Synthesis, anti-HIV activity, and resistance profile of thymidine phosphonomethoxy nucleosides and their bis-isopropyloxymethylcarbonyl (bisPOC) prodrugs

Article abstract of DOI:10.1016/j.bmc.2007.05.047

Phosphonomethoxy nucleoside analogs of the thymine containing nucleoside reverse transcriptase inhibitors (NRTIs), 3′-azido-2′,3′-dideoxythymidine (AZT), 2′,3′-didehydro-2′,3′-dideoxythymidine (d4T), and 2′,3′-dideoxythymidine (ddT), were synthesized. The anti-HIV activity against wild-type and several major nucleoside-resistant strains of HIV-1 was evaluated together with the inhibition of wild-type HIV reverse transcriptase (RT). Phosphonomethoxy analog of d4T, 8 (d4TP), demonstrated antiviral activity with an EC₅₀ value of 26 μM, whereas, phosphonomethoxy analogs of ddT, 7 (ddTP), and AZT, 6 (AZTP), were both inactive at concentrations up to 200 μM. Bis-isopropyloxymethylcarbonyl (bisPOC) prodrugs improved the anti-HIV activity of 7 and 8 by >150-fold and 29-fold, respectively, allowing for antiviral resistance to be determined. The K65R RT mutant virus was more resistant to the bisPOC prodrugs of 7 and 8 than bisPOC PMPA (tenofovir DF) 1. However, bisPOC prodrug of 7 demonstrated superior resistance toward the RT virus containing multiple thymidine analog mutations (6TAMs) indicating that new phosphonate nucleoside analogs may be suitable for targeting clinically relevant nucleoside resistant HIV-1 strains.

Full text of DOI:10.1016/j.bmc.2007.05.047

Bioorganic & Medicinal Chemistry 15 (2007) 5519–5528
Synthesis, anti-HIV activity, and resistance profile
of thymidine phosphonomethoxy nucleosides and their
bis-isopropyloxymethylcarbonyl (bisPOC) prodrugs
Richard L. Mackman,* Lijun Zhang, Vidya Prasad, Constantine G. Boojamra,
Janet Douglas, Deborah Grant, Hon Hui, Choung U. Kim, Genevieve Laflamme,
Jay Parrish, Antitsa D. Stoycheva, Swami Swaminathan, KeYu Wang and Tomas Cihlar
Gilead Sciences, Inc., 333 Lakeside Drive, Foster City, CA 94404, USA
Received 6 March 2007; revised 17 May 2007; accepted 18 May 2007
Available online 25 May 2007
Abstract—Phosphonomethoxy nucleoside analogs of the thymine containing nucleoside reverse transcriptase inhibitors (NRTIs),
30-azido-20,30-dideoxythymidine (AZT), 20,30-didehydro-20,30-dideoxythymidine (d4T), and 20,30-dideoxythymidine (ddT), were syn-
thesized. The anti-HIV activity against wild-type and several major nucleoside-resistant strains of HIV-1 was evaluated together
with the inhibition of wild-type HIV reverse transcriptase (RT). Phosphonomethoxy analog of d4T, 8 (d4TP), demonstrated anti-
viral activity with an EC50 value of 26 lM, whereas, phosphonomethoxy analogs of ddT, 7 (ddTP), and AZT, 6 (AZTP), were both
inactive at concentrations up to 200 lM. Bis-isopropyloxymethylcarbonyl (bisPOC) prodrugs improved the anti-HIV activity of 7
and 8 by >150-fold and 29-fold, respectively, allowing for antiviral resistance to be determined. The K65R RT mutant virus was
more resistant to the bisPOC prodrugs of 7 and 8 than bisPOC PMPA (tenofovir DF) 1. However, bisPOC prodrug of 7
demonstrated superior resistance toward the RT virus containing multiple thymidine analog mutations (6TAMs) indicating that
new phosphonate nucleoside analogs may be suitable for targeting clinically relevant nucleoside resistant HIV-1 strains.
ꢀ 2007 Elsevier Ltd. All rights reserved.
1. Introduction
bility to tenofovir prompted our search to identify new
N(t)RTIs that possess a superior resistance profile to
tenofovir DF.
Since the introduction of AZT-based antiretroviral ther-
apy, a suite of nucleoside and nucleotide HIV reverse
transcriptase inhibitors (N(t)RTIs) have been approved
for clinical use in the treatment of HIV.1 Tenofovir diso-
proxil fumarate (tenofovir DF, VireadTM) 1 is a prodrug
of PMPA 2, a unique member of this class of drugs since
it is the only nucleoside phosphonate and the only acy-
clic ribose analog (Fig. 1).2 The favorable resistance and
toxicity profile of tenofovir DF, in addition to the ben-
efits of a once-daily administration, has led to its wide-
spread use in antiretroviral combinations.3 However, a
subset of treatment-experienced patients with either
K65R mutation or multiple (P3) thymidine analog
mutations (TAMs), including M41L and/or L210W in
RT, respond sub-optimally to tenofovir DF treatment.4
The identification of HIV strains with reduced suscepti-
2. Strategy
The phosphonomethoxy moiety of 2 is a bioisostere of a
nucleoside monophosphate. However, compared to the
phosphate group, the phosphonate is chemically stable
and insensitive to enzymatic hydrolysis. Once delivered
into cells, via a prodrug (e.g., 1), the released phospho-
nate diacid 2 is effectively metabolized to the active
diphospho-phosphonate (triphosphate equivalent).5
Moreover, the high polarity and limited permeability
of the phosphonate diacid can be a significant advantage
when combined with a long intracellular half-life of the
active diphospho-phosphonate species. Together, these
properties favor the intracellular accumulation of the ac-
tive metabolite resulting in optimal pharmacodynam-
ics.5 For these reasons, we focused our design efforts
toward new nucleoside phosphonate analogs and their
prodrugs.
Keywords: Phosphonates; Nucleosides; Antiviral; Resistance.
