivatives of nucleosides was originally
proposed to require the esterase-mediated hydrolysis
of the carbomethoxy group associated with the alanine
side chain of stampidine.10,11,18–20,26–28 We hypothesized
that in various tissue microenvironments such com-
pounds may be metabolized through the action of other
hydrolytic enzymes in addition to esterase. This was
supported recently by a study showing the lipase and
protease activations of various phosphoramidate deriva-
tives.29 The focus of this structure–activity relationship
study was the role of the nucleoside group in a series
of phosphoramidate derivatives to understand enzyme
recognition and activation of this class of compounds.
2. Results and discussion
It has been postulated that phosphoramidate derivatives
of stavudine are activated by an enzymatic hydrolysis
mediated by esterases.26–28 No other enzymes have been
proposed as alternative mediators of the hydrolysis of
this promising class of anti-HIV compounds. Since lip-
ases have been shown to be capable of hydrolyzing other
chiral compounds with a phosphorous center,30–34 we
sought to determine if the phosphoramidate derivative
of d4T could be hydrolyzed to their corresponding
active metabolites by a lipase as well. We observed
In the presence of the protease subtilisin Carlsberg, all
the substituted derivatives (NO2, Cl, Br, OMe, and F)
underwent hydrolysis at a rate ranging from 1.83 to
1.45 hꢀ1 which was faster than the hydrolysis of the
unsubstituted phenyl analog of 3dT. Similar results were
obtained for all compounds, except for the phenyl-
substituted compound in the presence of the protease
enzyme, Carica papaya.
Table 1. First order rate constant values for 3dT-substituted phosphoramidate derivatives in the presence of enzymes
Compound (X)
Lipase
Esterase
Subtilisin Carlsberg
Carica papaya
NO2
Br
0.67 0.2
2.90 0.6
1.4 0.2
1.2 0.6
1.01 0.1
1.25 0.1
0.10 0.03
0.9 0.08
0.8 0.2
1.31 0.1
1.05 0.1
1.8 0.1
1.01 0.03
0.89 0.1
0.7 0.07
0.84 0.01
0.41 0.06
0.41 0.08
Cl
OMe
H
3.06 0.7
1.72 0.9
1.5 0.7
1.83 0.05
0.23 0.04
1.45 0.05
F
Rate constants are expressed per hour.

5410
T. K. Venkatachalam et al. / Bioorg. Med. Chem. 13 (2005) 5408–5423
Table 2 shows the first order rate constants of hydrolysis
observed for the phosphoramidate derivatives of d4T
using the same enzymes as those used with the 3dT
derivatives. As with the 3dT derivatives, the lipase-med-
iated hydrolysis showed the fastest rates for compounds
with electron withdrawing groups. However, the rate of
lipase-mediated hydrolysis was faster for the d4T deriv-
atives. For example, stampidine, the bromo-substituted
phenyl phosphoramidate derivative showed a rate value
of 3.10 hꢀ1 in comparison with 3dT (2.9 hꢀ1) for lipase-
mediated hydrolysis.
range of 0.5–0.9 hꢀ1. From the results, it appears that
subtilisin Carlsberg protease enzyme recognizes all of
the substituted phosphoramidate derivatives of d4T.
Table 3 shows the hydrolysis rate constants observed for
various phosphoramidate derivatives of AZT in the
presence of the enzymes examined. The rate for the li-
pase-mediated hydrolysis was comparatively slow for
these phosphoramidate derivatives. (0.35–0.8 hꢀ1). The
data in Table 3 show that the rate of lipase-mediated
hydrolysis for unsubstituted compound is minimal
(0.02 hꢀ1). Conversely, the esterase-mediated hydrolysis
gave a fast rate of hydrolysis. The rate of esterase-med-
iated hydrolysis was 5.73 and 4.71 hꢀ1 for bromo- and
OMe-substituted derivatives, respectively. Additionally,
the bromo-substituted the derivative displayed the high-
est rate of hydrolysis in the presence of the protease en-
zyme subtilisin Carlsberg (3.56 hꢀ1). The rate of
hydrolysis in the presence of cysteine protease Carica
papaya was nearly the same for most of the substituted
derivatives (0.63–0.81 hꢀ1). Compared with the other
derivatives the bromo-substituted compound had a
higher rate of hydrolysis with this protease.
The fluoro-substituted compound underwent faster
hydrolysis than the unsubstituted compound in the pres-
ence of subtilisin Carlsberg but not Carica papaya. In
the presence of the esterase enzyme, the OMe- and Br-
substituted compounds underwent faster hydrolysis
(rate of hydrolysis values of 3.90 and 3.3 hꢀ1, respective-
ly), than other compounds (rates of hydrolysis: 2.28–
3.04 hꢀ1). The most striking results, however, were
shown using serine protease. This enzyme gave the high-
est rate of hydrolysis, especially for the subtilisin Carls-
berg. For example, subtilisin Carlsberg showed
hydrolysis rate in the range of 4.2–8.5 hꢀ1 for all the
derivatives. On the other hand, the cysteine protease,
Carica papaya, caused slow hydrolysis of the com-
pounds. With the exception of stampidine (1.43 hꢀ1),
the hydrolysis rates for all the compounds were in the
2.1. Chiral selectivity at the phosphorus center
2.1.1. Nucleoside change. We examined the differences in
the chiral selectivity for the various nucleoside-substitut-
Table 2. First order rate constant values for d4T-substituted phosphoramidate derivatives in the presence of enzymes
Compound (X)
Lipase
Esterase
S. Carlsberg
Carica papaya
Br (stampidine)
Cl
3.1 0.5
3.2 2.28
1.24 0.8
1.81 0.3
1.75 0.7
3.3 0.5
3.04 0.2
3.9 1.6
8.55 0.7
4.21 0.3
7.07 2.2
4.46 2.8
4.64 1.8
1.43 0.4
0.7 0.1
OMe
H
0.51 0.05
0.83 0.04
0.92 0.1
2.28 0.9
2.78 0.1
F
Rate constants are expressed per hour.
