Discovery of GS-9131: Design, synthesis and optimization of amidate prodrugs of the novel nucleoside phosphonate HIV reverse transcriptase (RT) inhibitor GS-9148

Article abstract of DOI:10.1016/j.bmc.2010.03.041

GS-9148 [(5-(6-amino-purin-9-yl)-4-fluoro-2,5-dihydro-furan-2-yloxymethyl)phosph onic acid] 4 is a novel nucleoside phosphonate HIV-1 reverse transcriptase (RT) inhibitor with a unique resistance profile toward N(t)RTI resistance mutations. To effectively

Full text of DOI:10.1016/j.bmc.2010.03.041

Bioorganic & Medicinal Chemistry 18 (2010) 3606–3617
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Discovery of GS-9131: Design, synthesis and optimization of amidate
prodrugs of the novel nucleoside phosphonate HIV reverse transcriptase
(RT) inhibitor GS-9148
*
Richard L. Mackman , Adrian S. Ray, Hon C. Hui, Lijun Zhang, Gabriel Birkus, Constantine G. Boojamra,
Manoj C. Desai, Janet L. Douglas, Ying Gao, Deborah Grant, Genevieve Laflamme, Kuei-Ying Lin,
David Y. Markevitch, Ruchika Mishra, Martin McDermott, Rowchanak Pakdaman, Oleg V. Petrakovsky,
Jennifer E. Vela, Tomas Cihlar
Departments of Medicinal Chemistry, Biology, Drug Metabolism and Clinical Virology, Gilead Sciences, 333 Lakeside Drive, Foster City, CA94404, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
GS-9148 [(5-(6-amino-purin-9-yl)-4-fluoro-2,5-dihydro-furan-2-yloxymethyl)phosphonic acid] 4 is a
novel nucleoside phosphonate HIV-1 reverse transcriptase (RT) inhibitor with a unique resistance profile
toward N(t)RTI resistance mutations. To effectively deliver 4 and its active phosphorylated metabolite 15
into target cells, a series of amidate prodrugs were designed as substrates of cathepsin A, an intracellular
lysosomal carboxypeptidase highly expressed in peripheral blood mononuclear cells (PBMCs). The eth-
ylalaninyl phosphonamidate prodrug 5 (GS-9131) demonstrated favorable cathepsin A substrate proper-
ties, in addition to favorable in vitro intestinal and hepatic stabilities. Following oral dosing (3 mg/kg) in
Received 22 January 2010
Revised 13 March 2010
Accepted 16 March 2010
Available online 27 March 2010
Keywords:
Reverse transcriptase
Prodrugs
Beagle dogs, high levels (>9.0
drug design process and leading to the nomination of 5 as a clinical candidate.
Ó 2010 Elsevier Ltd. All rights reserved.
lM) of active metabolite 15 were observed in PBMCs, validating the pro-
HIV
Cathepsin A
Phosphonates
Nucleosides
1. Introduction
has now increased life expectancy among infected patients to >20
years, in effect, transforming HIV into a manageable chronic dis-
ease. In this context, the durability of therapy can be limited by
drug resistance, side effects as a result of chronic use, and drug-
drug interactions. Some of the adverse effects observed during
chronic therapy include neuropathy, hepatotoxicity, peripheral lip-
oatrophy and lactic acidosis. The ability of the active triphosphate
metabolites of currently approved N(t)RTIs to interfere with mito-
chondrial DNA replication are likely related to some of these side
effects.4 Therefore, we embarked on a program to identify a novel
N(t)RTI with improved resistance profile, safety, and long term tol-
erability that resulted in the identification of GS-9148 (4) and its
orally bioavailable amidate prodrug GS-9131 (5).5,6
In this report a second generation, more robust and scalable
synthesis of 4 is described followed by the design, synthesis and
SAR of an extensive array of amino acid based prodrugs, designed
as substrates for lysosomal carboxypeptidase cathepsin A (Cat A).
The high activity of Cat A in lymphoid cells provided a vehicle to
preferentially ‘target’ the novel phosphonate to tissues where
HIV replicates in vivo. Prodrugs were evaluated across a panel of
stability assays to identify the optimal oral candidates from which
prodrug 5 was selected as a clinical candidate.
The worldwide burden of HIV has continued to increase over
the last decade and revised WHO estimates now place the infected
population between 30.6 and 36.1 million.1 No vaccine or cure is
available for infected patients, but tremendous advances have been
made in the development of chemotherapeutic treatments to sup-
press viral replication. These efforts have resulted in the clinical
approval of more than two dozen HIV drugs, with diverse mecha-
nisms of action, and their use in combination therapy, now known
as, highly active anti-retroviral therapy (HAART).2 The selection of
two nucleos(t)ide HIV reverse transcriptase (RT) inhibitors
(N(t)RTIs) from this portfolio of drugs has emerged as the backbone
of choice in the most effective HAART regimens.3 Two examples of
commonly prescribed N(t)RTIs are TDF 1, an ester based prodrug of
the nucleoside phosphonate TFV 2, and FTC 3, which together with
the non-nucleoside inhibitor, efavirenz, form Atripla™, a once daily
(QD) fixed dose combination regimen recommended for treatment
naïve HIV-infected patients (Fig. 1). The success of HAART therapy
* Corresponding author. Tel.: +1 650 522 5258; fax: +1 650 522 5899.
E-mail address: Richard.Mackman@gilead.com (R.L. Mackman).
0968-0896/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2010.03.041

R. L. Mackman et al. / Bioorg. Med. Chem. 18 (2010) 3606–3617
3607
10, a key intermediate that allows for the stereocontrolled intro-
duction of the hydroxymethyl phosphonomethoxy synthon, 11,
cis to the purine. Unlike the first generation route, glycal interme-
diate 10 has a 6-methoxy substituted purine, rather than 6-amino-
purine. Introducing, the amino group early in the first generation
route, ultimately required multiple protection and deprotections
with trityl, formamidine and pivaloyl groups, thus contributing a
significant number of additional steps. The 6-methoxy group
avoided all these laborious steps and could be transformed to the
6-amino group late in the synthesis. In the forward route, fluoro
sugar 6 was converted via the anomeric bromide and addition of
6-chloropurine sodium salt, into 7.7 Using precipitation, the pure
N-9 substituted b-anomer 7 was readily isolated in 54% yield from
NH2
N
N
NH2
N
N
N
N
N
N
O
O
O
O
P
HO
P
O
RO
OH
OR
F
O
4. 2'-Fd4AP, GS-9148
1. R =
O
O
2. R = H (PMPA or TFV)
NH2
N
N
NH2
N
N
the
a-anomer and the regioisomeric N-7 isomers. Alternative 6-
O
P
F
N
O
O
aminopurine precursors, for example, 6-NHBz, 6-Br, and 6-SMe
purines were considered but none were found to be superior in
yield and isolation ease to 6-chloropurine. The benzoyl esters were
removed by treatment with potassium carbonate in methanol,
which also introduced the 6-methoxy group to provide nucleoside
8. Selective oxidation of the 50-hydroxyl was efficiently carried out
in one step using Jones reagent to generate acid 9 which was read-
ily recrystallized to remove all chromium impurities. Glycal forma-
tion at elevated temperatures using dimethylformamidine
dineopentyl acetal provided the desired intermediate 10 but only
in low yield.8 The high temperature required was well suited to
the synthesis of the non-fluorinated analog but the presence of
the 2-fluorine group in 10 led to elimination of HF. Milder methods
for effecting the decarboxylative dehydration were therefore ex-
plored. Activation under Mitsunobu conditions proved to be the
most effective at lower temperatures and allowed the glycal 10
to be formed in high yield.9 In the first generation synthesis phenyl
selenium chloride was used to activate the glycal for nucleophilic
addition of the phosphonate synthon 11, but in the interest of sca-
EtO2C
N
H
O
N
O
OH
O
F
S
3. FTC
5. GS-9131
Figure 1. N(t)RTI and NRTI inhibitors of HIV reverse transcriptase.
2. Chemistry
The first generation synthesis of 4 was derived from an earlier
route reported for the non-fluoro analog, but yielded only small
amounts of material over 16 steps.6 Clearly, prodrug development
of 4, especially in vivo optimization would require a more robust
and scalable synthesis. The second generation synthesis is shown
in Scheme 1 and includes some key distinctions from the first gen-
eration route. Commercially available fluoro sugar 6 was chosen as
the ideal starting material since it was readily available in kilogram
amounts. The route to 4 funnels through the glycal intermediate
OMe
OMe
Cl
N
N
N
N
N
N
N
F
O
N
O
a, b
OBz
F
O
c
N
F
O
d
N
F
N
O
BzO
BzO
HO
N
BzO
BzO
HO
HO
HO
9
7
8
6
OMe
N
N
OMe
N
N
N
N
OMe
N
N
N
f
g
O
e
N
N
O
Et2O3P
O
O
N
F
Et2O3P
O
F
HO PO3Et2
F
I
11
13
12
10
NH2
N
N
NH2
N
N
NH2
N
N
N
N
N
N
N
N
O
O
P
h
O
i
j
O
O
P
R
O
O
O
O
P
1 R2
EtO
RO
OH
OR
F
F
F
14
4 R = Na
15 R1=OH, R2= OP2O6H3
16 R1=R2=OCH2OC(O)OiPr (POC)
17 a-l (R1=R2=amino acid ester)
18 a-w (R1=OPh, R2=amino acid ester)
19 a-b (R1=OCH2CF3, R2=amino acid ester)
Scheme 1. Second generation synthesis of 4 and its prodrugs. Reagents and conditions: (a) 33% HBr/AcOH, CH2Cl2, quant.; (b) 60% NaH, CH3CN, 6-chloropurine, 54%; (c)
K2CO3, MeOH, 84%; (d) acetone, Celite545, Jones’ reagent, 67%; (e) PPh3, DIAD, CH2Cl2; (f) IBr, HOCH2PO3Et2, CH2Cl2, 55% two steps; (g) NaClO, MeOH, AcOH, 88%; (h) NH4OH
(aq), 110 °C; (i) TMSBr, 2,6-lutidine, CH3CN, 68%; (j) For 15 see Ref. 11a; for 16, N-methyl pyrrolidinone, TEA, ClCH2OCO2iPr (9 equiv), 7%; for 17a–l, amino ester, Et3N, 2,20-
dithiodipyridine, PPh3, pyridine, 11–73%; for 18a–w, amino ester, PhOH, Et3N, 2,20-dithiodipyridine, PPh3, pyridine, 13–52%; for 19a–b, amino ester, CF3CH2OH, Et3N, 2,20-
dithiodipyridine, PPh3, pyridine, 32–33%.