*
Corresponding author. Tel.: +1 650 522 5258; fax: +1 650 522
5899; e-mail: rmackman@gilead.com
0968-0896/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2007.05.047

5520
R. L. Mackman et al. / Bioorg. Med. Chem. 15 (2007) 5519–5528
O
N
O
NH2
NH
O
NH
O
N
N
N
RO
RO
O
O
N
HO
N
HO
P
O
O
R
1 R = Poc (Tenofovir DF)
2 R = H (PMPA)
3 R = N3 (AZT)
4 R = H (ddT)
5 d4T
Poc =
O
O
O
O
O
O
NH
O
NH
O
NH
O
H2O3P
N
O
N
H2O3P
O
X
O
N O
H2O3P
X
R
R
6 R = N3 (AZTP)
7 R = H (ddTP)
8 X = O (d4TP)
9 X = CH2
10 R = H, X = CH2
11 R = H, X = CF2
12 R = N3, X = CH2
Figure 1. Structures of phosphonate and nucleoside analogs.
In comparison with other clinically approved and inves-
tigational N(t)RTIs, the anti-HIV activity of thymidine
nucleosides AZT, ddT, and d4T (3–5) is only modestly
affected by the K65R mutation in cell culture assays.6,7
Evaluation of 20,30-dideoxy (dd) analogs of inosine, gua-
nosine, cytidine, and thymidine toward a recombinant
K65R virus indicated that the thymidine analog, ddT
4, was the least susceptible to K65R mutation.7 It was
therefore reasoned that nucleoside phosphonates of thy-
midine based HIV RT inhibitors AZT, ddT, and d4T,
for example, 6–8, may be highly effective against HIV
strains containing the K65R mutation. Phosphonometh-
oxy nucleoside 8 and several phosphonoalkyl thymidine
nucleosides (9–12) have been reported in the literature
but all, apart from 8, are inactive toward HIV.8–11 How-
ever, the presence of anti-HIV activity for phos-
phonomethoxy analog, 8, and lack of anti-HIV activity
for the corresponding phosphonoalkyl analog 9, sug-
gests that the phosphonomethoxy group may be critical
for anti-HIV potency. Therefore, our strategy was to
synthesize the thymidine based phosphonomethoxy ana-
logs 6–8, compare and contrast their respective anti-HIV
potency and resistance profiles, and evaluate prodrugs
accordingly.
yl protection of 30-OH was followed by TBAF mediated
50-O-silyl deprotection to afford 16 in high yield.12 The
50-OH was then efficiently oxidized to carboxylic acid
17 by treatment with catalytic TEMPO and stoichio-
metric bisacetoxyiodobenzene.14 Acid 17 was then
transformed into a and b anomeric acetates 18 by treat-
ment with Pb(OAc)4. The b anomer predominated but
the mixture of anomers was carried forward to the next
step. Treatment of the acetate anomers with diethyl
hydroxymethylphosphonate in the presence of Lewis
acid TMSOTf, followed by removal of the 30-benzoyl
group using ammonium hydroxide, resulted in the for-
mation of the desired phosphonate 19. Unfortunately,
the overall yield for this reaction was relatively poor
(28% from 18) compared to the literature.12 An equiva-
lent amount of the C-10-a-isomer was obtained, presum-
ably through TMSOTf-mediated isomerization over the
prolonged reaction time. The structures of 19 and its
C-10-a-isomer were established using 2D NMR spectros-
copy. Similar C-10-epimerizations in the presence of
TMSOTf have been reported.15 The N-benzyloxymethyl
group was then removed by hydrogenolysis followed by
treatment with ammonium hydroxide to provide 20.
Dehydration of 20 to give the anhydro intermediate 21
proceeded smoothly using DAST.16 Mitsunobu dehy-
dration conditions for the formation of the anhydro
intermediate 21 were not successful. Finally, acid pro-
moted opening of 21 by treatment with sodium azide
proved quite difficult, requiring high temperatures and
prolonged reaction times. Under the forcing conditions
required one of the phosphonate alkyl groups was
cleaved to give the mono acid 22 as the major product.
To complete the synthesis of 6, the remaining phospho-
nate ester group was removed by treatment with TMSBr
followed by isolation as the ammonium salt. Saturated
analog 7 was prepared from 8 by hydrogenation using
catalytic Pd/C (Scheme 2).17 The bisPOC prodrugs, 23
and 24, were prepared from the phosphonate diacids 7
and 8, respectively, by treatment with chloromethyl iso-
propyl carbonate in the presence of triethylamine.18
Diphospho-phosphonates, 25 and 26 were prepared
3. Synthesis
The phosphonomethoxy analog 6 was prepared from
thymidine in 13 steps (Scheme 1). The two key transfor-
mations in the sequence were first, the stereoselective
introduction of the phosphonate group in the b-config-
uration, and second, replacement of the 30-OH with an
azide. A Lewis acid-mediated glycosidation of a phos-
phonate alcohol onto an anomeric acetate such as 18
could be used to introduce the desired phosphonate as
reported previously.12 Stereoselectivity would be con-
trolled by anchiomeric assistance from the 30-OBz
group. To begin, 50-tert-butyl-diphenylsilyl-N3-benzyl-
oxymethyl 20-deoxythymidine, 14, was prepared from
thymidine, 13, following literature procedures.13 Benzo-

R. L. Mackman et al. / Bioorg. Med. Chem. 15 (2007) 5519–5528
5521
O
N
O
N
O
N
BOM
BOM
NH
O
N
N
(i), (ii)
O
O
(iv)
(v)
O
O
O
HO
TBDPSO
(iii)
HO
HO
13
RO
14 R = H
BzO
16
15 R = Bz
O
O
N
O
BOM
BOM
BOM
N
N
N
O
O
P
OEt
O
N
O
O
O
O
(ix), (x)
(vi)
O
N
AcO
O
EtO
(vii), (viii)
HO
BzO
HO
19
BzO
18
17
O
O
O
N
NH
O
O
RO P
R1O
NH
O
N
O
O
O
N
O
O
N
P
O
O
EtO
EtO
(xi)
(xii)
EtO
O
P
O
OEt
N3
HO
22 R = Et, R1 = H
6 R = R1 = NH4
20
21
(xiii)
Scheme 1. Reagents and conditions: (i) TBSCl, DMAP, pyridine, rt, 16 h; (ii) BOMCl, DBU, DMF, 2 h; (iii) BzCl, pyridine, DMAP; (iv) TBAF,
THF, 0 ꢁC, 1.5 h; (v) TEMPO, PhI(OAc)2, CH3CN, H2O, rt, 16 h; (vi) Pb(OAc)4, pyridine/DMF, 16 h; (vii) TMSOTf, hydroxymethyl phosphonic
acid diethyl ester, CH2Cl2, ꢀ25 ꢁC to rt, 16 h; (viii) NH4OH, MeOH, rt, 6 h; (ix) Pd(OH)2, H2, EtOH, rt, 16 h; (x) NH4OH, MeOH, 40 ꢁC, 16 h; (xi)
diethylaminosulfur trifluoride, CH2Cl2, ꢀ78 ꢁC, 30 min; (xii) NaN3, DMF, benzoic acid, 120 ꢁC, 36 h; (xiii) TMSBr, DMF, 2,6-lutidine, 0 ꢁC to rt,
16 h, then NH4OH.