Table 3. First order rate constant values for AZT-substituted phosphoramidate derivatives in the presence of enzymes
Compound (X)
Lipase
Esterase
Subtilisin Carlsberg
Carica papaya
NO2
Br
0.35 0.06
0.83 0.09
0.46 0.01
0.02
0.22 0.1
5.72 0.2
4.71 1.2
—
0.39 0.04
3.56 1.4
0.70 0.1
0.27 0.09
0.65 0.01
0.81 0.05
0.63 0.02
0.08
OMe
H
Rate constants are expressed per hour.

T. K. Venkatachalam et al. / Bioorg. Med. Chem. 13 (2005) 5408–5423
5411
ed phosphoramidate derivatives hydrolyzed by enzymes.
Figure 1 shows the HPLC profile comparison for
bromo-substituted aryl phosphoramidate derivatives
with varying nucleosides treated with lipase. Lipase
had the greatest preference for stereoselectivity with
the d4T-substituted aryl phosphoramidate derivatives in
comparison to the AZT and 3dT derivatives. For ester-
ase-mediated hydrolysis, we observed that AZT-substi-
tuted aryl phosphoramidate derivative had the highest
chiral selectivity (Fig. 2). This suggests that the esterase
discriminates between the diastereoisomers of AZT-
substituted arylphosphoramidate derivatives much bet-
ter than the 3dT and d4T compounds. Both the d4T
and 3dT derivatives indicated very little chiral selectivi-
ty, among the isomers (Fig. 2). Focusing on the chiral
selectivity observed for the protease subtilisin Carlsberg,
the highest chiral selectivity was noticed in the case of
d4T derivatives, followed by the AZT and 3dT analogs
(Fig. 3). Figure 4 illustrates the HPLC profile observed
for Carica papaya protease-mediated hydrolysis of the
p-bromophosphoramidate derivative. For the 3dT- and
AZT-substituted compounds, one of the isomers was
preferentially hydrolyzed compared to the other. How-
ever, the d4T-substituted derivative underwent faster
hydrolysis (1.01 for 3dT, 1.43 for d4T, and 0.65 for
AZT) with low chiral selectivity. This suggests that Cari-
ca papaya recognizes both the 3dT- and AZT-substitut-
ed analogs better than the d4T derivatives. The greatest
degree of chiral selectivity was observed for the AZT-
substituted aryl phosphoramidate derivative followed
by 3dT.
Table 4 shows the summary of the trends observed for
each of the nucleoside analogs. It shows the groups that
were found to yield the fastest rates of hydrolysis for the
individual enzymes. In general, the preferred substituent
on the phenyl moiety of these phosphoramidate deriva-
tives seems to be a bromo, chloro, or methoxy group.
However, for the esterase mediated hydrolysis of 3dT,
the methoxy derivative was the only compound that was
more readily hydrolyzed, although no chiral selectivity
is observed. This is an important observation that may
be crucial in deciding which nucleoside analog should be
used for antiviral therapy for a particular case.
2.2. Substitution on the phenyl ring and chiral selectivity
We also examined the influence of various para substitu-
tions on the phenyl ring on the chiral selectivity by indi-
vidual enzymes. Esterase showed very little chiral
selectivity irrespective of the substituents present in the
phenyl ring demonstrating that these substituents did
not affect the recognition of both.
Figure 1. Chiral selectivity of lipase-mediated hydrolysis of the
p-bromophenyl phosphoramidate derivatives of 3dT (A), d4T (B),
and AZT (C) as observed by HPLC at t = 15 min.
stituents on the phenyl ring in all three nucleoside deriv-
atives. As an example, the selectivity indices observed
for individual phosphorus isomers of d4T-substituted
phosphoramidate analogs treated with lipase and ester-
ase are shown in Table 5.29,41 The unprecedented selec-
tivity shown by lipase is rationalized due to the tight
binding pocket of the lipase compared to esterase.29
Recently, we also reported the selective indices for pro-
tease-mediated hydrolysis of phosphoramidate deriva-
tives of stavudine and found that substituents on the
phenyl ring have a profound effect on the selectivity.41
Examining the chiral selectivity of lipase, we found that
the chloro and bromo groups on the phenyl ring influ-
enced the chiral selectivity in 3dT and AZT derivatives
more than groups such as OMe. This may be attributed
to the size and/or polarity of the halo groups. The sub-
stituents on the phenyl ring also influenced the chiral
selectivity of d4T derivatives. In the case of protease en-
zymes, chiral selectivity by subtilisin Carlsberg39 as well
as Carcia papaya37,40 was influenced by some of the sub-

5412
T. K. Venkatachalam et al. / Bioorg. Med. Chem. 13 (2005) 5408–5423
Figure 2. Chiral selectivity of esterase-mediated hydrolysis of the
p-bromophenyl phosphoramidate derivatives of 3dT (A), d4T (B),
and AZT (C) as observed by HPLC at t = 15 min.