3608
R. L. Mackman et al. / Bioorg. Med. Chem. 18 (2010) 3606–3617
lability, less toxic alternatives were tried. Phenylsulfide activation
led to multiple products of the desired mass by LC–MS which
was interpreted to be the result of unselective facial activation of
the glycal due to the smaller sulfur reagent compared to selenium.
Iodide activation however, resulted in the observation of only one
product with the desired mass. Phosphonate 12 was obtained in
gram scale and a robust 55% yield when the intermediate glycal
10 was treated with IBr followed by the phosphonate synthon
11. Unsaturated intermediate 13 was obtained in 88% yield using
buffered sodium hypochlorite oxidation of the iodine, followed
by elimination.10 Triphenylphosphine oxide traces were removed
by dissolution of the product in EtOAc followed by filtration. With
all the essential groups of the target molecule introduced the
remaining steps involved conversion of the 6-OMe group to ade-
nine and removal of the phosphonate esters. Treatment of 13 with
concentrated ammonium hydroxide generated 6-aminopurine 14,
and removed one of the phosphonate esters. Subsequent dealkyla-
tion using TMSBr in CH3CN then afforded diacid 4 isolated as the
sodium salt in an overall yield of ꢀ10% over eight steps.
(MTC50 = 3.6 l
M).5a,6,11 The 20-fluoro analog was rationally de-
signed as a more selective RT inhibitor based on structural differ-
ences between HIV RT and a molecular model of mitochondrial
DNA polymerase c, in addition to literature reports demonstrating
reduced mitochondrial toxicity for 20-F modified nucleoside ana-
logs.12 Upon synthesis and evaluation, the fluoro analog 4 reduced
mitochondrial cytotoxicity in HepG2 cells (MTC50 >300
the inhibition of mitochondrial DNA polymerase by the active
metabolite 15 exceeded 100
M.5,6 The inhibition of wild type RT
by 15 was only marginally reduced by adding the fluorine leading
to a selectivity of >130-fold over mitochondrial DNA polymerase
lM) and
c
l
c
inhibition. Finally, extensive resistance profiling toward many clin-
ical isolates, demonstrated that analog 4 had a unique and promis-
ing resistance profile compared to the FDA approved N(t)RTIs.5a
Thus, the favorable RT inhibition, selectivity, and resistance profile
of 4 led to its selection for prodrug optimization to develop an or-
ally bioavailable RT inhibitor.
A series of phosphonamidate and phosphonic diamide prodrugs
were designed on the parent molecule to mask the diacid group, a
strategy that had been developed for phosphate prodrugs and was
recently applied to phosphonate 2 with considerable success.13,14
Phosphonate monoamidate 24 (GS-7340), a second generation pro-
drug of 2, was developed to improve systemic prodrug circulation
and undergo selective cleavage inside target lymphatic cells
(Fig. 2). In beagle dogs, oral dosing of 24, improved the PBMC-to-
plasma AUC0–24 ratio of 2 by 34-fold compared to that of bisPOC
prodrug 1.14 Furthermore, in pivotal phase 1 studies, 24 achieved
proof-of-concept for enhanced delivery of 2 to PBMCs in patients.
Recently, 24 was reported to be an excellent substrate for lyso-
somal carboxypeptidase cathepsin A (Cat A) which initiates the
intracellular breakdown of the prodrug inside PBMCs (Fig. 2).15
Once the initial ester is cleaved to acid 25 the subsequent break-
down is proposed to go through an cyclized intermediate 26 by
displacement of one phosphorous leaving group, in this case, phe-
nol. The cyclic intermediate is then hydrolyzed and the remaining
P–N bond cleaved to release the phosphonic diacid. Histidine triad
nucleotide binding proteins (Hints) have been implicated in the
cleavage of the P–N bond of phosphoramidate prodrugs.16 These
proteins or similar proteins may have a role in the phosphonate
P–N bond cleavage but this late step in the phosphonate prodrug
breakdown has not been elucidated yet. Prodrug design for lead
compound 4 centered on targeting cleavage by the initial enzyme
in the pathway, Cat A to preferentially deliver 4 to the PBMCs.
The prodrugs were initially evaluated in vitro by measuring pro-
drug cleavage rates under exposure to recombinant Cat A, cleavage
rates in human PBMC extracts, and also antiviral activity and tox-
icity in MT-2 cells (Table 1). Importantly, a positive correlation be-
tween Cat A activity and hPBMC extract activity (R2 = 0.88 for all
values represented in Table 1) was found for the prodrugs and pro-
vides support for Cat A being the dominant enzyme involved in the
initial breakdown of the prodrugs inside PBMCs. Table 1 also shows
that although many of the prodrugs demonstrated significantly
Diphosphate 15 and bisPOC prodrug 16 were prepared from the
diacid 4 using reported procedures.11a Phosphonic diamide (often
referred to as bisamidate) prodrugs 17a–l were also prepared from
4 using 2,20-dithiodipyridine and triphenylphosphine activation
followed by treatment with the desired amino acid ester. Monoam-
idate prodrugs were prepared by utilizing a mixture of phenol and
amino acid ester in the case of 18a–w, or trifluoroethanol and ami-
no acid ester in the case of 19a–b. The monoamidate methods all
afforded mixtures of isomers at phosphorus which, based on pro-
ton and phosphorus NMR analysis, were approximately 1:1. Sepa-
ration of the diastereoisomers could be achieved using chiral HPLC
chromatography as described in the experimental methods. The
chiral separation of mixture 18v resulted in the isolation of
18v(B) (5) as the second eluting isomer.
An alternative method for generation of the phenol based
monoamidate mixtures is shown in Scheme 2. In this alternative
procedure, diphenyl hydroxymethyl phosphonate synthon 20 re-
placed the previous diethyl phosphonate 11. Upon addition of 20
to the same glycal intermediate, diphenyl phosphonate 21 was ob-
tained. Oxidation of iodine and elimination of HI, followed by con-
version of the 6-methoxy to 6-aminopurine, provided the
monoacid 22 suitable for the direct incorporation of amino acid es-
ters to provide, for example, 18v as an isomeric mixture that con-
tains 5. Lactate prodrug 23 was prepared from 22 by substituting
the lactate ester for the amino acid ester.
2.1. Antiviral activity and cathepsin A cleavage of prodrugs
The 20-fluoro analog 4 was discovered following broad SAR
studies on a range of nucleoside phosphonates analogs.6,11 Earlier
reports had shown that the non-fluorinated analog of 4, although
an effective inhibitor of HIV RT with a promising resistance profile,
interfered with host mitochondrial DNA replication in HepG2 cells
NH2
N
N
OMe
N
N
NH2
N
N
N
N
N
O
P
OPh
O
O
O
O
N
O
N
F
Ph2O3P
O
O
P
N
R1
HO
c, d
e
a, b
OPh
9
F
I
F
OH
Ph2O3P
20
21
22
18v R1=(S)-NH-CH(CH3)CO2Et
23 R1=(S)-OCH(CH3)CO2Et
Scheme 2. Alternative synthesis of monoamidate prodrugs and lactate prodrug 23. Reagents and conditions: (a) PPh3, DIAD, CH2Cl2; (b) IBr, HOCH2PO3Ph2, CH2Cl2; (c) NaClO,
MeOH, AcOH; (d) NH4OH (aq); (e) for 18v, (
19%.
L
)-alanine amino acid ethyl esterꢁHCl, DMF, PyBOP, TEA, 30%; for 23, 2-(S)-hydroxypropionic acid ethyl ester, DMF, PyBOP, TEA,

R. L. Mackman et al. / Bioorg. Med. Chem. 18 (2010) 3606–3617
3609
NH2
N
N
NH2
N
N
NH2
N
N
N
N
N
N
N
N
O
P
Cathepsin A
O
O
P
O
O
O
O
P
O
HO2C
N
H
iPrO2C
N
H
O
NH
O
25
26
24
NH2
N
N
NH2
N
N
NH2
N
N
N
N
N
N
N
O
O
N
O
host kinases
O
P
OH
O
unknown
O
P
O
P
HO
O
P
P
HO
N
H
O
O
HO
OH
OH
O
OH
OH
2
27
Active intracellular
metabolite
Figure 2. Proposed amidate prodrug breakdown pathway utilizing cathepsin A. Cathepsin A cleaves the isopropyl ester to acid 25, which in turn cyclizes to 26 releasing
phenol. Hydrolysis of the P–O bond to 27 is then followed by cleavage of the P–N bond to generate 2.
improved antiviral activity compared to the parent 4, cell based
toxicities in MT-2 cells were mostly in the high micromolar range.
One isomer of the monoamidate ethyl ester mixture, 18v(B) (5), la-
ter chosen as the optimal prodrug, demonstrated good selectivity
ter group was linear, branched non-proximally, or proximally
branched in a small ring, the natural -configured amino acids
L
Ala or Phe provided sufficient cleavage by Cat A to result in antivi-
ral activities below 100 nM, more than two orders of magnitude
improved over the diacid 4. Examples 19a and 19b are interesting
analogs in which the phenol is replaced with trifluoroethoxy
group. Cat A breaks down the trifluoroethoxy prodrug 19a with
similar efficiency to the equivalent phenol analog 18d yet has
20-fold reduced potency in the antiviral assay? This difference in
antiviral activity, most likely, reflects differences in subsequent
steps of the prodrug breakdown, perhaps displacement of the tri-
fluoroethyl group by the amino acid in formation of the cyclic
intermediate (Fig. 2).
The more promising monoamidate mixtures, examples 18r
(Ala-cBu), and 18v (Ala-Et), were separated using chiral chroma-
tography and established that the second eluting isomer (B) was
in all cases the better substrate for Cat A resulting in greater anti-
viral activity (Table 1). To establish absolute configuration, purified
isomer 18v(A) was crystallized and the X-ray structure determined
(see Supplementary data). The prodrug 18v(A) has P(R) stereo-
chemistry so by inference the preferred Cat A substrate was the
P(S) isomer 18v(B) (5). The P(S) stereochemistry is the same stereo-
chemistry at phosphorous reported for monoamidate 24 confirm-
in MT-2 cells (CC50 >100 lM, Table 1).
The diamide prodrugs 17a–l were explored extensively in the
early stages of the optimization process since they avoided the
generation of a phosphorous chiral centre. Lipophilic phenylala-
nine (Phe) ester prodrugs 17a–c were the best substrates for Cat
A, superior in some cases to 24! Consistent with the high cleavage
rate these prodrugs demonstrated potent antiviral activity below
10 nM, >10-fold more potent than the bisPOC prodrug 16. The less
hydrophobic alanine (Ala) esters 17g–j were weaker Cat A sub-
strates, but nevertheless, 17g, for example, still demonstrated im-
proved antiviral activity over 16. Unnatural amino acid, 2-L-
aminobutyric acid (ABA) analogs 17d–f, and glycine (Gly) analogs,
for example, 17k–l were the poorest substrates.
A large array of monoamidate prodrugs (diastereomeric mix-
tures) bearing phenol, 18a–w, or trifluoroethoxy, 19a–b, and one
amino acid ester were also evaluated. Comparing four different iso-
butyl esters 18d, 18k, 18q and 18w established a similar trend to
the diamides in that the Phe and Ala based prodrugs are superior
Cat A substrates than ABA and Gly, leading to more potent antiviral
activity. Similarly, the more lipophilic Phe esters 18a–i had the
highest Cat A activities and the most potent antiviral activity
(EC50 range 3.8–199 nM) compared to the Ala esters 18l–v. Proxi-
mal branching of the ester group in the Phe series, examples
18a–b (sec-butyl(S) and sec-butyl(R), respectively) and 18f (iPr)
led to significantly reduced Cat A cleavage rates and reduced anti-
viral activities relative to the linear n-alkyl esters. In the alanine
series the proximally branched 3-pentyl ester 18l and isopropyl es-
ter 18s also had very poor Cat A cleavage rates. The fact that the
alanine isopropyl ester was a poor Cat A substrate whereas the
same prodrug 24 (albeit a single isomer) is an excellent substrate,
suggests that the parent nucleoside phosphonate is an important
recognition aspect for Cat A. The purpose for exploring proximally
branched esters was to try and modify the balance between Cat A
activity and in vitro hepatic stability (see next section). Given the
poor Cat A activity, constrained, proximally branched esters, for
example, 18r (cyclo-butyl) were attempted and found to be sur-
prisingly good substrates. The lactate ester 20 was not a good sub-
strate for Cat A and established that an amino acid is preferred for
at least one of the two prodrug groups. In general, provided the es-
ing the ability of Cat
A to preferentially cleave a single
phosphorus isomer.17
2.2. In vitro stability profiling of amidate prodrugs
Plasma stability, hepatocyte stability and intestinal S9 stability
were evaluated in both dog and human (Table 1) fractions to deter-
mine the best prodrugs for in vivo oral evaluation. Rodents were
avoided due to their high plasma esterase activities. Tenofovir
monoamidate 24 served as a good comparitor prodrug for profiling
both in vitro and in vivo stability since the clinical phase 1 studies
had demonstrated good oral loading of PBMCs using this prodrug.
Clearly, the goal for amidate prodrugs of 4 was to try and match, or
exceed the stability properties exhibited by 24. It was found that
all amidate prodrugs of 4 demonstrated good plasma stability over
60 min in both dog and human species (Table 1). The best Cat A
substrates in the diamide series of prodrugs, 17a–l demonstrated
low hepatic stability, and in general a correlation between high
Cat A activity and low hepatic stability was observed throughout
all the prodrugs tested. Therefore, designing the optimal prodrug