O
N
O
N
O
N
NH
O
NH
O
NH
O
O
O
(PocO)2OP
(i)
O
O
H2O3P
(ii)
O
O
Na2O3P
23
8
7
(ii)
(iii)
(iii)
O
O
O
N
NH
NH
NH
O
O
O
O
O
(PocO)2OP
O
N
O
O
N
O
H4O9P3
H4O9P3
26
25
24
Scheme 2. Reagents and conditions: (i) Pd/C, H2, EtOH, rt, 2 h; (ii) chloromethylisopropyl carbonate, Et3N, DMF, 70 ꢁC, 3 h, then rt, 16 h; (iii)
bis(tributylammonium)pyrophosphate, CDI, DMF, rt, 16 h.
from diacids 8 and 7, respectively, following literature
procedures (Scheme 2).10
less active toward HIV RT than the d4 analog 25. Thus,
the weak antiviral activity of dd analog 7 can be attrib-
uted, at least in part, to the weaker inhibition of HIV
RT by the active metabolite produced inside the cell.
It is worth noting that both phosphonates are poorer
inhibitors of RT than the corresponding nucleosides
ddT and d4T (Table 1). The phosphonate bioisostere
replacement for the nucleoside phosphate group is
clearly inferior with respect to RT inhibition. The trends
in RT activity between analogs 7 and 8, (the d4 analog
being more potent than the dd analog) are mirrored
by the respective nucleoside triphosphates of d4T and
ddT (Table 1). Favorable p–p interactions between tyro-
sine 115 in the active site of HIV RT and the double
bond of the d4 analogs are proposed to be responsible
for the superior activity.19 The parallel activity trends
4. Results and discussion
The anti-HIV activity of the thymidine phosphonates,
6–8, is reported in Table 1. Neither 6 nor 7 displayed
anti-HIV activity at concentrations up to 200 lM. How-
ever, the unsaturated 20,30-didehydro-20,30-dideoxy (d4)
analog 8, consistent with earlier reports,8 demonstrated
low micromolar activity, only 9-fold less active than
PMPA 2. To probe the basis for the different antiviral
activities of 7 and 8, their respective active metabolites,
diphospho-phosphonates, 26 and 25 were prepared.
20,30-Dideoxy (dd) analog 26 was approximately 14-fold

5522
R. L. Mackman et al. / Bioorg. Med. Chem. 15 (2007) 5519–5528
Table 1. HIV antiviral activity, cytotoxicty, RT inhibition, and resistance profile
Compound Diphospho-phosphonate Code
WT HIVa EC50 HIV RTb IC50 MT-2c CC50 6TAMsd fold K65R fold M184V fold
2
3
PMPA
AZT
ddT
3.6 (1.5)
0.16 (0.12)
>200
0.38 (0.20)
0.05 (0.03)
0.12 (0.09)
0.06 (0.01)
—
>200
>200
—
8.8 (3.7)
>50
—
4.3 (1.5)
1.2 (0.7)
—
0.7 (0.2)
0.7 (0.2)
—
4
5
d4T
4.8 (2.3)
>200
>200
286 (122)
—
5.5 (2.5)
—
—
2.3 (1.1)
—
—
1.2 (0.2)
—
—
6
AZTP
ddTP
d4TP
7
26
25
5.4 (2.1)
0.39 (0.08)
—
>500
>500
>50
8
26 (8.8)
13 (3.9)
7.7 (2.5)
2.6 (0.5)
6.8 (0.7)
5.2 (0.3)
3.1 (0.4)
5.9 (0.6)
6.7 (2.6)
1.1
—
1
BPe-PMPA 0.015 (0.007)
BPe-ddTP 1.3 (0.8)
BPe-d4TP 0.91 (0.42)
23
24
—
—
>50
10
—
—
Results are expressed in lM with standard deviations in parentheses.
a Antiviral activity in MT-2 cells using HIV-IIIa.
b Diphospho-phosphonates or nucleoside triphosphates were used in this assay.
c Cytotoxicity measured in MT-2 cells.
d 6TAMs virus contains M41L/D67N/K70R/L210W/T215Y/K219Q mutations in reverse transcriptase.
e BP, bis-isopropyloxymethylcarbonyl prodrug.