Based on our earlier studies we tentatively assigned the
stereochemistry of the isomers that under go hydrolysis
in the presence of enzymes for d4T derivatives.29 Lipase
preferably hydrolyzes the (S) isomers of all the phospho-
ramidate derivatives faster than the (R) isomers. Ester-
ase showed no isomeric preference. The protease
subtilisin Carlsberg preferably hydrolyzes (R) isomers
and the cysteine protease prefers to hydrolyze the (S)
isomers. This point is further illustrated by examining
a mixture of phosphoramidate derivatives treated with
various enzymes. Figure 5 shows a series of representa-
tive HPLC chromatograms of a mixture of phospho-
ramidate treated with lipase over time. The peaks at
6.0 and 6.5 min represent the phosphorus diastereoiso-
mer [Rp or Sp] of 2-bromophenyl phosphoramidate of
stavudine (DDE605). The peaks at 4.3 and 4.59 corre-
spond to the phosphorus diastereoisomers [Rp or Sp]
Figure 3. Chiral selectivity of subtilisin Carlsberg mediated hydrolysis
of the p-bromophenyl phosphoramidate derivatives of 3dT (A), d4T
(B), and AZT (C) as observed by HPLC at t = 15 min.
of 4-methoxy phenyl alaninyl phosphoramidate of
stavudine (DDE598). At 18 min of the reaction, addi-
tional peaks are visible in the HPLC chromatogram.
The peaks at 1.98 min represent stavudine (d4T), fol-
lowed by the peak at 3.1 min of p-methoxyphenol. The
peak at 11.3 min correspond to that of 2-bromophenol.
The isomers are hydrolyzed preferentially, demonstrat-
ing the chiral selectivity of lipase. Figure 6 shows the
selectivity indices for the various phosphoramidate
derivatives of stavudine treated with serine and cysteine
proteases. Cysteine protease shows the highest degree of

T. K. Venkatachalam et al. / Bioorg. Med. Chem. 13 (2005) 5408–5423
5413
Table 5. Selectivity index of various arylphosphoramidate derivatives
of d4T29,41
X
SI
Lipase
Esterase
Carlsberg
Carica papaya
H
4Br
0.7
4.0
5.4
1.9
2.7
1.0
1.0
1.5
1.3
1.3
0.9
1.3
0.1
0.4
0.2
0.4
0.5
—
6.1
1.4
2.0
5.2
0.8
—
4Cl
4F
4OMe
2,6-OMe
SI denotes the selectivity index is the ratio of rate of hydrolysis
observed for each of the isomers.
an IC50 value of 0.006 lM. The other derivatives showed
only marginal to low antiviral activity (0.048–0.057 lM).
Among the groups of nucleoside analogs studied, we
found that the d4T derivatives were 5–40 times more po-
tent than the AZT and 3dT compounds. On the other
hand, most d4T phosphoramidate derivatives showed
considerable potency (0.001–0.004 lM). However, all
3dT derivaties were shown to possess micromolar activi-
ty, indicating that saturation of the double bond in the
structure of the sugar ring is not beneficial.
In the previous discussion, we have indicated that most of
the enzymes hydrolyzed d4T derivatives more efficiently
compared to the corresponding AZT and 3dT analogs.
Relating the antiviral activity to the rate of enzymatic
hydrolysis, we conclude that in order for a drug to be effec-
tive we need to have a structural framework that can be
recognized by a variety of enzymes. Also it shows that
hydrolysis may be one of the key factors for the biological
activity shown by these nucleoside analogs. In addition,
we also point out that the metabolite resulting from the
hydrolysis of the nucleoside analogs needs to be further
recognized by kinases to form the active compounds.
Figure 4. Chiral selectivity of Carica papaya mediated hydrolysis of
the p-bromophenyl phosphoramidate derivatives of 3dT (A), d4T (B),
and AZT (C) as observed by HPLC at t = 15 min.
Table 4. Fast rate of hydrolysis of the best substituents present on the
phenyl ring of phosphoramidate derivatives of nucleosides
Nucleoside Esterase
Lipase Carlsberg Carica
papaya
2.4. Role of the amino acid
d4T
AZT
3dT
OMe, Br, Cl Br, Cl Br
Br
OMe, Br
OMe
Br
Br
Br
Cl
Br
Br
To evaluate the influence of the amino acid of these
phosphoramidate derivatives on their anti-HIV activity
we prepared AZT analogs with different amino acids.
We chose the 4-bromo substitutions primarily because
the bromo derivatives were found to be more active as
well as more easily hydrolyzed. Table 7 shows the rate
of hydrolysis observed for the compounds in the pres-
ence of the various enzymes examined. Among the ami-
no acids chosen alanine was found to be the best amino
acid with the highest rate of hydrolysis for all the en-
zymes. However, phenyl alanine-substituted aryl phos-
phoramidate derivative of AZT showed a faster rate
for both esterase and subtilisin Carlsberg-mediated
hydrolysis. The proline, tryptophan, leucine, and phenyl
chiral selectivity among the substituents as compared to
the other protease.
2.3. Antiviral activity
We focused our attention on the antiviral activity of the
aryl-substituted phosphoramidate derivatives belonging
to one of the three classes of nucleoside analogs: AZT,
d4T, and 3dT. Table 6 shows the antiviral activity shown
by various analogs. In the case of the AZT derivatives, the
bromo-substituted compound was the most potent with

5414
T. K. Venkatachalam et al. / Bioorg. Med. Chem. 13 (2005) 5408–5423
Figure 5. HPLC profiles of a mixture of the 4-methoxy- (DDE598) and the 2-bromophenyl phosphoramidate (DDE605) derivatives of d4T treated
with lipase at various intervals of time.
alanine substituted phosphoramidate derivatives all had
a slower rate of hydrolysis with lipase. The antiviral
potency shown by the alanine-substituted phosphoram-
idate derivative was 10-fold higher compared to other
amino acid derivatives.35
AZT-triphosphate, is the conversion of the monophos-
phate derivative to the diphosphate derivative. By con-
trast, the rate limiting step for the intracellular
generation of the bioactive stavudine metabolite, stavu-
dine-triphosphate, was reported to be the conversion of
the nucleoside to its monophosphate derivative.9–11 The
difference in the rate limiting steps for these two nucleo-
sides indirectly points out that the kinases differentiate be-
tween the structural differences of the two compounds. In
other words, the change of the nucleoside may impart a
change in the binding characteristics with kinases. We
2.5. Mechanism of enzyme mediated hydrolysis of phos-
phoramidate derivatives of nucleosides
It has been proposed that the rate limiting step for
the conversion of AZT to its bioactive metabolite,

T. K. Venkatachalam et al. / Bioorg. Med. Chem. 13 (2005) 5408–5423
5415
ate between these three nucleosides as observed from their
hydrolysis profile. It is essential that the phosphoramidate
derivatives be hydrolyzed first to remove the phenoxy
moiety in order to be activated by kinases. As such, the
proposed hydrolysis pathway may be playing the key role
for initiating the metabolism of these compounds. For
this to happen the necessary enzymes in the cells need to
recognize the substrate structure. To illustrate this, we
propose the following pathway for hydrolytic profile of
phosphoramidate derivatives of nucleosides (Scheme 1).