3610
R. L. Mackman et al. / Bioorg. Med. Chem. 18 (2010) 3606–3617
Table 1
NH2
In vitro stability profiles of amidate prodrugsa
N
N
N
O
P
O
O
N
R
1 R2
F
b
b
Compd
R1
H
R2
WT HIV
EC50 (nM)
MT-2
CC50 (nM)
c log Dc Human Human PBMC Dog/human
Cat Ad extractd
Plasma% remain
Dog/human
hepatocytes T1/2 (min) intestinal S9 T1/2 (min)
Dog/human
e
4
2
16
H
10,600
13 (9)
105 (33)
>100,000
>50,000
—
4800
—
—
>100,000
—
—
51,000
>100,000
—
—
ꢂ5.7
0.98
1.80
6.49
5.47
4.45
4.30
3.87
3.33
3.28
2.85
2.26
1.24
1.78
1.24
—
—
—
—
—
—
—/—
—/—
—/—
—/—
—/—
—/—
—/—
—/—
—/—
—/—
—/—
—/—
—/—
—/—
—/—
—/—
—/—
—/—
—/—
—/—
POC
POC
POC
POC
—/—
10/—
—/—
17a
17b
17c
17d
17e
17f
17g
17h
17i
17j
17k
17l
Phe-nBu
Phe-nPr
Phe-Et
ABA-nBu
ABA-iBu
ABA-iPr
Ala-nBu
Ala-iBu
Ala-nPr
Ala-Et
Phe-nBu 8.4 (5.6)
Phe-nPr 5.2 (1.5)
Phe-Et
ABA-nBu 90 (28)
ABA-iBu 128 (38)
ABA-iPr
66,680 50
27,642 20
41,020 40
1872
867
0
17,355 18
1770
—
96/99
99/100
95/100
100/99
—/—
100/100
100/96
100/100
—/—
8.9 (4.9)
—/—
1.3
1.3
0.05
73/—
71/—
690/—
48/—
193/—
—/—
>2000
Ala-nBu 68 (32)
Ala-iBu
Ala-nPr
Ala-Et
Gly-iBu
Gly-iPr
119 (30)
1065 (21)
723 (258)
1610
2.1
—
—
0
—
0
—
—/—
—/—
—/—
—/—
—/—
—/—
Gly-iBu
Gly-iPr
>100,000
—
>2000
—
24
Ala-iPr (B)
OPh
9.2 (7.9)
139,000
2.16
35,000 20
100/—
103/159
15/106
18a
18b
18c
18d
18e
18f
18g
18h
18i
19a
18j
18k
18l
18m
18n
18o
18p
18q
18r
18s
18t
18u
18v
19b
23
Phe-sBu(S)
Phe-sBu(R)
Phe-CH2cBu OPh
Phe-iBu
Phe-cBu
Phe-iPr
Phe-nPr
Phe-CH2cPr OPh
Phe-Et
Phe-iBu
ABA-nBu
ABA-iBu
OPh
OPh
199 (2)
141 (7)
3.8 (2.2)
6.0 (3.8)
20 (8)
63 (32)
4.9 (0.5)
7.5 (3.7)
33 (11)
>100,000
>100,000
46,500
5.10
5.10
4.92
4.86
4.65
4.59
4.56
4.42
4.05
4.05
3.98
3.76
4.00
3.98
3.55
3.47
3.32
3.25
3.05
2.98
2.96
2.82
2.45
1.64
2.77
2.72
504
1510
37,017 24
29,655 31
30,786 25
0.67
2.1
84/83
84/74
80/51
84/52
90/64
100/85
98/65
85/63
78/74
84/91
95/100
95/100
—/—
79/91
58/54
100/89
71/69
93/81
87/84
78/97
100/96
64/69
84/88
99/95
—/—
79/—
56/—
23/62
28/—
27/61
98/720
14/37
17/40
37/78
27/—
91/—
127/—
—/—
—/—
—/—
3.4/—
8.3/—
5.4/—
—/—
—/—
—/—
—/—
—/—
—/—
—/—
—/—
—/—
—/—
—/—
63/143
—/—
25/122
—/—
OPh
OPh
OPh
OPh
38,300
>100,000
>100,000
>100,000
>100,000
25,000
>100,000
70,000
>100,000
8350
27,000
>100,000
91,000
>100,000
>100,000
57,000
1417
2.8
33,250 21
39,270 41
39,246 35
30,107 22
1377
728
OPh
OCH2CF3 119 (106)
OPh
OPh
88 (17)
103 (18)
662 (214)
33 (3.3)
63 (22)
92 (21)
25 (12)
58 (45)
36 (13)
728 (214)
52 (71)
59 (34)
96 (64)
1.4
1.9
0
6.7
3.0
7.3
3.0
4.0
Ala-3-Pentyl OPh
Ala-nPentyl OPh
Ala-cPentyl
Ala-nBu
0
5088
5688
6183
5311
7916
18,150 7.7
216 0.3
19,103 26
28,235 28
7280
1132
—/—
226
57/—
—/—
63/—
OPh
OPh
Ala-CH2cBu OPh
132/92
38/86
100/98
88/127
88/—
85/69
165/189
86/185
—/—
Ala-iBu
Ala-cBu
Ala-iPr
Ala-nPr
Ala-CH2cPr
Ala-Et
Ala-Et
Ala(O)-Et
Gly-iBu
OPh
OPh
OPh
OPh
OPh
OPh
—
31,000
>100,000
74,000
—/—
50/98
146/318
—/—
—/—
—/—
7.2
OCH2CF3 1050 (212) >100,000
OPh
OPh
0.69
—/—
0.72
>2000
1120
—
22,000
18w
—/—
—/—
18r(A)
18r(B)
18v(A)
Ala-cBu(A)f
Ala-cBu(B)f
Ala-Et(A)f
OPh
OPh
OPh
OPh
45 (16)
—
—
3.05
3.05
2.45
2.45
2527
23,997 19
645
12,371 5.8
3.2
80/84
60/82
57/94
56/84
140/—
52/—
165/—
71/43
32/—
3.9/—
108/—
19/—
21 (4.9)
262 (94)
157 (23)
>100,000
>100,000
0.87
18v(B)(5) Ala-Et(B)f
a
Values for active compounds are results of at least two experiments in MT-2 cells, standard deviation is given in parentheses.
Disoproxil (POC); phenylalanine (Phe); 2-(L)-aminobutyric acid (ABA); alanine (Ala); 2-(L)-hydroxypropionic acid (Ala(O)); glycine (Gly); cyclopentyl (cPentyl); n-butyl
b
(nBu); n-propyl (nPr); ethyl (Et); iso-butyl (iBu); iso-propyl (iPr); (S)-sec-butyl (sBu(S)); (R)-sec-butyl (sBu(R)); cyclobutyl (cBu); cyclopropyl (cPr); phenol (OPh).
c
c log D was calculated using Pallas 3.0 from CompuDrug International, Inc., Sedona, AZ.
d
Activity is represented as pmol compound cleaved/min/
% prodrug remaining after 60 min incubation.
lg of recombinant Cat A enzyme or PBMC extract.
e
f
A refers to first eluting isomer and B the second eluting isomer from the chiral chromatography method described in the experimental.
required balancing rapid Cat A mediated cleavage, with broader
esterase stabilities in hepatocytes and intestinal fractions. For
example, 17f (ABA-iPr) had very high hepatic stability but no
cleavage by Cat A leading to poor antiviral activity. In contrast,
17g (Ala-nBu) showed promising Cat A mediated cleavage result-
ing in potent antiviral activity, and hepatic stability approximately
half that of 24 leading to its later selection for in vivo evaluation.
Amongst the Phe monoamidates (18a–i) hepatocyte stability
rarely reached that of 24 unless a proximally branched ester was
used, for example, 18f (Phe-iPr). Despite the low Cat A activity,
18f demonstrated good antiviral activity and was also selected
for in vivo evaluation. In the Ala series of monoamidates, plasma
stability and hepatocyte stability were generally improved, provid-
ing profiles that more closely resembled 24. Some of the most
promising prodrugs, for example, 18p and 18v (Ala-CH2cBu and
Ala-Et respectively) demonstrated superior dog hepatocyte stabil-
ity than 24! The desire to obtain the best possible hepatic stabili-
ties, yet also have optimal Cat A activity, drove the exploration of
the constrained aliphatic esters such as cyclobutyl and cyclopentyl.
The cyclobutyl analog 18r was remarkable in that the constraint of
the ring improved the Cat A cleavage rate substantially compared
to the isopropyl ester 18s, but hepatic stability did not decrease.

R. L. Mackman et al. / Bioorg. Med. Chem. 18 (2010) 3606–3617
3611
M)
Table 2
In vivo pharmacokinetics of selected prodrugs 17 and 18 in Beagle dogs
a
Compd
Prodrug
IV clearance (l/h/kg)
PBMC Cmax 4 (
lM)
Loading efficiencyb
Oral plasma prodrug %F
Oral PBMC Cmax 15 (l
24
17g
18d
18f
18p
18q
18r
18u
Ala-iPr
Bis Ala-nBu
Phe-iBu
Phe-iPr
Ala-CH2cBu
Ala-iBu
Ala-cBu
Ala-CH2cPr
Ala-Et
Ala-Et(A)
Ala-Et(B)
Ala-cBu(A)
Ala-cBu(B)
2.23 (0.25)
3.22 (1.17)
3.70 (0.30)
1.35 (0.31)
1.63 (0.4)
2.62 (0.54)
2.76 (0.88)
1.56 (0.63)
1.71 (0.41)
—
nd
0.58
0.05
1.94
0.10
0.30
0.16
0.62
0.48
0.47
—
21 (3)
—
2.6 (0.9)
—
—
—
—
—
—
—
—
—
—
—
0.94 (0.26)
22 (7.3)
3.0 (0.5)
9.0 (2.5)
3.8 (0.9)
21 (5.3)
18 (5.6)
17 (1.5)
—
—
6.3 (4.8)
12 (3.5)
9.8 (6.8)c
18 (13)c
7.7 (4.6)c
9.1 (4.5)c
18v
7.8 (3.0)
5.2 (1.0)
9.0 (2.3)
14 (5.4)
9.4 (4.3)
18v(A)
18v(B) (5)
18r(A)
18r(B)
—
—
—
—
—
—
—
—
—
All values are the mean SD in parentheses from 3 dogs and dose normalized.
a
Quantity of diacid 4 measured after dephosphorylation of samples with calf intestinal phosphatase.
b
c
Loading efficiency ratio of [AUC]0–24 of 4 and its metabolites inside PBMCs divided by [AUC]0–24 plasma prodrug.
h
h
Estimated based on IV parameters for isomeric mixture.
Several of the promising alanine ester prodrugs were selected for
in vivo profiling.
2.3. In vivo pharmacological profiling of amidate prodrugs
Initial in vivo profiling was performed using a 30 min IV infu-
sion dosing protocol in beagle dogs to establish plasma clearance
of the prodrug and evaluate total PBMC exposure to 4 (and its
metabolites) over 24 h (Table 2).5 These measurements provided
insight into the pharmacokinetics that would drive PBMC cell load-
ing, that is, Cat A cleavage versus hepatic and other clearance
routes of the prodrug. The ratio of PBMC cell loading of 4 over
24 h ([AUC]0–24 h of 4 and its metabolites inside PBMCs) compared
to the plasma prodrug exposure over 24 h ([AUC]0–24 h) was de-
fined as the loading efficiency and provided a useful parameter
to rate the best candidates for oral studies. It should be noted that
IV dosing of diacid 4 resulted in low levels of 4 inside PBMCs,
thereby establishing that circulating levels of 4 in the plasma fol-
lowing prodrug cleavage in sites other than PMBCs, would not con-
tribute significantly to the intracellular PBMC levels of 4 and its
metabolites (data not shown). Phosphonic diamide 17g (Ala-nBu)
had rapid clearance approaching that of liver blood flow in dogs,
a low PBMC Cmax for 4 and more than a 10-fold reduced loading
efficiency compared to 24. This result suggested that hepatic
extraction and loading of other tissues was more rapid than PBMC
cell loading. In contrast, Phe monoamidate 18d (iBu) which also
showed a high clearance rate had the highest Cmax and the highest
loading efficiency. The difference can be attributed to the faster
cleavage by Cat A, essentially redressing the balance between rapid
PBMC cell loading versus other clearance routes. Unfortunately,
first pass hepatic extraction via oral administration was extensive
for 18d, resulting in an oral bioavailability of 2.6% and minimal
intracellular delivery of 4 and its metabolites. The Phe-iPr ester
18f, showed reduced clearance, as expected based on its improved
hepatocyte stability profile, but demonstrated a poor loading effi-
ciency compared to 24 and was therefore not considered for oral
evaluation. More success was achieved in the Ala series of mono-
amidates due to their improved hepatocyte stabilities. Prodrugs,
18u, 18v, and 18r (CH2cPr, Et and cBu respectively) all demon-
strated lower IV clearance than the Phe monoamidates and compa-
rable loading efficiencies to 24. Consistent with the low clearance
and excellent in vitro intestinal S9 stability of 18v (Table 1) good
oral bioavailability was observed for 18v leading to intracellular
PBMC levels of diphosphate 15 close to 8 micromolar. The sepa-
rated diastereoisomers for both 18v and 18r were then evaluated
Figure 3. Oral plasma and PBMC pharmacokinetics of prodrug 5 in beagle dogs
(n = 3).
orally to identify 5 (18v(B)) as the most promising isomer based
on oral bioavailability and intracellular levels of active metabolite
15.
The full pharmacokinetic profile of 5 upon oral administration
to beagle dogs (3 mg/kg) is shown in Figure 3. The prodrug is
cleared from circulation in the first 1–2 h with a half life of less
than 20 min, but during this time the PBMCs are efficiently loaded
with the parent diacid 4 to levels 10-fold higher than plasma levels.
The early Cmax for 4 inside PBMCs, is consistent with the rapid pro-
drug clearance from the plasma and supports that only the plasma
prodrug is driving the intracellular levels of 4. Critically, the intra-
cellular accumulation of the active diphosphate metabolite 15,
peaks at ꢀ5 h with a concentration of >9.0
lM, a level well above
the RT IC50 of 1.9 M. Furthermore, metabolite 15 persists inside
l
the cell for periods in excess 24 h, supporting a QD dosing regimen
for the prodrug. Based on this promising pharmacokinetic data,
and the resistance and selectivity data, prodrug 5 was chosen as
a candidate for preclinical toxicological evaluation in vivo.5
3. Conclusion
The design of amidate prodrugs as substrates for the lysosomal
carboxypeptidase, cathepsin A, in combination with hepatocyte
and plasma stability testing allowed for the rapid identification
of an oral prodrug candidate of the novel nucleoside phosphonate
RT inhibitor 4. Cathepsin A recognized both phosphonic diamide
and monoamidate prodrugs, and the best substrates demonstrated
antiviral activity below 10 nM. The optimal prodrug, 5, contained
an ethyl ester of alanine and phenol, as the two masking moieties
of the phosphonic acid. Monoamidate prodrug 5 effectively loaded