H
O
H
also suggest that the weaker IC50 of the phosphonates is
not the result of gross differences in binding mode in the
active site of RT but, more likely, differences in the poly-
merization step. Indeed AZT phosphonate 12, when
compared to AZT using pre-steady state kinetics, was
found to have a significantly reduced rate of polymeriza-
tion but a similar binding constant.10,20
HO
H
Base
Base
H
H
O
HO
H
H
H
H
H
H
H
Northern (P=0o)
Southern(P=180o)
Despite the parallel RT inhibition trends, the dd analog
7 is a surprisingly weak RT inhibitor, being 45-fold
weaker than the corresponding nucleoside ddT, 4, tri-
phosphate, whereas the unsaturated d4 analog 8 is only
7-fold weaker than d4T (5) triphosphate. This incremen-
tal increase in the IC50 for the dd analog 7 implies there
may be an additional factor influencing the RT activity
of this particular type of phosphonate. One key struc-
tural difference between the phosphonomethoxy analogs
and the corresponding nucleosides is the generation of a
second anomeric center. To probe the impact of the
additional anomeric center, the solution conformation
Figure 2. NMR derived conformation of 7 (P = 246ꢁ, w = 20ꢁ) and
schematic showing ribose conformations defined as Northern (P = 0ꢁ)
and Southern (P = 180ꢁ).
1
of 7 was determined from the H NMR J couplings
and NOEs in D2O at 25 ꢁC (see experimental Section
6.3).21 20,30-Dideoxy nucleosides typically have a pre-
dominantly Northern (P = 0–20ꢁ; 30-endo, 20-exo) con-
leading to similar IC50 trends and conformational differ-
ences (data not shown).
formation.21–23 In contrast
7 has a dramatically
In addition to enzymatic potency, poor cellular uptake
and/or intracellular phosphorylation can also contribute
to poor antiviral activity. Therefore, to evaluate the
effects of poor cellular uptake on the antiviral activity,
bisPOC prodrugs, 23 and 24, were prepared from 7
and 8, respectively (Scheme 2). The bisPOC prodrugs
improved antiviral potency of 8 by 29-fold and 7 by
>150-fold, whereas the bisPOC prodrug of PMPA 2
improved activity by 250-fold (Table 1). Interestingly,
bis(POC) prodrug 23 now displays comparable antiviral
activity to bis(POC) prodrug 24 despite the lower inhibi-
tion of HIV RT. Thus, the prodrugs can compensate for
reduced RT activity by facilitating permeability of the
phosphonate across the cell membrane leading to the
generation of higher intracellular levels of active diphos-
phate (triphosphate equivalent) species. Further studies
on AZT analog 6 to probe the basis for its poor antiviral
potency have been hampered by a lack of sufficient
different (P = 246ꢁ) conformation (Fig. 2). The anomeric
effect drives the phosphonomethoxy group closer to a
pseudoaxial position unlike the psuedoequatorial orienta-
tion found in a typical Northern nucleoside conforma-
tion. As a result, it is likely that there is an
unfavorable energy associated with dd analog 7 adopt-
ing a Northern conformation similar to ddT. This is
important, because the Northern conformation has been
proposed to be the required conformation for substrate
binding to HIV RT and is supported by the X-ray struc-
ture and reported data on constrained carbocyclic inhib-
itors.19,24 In the case of the d4 phosphonate analog 8,
the sp2 carbonsrender the furanose ring planar and more
rigid, thus minimizing the impact of the additional ano-
meric center on the conformation.25 It is anticipated that
the conformation of the d4 phosphonate analog 8 is
therefore similar to nucleoside d4T, 5. Further studies
have been performed on other dd phosphonate analogs

R. L. Mackman et al. / Bioorg. Med. Chem. 15 (2007) 5519–5528
5523
material. However, given that dd analog 7 is a poor
inhibitor of HIV RT, it is likely that 6, being a saturated
ribose analog, also suffers from poor RT potency. Fur-
thermore, intracellular phosphorylation of 6 may be lim-
phonate group unfavorably for the excision mechanism
to take place. Further kinetic studies will be required to
understand the basis for the improved TAMs profile.
ited,
similar
to
that
observed
for
AZT
The M184V mutation is located in the active site of RT
underneath the primer and in proximity to the ribose
oxygen. It is primarily selected by L-nucleosides, that
is, 3TC and FTC, and represents the most prevalent
NRTI mutation occurring in >40% of all NRTI-experi-
enced patients.32 Similar to PMPA, and many natural
d-configuration nucleosides with a 10-O atom, the activ-
ity of analog 8 was not significantly affected by M184V
(1.1-fold, Table 1). The level of M184V resistance was
essentially identical to that of d4T 5.
monophosphate.26
The resistance profile toward viruses containing six thy-
midine analog mutations (6TAMs), K65R or M184V
was evaluated for analog 8 and its bis(POC) prodrug
23, and also bis(POC) prodrug 24 (Table 1). The suscep-
tibility of many clinically approved and investigational
N(t)RTI toward K65R has been reported in the litera-
ture.6,7 Thymidine analogs, for example, AZT and
d4T, have, in general, been found to be less susceptible
in cell culture to the K65R recombinant viruses than
other inhibitors, including PMPA 2.6 In this study,
nucleoside d4T 5 was found to have a 2.3-fold drop in
potency toward K65R, whereas, d4 analog 8 and its bis-
POC prodrug 24 both exhibited 5- to 7-fold reduced
activity against K65R recombinant virus. A comparison
cannot be made between ddT 4 and the dd analog 7 due
to its poor antiviral potency, but the bis(POC) prodrug
23 showed a 5.9-fold reduction in the potency against
K65R compared to 2.2-fold reported for ddT.7 The
phosphonate therefore introduces a small unfavorable
contribution toward K65R resistance compared to the
nucleosides. The K65 residue lies in the fingers domain
of HIV RT and H-bonds to the c-phosphate of the
incoming dNTP substrate, thereby aligning it correctly
for incorporation.19 Mutation to arginine is proposed
to reduce the conformational mobility of RT active site,
and, in turn, reduce binding and incorporation of the
dNTP.27,28 In the case of phosphonate analogs 7 and
8, the increased K65R resistance is likely due to subtle
changes in the binding mode that lead to further drops
in incorporation efficiency. The O–C–P atom arrange-
ment compared to the natural C–O–P is often regarded
as isosteric but a more strict analysis indicates that the
bond lengths and angles are subtly different.29 More-
over, the ether oxygen creates an anomeric center at
C40 which was shown above to impact the ribose confor-
mation. Thus, the increased ability of the K65R mutant
to resist 7 and 8 could easily be caused by subtle differ-
ences in the positioning of the a-phosphonate for attack
by the primer 30-hydroxyl.