We chose only d4T and AZT derivatives as these phos-
phoramidate derivatives showed the most promising
antiviral activity. In the first case, we observed that the
rate of hydrolysis by the bacterial enzyme serine prote-
ase, subtilisin Carlsberg, yielded the fastest rate of
hydrolysis for all the stavudine derivatives examined.
Both lipase and esterase showed similar trends, however
at a lower rate of hydrolysis. We also point out that cys-
teine protease Carcia papaya showed a 3- to 5-fold lower
hydrolysis rate, implying that it is not being recognized
as well as the d4T phosphoramidates.
Figure 6. Selectivity index for two proteases for varying substituents
on the phenyl ring of phosphoramidate derivatives of stavudine. Left
to right increasing value of Hamett sigma.
Table 6. Antiviral activity of various phosphoramidate derivatives of
nucleoside analogs
In the case of AZT derivatives, esterase showed consid-
erable activation as compared to lipase and Carica
papaya. However, in the case of subtilisin Carlsberg,
all four derivatives (leucine, tryptophan, phenyl alanine,
and alanine) showed moderate rate of hydrolysis. Based
on the result, we propose that bromo- and methoxy-
substituted AZT derivatives of phosphoramidate deriva-
tives are well recognized by esterase and subtilisin
Carlsberg as compared to other enzymes. Lipase showed
very low rate of hydrolysis in the case of AZT derivatives
as compared to both d4T and 3dT analogs. This is under-
standable, given that the binding pocket of lipase is very
small 29 and the azido group in AZT present in the sugar
moiety is not able to enter the binding pocket. In the case
of 3dT presence of two hydrogen atoms in the sugar
moiety may affect the binding although to a lesser extent
than the azido group. On the other hand, esterase has a
larger area for binding and we observe a faster rate of
hydrolysis in each of the nucleoside analogs. A similar
trend was observed with subtilisin Carlsberg. In summa-
ry, we hypothesize that each structural modification of
the phosphoramidate derivatives of nucleosides has a sig-
nificant role toward activation by different enzymes.
Substituents
AZT
d4T18
3dT18
H
0.05517
0.002
0.004
2.1
1.3
3.1
2.1
1.2
OMe
F
Cl
0.05717, 0.05718
0.04817, 0.02918
—
0.001
0.001
0.001 (stampidine)
Br
0.006
Antiviral activity expressed in lM.
The viral strain used for the above studies was HTLVIIIB. HIV-1
strain. The applied p24 enzyme immunoassay (EIA) was unmodified
kinetic assay commercially available from Coulter Corporation/
Immunotech, Inc. (Westbrooke, ME), which utilizes a murine mAb to
HIV core protein coated on to microwell strips to which the antigen
present in the test culture supernatant samples binds. Percent viral
inhibition was calculated by comparing the p24 values from untreated
infected cells (i.e., virus controls).
speculate that a similar trend occurs for the 3dT deriva-
tives, although detailed studies have not been initiated
due to the low activity profile of this compound. This
analogy is supported by how bacterial enzymes differenti-
Table 7. Rate of hydrolysis of phosphoramidate derivatives of AZT with enzymes
Amino Acid (L) (R1)
Lipase
Subtilisin Carlsberg
Esterase
Carica papaya
Leucine
0.03 0.02
0.05 0.04
0.31 0.2
0.83 0.09
2.8 2.5
2.8 0.08
8.6 4.6
3.6 1.4
5.06 0.9
5.65 0.5
11.23 0.8
5.72 0.2
0.17 0.1
0.23 0.2
0.17 0.1
0.81 0.05
Tryptophan
Phenyl alanine
Alanine
Rate constants are expressed per hour.

5416
T. K. Venkatachalam et al. / Bioorg. Med. Chem. 13 (2005) 5408–5423
Scheme 1. Enzymatic hydrolysis pathway of phosphoramidate derivatives of stavudine.
2.6. Tentative pathway for the activation of phospho-
ramidate derivatives of stavudine with various enzymes
specificity of the HIV-1 protease. These proteases do
not accept a broad spectrum of compounds. However,
our results demonstrate that the HIV-1 protease is not
directly involved in the activation step depicted in our
flowchart. Based on this result we propose that cellular
proteases may be involved instead of HIV-1 protease
in the activation step illustrated in Scheme 1. Further
work is in progress to evaluate the role of intra/extracel-
lular proteases with these phosphoramidate derivatives.
Based on the results obtained from the enzymes exam-
ined we propose that there are two steps involved in
the conversion of these phosphoramidate derivatives to
their active metabolites (Fig. 7). We selected stampidine,
a 4-bromophenylmethoxylalaninyl phosphoramidate of
stavudine, as the example compound represented in
the pathway. In the drug activation step, either one or
all the three enzymes could act on the compound to con-
vert it into the active metabolite following the mecha-
nism depicted in Scheme 1. Based on the rate of
reaction we hypothesize that first protease may be in-
volved (due to the highest rate of reaction), followed
by other enzymes. Under these circumstances, the drug
is a competitive substrate for hydrolysis between these
three broad classes of enzymes. Once the active metabo-
lite is formed, a subsequent step involves the conversion
of the above by thymidine kinase in an ATP-dependent
mechanism to inhibit the viral reverse transcriptase
(viral inhibition step, Fig. 7).