3612
R. L. Mackman et al. / Bioorg. Med. Chem. 18 (2010) 3606–3617
lymphoid cells following oral dosing in dogs, and was efficiently
phosphorylated to its active metabolite 15.
79.2, 66.9, 64.9. 19F NMR (282.2 MHz, D2O) d ꢂ134.6 (m). 31P
NMR (D2O) d 12.4. LC–MS m/z 332 [M+H]+. Anal. (C10H9FN5Na2O5Pꢁ
2H2O) C, H, N.
4. Experimental and methods
4.2. 6-Chloro-9-(2-deoxy-2-fluoro-3,5-di-O-benzoyl-b-
arabinofuranosyl)-9H-purine (7)7
D-
All reactions were performed with reagent-grade materials un-
der an atmosphere of nitrogen. Solvents were reagent-grade or bet-
ter. Silica gel chromatography was performed using EMD
Chemicals Silica Gel 60. 1H, 13C, 31P and 19F NMR spectra were re-
corded on a Varian 300 MHz NMR and are reported in parts per
million (ppm) and signals reported as singlet (s), doublet (d), trip-
let (t), quartet (q), broad (br), and multiplet (m). 1H, and 13C NMR
shifts are reported relative to solvent residual. 19F NMR are re-
ported relative to trifluoroacetic acid external standard
(d = ꢂ77 ppm). 31P NMR are reported relative to phosphate ion in
D2O external standard d = 0 ppm. Low-resolution LC mass spectra
(LC–MS) were recorded on a Thermo Finnigan LCQ Advantage, elec-
trospray ionization mass spectrometer. The LC–MS was equipped
with a Phenomenex Synergi 4 micron Polar-RP C18 column
(30 ꢃ 4.6 mm), eluting with a gradient of 5–100% solvent B in sol-
vent A over 2.5 min at a flow rate of 2 mL/min (solvent A = 0.1%
AcOH in water, solvent B = 0.1% AcOH in CH3CN). High resolution
mass spectra were provided by Gilead Sciences Analytical Depart-
ment using ESI. Melting points are uncorrected. Elemental analysis
was carried out by Galbraith Laboratories Inc., Knoxville, TN. Ana-
lytical HPLC was carried out on a Agilent 1100 series analytical
A solution of 2-fluoro-1,3,5-tri-O-benzoyl-D-arabinofuranose 6
(600 g, 1.3 mol) in CH2Cl2 (2 L) was treated with 33% HBr
(1.5 equiv) in acetic acid (340 mL). The mixture was stirred at rt
for 16 h, then cooled with ice water, and slowly neutralized over
1–2 h with NaHCO3. The organic phase was separated and washed
with NaHCO3. The organic phase was dried over Na2SO4, filtered
and concentrated under reduced pressure to yield the anomeric
bromide (ꢀ540 g, 1.3 mol) as a viscous yellow oil. NMR data was
consistent with data reported.7 A suspension of NaH (56.4 g, 60%,
1.11 equiv) in CH3CN (2 L), was treated with 6-chloropurine
(208 g, 1.05 equiv) over 2–3 h. The mixture was stirred at rt for
1.5 h and then a solution of the anomeric bromide in CH3CN
(500 mL) was added over 1 h. The resultant mixture was stirred
at rt for a further 16 h. The mixture was carefully quenched with
acetic acid, and the solid collected by filtration. The solid was
washed with water (3ꢃ), MeOH (3ꢃ), and then dried under high
vacuum to provide the title product 7. The mother liquor, following
the acetic acid quench, was concentrated under reduced pressure
and allowed to stand overnight to precipitate further solid mate-
rial. The solid was washed with water (3ꢃ), MeOH (3ꢃ) and dried
under vacuum to provide a second crop of 7 (total yield, 345.5 g,
0.70 mol, 54%). 1H NMR (300 MHz, CDCl3) d 8.75 (s, 1H), 8.41 (d,
J = 2.7 Hz, 1H), 8.08 (d, J = 7.2 Hz, 4H), 7.4–7.7 (m, 6H), 6.7 (dd,
J = 22, 2.7 Hz, 1H), 5.8 (dd, J = 16.5, 2.5 Hz, 1H), 5.4 (dd, J = 2.5,
50 Hz, 1H), 4.83 (d, J = 4.6 Hz, 2H), 4.64 (dd, J = 4.6, 4.6 Hz, 1H).
13C NMR (77 MHz, CDCl3) d 166.1, 165.1, 152.2, 151.4, 151.1,
144.5, 144.4, 134.3, 133.5, 129.9, 129.7, 129.2, 128.7, 128.6,
128.0, 93.8, 91.2, 84.0, 83.7, 81.5, 77.5, 76.8, 76.4, 63.1. 19F NMR
(282.2 MHz, CDCl3) d ꢂ198.3 (m). LC–MS m/z 497 [M+H]+.
HPLC, equipped with
a DIONEX DNAPac PA-100 column
(250 ꢃ 4 mm) eluting with a gradient of 0–50% solvent B in solvent
A over 12 min followed by 100% B for 5 min (solvent A = water, sol-
vent B = 0.5 mM triethylammonium bicarbonate buffer pH 8.5.
Compounds were >95% pure at 254 nM and 214 nM for bioassay.
Preparative HPLC chromatography was carried out on a Gilson Li-
quid Handler 215 equipped with 322 pumps, a UV/Vis-156 detec-
tor and a Phenomemex Synergi Combi-HTS 4 micron hydro-RP C18
column (150 ꢃ 30 mm). The flow rate was 20 mL/min and a linear
gradient of 5–100% ACN in water was typically used. Amino acids
are abbreviated as follow glycine (Gly), alanine (Ala), phenylala-
nine (Phe), L-2-aminobutyric acid (ABA), and L-2-hydroxypropionic
4.3. 6-Methoxy-9-(2-deoxy-2-fluoro-b-D-arabinofuranosyl)-9H-
acid (Ala(O)). X-ray structure of 18v(A) was determined by Rigaku
purine (8)
Americas, TX, USA
A suspension of 7 (345.5 g, 0.70 mol) in MeOH (6 L) was trea-
ted with K2CO3 (384 g, 2.78 mol). The mixture was mechanically
stirred at rt for 13 h, and then filtered. The eluant was neutral-
ized with acetic acid to pH 7 and concentrated under reduced
pressure (not to dryness). The resultant residue was partitioned
between hexane and water. The aqueous layer was separated
and concentrated under reduced pressure to precipitate a solid
which was collected by filtration. The solid was washed with
ice water to provide 8 (167 g, 0.59 mol, 84%) as a white solid.
1H NMR (300 MHz, CD3OD) d 8.5 (s, 2H), 6.55 (dd, J = 15.6,
4.0 Hz, 1H), 5.18 (dq, J = 52, 4.0, 3.0 Hz, 1H), 4.55 (dq, J = 18,
4.5, 3.0 Hz, 1H), 4.18 (s, 3H), 4.02 (m, 1H), 3.85 (m, 2H). 13C
NMR (77 MHz, CD3OD) d 160.7, 152.1, 151.2, 142.2, 120.1, 96.7,
94.2, 84.3, 83.2, 83.0, 73.4, 73.0, 60.6, 53.5. 19F NMR
(282.2 MHz, CD3OD) d ꢂ200.0 (m). LC–MS m/z 285 [M+H]+.
Recrystallization from EtOAc, mp = 146 °C. Anal. (C11H13FN4O4)
C, H, N.
4.1. 5-[(6-Amino-purin-9-yl)-4-fluoro-2,5-dihydro-furan-2-
yloxymethyl] phosphonic acid disodium salt (4)
Compound 13 (62.4 g, 155 mmol) was treated with ammonium
hydroxide (28–30%) (150–200 mL) and vigorously agitated for
20 min to dissolve the oily material. The resultant mixture was
heated at 110–120 °C for 3 h in a steel bomb, allowed to cool and
then concentrated under reduced pressure to approximately 50–
100 mL volume. The residue was filtered through a plug of C18
to remove insoluble material and then concentrated to near dry-
ness. MeOH (200 mL) was added and the mixture filtered once
again to remove insoluble material. The eluant was concentrated
to dryness (ꢀ60 g) and the residue containing monoacid 14 dried
under high vacuum for 24 h. The resultant solid was suspended
in CH3CN (750 mL) and then treated with 2,6-lutidine (35 mL,
300 mmol) followed by TMSBr (115 g, 750 mmol, 5 equiv). The
mixture was heated for 1 h at 40 °C and then cooled to 0 °C. The
mixture was poured into cold aqueous NaHCO3 and then evapo-
rated under reduced pressure. The crude residue was purified on
4.4. 6-Methoxy-9-(2-deoxy-2-fluoro-5-carboxy-b-D-arabinofur-
anosyl)-9H-purine (9)
C-18 (40 lm) column, eluting with water. The aqueous fractions
were freeze dried to give disodium salt 4 (40 g, 106 mmol, 68%).
1H NMR (300 MHZ, D2O) d 8.22 (s, 1H), 8.02 (s 1H), 6.6 (d,
J = 2.7 Hz, 1H), 5.92 (d, J = 4.8 Hz, 1H), 5.89 (s, 1H), 3.59 (dd,
J = 12.6, 9 Hz, 1H), 3.40 (dd, J = 12.6, 9 Hz, 1H). 13C NMR (77 MHz,
D2O) d 158.1, 155.4, 154.3, 152.8, 148.6, 140.2, 118.3, 106.2, 79.6,
Compound 8 (100 g, 0.35 mol) was dissolved in acetone (1.4 L)
and treated with 200 g of vigorously agitated acetone-washed Celite
545. The mixture was cooled to 0 °C and Jones’ reagent (2.63 M chro-
mium(VI), 2 mol equiv or 1.5 redox equiv, 268 mL) was added drop-
wise with mechanical stirring over 2 h. The mixture was warmed to