5. Conclusions
The synthesis of several thymidine phosphonomethoxy
nucleosides has established that phosphonates 7 (ddT
analog) and 8 (d4T analog) are both inhibitors of HIV
RT, albeit weaker than the corresponding nucleosides.
The second anomeric center in the saturated dd analogs
promotes an unfavorable ribose ring conformation
which may account for the greater loss in RT inhibition
for the dd analog. BisPOC prodrugs effectively im-
proved cellular uptake leading to low micromolar antivi-
ral activity and allowed the resistance profile toward
K65R, 6TAMs, and M184V to be determined. Although
bis(POC) prodrug 23 shows a superior TAMs profile
than bis(POC) PMPA 1, the poor K65R resistance pro-
file has reduced the potential for further development.
Efforts to identify alternative phosphonates with excel-
lent antiviral activity and resistance profiles continue.
6. Experimental
6.1. General
All reactions were performed with reagent-grade materi-
als under an atmosphere of nitrogen. Solvents were
reagent-grade or better. Silica gel chromatography was
1
performed using EMD Chemicals Silica Gel 60. H and
31P NMR spectra were recorded on a Varian 300- or
500-MHz NMR. NMR shifts are reported relative to an
internal standard as noted. Low-resolution LC mass spec-
tra (LR LCMS) were recorded on two different Thermo
Finnigan LCQ Advantage, electrospray ionization (ESI)
mass spectrometers. LR LCMS(A) was equipped with a
Phenomenex Synergi 4 micron Polar-RP C18 column
(30 · 4.6 mm), eluting with a gradient of 5–100% solvent
B in solvent A over 2.5 min at a flow rate of 2 mL/min (sol-
vent A = 0.1% AcOH in water, solvent B = 0.1% AcOH in
acetonitrile). LR LCMS(B) was equipped with a Phenom-
enex Luna C18 (30 · 2.0 mm) column eluting with a gra-
dient of 0–100% solvent B in solvent A over 8 min at a flow
rate of 1.75 mL/min (solvent A = 1% AcOH in water, sol-
vent B = 1% AcOH in MeOH). High resolution mass
spectroscopy (HRMS) was determined by UC Berkeley
mass spectrometry facility using FAB ionization. Re-
versed phase analytical HPLC for phosphonate diacids
and bisPOC prodrugs was performed on a Agilent 1100
The 6TAMs mutant promotes ATP- or pyrophosphate-
mediated unblocking of the N(t)RTI terminated primer
in the Nucleotide-site (N-site) of RT.30 It has been pro-
posed that some NRTIs, for example, AZT, preferen-
tially reside in the N-site leading to more efficient
excision by RT containing TAMs mutations.31 The
6TAMs mutant was marginally more resistant toward
analog 8 than the corresponding nucleoside d4T 5.
Bis(POC) prodrug 24 was not substantially different
given the experimental error in the antiviral assay. The
6TAMs mutant, however, is only 2.6-fold resistant to-
ward dd prodrug 23. Indeed the 6TAMs profile for 23
is superior to all the other phosphonate and nucleoside
thymidine analogs shown in Table 1. The underlying
reason for the improved profile may lie in the more
puckered dd ring that in turn positions the alpha phos-

5524
R. L. Mackman et al. / Bioorg. Med. Chem. 15 (2007) 5519–5528
series, equipped with a Phenomenex Luna C18 5 lm
100 A column (250 · 4.6 mm), eluting with a gradient of
2–70% solvent B in solvent A over 30 min (solvent
A = 0.1% TFA in water; solvent B = 0.1% TFA in aceto-
nitrile). Ion exchange HPLC analysis for diph-
osphophosphonates was performed on a Agilent 1100
series analytical HPLC, equipped with a DIONEX DNA-
Pac PA-100 column (250 · 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, solvent B = 0.5 mM
triethylammonium bicarbonate buffer, pH 8.5).
1H NMR (300 MHz, CDCl3) d 8.05 (dd, J = 1.2, 8.6 Hz,
2H), 7.22–7.63 (m, 9H), 6.39 (dd, J = 6.2, 8.0 Hz, 1H),
5.60 (m, 1H), 5.51 (s, 2H), 4.71 (s, 2H), 4.25 (d, J =
2.1 Hz, 1H), 4.00 (d, J = 2.4 Hz, 2H), 2.46–2.60 (m, 2H),
1.94 (d, J = 1.2 Hz, 3H). LR LCMS(B) (M+Na) 489.3.
6.2.4.
(2S,3S,5R)-Benzoyloxy-5-(3-benzyloxymethyl-5-
methyl-2,4-dioxo-3,4-dihydro-2H-pyrimidin-1-yl)-tetrahy-
dro-furan-2-carboxylic acid (17). The alcohol 16 (0.83 g,
1.78 mmol) was dissolved in acetonitrile (8 mL). BAIB
(1.26 g, 3.92 mmol) was added to the solution followed
by water (8 mL). The mixture was then treated with TEM-
PO (0.056 g, 0.36 mmol) and stirred for 16 h. The mixture
was then concentrated under reduced pressure and the
crude product subjected to silica gel chromatography,
eluting with a gradient of 30–50% ethyl acetate in hexanes,
to give the title product 17 (0.65 g, 76%). Compound 17:
1H NMR (CDCl3) d 8.05–8.10 (m, 3H), 7.25–7.64 (m,
9H), 6.57 (dd, J = 5.2, 9.2 Hz, 1H), 5.77 (d, J = 5.2 Hz,
1H), 5.53 (s, 2H), 4.80 (s, 1H), 4.72 (s, 2H), 2.66 (dd,
J = 5.5, 14.1 Hz, 1H), 2.20–2.29 (m, 1H), 1.97 (s, 3H).
LR LCMS(B) (MꢀH) 479.7.