2.7. Biological relevance of bacterial enzymes used in the
present study
To compare our results with the bacterial enzymes39,40
to biologically relevant systems, we have conducted a
series of experiments using a variety of cells: we used five
different cell lines to examine the intracellular metabo-
lism of stampidine, namely Jurkat (T-cell line), Cos-7,
(monkey kidney cell line), RAW 264 (macrophage-like
cell line), CEM (T-leukemia cell line), and LL2 (mouse
lung cancer cell line). Cells in phosphate buffer were
treated with stampidine and aliquots of the supernatant
were collected at various time intervals and analyzed by
HPLC. A guard column was used in order to avoid any
protein or particulate from entering the column. In
15 min, treatmentof RAW 264 cells with stampidine
resulted in the formation of d4T and p-bromophenol.
In 30 min, we observed an increase in the amount of
d4T and p-bromophenol; however, one of the isomers
was still present in intact form (Fig. 8). Similar results
were observed with Cos-7 cells. Furthermore, one of
the isomers of stampidine was found to get hydrolyzed
preferentially, implying that similar enzymes in RAW
264 and Cos-7 cell lines may be responsible for the chiral
selectivity of hydrolysis. (Table 8).
After having established that proteases and other bacte-
rial enzymes could be involved in conversion of these
phosphoramidate derivatives of stavudine to the active
metabolite, the next question was to determine whether
cellular or viral proteases could also result in drug acti-
vation. To address this, we conducted experiments that
involved the treatment of stampidine with commercially
available HIV-1 protease, recombinant, expressed in
Escherichia coli suitable for cleaving the HIV substrate.
We examined the reaction between this protease and
stampidine by HPLC and monitored the reaction over
time. We found that there was no reaction between the
para-bromo-substituted phosphoramidate derivative of
stavudine and this protease under our experimental
conditions. This could possibly be explained due to the
For the next set of experiments, we used Jurkat cells.
Treatment of Jurkat cells with stampidine resulted in

T. K. Venkatachalam et al. / Bioorg. Med. Chem. 13 (2005) 5408–5423
5417
Figure 7. Conversion of stampidine into active metabolites. Stampidine is thought to be a competitive substrate for hydrolysis by three broad classes
of enzymes: proteases, lipases, and carboxyesterases. The product of this drug activation step is ala-stv-monophosphate. The in vitro studies suggest
that proteases are the most active in hydrolyzing stampidine (K1 > K2, K3). Thymidine kinase then converts ala-stv-monophosphate to ala-stv-
triphosphate in an ATP dependent mechanism to inhibit the viral reverse transcriptase (viral inhibition step). Stampidine was 100-fold more potent
than stavudine and 2-fold more potent than zidovudine against primary clinical HIV-1 isolates of non-B envelope subtype originating from South
America, Asia, and sub-Saharan Africa.20,21 It is well known that both in stavudine and zidovudine, the active component is primarily due to the
formation of triphosphate. Our experimental results suggest that the potency of stampidine against these clinical isolates indirectly points out that
alanine stavudine triphosphate could be formed instead of stavudine triphosphate. Additionally, we have shown in our pharmacokinetic studies
indeed alanine stavudine monophosphate is formed in vivo studies.37,38
the intracellular hydrolysis of stampidine within 30 min.
One isomer underwent hydrolysis faster than the other
(Fig. 9). Additionally, it was evident that this isomer
was different from the isomer undergoing hydrolysis in
RAW 264 and Cos-7 cells. Thus, two different types of
enzymes each preferentially recognizing one of the iso-
mers may be operative in different cell types. The hydro-
lysis of stampidine in CEM Cells and LL2 mouse lung
cells (Figs. 10 and 11) mimicked its hydrolysis in Jurkat
cells.
capable of hydrolyzing stampidine apart from esterase.
Overall, the results demonstrated that the hydrolysis oc-
curs immediately after incubating the compound with
cells. In addition, enzyme inhibitors were found to slow
down the rate of hydrolysis. We used both protease and
esterase inhibitors to demonstrate the inhibition. In the
future, we plan to extend our studies to other enzymes
that are potentially involved in the metabolism of these
phosphoramidate derivatives by examining the hydroly-
sis profile in enzyme deficient cells.
2.8. Comparison of the rate of hydrolysis
3. Conclusion
The fastest rate of hydrolysis for stampidine was observed
in RAW 264 cells and the slowest rate of hydrolysis was
observed with CEM cells (Fig. 12). The rate of hydrolysis
of stampidine in RAW 264 cells was 3-fold faster than its
rate of hydrolysis in Cos-7 cells and 7-fold faster than its
rate of hydrolysis in Jurkat cells. Similarly the rate of
hydrolysis in RAW 264 cells was 5-fold faster than in
LLL2 cells and ꢁ15 times faster than in CEM cells.
In summary, the data support the proposal that a
change in the nucleoside structure in these phosphoram-
idate derivatives causes significant alteration in the chi-
ral selectivity, hydrolysis rate, and the antiviral activity.
4. Experimental
4.1. Compounds
Depending on the cell line utilized we observed varying
chiral selectivity in the hydrolysis of stampidine isomers.