R. L. Mackman et al. / Bioorg. Med. Chem. 18 (2010) 3606–3617
3613
rt and stirred for a further 12 h. The mixture was then cooled to 0 °C
4.8. 9-(R)-[5-(R)-[(Bis-isopropyloxycarbonyloxymethyl)
phosphonyl]methoxy-3-fluoro-2,5-dihydrofuran-2-yl]-9H-
adenine (16)
and the remaining chromium(VI) quenched by the addition of iPrOH
(2 equiv, 54 mL)followedbystirringfor2 h. Themixturewasfiltered
through a Celite pad and the pad washed generously with THF. The
filtrate was reduced in volume and the residue re-dissolved in
THF/EtOAc. The mixture was extracted with brine using color as
indicatorfor thecompletionof extraction. Theorganiclayer was sep-
arated, dried over anhydrous Na2SO4, filtered and concentrated un-
der reduced pressure to provide a white foam. The foam was
crystallized from MeOH:H2O to provide 9 (70 g, 0.235 mol, 67%) as
a white solid. 1H NMR (300 MHz, CD3OD) d 8.74 (d, J = 2.4 Hz, 1H),
8.5 (s, 1H), 6.76 (dd, J = 22.5, 2.4 Hz, 1H), 5.1 (d, J = 48 Hz, 1H), 4.8
(d, J = 9.6 Hz, 1H), 4.7 (s, 1H), 4.18 (s, 3H). 13C NMR (77 MHz, CD3OD)
d 171.1, 160.7, 152.1, 151.0, 143.2, 120.1, 94.1, 91.6, 86.1, 85.9, 83.6,
77.3, 76.9, 53.5. 19F NMR (282.2 MHz, CD3OD) d ꢂ199.5 (m). LC–MS
m/z 299 [M+H]+. Recrystallized from H2O, deposited as a fine flakes,
mp = 158 °C (dec.). Anal. (C11H11FN4O5) C, H, N.
Disodium salt 4 (20 mg, 0.060 mmol) was dissolved in anhy-
drous N-methyl pyrrolidinone (3 mL) and treated with triethyl-
amine (0.03 mL, 0.21 mmol). The mixture was heated at 60 °C
under nitrogen and then treated with isopropyloxycarbonyloxym-
ethylchloride (0.082 g, 0.53 mmol). The mixture was stirred at
60 °C for 3 h and then cooled, diluted with EtOAc and washed with
brine. The organic layer was dried over Na2SO4, filtered and con-
centrated under reduced pressure. The resultant residue was then
subjected to preparative HPLC (linear gradient, 5–95% CH3CN in
water) to afford compound 16 (2 mg, 7%): 1H NMR (300 MHz,
CDCl3) d 8.40 (s, 1H), 8.08 (s, 1H), 6.81 (d, J = 2.4 Hz, 1H), 5.90 (d,
J = 4.5 Hz, 1H), 5.84 (s, 1H), 5.65 (m, 6H), 4.94 (m, 2H), 4.05 (m,
2H), 1.32 (t, J = 5.4 Hz, 12H). 31P NMR (121.4 MHz, CDCl3) d 20.5.
LC–MS m/z 564 [M+H]+
4.5. Diethyl 5-(R)-[(6-methoxy-purin-9-yl)-4-(R)-fluoro-3-(R)-
iodo-tetrahydro-furan-2-(R)-yloxymethyl]phosphonic acid (12)
4.9. Phosphonic diamide example procedure: 9-(R)-[5-(R)-[[[(bis-
(S)-1-(n-propyloxycarbonyl)-ethyl]amino]phosphonyl]methoxy-
3-fluoro-2,5-dihydrofuran-2-yl]-9H-adenine (17i)
A solution of 9 (23 g, 71 mmol) in CH2Cl2 (1 L) was treated with
PPh3 (24.3 g, 92.6 mmol) and DIAD (18.8 g, 93 mmol). The resultant
mixture was refluxed for 0.5 h to generate glycal 10 and then cooled
to ꢂ40 °C. Diethyl (hydroxymethyl)phosphonate 11 (51.9 g,
309 mmol) was added followed by the dropwise addition of IBr
(31.95 g, 154 mmol) in CH2Cl2 (100 mL). The reaction mixture was
warmed to 0 °C and then stirred for 2 h. The reaction mixture was
poured into aqueous NaHCO3 (100 g) containing Na2S2O3 (30 g).
The organic phase was separated, dried over MgSO4, and concen-
trated under reduced pressure to dryness. The reaction mixture
was subjected to silica gel chromatography eluting with 0–100%
EtOAc in CH2Cl2, followed by 0–10% MeOH in CH2Cl2, to provide 12
(22.5 g, 55%). 1H NMR (300 MHz, CDCl3) d 8.5 (s, 1H), 8.21 (d,
J = 3.9 Hz, 1H), 7.05 (dd, J = 18.5, 3.9 Hz, 1H), 5.6 (s, 1H), 5.4 (dd,
J = 50, 4.0 Hz, 1H), 4.55 (d, J = 16.5 Hz, 1H), 4.05–4.2 (m, 4H), 4.15
(s, 3H), 4.03 (dd, J = 13.7, 10.8 Hz, 1H), 3.82 (dd, J = 13.7, 9.3 Hz,
1H), 1.3 (m, 6H). 13C NMR (77 MHz, CDCl3) d 166.0, 152.4, 151.7,
142.4, 142.3, 120.7, 111.7, 111.5, 96.1, 93.4, 84.0, 83.8, 62.7–63.3
(m), 61.0, 54.4, 21.3, 21.0, 16.5, 16.4. 19F NMR (282.2 MHz, CDCl3) d
ꢂ165.6 (m). 31P NMR (121.4 MHz, CDCl3) d 19.3. LC–MS m/z 531
[M+H]+. Anal. (C15H21FN4O6P) C, H, N.
Disodium salt 4 (12 mg, 0.032 mmol) and (S)-alanine-O-n-Pr es-
ter hydrochloride (32 mg, 6 equiv, 0.192 mmol) were mixed in
anhydrous pyridine (1 mL). Et3N (53 lL, 0.384 mmol) was added
and the resultant mixture stirred at 60 °C under N2. In a separate
flask, 2,20-dithiodipyridine (59 mg, 0.224 mmol) and PPh3 (49 mg,
0.224 mmol) were dissolved in anhydrous pyridine (0.5 mL) and
the resultant yellow solution was stirred for 15–20 min. The solu-
tion was then added to the 60 °C solution of 4 in one portion. The
combined mixture was stirred at 60 °C under nitrogen for 16 h to
give a clear yellow to light brown solution. The mixture was then
concentrated under reduced pressure. The resultant oil was dis-
solved in CH2Cl2 and purified by silica gel chromatography (eluting
with a linear gradient of 0–5% MeOH in CH2Cl2). The resultant oil
was dissolved in CH3CN and water and then purified by prepara-
tive reverse phase HPLC (linear gradient, 5–95% CH3CN in water).
Pure fractions were combined and freeze dried to give phosphonic
diamide 17i as a white powder (7.0 mg, 39%). 1H NMR (300 MHz,
CD3OD) d 8.29 (s, 1H), 8.25 (s, 1H), 6.84 (m, 1H), 6.00 (s, 1H),
5.96 (m, 1H), 4.04 (m, 8H), 1.66 (m, 4H), 1.38 (m, 6H), 0.98 (m,
6H). 31P NMR (121.4 MHz, CD3OD) d 22.8. LC–MS m/z 558 [M+H]+.
The following phosphonic diamide prodrugs were prepared
using the general procedure for preparation of 17i except substi-
tuting the appropriate amino acid.
4.6. Diethyl 5-[(6-methoxy-purin-9-yl)-4-fluoro-2,5-dihydro-
furan-2-yloxymethyl] phosphonic acid (13)
A solution of compound 12 (6 g, 11.3 mmol) in acetic acid
(2.5 mL) and MeOH (50 mL), was treated with the dropwise addi-
tion of NaClO (10–13%) (50 mL). The reaction mixture was stirred
for 0.5 h and then concentrated under reduced pressure. The resi-
due was treated with EtOAc and then filtered to remove solids. The
filtrate was concentrated and the residue was purified by silica gel
chromatography to give 13 (4.0 g, 88%). 1H NMR (300 MHz, CDCl3)
d 8.53 (s, 1H), 8.16 (s 1H), 6.82 (dd, J = 4.0, 1.2 Hz, 1H), 5.84 (dd,
J = 4.0, 1.2 Hz, 1H), 5.78 (d, J = 1.2 Hz, 1H), 4.0–4.2 (m, 5H), 4.18
(s, 3H), 3.82 (m, 1H), 1.3 (m, 6H). 13C NMR (77 MHz, CDCl3) d
161.1, 160.0, 156.2, 152.7, 151.9, 140.2, 121.4, 105.0–105.6 (m),
79.7, 79.3, 62.5–63.1 (m), 60.9, 54.3, 16.5, 16.4, 16.3. 19F NMR
(282.2 MHz, CDCl3) d ꢂ132.6 (m). 31P NMR (121.4 MHz, CDCl3) d
19.9. LC–MS m/z 403 [M+H]+.
4.10. Prodrug 17a (20-Fd4AP bis-Phe-n-butyl ester)
Yield = 42%. 1H NMR (300 MHz, CD3OD) d 8.25 (s, 1H), 8.17 (s,
1H), 7.21 (m, 10H), 6.80 (m, 1H), 5.91 (s, 1H), 5.72 (m, 1H), 4.04
(m, 6H), 3.50 (m, 2H), 2.90 (m, 4H), 1.47 (m, 8H), 0.92 (m, 6H).
31P NMR (121.4 MHz, CD3OD) d 22.3. LC–MS m/z 738 [M+H]+.
4.11. Prodrug 17b (20-Fd4AP bis-Phe-n-propyl ester)
Yield = 60%. 1H NMR (300 MHz, CD3OD) d 8.25 (s, 1H), 8.16 (s,
1H), 7.22 (m, 10H), 6.81 (m, 1H), 5.90 (s, 1H), 5.72 (m, 1H), 4.02
(m, 6H), 3.63 (m, 2H), 2.90 (m, 4H), 1.58(m, 4H), 0.87(m, 6H). 31P
NMR (121.4 MHz, CD3OD) d 22.3. LC–MS m/z 710 [M+H]+.
4.7. 5-[(6-Amino-purin-9-yl)-4-fluoro-2,5-dihydro-furan-2-
yloxymethyl] phosphonic acid diphosphate (15)
4.12. Prodrug 17c (20-Fd4AP bis-Phe-ethyl ester)
Yield = 11%. 1H NMR (300 MHz, CD3OD) d 8.25 (s, 1H), 8.16 (s,
The diphosphate was prepared from the disodium salt 4 by Tri-
1H), 7.24 (m, 10H), 6.84 (m, 1H), 5.91 (s, 1H), 5.75 (m, 1H), 4.08
link Technologies, Inc, San Diego.