6.2. Chemistry
6.2.1. (2R,5R)-[5-(5-Methyl-2,4-dioxo-3,4-dihydro-2H-pyr-
imidin-1-yl)-2,5-dihydro-furan-2-yloxymethyl]-phosphonic
acid disodium salt (8). Compound 8 was prepared accord-
ing to the method of Kim et al.17 Compound 8: 1H NMR
(300 MHz, D2O) d 7.30 (s, 1H), 6.68 (s, 1H), 6.34 (d,
J = 8.7 Hz, 1H), 6.05 (d, J = 8.7 Hz, 1H), 5.85 (s, 1H),
3.40–3.62 (m, 2H), 1.75 (s, 3H). 31P NMR (121.4 MHz,
D2O) d 12.72. LR LCMS(A) (M+Na) 327.2 (0.25 min).
FABHRMS (M+Na) calc. for C10H13N2O7NaP
327.0358, found 327.0354. HPLC 8.62 min, 99.5%.
6.2.5. (2RS,3S,5R)-Benzoic acid 2-acetoxy-5-(3-benzyl-
oxymethyl-5-methyl-2,4-dioxo-3,4-dihydro-2H-pyrimidin-1-
yl)-tetrahydro-furan-3-yl ester (18). The acid 17 (1.89 g,
3.94 mmol) was dissolved in DMF (10 mL) and treated
with pyridine (1.15 mL, 15.8 mmol). The reaction mixture
was flushed with nitrogen and then treated with lead tetra-
acetate (8.5 g, 19.7 mmol). The reaction mixture was pro-
tected from light and stirred for 16 h. The mixture was
then concentrated under reduced pressure and the crude
residue subjected to silica gel chromatography, eluting with
a gradient of 30–60% ethyl acetate in hexanes, to give the ti-
tle product 18 (1.67 g, 86%) as a mixture of isomers (pre-
dominantly b-isomer). A sample of the b-isomer was
purified for analysis. Compound b-18: 1H NMR
(300 MHz, CDCl3) d 8.06 (d, J = 7.7 Hz, 1H), 7.23–7.66
(m, 9H), 6.74 (dd, J = 6.4, 8.2 Hz, 1H), 6.43 (s, 1H), 5.54
(d, J = 5.1 Hz, 1H), 5.51 (s, 2H), 4.71 (s, 2H), 2.77 (dd,
J = 6.3, 15.2 Hz, 1H), 2.29–2.39 (m, 1H), 2.20 (s, 3H),
1.97 (s, 3H). LR LCMS(B) (M+Na) 517.2.
6.2.2. (2R,3S,5R)-Benzoic acid 5-(3-benzyloxymethyl-5-
methyl-2,4-dioxo-3,4-dihydro-2H-pyrimidin-1-yl)-2-(tert-
butyl-diphenyl-silanyloxymethyl)-tetrahydro-furan-3-yl ester
(15). The protected thymidine 1411 (5.0 g, 8.3 mmol) was
dissolved in pyridine (10 mL) and the mixture cooled to
0 ꢁC. The mixture was then treated with benzoyl chlo-
ride (1.45 mL, 12.5 mmol) followed by DMAP (0.31 g,
2.5 mmol). The resultant mixture was stirred for 3 h
and then concentrated under reduced pressure. The
crude solid was redissolved in ethyl acetate and the
insoluble material was collected by filtration. The solu-
tion was then concentrated under reduced pressure
and the resultant solid was subjected to silica gel chro-
matography eluting with a gradient of 15–30% ethyl ace-
tate in hexanes to give the title product 15 (5.56 g, 94%).
Compound 15: 1H NMR (300 MHz, CDCl3) d 8.07 (dd,
J = 1.2, 8.6 Hz, 2H), 7.24–7.73 (m, 9H), 6.56 (dd,
J = 5.1, 9.3 Hz, 1H), 5.71 (d, J = 6.1 Hz, 1H), 5.53 (s,
2H), 4.73 (s, 2H), 4.26 (d, J = 1.9 Hz, 1H), 4.08 (d,
J = 1.9 Hz, 2H), 2.64 (dd, J = 5.4, 13.9 Hz, 1H), 2.32–
2.42 (m, 1H), 1.60 (d, J = 1.0 Hz, 3H), 1.41 (s, 9H).
LR LCMS(B) (M+H) 705.3.
6.2.6. (2R,3S,5R)-[5-(3-Benzyloxymethyl-5-methyl-2,4-
dioxo-3,4-dihydro-2H-pyrimidin-1-yl)-3-hydroxy-tetrahy-
dro-furan-2-yloxymethyl]-phosphonic acid diethyl ester
(19). The acetate 18 (1.67 g, 3.38 mmol) was dissolved
in DCM (50 mL) and then cooled to ꢀ25 ꢁC. The cooled
solution was then treated with hydroxyethyl phosphonic
acid diethyl ester (1.5 mL, 10.1 mmol) followed by
TMSOTf (2.45 mL, 13.5 mmol). The mixture was stirred
for 16 h and then quenched with triethylamine (2.1 mL,
15 mmol). The resultant solution was concentrated un-
der reduced pressure and the crude residue used in the
next reaction. The crude residue was dissolved in meth-
anol (4 mL) and ammonium hydroxide (4 mL) and the
reaction mixture stirred for 6 h. The reaction mixture
was concentrated under reduced pressure and the crude
residue subjected to HPLC chromatography to isolate
the title product 5R-19 (232 mg, 14%). Eluant was a
gradient of 20–90% solvent B in solvent A over 26 min
(sovent A = water, solvent B = acetonitrile). An equiva-
lent amount (14%) of the 5S-19 was also obtained.