This observation indicates that different enzymes may be
operative in the hydrolysis of stampidine in different cell
types. We also propose that various other enzymes are
All the compounds in the present study were prepared
using known literature procedures and were character-
ized using analytical techniques, as reported in our

5418
T. K. Venkatachalam et al. / Bioorg. Med. Chem. 13 (2005) 5408–5423
Figure 8. HPLC profile of extracts from RAW 264 cells treated with stampidine.
earlier publications.18–20 Lipase (Candida Antartica-B,
cat #ICR-110), protease (subtilisin Carlsberg, cat
#ICR–119), as well as protease (Carica papaya, cat
#ICR-124) were obtained from Biocatalytics Inc., Pasa-
dena, CA 91106. The solid enzymes were used as such
without further purification.
the contents were vortexed. Separately the enzyme stock
solution was prepared using 5 mg of solid lipase powder
and was transferred to a volumetric flask. To this was add-
ed 8 mL of water and the contents were shaken to dissolve
the enzyme. The reaction mixture was prepared as follows
for the kinetic study. From the stock solution of the com-
pound as mentioned above, 500 lL of the methanolic
solution of the phosphoramidate derivative was pipetted
out into another glass vial and to this 500 lL of enzyme
solution was added. The contents were shaken to form a
homogeneous solution. From this reaction mixture,
50 lL was used for HPLC analysis. The column used
was a Lichrospher RP-18 analytical column (5 lm,
4 · 250 mm). The eluent used for HPLC was water
(0.1% TFA/TEA) and acetonitrile in the ratio of 65:35.
The column was maintained at room temperature. The
flow rate was maintained at 1 mL/min, the detection
4.2. Experimental conditions for lipase study
For the kinetic study, a known amount of a phosphoram-
idate derivative was carefully weighed (5–7 mg) using an
analytical balance and transferred into a 20 mL scintilla-
tion glass vial. Using a pipette man, 3 mL of methanol was
added and the contents were vortexed for 2 min until a
homogeneous solution resulted. Using another pipet
man, 100 lL of that stock solution was transferred into
another scintillation vial along with 900 lL of water and

T. K. Venkatachalam et al. / Bioorg. Med. Chem. 13 (2005) 5408–5423
5419
Table 8. Chiral selectivity of phosphorus stereoisomers of phosphoramidate derivative stampidine with different cell treatment
Cells used
Fastest hydrolyzing isomer
Rate per hour
RAW 264 (macrophages)
Cos-7 (monkey kidney cells)
Jurkat cells
ÔRÕ isomer
ÔRÕ isomer
ÔSÕ isomer
ÔSÕ isomer
ÔSÕ isomer
4.03
1.41
0.69
0.28
0.75
CEM cells
LL2-lung cells
Tentative stereochemistry assignment based on previous publication.29
Figure 9. HPLC profile of extracts from Jurkat cells treated with stampidine.

5420
T. K. Venkatachalam et al. / Bioorg. Med. Chem. 13 (2005) 5408–5423
Figure 10. HPLC profile of extracts from CEM cells treated with stampidine.
wavelength was adjusted to 265 nm, and the reference
wavelength was kept at 400 nm. Aliquot of the sample
was drawn at various intervals of time from the reaction
vial and analyzed. For fast reactions, two HPLC instru-
ments were used simultaneously to obtain the rates.
Hydrolysis rates were determined by fitting single expo-
nential decay equations to the disappearance of each iso-
mer substrate in the presence of enzyme.
an analytical balance and transferred into a scintillation
glass vial. Using a pipette man, 3 mL of methanol was
added and the contents were vortexed for 2 min until a
homogeneous solution resulted. Using another pipet-
man, 100 lL of the compound stock solution was trans-
ferred into another scintillation vial and to this was
added 900 lL of water and the contents were vortexed.
Separately the enzyme stock solution was prepared
using 5.8 mg of solid protease powder and transferred
to a volumetric flask, diluted with 8 mL of water, and
the contents were shaken to dissolve the enzyme. The
reaction mixture was prepared as follows for the kinetic
study. From the stock solution of the compound 150 lL
4.3. Experimental conditions for protease study
For the kinetic study, a known amount of the phospho-
ramidate derivative was carefully weighed (3 mg) using

T. K. Venkatachalam et al. / Bioorg. Med. Chem. 13 (2005) 5408–5423
5421
Figure 11. HPLC profile of extracts from LL2 mouse lung cells treated with stampidine.
was pipetted into another glass vial. To this, 150 lL of
protease stock solution was added and the contents were
shaken to form a homogeneous solution. From this
reaction mixture, 20 lL was used for HPLC analysis.
vial. To this, was added 40 lL of water followed by
10 lL of esterase enzyme solution.
4.5. Estimation of products
4.4. Experimental conditions for esterase study
The amount of products observed during the reactions
was estimated from the area obtained from the HPLC
profiles. In addition, authentic samples of the products
when possible were run to identify the peaks observed
during the reaction. The rate of reaction was computed
by using first order rate constants and an average of eight
to nine time points were used for this estimate. The rate
constants reported refer to rate per hour, since some of
the reactions were too slow to obtain meaningful results.
For the esterase studies, 10 lL of esterase solution was
transferred into a vial and 40 lL of water was added
and the contents were vortexed for 2 min. The sample
was prepared using a similar procedure as reported in
other studies (3 mg/3 mL methanol, 100 lL taken, and
diluted with 900 lL of water). For the kinetic study,
40 lL of the diluted compound was transferred into a

5422
T. K. Venkatachalam et al. / Bioorg. Med. Chem. 13 (2005) 5408–5423
Figure 12. Rate of hydrolysis of stampidine treated with various cells: 1, CEM; 2, Jurkat; 3, Cos-7; 4, LL2 Lung; 5, RAW 264.