3614
R. L. Mackman et al. / Bioorg. Med. Chem. 18 (2010) 3606–3617
(m, 6H), 3.60 (m, 2H), 2.90 (m, 4H), 1.21 (m, 6H). 31P NMR
Anhydrous pyridine (35 mL) was added followed by Et3N (6.7 mL,
48 mmol). The mixture was stirred at 60 °C under nitrogen for
15–20 min. The 2,20-dithiodipyridine (7.3 g) was mixed in a sepa-
rate flask with PPh3 (6.2 g) in anhydrous pyridine (5 mL) and the
resultant mixture was stirred for 10–15 min to give a clear light
yellow solution. The solution was then added to the solution of 4
and the combined mixture stirred overnight at 60 °C. The mixture
was concentrated under reduced pressure and the residue dis-
solved in EtOAc, washed twice with saturated aqueous sodium
bicarbonate (2ꢃ), and then brine. The organic layer was separated,
dried over Na2SO4, filtered and concentrated under reduced pres-
sure. The resultant oil was dissolved in CH2Cl2 and subjected to sil-
ica gel chromatography eluting with 0–5% MeOH in CH2Cl2. The
product foam was dissolved in CH3CN and further purified by pre-
parative HPLC (linear gradient, 5–95% CH3CN in water) to provide
18v (0.95 g, 42%) as a mixture of P(R) and P(S) isomers (approx.
3:2). 1H NMR (300 MHz, CD3OD) d 8.29 (s, 1H), 8.24 (m, 1H), 7.30
(m, 5H), 6.84 (m, 1H), 6.04 (m, 2H), 4.05 (m, 5H), 1.24 (m, 6H).
13C NMR (77 MHz, DMSO-d6) d 173.3, 157.2, 156.1, 153.2, 150.1,
149.4, 138.4, 129.5, 124.5, 120.6, 118.6, 105.6, 104.9, 79.0, 63.2,
60.5, 49.1, 20.3, 13.9. 31P NMR (121.4 MHz, CD3OD) d 23.6, 22.3.
LC–MS m/z 507.4 [M+H]+.
(121.4 MHz, CD3OD) d 22.4. LC–MS m/z 682 [M+H]+.
4.13. Prodrug 17d (bis-ABA-n-butyl ester)
Yield = 73%. 1H NMR (300 MHz, CDCl3) d 8.38 (s, 1H), 8.12 (s,
1H), 6.80 (m 1H), 5.93 (m, 1H), 5.79 (s, 1H), 4.02 (m, 6H), 3.42
(m, 1H), 3.21 (m, 1H), 1.65 (m, 4H), 1.35 (m, 8H), 0.92 (m, 12H).
31P NMR (121.4 MHz, CDCl3) d 20.5. LC–MS m/z 614 [M+H]+.
4.14. Prodrug 17e (bis-ABA-iso-butyl ester)
Yield = 35%. 1H NMR (300 MHz, CDCl3) d 8.39 (s, 1H), 8.12 (s,
1H), 6.82 (m, 1H), 5.81–5.96 (m, 4H), 3.79–4.03 (m, 10H), 3.49 (s,
1H), 3.2 (m, 2H), 1.69–1.96 (m, 10H), 1.26 (m, 4H), 0.91 (m,
12H). 31P NMR (121.4 MHz, CDCl3) d 20.4. LC–MS m/z 614 [M+H]+.
4.15. Prodrug 17f (bis-ABA acid-iso-propyl)
Yield = 59%. 1H NMR (300 MHz, CDCl3) d 8.39 (s, 1H), 8.13 (s,
1H), 6.81 (m, 1H), 5.95 (m, 1H), 5.81(s, 1H), 4.98 (m, 2H), 3.90
(m, 2H), 3.37 (m, 1H), 3.19 (m, 1H), 1.71 (m, 4H), 1.25 (m, 12H),
0.90 (m, 6H). 31P NMR (121.4 MHz, CDCl3) d 20.4. LC–MS m/z 586
[M+H]+.
4.22. Monoamidate example procedure 2: Preparation of 18v
from monophenoxy ester 22
4.16. Prodrug 17g (bis-Ala-n-butyl ester)
A solution of monoacid 22 (22 mg, 0.05 mmol) in DMF (2 mL) was
treated with PyBoP (56 mg, 0.11 mmol) and stirred at rt for 30 min.
The mixture was then treated with ethyl alanine esterꢁHCl salt
Yield = 55%. 1H NMR (300 MHz, CDCl3) d 8.37 (s, 1H), 8.17 (s, 1H),
6.80 (m, 1H), 6.18 (s, 1H), 5.93 (m, 1H), 5.79 (s, 1H), 4.02 (m, 6H), 3.46
(m, 1H), 3.37 (m, 1H), 1.61 (m, 4H), 1.32 (m, 10H), 0.92 (m, 6H). 31P
NMR (121.4 MHz, CDCl3) d 20.0. LC–MS m/z 586 [M+H]+.
(12 mg, 0.1 mmol) and Et3N (10 lL, 0.072 mmol). The resultant mix-
ture was stirred overnight and then diluted with EtOAc. The solution
was washed with NaHCO3, brine, dried over MgSO4, filtered and con-
centratedto givea crudesolid. ThecrudesolidwassubjectedtoHPLC
chromatography(lineargradient, 5–95%CH3CN inwater)togive18v
(8 mg, 30%) as a mixture of P(R) and P(S) isomers (approx. 3:2).
The following monoamidates were prepared from diacid 4 using
the general procedure 1 shown above, except substituting the
appropriate amino acid and phenol, or trifluoroethanol.
4.17. Prodrug 17h (bis-Ala-iso-butyl ester)
Yield = 47%. 1H NMR (300 MHz, CDCl3) d 8.38 (s, 1H), 8.12 (s,
1H), 6.80 (m, 1H), 5.93 (m, 2H), 5.80 (s, 1H), 3.91 (m, 6H), 3.42
(m, 1H), 3.30 (m, 1H), 1.91 (m, 2H), 1.40 (m, 6H), 0.90 (m, 12H).
31P NMR (121.4 MHz, CDCl3) d 19.9. LC–MS m/z 586 [M+H]+.
4.18. Prodrug 17j (bis-Ala-ethyl ester)
4.23. Prodrug 18a (Phe-sec-(S)-butyl, OPh ester)
Yield = 22%. 1H NMR(300 MHz, CD3OD)d 8.30(s, 1H), 8.25 (s, 1H),
6.84 (m, 1H), 6.00 (s, 1H), 5.95 (m, 1H), 4.06 (m, 8H), 1.31 (m, 12H).
31P NMR (121.4 MHz, CD3OD) d 22.8. LC–MS m/z 530 [M+H]+.
Yield = 37%. 1H NMR (300 MHz, CD3OD) d 8.25 (m, 2H), 7.20 (m,
9H), 6.96 (m, 1H), 6.81 (m, 1H), 5.97 (m, 2H), 5.73 (m, 1H), 4.71 (m,
1H), 4.05 (m, 2H), 3.60 (m, 1H), 3.02 (m, 1H), 2.81 (m, 1H), 1.49 (m,
2H) 1.07 (m, 3H), 0.82 (m, 3H). 31P NMR (121.4 MHz, CD3OD) d
+
4.19. Prodrug 17k (bis-Gly-iso-butyl ester)
23.2, 22.5. LC–MS m/z 611 [M+H]+, 609 [MꢂH]
.
Yield = 73%. 1H NMR (300 MHz, CD3OD) d 8.31 (s, 1H), 8.25 (s,
1H), 6.84 (m 1H), 6.01 (s, 1H), 5.98 (m, 1H), 4.03 (m, 2H), 3.86
(m, 4H), 3.68 (m, 4H), 1.92 (m, 2H), 0.93 (m, 12H). 31P NMR
(121.4 MHz, CD3OD) d 25.5. LC–MS m/z 558 [M+H]+.
4.24. Prodrug 18b (Phe-sec-(R)-butyl, OPh ester)
Yield = 16%. 1H NMR (300 MHz, CD3OD) d 8.25 (m, 2H), 7.20 (m,
9H), 6.96 (m, 1H), 6.81 (m, 1H), 5.97 (m, 2H), 5.73 (m, 1H), 4.71 (m,
1H)), 4.05 (m, 2H), 3.60 (m, 1H), 3.02 (m, 1H), 2.81 (m, 1H), 1.49
(m, 2H) 1.07 (m, 3H), 0.82 (m, 3H). 31P NMR (121.4 MHz, CD3OD)
d 23.3, 22.6. LC–MS m/z 611 [M+H]+.
4.20. Prodrug 17l (bis-Gly-iso-propyl ester)
Yield = 59%. 1H NMR (300 MHz, CD3OD) d 8.31 (s, 1H), 8.25 (s,
1H), 6.84 (m 1H), 6.02 (s, 1H), 5.98 (m, 1H), 4.98 (m, 2H), 4.01
(m, 2H), 3.66 (m, 4H), 1.23 (m, 12H). 31P NMR (121.4 MHz, CD3OD)
d 25.5. LC–MS m/z 530 [M+H]+.
4.25. Prodrug 18c (Phe-CH2-cyclobutyl, OPh ester)
Yield = 44%. 1H NMR (300 MHz, CD3OD) d 8.25 (m, 2H), 7.12 (m,
10H), 6.83 (m, 1H), 5.99 (m, 2H), 5.72 (m, 1H), 4.10 (m, 4H), 3.65
(m, 1H), 3.02 (m, 1H), 2.79 (m, 1H), 2.50 (m, 1H), 1.89 (m, 6H).
31P NMR (121.4 MHz, CD3OD) d 23.3, 22.5. LC–MS m/z 623 [M+H]+.
4.21. Monoamidate example procedure 1, from disodium salt 4.
9-(R)-[5-(R)-[[(S)-1-
[(ethoxycarbonyl)ethyl]amino]phenoxyphosphonyl]methoxy-
3-fluoro-2,5-dihydrofuran-2-yl]-9H-adenine (18v)
4.26. Prodrug 18d (Phe-iso-butyl, OPh ester)
Disodium salt 4 (1.5 g, 4.52 mmol) was mixed with alanine
Yield = 32%. 1H NMR (300 MHz, CDCl3) d 8.40 (m, 1H), 8.09 (m,
ethyl ester HCl salt (1.23 g, 8 mmol) and PhOH (1.88 g, 20 mmol).
1H), 7.27–6.74 (m, 8H), 5.30–5.93 (m, 4H), 4.39 (m, 1H), 4.14–3.77

R. L. Mackman et al. / Bioorg. Med. Chem. 18 (2010) 3606–3617
3615
(m, 4H), 3.58 (m, 2H), 2.95 (m, 2H), 1.90 (m, 3H), 1.26 (m, 1H), 0.85
4.35. Prodrug 18m (Ala-n-pentyl, OPh ester)
(m, 6H). 31P NMR (121.4 MHz, CDCl3) d 21.0, 20.0. LC–MS m/z 611
[M+H]+.
Yield = 40%. 1H NMR (300 MHz, CDCl3) d 8.39 (s, 1H), 8.13 (s,
1H), 7.27–7.14 (m, 5H), 6.85 (s, 1H), 5.97–5.77 (m, 4H), 4.19–
4.05 (m, 7H), 1.60 (m, 3H), 1.29 (m, 7H), 0.90 (m, 3H). 31P NMR
(121.4 MHz, CDCl3) d 20.7, 19.8. LC–MS m/z 549 [M+H]+.
4.27. Prodrug 18e (Phe-cyclobutyl, OPh ester)
Yield = 31%. 1H NMR (300 MHz, CD3OD) d 8.23 (m, 2H), 7.18 (m,
10H), 6.96 (m, 1H), 6.81 (m, 1H), 5.94 (m, 2H), 5.72 (m, 1H), 4.81
(m, 1H), 4.05 (m, 2H), 3.60 (m, 1H), 3.02 (m, 1H), 2.81 (m, 1H),
2.25 (m, 2H) 1.81 (m, 4H). 31P NMR (121.4 MHz, CD3OD) d 23.3,
22.5. LC–MS m/z 609 [M+H]+.
4.36. Prodrug 18n (Ala-cyclopentyl, OPh ester)
Yield = 34%. 1H NMR (300 MHz, CD3OD) d 8.28 (m, 1H), 8.24 (m,
1H), 7.17 (m, 5H), 6.85 (m, 1H), 6.04 (m, 2H), 5.15 (m, 1H), 4.14 (m,
3H), 1.64 (m, 8H), 1.25 (m, 3H). 31P NMR (121.4 MHz, CD3OD) d
23.6, 22.3. LC–MS m/z 547 [M+H]+.
4.28. Prodrug 18f (Phe-iso-propyl, OPh ester)
Yield = 40%. 1H NMR (300 MHz, CD3OD) d 8.25 (m, 2H), 7.20 (m,
9H), 6.96 (m, 1H), 6.81 (m, 1H), 5.97 (m, 2H), 5.73 (m, 1H), 4.05 (m,
2H), 3.60 (m, 1H), 3.02 (m, 1H), 2.81 (m, 1H), 1.13 (m, 6H). 31P NMR
(121.4 MHz, CD3OD) d 23.3, 22.6. LC–MS m/z 598 [M+H]+.
4.37. Prodrug 18o (Ala-n-butyl, OPh ester)
Yield = 25%. 1H NMR (300 MHz, CD3OD) d 8.25–8.27 (m, 2H),
7.25–7.34 (m, 2H), 7.10–7.19 (m, 3H), 6.83–6.84 (m, 1H), 5.93–
6.02 (m, 2H), 3.95–4.26 (m, 5H), 1.48–1.58 (m, 3H), 1.24–1.38
(m, 6H), 0.85–0.95 (m, 3H). 31P NMR (121.4 MHz, CD3OD) d 23.6,
22.2. LC–MS m/z 535 [M+H]+.
4.29. Prodrug 18g (Phe-n-propyl, OPh ester)
Yield = 13%. 1H NMR (300 MHz, CD3OD) d 8.25 (m, 2H), 7.22 (m,
8H), 6.95 (m, 1H), 6.82 (m, 1H), 5.90 (m, 2H), 5.72 (m, 1H), 3.95 (m,
4H), 3.63 (m, 1H), 3.07 (m, 1H), 2.81 (m, 1H), 1.55 (m, 2H), 0.86 (m,
3H). 31P NMR (121.4 MHz, CD3OD) d 23.3, 22.5. LC–MS m/z 597
[M+H]+.
4.38. Prodrug 18p (Ala-CH2-cyclobutyl, OPh ester)
Yield = 37%. 1H NMR (300 MHz, CD3OD) d 8.25 (m, 2H), 7.15 (m,
5H), 6.83 (m, 1H), 5.98 (m, 2H), 4.10 (m, 5H), 2.57 (m, 1H), 1.80 (m,
6H), 1.25 (m, 3H). 31P NMR (121.4 MHz, CD3OD) d 23.6, 22.2. LC–
MS m/z 548 [M+H]+.
4.30. Prodrug 18h (Phe-CH2-cyclopropyl, OPh ester)
4.39. Prodrug 18q (Ala-iso-butyl, OPh ester)
Yield = 33%. 1H NMR (300 MHz, CD3OD) d 8.25 (m, 2H), 7.15 (m,
10H), 6.82 (m, 1H), 5.99 (m, 2H), 5.73 (m, 1H), 3.99 (m, 4H), 3.65
(m, 1H), 3.05 (m, 1H), 2.85 (m, 1H), 1.02 (m, 1H), 0.51 (m, 2H),
0.20 (m, 2H). 31P NMR (121.4 MHz, CD3OD) d 23.4, 22.5. LC–MS
m/z 609 [M+H]+.
Yield = 17%. 1H NMR (300 MHz, CDCl3) d 8.39 (s, 1H) d 8.13 (s,
1H), 7.27–7.11 (m, 5H), 6.82 (s, 1H), 5.97–5.77 (m, 4H), 4.14–
3.79 (m, 6H), 3.64 (m, 1H), 2.00–1.88 (bm, 4H), 1.31 (m, 3H),
0.91 (m, 6H). 31P NMR (121.4 MHz, CDCl3) d 20.8, 19.8. LC–MS m/
z 535 [M+H]+.
4.31. Prodrug 18i (Phe-ethyl, OPh ester)
4.40. Prodrug 18r (Ala-cyclobutyl, OPh ester)
Yield = 38%. 1H NMR (300 MHz, CD3OD) d 8.24 (s, 1H), 8.20 (m,
1H), 7.10 (m, 9H), 6.94 (m, 1H), 6.82 (m, 1H), 5.94 (m, 1H), 5.73 (m,
1H), 4.05 (m, 4H), 3.63 (m, 1H), 3.03 (m, 1H), 2.92 (m, 1H), 1.14 (m,
3H). 31P NMR (121.4 MHz, CD3OD) d 23.3, 22.5. LC–MS m/z 583
[M+H]+.
Yield = 37%. 1H NMR (300 MHz, CD3OD) d 8.29 (m, 1H), 8.25 (m,
1H), 7.20 (m, 5H), 6.85 (m, 1H), 5.97 (m, 2H), 4.85 (m, 1H), 4.15 (m,
2H), 3.95 (m, 1H), 2.28 (m, 2H), 1.99 (m, 2H), 1.77 (m, 2H) 1.26 (m,
3H). 31P NMR (121.4 MHz, CD3OD) d 23.6, 22.2. LC–MS m/z 533
[M+H]+.
4.32. Prodrug 18j (ABA-n-butyl ester, OPh ester)
4.41. Prodrug 18s (Ala-iso-propyl, OPh ester)
Yield = 39%. 1H NMR (300 MHz, CDCl3) d 8.39 (m, 1H), 8.11 (m,
1H), 7.10–7.28 (m, 5H), 6.82 (s, 1H), 5.76–5.98 (m, 3H), 3.56–4.18
(m, 4H), 3.59 (m, 1H), 0.70–1.74 (m, 12H). 31P NMR (121.4 MHz,
CDCl3) d 21.0, 20.1. LC–MS m/z 549 [M+H]+.
Yield = 43%. 1H NMR (300 MHz, CD3OD) d 8.22–8.27 (m, 2H),
7.25–7.34 (m, 2H), 7.10–7.18 (m, 3H), 6.83–6.84 (m, 1H), 5.94–
6.02 (m, 2H), 4.90–5.05 (m, 1H), 3.90–4.26 (m, 3H), 1.15–1.31
(m, 9H). 31P NMR (121.4 MHz, CD3OD) d 23.5, 22.3. LC–MS m/z
521 [M+H]+.
4.33. Prodrug 18k (ABA-iso-butyl ester, OPh ester)
Yield = 40%. 1H NMR (300 MHz, CD3OD) d 8.22–8.26 (m, 2H),
7.22–7.34 (m, 2H), 7.10–7.16 (m, 3H), 6.83–6.84 (m, 1H), 5.9–
6.01 (m, 2H), 4.05–4.22 (m, 2H), 3.76–3.86 (m, 3H), 1.79–1.88
(m, 1H), 1.68–1.74 (m, 1H), 1.51–1.68 (m, 1), 0.72–0.99 (m,
9H). 31P NMR (121.4 MHz, CD3OD) d 23.8, 22.6. LC–MS m/z 549
[M+H]+.
4.42. Prodrug 18t (Ala-n-propyl, OPh ester)
Yield = 52%. 1H NMR (300 MHz, CDCl3) d 8.38 (m, 1H), 8.11 (s,
1H), 7.11–7.31 (m, 5H), 6.82 (s, 1H), 5.76–5.96 (m, 4H), 3.63–
4.22 (m, 6H), 2.17 (bm, 2H), 1.65 (m, 2H), 1.30 (m, 4H), 0.88 (m,
3H). 31P NMR (121.4 MHz, CDCl3) d 20.8, 19.8. LC–MS m/z 521
[M+H]+.
4.34. Prodrug 18l (Ala-3-pentyl, OPh ester)
4.43. Prodrug 18u (Ala-CH2-cyclopropyl, OPh ester)
Yield = 14%. 1H NMR (300 MHz, CD3OD) d 8.25 (m, 2H), 7.17 (m,
5H), 6.85 (m, 1H), 5.99 (m, 2H), 4.66 (m, 1H), 4.12 (m, 3H), 1.56 (m,
4H), 1.28 (m, 3H), 0.88 (m, 6H). 31P NMR (121.4 MHz, CD3OD) d
23.5, 22.4. LC–MS m/z 549 [M+H]+.
Yield = 28%. 1H NMR (300 MHz, CD3OD) d 8.29 (s, 1H), 8.24 (m,
1H), 7.28 (m, 5H), 6.85 (m, 1H), 6.00 (m, 2H), 4.16 (m, 5H), 1.05 (m,
1H), 0.48 (m, 3H), 0.25 (m, 4H). 31P NMR (121.4 MHz, CD3OD) d