6.2.3. (2R,3S,5R)-Benzoic acid 5-(3-benzyloxymethyl-5-
methyl-2,4-dioxo-3,4-dihydro-2H-pyrimidin-1-yl)-2-hydroxy-
methyl-tetrahydro-furan-3-yl ester (16). The benzoyl
protected thymidine 15 (1.55 g, 2.2 mmol) was dissolved
in THF (20 mL) and the mixture cooled to 0 ꢁC. The
cooled solution was then treated with a solution of TBAF
in THF (1 M, 2.8 mL, 2.8 mmol). The reaction mixture
was stirred for 1.5 h and then quenched by the addition
of saturated aqueous ammonium chloride. The mixture
was extracted with ethyl acetate (3· 20 mL) and the com-
bined organic extracts dried over sodium sulfate, filtered,
and concentrated under reduced pressure. The crude
product was subjected to silica gel chromatography,
eluting with a gradient of 15–70% ethyl acetate in hexanes,
to give the title product 16 (0.83 g, 81%). Compound 16:

R. L. Mackman et al. / Bioorg. Med. Chem. 15 (2007) 5519–5528
5525
Compound 5R-19: 1H NMR (300 MHz, CDCl3) d 7.21–
7.45 (m, 6H, Ph and 6-H), 6.78 (dd, J = 6.5, 8.0 Hz, 1H,
20-H), 5.49 (s, 2H, NCH2O), 5.05 (s, 1H, 50-H), 4.69 (s,
2H, OCH2Ph), 4.41 (d, J = 4.6 Hz, 1H, 40-H), 4.10–4.24
(m, 4H, Ethyl-CH2), 3.96 (dd, J = 9.5, 13.7 Hz, 1H,
CH2P), 3.76 (dd, J = 9.5, 13.7 Hz, 1H, CH2P), 2.38
(dd, J = 6.4, 14.0 Hz, 1H, 30-Ha), 2.08 (ddd, J = 5.3,
8.6, 14.3 Hz, 1H, 30-Hb), 1.98 (s, 3H, 5-CH3), 1.35 (t,
J = 7.0 Hz, 6H, Ethyl-CH3). 31P NMR (121.4 MHz,
CDCl3) d 20.72.
6.2.9.
(2R,3S,5R)-[3-Azido-5-(5-methyl-2,4-dioxo-3,4-
dihydro-2H-pyrimidin-1-yl)-tetrahydro-furan-2-yloxymeth-
yl]-phosphonic acidmonoethyl ester (22). Theanhydrocom-
pound 21 (33 mg, 90 lmol) was dissolved in DMF (8 mL)
and then treated with sodium azide (18 mg, 275 lmol) fol-
lowedbybenzoicacid(11 mg, 90 lmol). Theresultantmix-
ture was heated at 120 ꢁC for 36 h and then concentrated
under reduced pressure. The crude residue was subjected
to silica gel chromatography eluting with a gradient of
0–15% methanol in dichloromethane to give the title prod-
uct 22 (24 mg, 71%). Compound 22: 1H NMR (300 MHz,
D2O) d 7.43 (d, J = 1.2 Hz, 1H), 6.41 (dd, J = 6.7, 11.3 Hz,
1H), 5.07 (s, 1H), 4.34 (d, J = 4.9 Hz, 1H), 3.85 (m, 1H),
3.73 (dd, J = 8.9, 13.2 Hz, 1H), 3.56 (dd, J = 9.6, 13.3 Hz,
1H), 2.30–2.47 (m, 2H), 1.81 (s, 3H), 1.10 (t, J = 4.0 Hz,
3H). 31P NMR (121.4 MHz, D2O) d 16.26. LR LCMS(B)
(M+H) 376.1.
Compound 5S-19: 1H NMR (300 MHz, CD3CN) d 7.74
(d, J = 1.2 Hz, 1H, 6-H), 7.30-7.39 (m, 5H, Ph), 6.36
(dd, J = 2.8, 8.6 Hz, 1H, 20-H), 5.44 (s, 2H, NCH2O),
5.21 (s, 1H, 50-H), 4.66 (s, 2H, OCH2Ph), 4.27 (dd,
J = 3.4, 5.5 Hz, 1H, 40-H), 4.05-4.16 (m, 4H, Ethyl-
CH2), 3.98 (dd, J = 9.7, 14.0 Hz, 1H, CH2P), 3.82 (dd,
J = 12.2, 14.1 Hz, 1H, CH2P), 2.63-2.73 (m, 1H, 30-
Hb), 1.85-1.91 (m, 1H, 30-Ha), 1.88 (d, J = 1.2 Hz, 3H,
5-CH3), 1.33 (m, 6H, Ethyl-CH3). 31P NMR
(121.4 MHz, CD3CN) d 20.61.
6.2.10.
(2R,3S,5R)-[3-Azido-5-(5-methyl-2,4-dioxo-3,4-
dihydro-2H-pyrimidin-1-yl)-tetrahydro-furan-2-yloxymeth-
yl]-phosphonic acid ammonium salt (6). The monoacid 22
(8.9 mg, 24 lmol) was dissolved in DMF (2 mL) and evap-
orated (2·). The solid was then dissolved in dry DMF
(1.5 mL) and cooled to 0 ꢁC. The solution was then treated
with 2,6-lutidine (30 lL, 240 lmol) followed by the drop-
wise addition of TMSBr (15 lL, 120 lmol). The solution
was allowed to warm to RT and stirred for 16 h. The reac-
tion was then cooled to ꢀ78 ꢁC and treated with methanol
(30 lL). The mixture was allowed to warm to room tem-
perature, treated with ammonium hydroxide, and then
concentrated under reduced pressure. The residue was trit-
urated withacetone and the solid that formed was thencol-
lected by filtration. The crude solid was subjected to C18
chromatographyelutingwithwatertogivethetitleproduct
6 as the bis ammonium salt (2 mg, 22%). Compound 6: 1H
NMR (300 MHz, D2O) d 7.46 (d, J = 1.2 Hz, 1H), 6.42 (t,
J = 7.3 Hz, 1H), 5.08 (s, 1H), 4.35 (d, J = 4.2 Hz, 1H), 3.67
(dd, J = 8.8, 12.8 Hz, 1H), 3.47(dd, J = 10.1, 12.8 Hz, 1H),
2.30–2.45 (m, 2H), 1.80 (s, 3H). 31P NMR (121.4 MHz,
D2O) d 14.21. LR LCMS(B) (MꢀH) 346.3 (1.43 min).
FABHRMS (M+H) calc. for C10H15N5O7P 348.0709,
found 348.0718.