4.6. Lymphocyte experimental protocol
4.7. In vitro assays of anti-HIV activity
The cells were propagated in RPMI 1640 medium
with 2 mM L-glutamine adjusted to contain 1.5 g/L
sodium bicarbonate, 4.5 g/L glucose, 10 mM HEPES,
and 1.0 mM sodium pyruvate, 90% and fetal bovine
serum, 10%, at 37 ꢁC in a humidified atmosphere
containing 5% CO2. The cell line was obtained from
the Cell Biology Laboratory of the Parker Hughes
Cancer Center. To obtain these cells, 30 million fro-
zen cells were plated in a T-150 flask with 20 mL of
medium and incubated at 37 ꢁC for 48 h. The cells
were collected in a 50 mL conical tube and centri-
fuged at 1000 rpm for 5 min, the supernatant decant-
ed off and the cells were re-suspended in 90 mL of
medium and grown further in three T-150 flasks with
30 mL each at 37 ꢁC for another 48 h to give 120
million cells. The cells were collected by centrifuga-
tion at 1000 rpm for 5 min, subsequently washed
thrice in 1· PBS and then resuspended in 1 mL of
1· PBS.
Normal human peripheral blood mononuclear cells
(PBMNC) from HIV-negative donors (HTLV IIIB
strain) were cultured 72 h in RPMI 1640 supplemented
with 20% (v/v) heat-inactivated fetal bovine serum
(FBS), 3% interleukin-2, 2 mM L-glutamine, 25 mM
HEPES, 2 g/L NaHCO3, 50 mg/mL gentamicin, and
4 mg/mL phytohemagglutinin prior to exposure to
HIV-1 at a multiplicity of infection (MOI) of 0.1 dur-
ing a 1 h adsorption period at 37 ꢁC in a humidified 5%
CO2 atmosphere. Subsequently, cells were cultured in
96-well microtiter plates (100 mL/well; 2 · 106 cells/
mL) in the presence of various concentrations of phos-
phoramidates and aliquots of culture supernatants
were removed from the wells on the seventh day after
infection for p24 antigen assays, as previously de-
scribed.36 The applied p24 enzyme immunoassay
(EIA) was unmodified kinetic assay commercially avail-
able from Coulter Corporation/Immunotech., Inc.
(Westbrooke, ME), which utilizes a murine mAb to
HIV core protein coated onto microwell strips to which
the antigen present in the test culture supernatant sam-
ples binds. Percent viral inhibition was calculated by
comparing the p24 values from untreated infected cells
(i.e., virus controls).
For the HPLC measurements, 0.5 mL of the suspended
cell mixture in PBS buffer was taken, treated with
stampidine, and incubated at 37 ꢁC. At various time
intervals, an aliquot of the supernatant was drawn and
assayed using HPLC. Authentic samples were used to
identify the products during cell mediated hydrolysis
of stampidine. A lichrospher-RP-18 column (5 lm,
4 · 250 mm) was used for this purpose and the eluent
was a mixture consisting of 65% water containing
0.1% TEA and TFA, and 35% of acetonitrile. The flow
rate was maintained at 1 mL per minute and the column
was maintained at room temperature throughout the
analysis. The rate of reaction was calculated using a first
order kinetic equation.
References and notes
1. Mansuri, M. M.; Starrett, J. E., Jr.; Ghazzouli, I.;
Hitchcock, M. J. M.; Sterzycki, R. Z.; Brankovan, V.;
Lin, T. S.; August, E. M.; Prussoff, W. H.; Sommadossi, J.
P.; Martin, J. C. J. Med. Chem. 1989, 32, 462.
2. Horwitz, J. P.; Chua, J.; DaRooge, M. A.; Noel, M.;
Klindt, I. L. J. Org. Chem. 1996, 31, 205.

T. K. Venkatachalam et al. / Bioorg. Med. Chem. 13 (2005) 5408–5423
5423
3. Bell, F. W.; Cantrell, A. S.; Hogberg, M.; Jaskunas, S. R.;
Johansson, N. G.; Jordan, C. L.; Kinnick, M. D.; Lind, P.;
Morin, J. M., Jr.; Noreen, R.; Oberg, B.; Palkowitz, J. A.;
Parrish, C. A.; Pranc, P.; Sahlberg, C.; Ternansky, R. T.;
Vasileff, R. T.; Vrang, L.; West, S. J.; Zhang, H.; Zhou, X.
X. J. Med. Chem. 1995, 38, 4929.
4. Cantrell, A. S.; Engelhardt, P.; Hogberg, M.; Jaskunas, S.
R.; Johansson, N. G.; Jordan, C. L.; Kangasmetsa, J.;
Kinnick, M. D.; Lind, P.; Morin, J. M., Jr.; Muesing, M.
A.; Noreen, R.; Oberg, B.; Pranc, P.; Sahlberg, C.;
Ternansky, R. J.; Vasileff, R. T.; Vrang, L.; West, S. J.;
Zhang, H. J. Med. Chem. 1996, 39, 4261.
5. Vig, R.; Mao, C.; Venkatachalam, T. K.; Tuel-Ahlgren,
L.; Sudbeck, E. A.; Uckun, F. M. Bioorg. Med. Chem.
1998, 6, 1789.
6. Mao, C.; Vig, R.; Venkatachalam, T. K.; Sudbeck, E. A.;
Uckun, F. M. Bioorg. Med. Chem. Lett. 1998, 8, 2213.
7. Furman, P. A.; Fyfe, J. A.; St. Clair, M. H.; Weinhold, K.;
Rideout, J. L.; Freeman, G. A.; Nusinoff, L. S.; Bolognesi,
D. P.; Broder, S.; Mitsuya, H.; Barry, D. W. Proc. Natl.
Acad. Sci. U.S.A. 1986, 83, 8333.
20. Uckun, F. M.; Pendergrass, S.; Venkatachalam, T. K.;
Qazi, S.; Richman, D. Antimicrob. Agents Chemother.
2002, 46, 3613.
21. Uckun, F. M.; Qazi, S.; Pendergrass, S.; Lisowaski, E.;
Waurzyniak, B.; Chen, C. L.; Venkatachalam, T. K.
Antimicrob. Agents Chemother. 2002, 46, 3428.