3616
R. L. Mackman et al. / Bioorg. Med. Chem. 18 (2010) 3606–3617
23.6, 22.2. LC–MS m/z 533 [M+H]+. HRMS (ESI) m/z Calc. for
C23H26FN6O6P, 532.1635. Found, 532.1662.
(m, 2H), 2.02 (m, 2H), 1.62–1.80 (m, 4H) 1.28 (d, J = 6.9 Hz, 3H).
31P NMR (121.4 MHz, CDCl3) d 20.7. HRMS (ESI) m/z Calc. for
C23H26FN6O6P, 532.1635. Found, 532.1659.
4.44. Prodrug 18w (Gly-iso-butyl, OPh ester)
4.51. Prodrug 18r(B) (Ala-cyclo-butyl, OPh ester,
diastereoisomer B, 36% recovery)
Yield = 43%. 1H NMR (300 MHz, CDCl3) d 8.39 (s, 1H), 8.17 (m,
1H), 6.82–7.32 (m, 5H), 6.82 (s, 1H), 5.81–5.98 (m, 3H), 3.64–
4.27 (m, 6H), 1.94 (m, 1H), 0.90 (m, 6H). 31P NMR (121.4 MHz,
CDCl3) d 21.5, 21.4. LC–MS m/z 521 [M+H]+.
1H NMR (300 MHz, CDCl3) d 8.39 (s, 1H), 8.10 (s, 1H), 7.14–7.30
(m, 5H), 6.80 (m, 1H), 5.86 (m, 1H), 5.77 (m, 2H), 4.98 (m, 1H),
3.92–4.22 (m, 4H), 3.72 (m, 1H), 2.32 (m, 2H), 2.04 (m, 2H),
1.60–1.90 (m, 4H) 1.33 (d, J = 7.8 Hz, 3H). 31P NMR (121.4 MHz,
CDCl3) d 19.7.
4.45. Prodrug 19a (Phe-iso-butyl, OCH2CF3 ester)
Yield = 33%. 1H NMR (300 MHz, CD3OD) d 8.20 (m, 2H), 7.25 (m,
6H), 6.82 (m 1H), 5.95 (m, 2H), 5.68 (m, 1H), 3.93 (m, 6H), 3.50 (m,
1H), 3.20 (m, 1H), 2.81 (m, 1H), 1.90 (m, 1H), 0.95 (m, 6H). 31P NMR
(121.4 MHz, CD3OD) d 27.4, 26.8. LC–MS m/z 617 [M+H]+.
4.52. Diphenyl-hydroxymethylphosphonate (20)
A suspension of paraformaldehyde (1.6 g, 53 mmol) in CH2Cl2
(25 mL) at 0 °C was treated with TMSI (8 mL, 56 mmol). The resul-
tant mixture was stirred at 0 °C for 30 min. The ice bath was re-
moved, and a solution of diphenylphosphite (12.5 g, 50 mmol) in
CH2Cl2 (25 mL) was added dropwise. The resultant solution was
stirred in the dark for 16 h. The reaction mixture was then parti-
tioned between CH2Cl2 and cold saturated aqueous NaHCO3 solu-
tion. The organic phase was separated and washed with brine.
The organic phase was concentrated under reduced pressure and
the crude residue subjected to silica gel chromatography eluting
with 0–40% EtOAc in CH2Cl2 to afford 20 (7.3 g, 55%). 1H NMR
(300 MHz, CDCl3) d 7.34–7.38 (m, 4H), 7.20–7.33 (m, 6H), 4.21
(m, 2H). 31P NMR (121.4 MHz, CDCl3) d 17.1.
4.46. Prodrug 19b (Ala-ethyl, OCH2CF3 ester)
Yield = 32%. 1H NMR (300 MHz, CD3OD) d 8.24 (m, 2H), 6.85 (m,
1H), 6.01 (m, 2H), 4.43 (m, 2H), 4.09 (m, 5H), 1.38 (m, 3H) 1.23 (m,
3H). 31P NMR (121.4 MHz, CD3OD) d 28.0, 26.9. LC–MS m/z 513
[M+H]+.
4.47. Separation of monoamidate phosphorus diastereoisomers,
example prodrug 18v
The isomer mixture of monoamidate 18v (950 mg, 1.87 mmol)
was dissolved in a small amount of CH3CN and allowed to stand
at rt overnight. The solid that formed (isomer A) was collected by
filtration and washed with a small amount of cold CH3CN. The fil-
trate was reduced under vacuum and then loaded onto Chiralpak
AS-H column equilibrated in Buffer A, 2% ethanol in CH3CN. Isomer
A, was eluted out with Buffer A at 10 mL/min for 17 min. The elut-
ing buffer was then changed to Buffer B, 50% MeOH in CH3CN, and
isomer B eluted from the column in 8 min. All solvent was removed
under reduced pressure, and then the samples re-dissolved in
CH3CN and water. Samples were freeze dried to provide isomer A
(475 mg, 50%) and isomer B (348 mg, 37%).
4.53. Diphenyl 5-(R)-[(6-methoxy-purin-9-yl)-4-(R)-fluoro-3-
(R)-iodo-tetrahydro-furan-2-(R)-yloxymethyl]phosphonic acid
(21)
A suspension of carboxylic acid 9 (4.0 g, 13.4 mmol) in CH2Cl2
(100 mL) was heated to 40 °C. The mixture was treated with PPh3
(4.2 g, 16.1 mmol) and DIAD (3.3 g, 16.1 mmol). The resultant mix-
ture was heated at 40 °C for 2 h and then cooled to ꢂ50 °C. Diphe-
nyl-(hydroxymethyl)phosphonate 20 (7.3 g, 27 mmol) was added
followed by the dropwise addition of a solution of IBr (6.8 g,
33 mmol) in CH2Cl2 (30 mL). The resultant mixture was stirred at
ꢂ20 °C for 30 min and then at rt for 2 h. The reaction mixture
was poured into aqueous NaHCO3 containing Na2S2O3. The organic
phase was separated, dried over MgSO4, and concentrated under
reduced pressure. The residue was purified by silica gel column
chromatography eluting with EtOAc in hexanes to afford 21
(3.5 g, 40%). 1H NMR (300 MHz, CDCl3) d 8.56 (s, 1H), 8.17 (d,
J = 3.9 Hz, 1H), 7.20–7.38 (m, 10H), 7.10 (dd, J = 19.3, 4.0 Hz, 1H),
5.67 (s, 1H), 5.4 (dd, J = 49.2, 3.9 Hz, 1H), 4.55 (d, J = 16.5 Hz, 1H),
4.35–4.42 (m, 4H), 4.19 (s, 3H), 4.13–4.19 (m, 1H). 13C NMR
(77 MHz, CDCl3) d 166.1, 152.5, 151.7, 150.0, 149.6, 142.1, 142.0,
130.0, 129.9, 125.8, 125.7, 120.7, 120.6, 111.8, 111.6, 96.1, 93.4,
84.1, 84.8, 62.9, 60.6, 57.6, 54.4. 19F NMR (282.2 MHz, CDCl3) d
ꢂ166.14(m). 31P NMR (121.4 MHz, CDCl3) d 12.12.
4.48. Prodrug 18v(A) (Ala-ethyl, OPh ester, diastereoisomer A,
50% recovery)
1H NMR (300 MHz, CD3OD) d 8.29 (s, 1H), 8.24 (s, 1H), 7.30 (m,
5H), 6.84 (d, J = 2.7 Hz, 1H), 6.04 (s, 1H), 5.99 (d, J = 5.1 Hz, 1H),
4.05 (m, 5H), 1.26 (d, J = 7.2 Hz, 3H) 1.19 (t, J = 7.2 Hz, 3H). 31P
NMR (121.4 MHz, CD3OD) d 23.6. HRMS (ESI) m/z Calc. for
C21H24FN6O6P, 506.1479. Found, 506.1480
4.49. Prodrug 5 or 18v(B) (Ala-ethyl, OPh ester, diastereoisomer
B, 37% recovery)
1H NMR (300 MHz, CD3OD) d 8.29 (s, 1H), 8.25 (s, 1H), 7.30 (m,
5H), 6.85 (m, 1H), 5.98 (s, 1H), 5.95 (d, J = 4.5 Hz, 1H), 4.11 (m, 5H),
1.28 (d, J = 6.9 Hz, 3H) 1.21 (t, J = 7.5 Hz, 3H). 31P NMR (121.4 MHz,
CD3OD) d 22.3. HRMS (ESI) m/z calcd for C21H24FN6O6P, 506.1479.
Found, 506.1492.
4.54. Phenyl 5-(R)-[(6-amino-purin-9-yl)-4-fluoro-2,5-dihydro-
furan-2-(R)-yloxymethyl]phosphonic acid (22)
Prodrug 18r was isolated as pure isomers at phosphorous using
the same chiral separation procedure reported above for 18v.
Diphenyl ester 21 (2 g, ꢀ3 mmol) was dissolved in acetic acid
(2 mL) and MeOH (20 mL). The mixture was cooled to 0 °C and
then treated by dropwise addition with NaClO (10–13%) (13 mL).
The reaction mixture was stirred for 1 h and then concentrated un-
der reduced pressure. The residue was partitioned between CH2Cl2
and saturated aqueous NaHCO3. The organic phase was separated,
dried over MgSO4, filtered and concentrated under reduced pres-
sure. The crude product was redissolved in 1,4-dioxane (8 mL)
and. NH4OH (20 mL). The mixture was heated at 100 °C for 5 h
4.50. Prodrug 18r(A) (Ala-cyclo-butyl, OPh ester,
diastereoisomer A, 48% recovery)
1H NMR (300 MHz, CDCl3) d 8.39 (s, 1H), 8.10 (s, 1H), 7.21 (m,
5H), 6.82 (m, 1H), 5.97 (m, 1H), 5.79 (d, J = 1.2 Hz, 1H), 5.68 (br s,
1H), 4.96 (m, 1H), 4.01–4.20 (m, 4H), 3.60 (t, J = 9.9 Hz, 1H), 2.29