22. Uckun, F. M.; Chen, C. L.; Samuel, P.; Pendergrass, S.;
Venkatachalam, T. K.; Waurzyniak, B.; Qazi, S. Antimic-
rob. Agents Chemother. 2003, 47, 1233.
23. Uckun, F.M.; Vig, R.; U.S. patent 6,030,957. 2000.
24. Uckun, F. M.; Chen, C. L.; Lisowski, E.; Mitcheltree, G.;
Venkatachalam, T. K.; Eerbeck, D.; Chen, H.; Waurzy-
niak, B. Arzneim.-Forsch. 2003, 53, 357.
25. Uckun, F. M.; Samuel, P.; Qazi, S.; Chen, C. L.;
Pendergrass, S.; Venkatachalam, T. K. Antiviral Chem.
Chemother. 2002, 13, 73.
26. Siddiqui, A. Q.; McGuican, C.; Ballatore, C.; Wedgwood,
O.; De Clercq, E.; Balzarini, J. Bioorg. Med. Chem. Lett.
1999, 9, 2555.
27. Siddiqui, A. Q.; Ballatore, C.; McGuigan, C.; De Clercq,
E.; Balzarini, J. J. Med. Chem. 1999, 42, 393.
28. Wagner, C. R.; McIntee, E. J.; Schinazi, R. F.; Abraham,
T. W. Bioorg. Med. Chem. Lett. 1995, 5, 1819.
29. Venkatachalam, T. K.; Samuel, P.; Li, P.; Mao, C.; Qazi,
S.; Uckun, F. M. Bioorg. Med. Chem., 2004, 12, 3371.
30. Cygler, M.; Grochulski, P.; Kazlauskas, R.; Schrag, J. D.;
Bouthillier, F.; Rubn, B.; Serreqi, A. N.; Gupta, A. K. J.
Am. Chem. Soc. 1989, 116, 3180.
31. Tsi, M. D.; Chang, T. T. J. Am. Chem . Soc. 1980, 102, 5416.
32. Bruzik, K.; Tsi, M. D. J. Am. Chem. Soc. 1982, 104, 863.
33. Bruzik, K.; Jiang, R. T.; Tsi, M. D. J. Am. Chem. Soc.
1983, 105, 5455.
8. Hao, Z.; Cooney, D. A.; Hartman, N. R.; Perno, C. F.;
Fridland, A.; De Vico, A. L.; Sarngadharan, M. G.;
Border, S.; Johns, D. G. Mol. Pharmacol. 1988, 34, 431.
9. Balzarini, J.; Cooney, D. A.; Dalal, M.; Kang, G. J.;
Cupp, J. E.; De Clercq, E.; Broder, S.; Johns, D. G. Mol.
Pharm. 1987, 32, 798.
10. McGuigan, C.; Cahard, D.; Sheeka, H. M.; De Clercq, E.;
Balzarini, J. J. Med. Chem. 1996, 39, 1748.
11. McGuigan, C.; Cahard, D.; Sheeka, H. M.; De Clercq, E.;
Balzarini, J. Bioorg. Med. Chem. Lett. 1996, 6, 1183.
12. Starnes, M. G.; Gheng, Y. C. J. Biol. Chem. 1987, 32, 798.
13. Hao, Z.; Cooney, D. A.; Farquhar, D.; Perno, C.
F.; Zhang, K.; Masood, R.; Wilson, Y.; Hartman,
N. R.; Balzarini, J.; Johns, D. G. Mol. Pharm. 1990,
37, 157.
34. Jiang, R. T.; Shyy, Y. J.; Tsi, M. D. Biochemistry 1984, 23,
1661.
35. Venkatachalam, T. K.; DÕCruz, O.; Uckun, F. M. Anti-
viral Chem. Chemother. 2000, 11, 31.
14. Johnson, M. A.; Ahluwalia, G.; Connelly, M. C.; Cooney,
D. A.; Broder, S.; Johns, D. G.; Fridland, A. J. Biol.
Chem. 1988, 263, 15354.
36. Uckun, F. M.; Chelstrom, L.; Tuel-Ahlgren, L.; Dibirdik,
I.; Irvin, J. D.; Langlie, M. C.; Myers, D. E. Antimicrob.
Agents Chemother. 1998, 42, 383.
15. Johnson, M. A.; Fridlan, A. Mol. Pharm. 1989, 36, 291.
16. McGuigan, C.; Pathirana, R. N.; Mahmood, N.; Devine,
K. G.; Hay, A. J. Antivir. Res. 1992, 17, 311.
17. McGuigan, C.; Pathirana, R. N.; Balzarini, J.; De Clercq,
E. J. Med. Chem. 1993, 36, 1048.
18. Vig, R.; Venkatachalam, T. K.; Uckun, F. M. Antiviral
Chem. Chemother. 1998, 9, 445.
19. Venkatachalam, T. K.; Tai, T. L.; Vig, R.; Chen, C. L.;
Jan, S. T.; Uckun, F. M. Bioorg. Med. Chem. Lett. 1998, 8,
3121.
37. Chen, C. L.; Yu, G.; Venkatachalam, T. K.; Uckun, F. M.
Drug Metab. Dispos. 2002, 30, 1523.
38. Chen, C. L.; Venkatachalam, T. K.; Zhu, Z. H.; Uckun, F.
M. Drug Metab. Dispos. 2001, 29, 1035.
39. Groves, M. R.; Taylor, M. A.; Scott, M.; Cummings, N. J.;
Pickersgill, R. W.; Jenkins, J. A. Structure 1996, 4, 1193.
40. Chen, S. C.; Bahar, I. Bioinformatics 2004, 20(Suppl. 1),
I77–I85.
41. Venkatachalam, T. K.; Samuel, P.; Qazi, S.; Uckun, F.
M.; Europ. J. Med. Chem., 2005, 40, 452.