R. L. Mackman et al. / Bioorg. Med. Chem. 18 (2010) 3606–3617
3617
and then cooled to rt. The resultant mixture was concentrated un-
der reduced pressure and the residue subjected to silica gel chro-
matography eluting with 50% EtOAc in CH2Cl2 followed by 15%
water in CH3CN. The collected product was then further purified
by HPLC chromatography (linear gradient, 5–95% CH3CN in water)
to give the title product 22 (240 mg, 18%). 1H NMR (300 MHz,
CD3OD) d 8.38(s, 1H), 8.27 (s, 1H), 6.98–7.20 (m, 5 H), 6.82 (d,
J = 3.0 Hz, 1H), 5.96–5.98 (m, 2H), 3.96–4.01 (m, 1H), 3.80–3.86
(m, 1H). 31P NMR (121.4 MHz, CDCl3) d 12.1. LC–MS m/z 408
[M+H]+.
MD). Briefly, 2 lM of the prodrugs were incubated with S9 extract
at 37 °C and samples taken over 90 min. Hepatic S9 studies were
done in the presence of mixed cofactors (NADPH regenerating sys-
tem, UDPGA and alamethicin). Amounts of remaining prodrug at
each collected time point were quantified by LC/MS/MS and the
half life determined by fitting the data to an exponential decay
curve. The methods for the in vivo evaluation of prodrugs have
been reported.4
Acknowledgements
4.55. Prodrug 23 (ethyl-(S)-2-hydroxypropionic acid ester, OPh
ester)
The authors would like to thank Richard Polniaszek (Process re-
search), and the department of analytical chemistry at Gilead for
contributions to the chemistry program and analytical data.
A solution of monoacid 22 (22 mg, 0.05 mmol) and PyBoP
(52 mg, 0.1 mmol) in DMF (1.5 mL) was stirred at rt for 30 min.
The mixture was then treated with ethyl, 2-(S)-hydroxypropionic
acid ester (18 mg, 0.15 mmol) and Et3N (7.6 mg, 0.075 mmol).
The resultant mixture was stirred overnight and then diluted with
EtOAc. The solution was washed with saturated NaHCO3, brine,
dried over MgSO4, filtered and concentrated to give a crude solid.
The crude solid was subjected to HPLC chromatography (linear gra-
dient, 5–95% CH3CN in water) to provide 23 (5 mg, 19%). 1H NMR
(300 MHz, CDCl3) d 8.40 (s, 1H), 8.11 (s, 1H), 7.22 (m, 5H), 6.82
(d, J = 2.4 Hz, 1H), 5.97 (d, J = 3.3 Hz, 1H), 5.81 (s, 1H), 5.65 (s,
2H), 5.03 (m, 1H), 4.19 (m, 5H), 1.44 (d, J = 6.9 Hz, 3H), 1.23 (t,
J = 6.9 Hz, 3H). 31P NMR (121.4 MHz, CDCl3) d 18.2. LC–MS m/z
508 [M+H]+.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmc.2010.03.041.
References and notes
1. http://www.who.int/hiv/data/2008_global_summary_AIDS_ep.png.
2. De Clercq, E. Int. J. Antimicrob. Agents 2009, 33, 307.
3. Hammer, S. M.; Saag, M. S.; Schechter, M.; Montaner, J. S. G.; Schooley, R. T.;
Jacobsen, D. M.; Thompson, M. A.; Carpenter, C. C. J.; Fischl, M. A.; Gazzard, B.
A.; Gatell, J. M.; Hirsch, M. S.; Katzenstein, D. A.; Richman, D. D.; Vella, S.; Yeni,
P. G.; Volberding, P. A. J. Am. Med. Assoc. 2006, 7, 827.
4. (a) Carr, A.; Miller, J.; Law, M.; Cooper, D. A. AIDS 2000, 14, F25; (b) Moyle, G.
Clin. Ther. 2000, 22, 911; (c) Kakuda, T. N. Clin. Ther. 2000, 22, 685.
5. (a) Cihlar, T.; Ray, A. S.; Boojamra, C. G.; Zhang, L.; Hui, H.; Laflamme, G.; Vela, J.
E.; Grant, D.; Chen, J.; Myrick, F.; White, K. L.; Gao, Y.; Lin, K.-Y.; Douglas, J. L.;
Parkin, N. T.; Carey, A.; Pakdaman, R.; Mackman, R. L. Antimicrob. Agents
Chemother. 2008, 52, 655; (b) Ray, A. S.; Vela, J. E.; Boojamra, C. G.; Zhang, L.;
Hui, H.; Callebaut, C.; Stray, K.; Lin, K.-Y.; Gao, Y.; Mackman, R. L.; Cihlar, T.
Antimicrob. Agents Chemother. 2008, 52, 648.
4.56. Antiviral activity
The methods used for the evaluation of the antiviral activity
have been described previously.5
6. Boojamra, C. G.; Mackman, R. L.; Markevitch, D. Y.; Prasad, V.; Ray, A. S.;
Douglas, J.; Grant, D.; Kim, C. U.; Cihlar, T. Bioorg. Med. Chem. Lett. 2008, 18,
1120.
4.56.1. In vitro Cat A, PBMC, hepatic and intestinal S9 stabilities
and in vivo methods
The method describing the profiling of compounds toward hu-
man recombinant cathepsin A has been reported.15 Extracts from
human PBMC were prepared as described previously.18 The pro-
7. (a) Howell, H. G.; Brodfuehrer, P. R.; Brundidge, S. P.; Benigni, D. A.; Sapino, C.,
Jr. J. Org. Chem. 1988, 53, 85; (b) Tann, C. H.; Brodfuehrer, P. R.; Brundidge, S. P.;
Sapino, C., Jr.; Howell, H. G. J. Org. Chem. 1985, 50, 3644; (c) Ma, T.; Lin, J. S.;
Gary Newton, M.; Cheng, Y.-C.; Chu, C. K. J. Med. Chem. 1997, 40, 2750.
8. Kim, C. U.; Luh, B. Y.; Martin, J. C. J. Org. Chem. 1991, 56, 2642.
9. Mulzer, J.; Lammer, O. Angew. Chem., Int. Ed. Engl. 1983, 22, 628.
10. Knapp, S.; Naughton, A. B. J.; Dhar, T. G. M. Tetrahedron Lett. 1992, 33, 1025.
11. (a) Mackman, R. L.; Zhang, L.; Prasad, V.; Boojamra, C. G.; Douglas, J.; Grant, D.;
Hui, H.; Kim, C. U.; Laflamme, G.; Parrish, J.; Stoycheva, A. D.; Swaminathan, S.;
Wang, K.; Cihlar, T. Bioorg. Med. Chem. 2007, 15, 5519; (b) Mackman, R. L.;
Boojamra, C. G.; Prasad, V.; Zhang, L.; Lin, K.-Y.; Petrakovsky, O.; Babusis, D.;
Chen, J.; Douglas, J.; Grant, D.; Hui, H. C.; Kim, C. U.; Markevitch, D. Y.; Vela, J.;
Ray, A.; Cihlar, T. Bioorg. Med. Chem. Lett. 2007, 17, 6785; (c) Mackman, R. L.;
Lin, K.-Y.; Boojamra, C. G.; Hui, H.; Douglas, J.; Grant, D.; Petrakovsky, O.;
Prasad, V.; Ray, A. S.; Cihlar, T. Bioorg. Med. Chem. Lett. 2008, 18, 1116.
12. Tsai, C.-H.; Doong, S.-L.; Johns, D. G.; Driscoll, J. S.; Cheng, Y.-C. Biochem.
Pharmacol. 1994, 48, 1477.
13. Cahard, D.; McGuigan, C.; Balzarini, J. Mini-Rev. Med. Chem. 2004, 4, 371.
14. (a) Lee, W. A.; He, G.-X.; Eisenberg, E.; Cihlar, T.; Swaminathan, S.; Mulato, A.;
Cundy, K. C. Antimicrob. Agents Chemother. 2005, 49, 1898; (b) Lee, W. A. XVII
International Roundtable of Nucleosides, Nucelotides and Nucleic acids, Bern,
Switzerland, September 3–7, 2006.
15. Birkus, G.; Wang, R.; Liu, X.; Kutty, N.; MacArthur, H.; Cihlar, T.; Gibbs, C.;
Swaminathan, S.; Lee, W.; McDermott, M. Antimicrob. Agents Chemother. 2007,
51, 543.
drugs (30 lM) were incubated with PBMC extracts (134 mg/mL)
for 5, 10, 30, and 120 min at 37 °C in a reaction buffer containing
25 mM Mes-Na, pH 6.5, 100 mM NaCl, 0.1% Nonidet P-40 and
1 mM DTT. After the incubation, the reactions were quenched by
adding ice-cold methanol (final concentration, 70%) and centri-
fuged at 13,000g for 30 min. The supernatants were evaporated,
re-solubilized in 100 lL of buffer A (25 mM potassium phosphate,
pH 6.0, 5 mM tetrabutylammonium bromide), and injected onto a
C18 reverse-phase column (Prodigy 5 micron ODS(2); Phenomenex,
Torrance, CA) equilibrated with buffer A. Bound prodrugs and
metabolites were resolved using a linear gradient of acetonitrile
(0–70%, 11 min, 0.25 mL/min) with buffer B (25 mM KPO4, pH
6.0, 5 mM tetrabutylammonium bromide, 70% acetonitrile). The
peak areas of parent nucleotide, parent nucleotide–amino acid
conjugate and prodrug was used for the calculation of the reaction
rates. The activity was expressed as picomoles of metabolites pro-
16. Chou, T. F.; Baraniak, J.; Kaczmarek, R.; Zhou, X.; Cheng, J.; Ghosh, B.; Wagner,
C. R. Mol. Pharm. 2007, 4, 208.
17. Chapman, H.; Kernan, M.; Prisbe, E.; Rohloff, J.; Sparacino, M.; Terhorst, T.; Yu,
R. Nucleosides Nucleotides Nucleic Acids 2001, 20, 621.
18. Birkus, G.; Kutty, N.; He, G. X.; Mulato, A.; Lee, W.; McDermott, M.; Cihlar, T.
Mol. Pharmacol. 2008, 74, 92.
duced per minute per minute per lg protein. The intestinal and he-
patic stability of prodrugs were assessed in pooled intestinal and
hepatic cytosolic supernatant fractions collected after centrifuga-
tion at 9000g (S9) obtained from In Vitro Technologies (Baltimore