6.2.7. (2R,3S,5R)-[3-Hydroxy-5-(5-methyl-2,4-dioxo-3,4-
dihydro-2H-pyrimidin-1-yl)-tetrahydro-furan-2-yloxy-
methyl]-phosphonic acid diethyl ester (20). The phospho-
nate 19 (232 mg, 0.47 mmol) was dissolved in ethanol
(4 mL) and then treated with Pd(OH)2 (160 mg). The
mixture was stirred under an atmosphere of hydrogen
for 16 h and then the catalyst was removed by filtration.
The eluant was concentrated under reduced pressure
and the crude residue dissolved in methanol (1.6 mL)
and ammonium hydroxide (1.6 mL). The resultant solu-
tion was heated at 40 ꢁC for 16 h. The mixture was con-
centrated under reduced pressure and the crude residue
subjected to silica gel chromatography, eluting with a
gradient of 0–5% methanol in dichloromethane, to give
1
the title product 20 (110 mg, 63%). Compound 20: H
NMR (300 MHz, CDCl3) d 9.15 (br s, 1H), 7.33 (s,
1H), 6.77 (dd, J = 6.8, 7.9 Hz, 1H), 5.10 (s, 1H), 4.45
(d, J = 4.6 Hz, 1H), 4.19 (m, 4H), 3.99 (dd, J = 9.3,
13.6 Hz, 1H), 3.81 (dd, J = 9.1, 13.7 Hz, 1H), 2.42 (dd,
J = 6.3, 14.2 Hz, 1H), 2.08–2.19 (m, 1H), 1.98 (s, 3H),
1.36 (t, J = 7.2 Hz, 6H). 31P NMR (121.4 MHz, CDCl3)
d 20.63.
6.2.11. (2R,5R)-[5-(5-Methyl-2,4-dioxo-3,4-dihydro-2H-
pyrimidin-1-yl)-tetrahydro-furan-2-yloxymethyl]-phosphon-
ic acid disodium salt (7). Disodium salt of 8 (50 mg,
0.14 mmol) was dissolved in ethanol (5 mL) and then trea-
ted with 10% Palladium on carbon (5 mg). The mixture was
stirred under an atmosphere of hydrogen for 2 h. The
mixturewasfilteredoverCeliteand the eluant concentrated
under reduced pressure to give the title product 7 (40 mg,
6.2.8.
(4-Methyl-5-oxo-8,11-dioxa-2,6-diaza-tricyclo-
[7.2.1.02,7]dodeca-3,6-dien-10-yloxymethyl)-phosphonic acid
diethyl ester (21). The phosphonate 20 (21 mg, 55 lmol)
was dissolved in DCM (1 mL) and cooled to ꢀ78 ꢁC.
The mixture was then treated with DAST (9 lL,
69 lmol) and stirred for 30 min before being warmed
to RT over 30 min. The mixture was treated with sat-
urated sodium hydrogen carbonate (0.35 mL) and then
concentrated under reduced pressure. The crude
residue was subjected to silica gel chromatography,
eluting with a gradient of 0–15% methanol in dichlo-
romethane, to give the title product 21 (15 mg,
75%). Compound 21: 1H NMR (300 MHz, CD3CN)
d 7.26 (d, J = 0.9 Hz, 1H), 5.62 (d, J = 3.6 Hz, 1H),
5.47 (m, 1H), 5.12 (m, 1H), 3.91–4.10 (m, 6H), 2.60
(d, J = 13.1 Hz, 1H), 2.49 (dd d, J = 3.1, 3.6,
13.1 Hz, 1H), 1.87 (d, J = 1.2 Hz, 3H), 1.21–1.28 (m,
6H).
1
80%). Compound 7: H NMR (300 MHz, D2O) d 7.56
(d, J = 1.2 Hz, 1H), 6.20 (m, 1H), 5.23 (m, 1H), 3.57 (dd,
J = 8.7, 12.9 Hz, 1H), 3.35 (dd, J = 9.6, 12.9 Hz, 1H),
2.25 (m, 1H), 2.04 (m, 1H), 1.81 (d, J = 1.2 Hz, 3H). 31P
NMR (121.4 MHz, D2O) d 13.6. LR LCMS(B) (MꢀH)
304.9 (0.28 min). FABHRMS (M+Na) Calc. for C10H15
N2O7NaP 329.0515, found 329.0510. HPLC 8.92 min,
99.4%.
6.2.12. (2R,5R)-[5-(5-Methyl-2,4-dioxo-3,4-dihydro-2H-
pyrimidin-1-yl)-tetrahydro-furan-2-yloxymethyl]-phosphonic
acid diisopropoxycarbonyloxymethyl ester (23). The diso-

5526
R. L. Mackman et al. / Bioorg. Med. Chem. 15 (2007) 5519–5528
dium salt 7 (20 mg, 0.057 mmol) was dissolved in acetoni-
trile (2 mL) and water (0.4 mL) and cooled to 0 ꢁC. The
stirred solution was treated with Dowex-H+ resin for
10 min and then the resin was removed by filtration. The
resin was washed with methanol and the combined eluants
were treated with triethylamine (0.1 mL). The solution was
concentrated under reduced pressure and then redissolved
in DMF (1 mL) and triethylamine (0.1 mL). The solution
was concentrated under reduced pressure once again and
the residue then treated with DMF (1 mL) and triethyl-
amine (0.07 mL). The mixture was stirred and chlorometh-
ylisopropylcarbonate (0.06 mL) was added. The mixture
was heated at 70 ꢁC for 3 h and then stirred at room tem-
perature overnight. The mixture was concentrated under
reduced pressure and the residue added to ethyl acetate
(5 mL) and water (0.2 mL). The mixture was filtered and
the eluant was then dried over sodium sulfate, filtered,
and concentrated under reduced pressure. The residue
was subjected to silica gel chromatography eluting with a
gradient of 50–100% ethyl acetate in hexanes to give the
ture was stirred for 2 h and then treated with bis(tributy-
lammonium)pyrophosphate (190 mg, 0.4 mmol). The
resultant mixture was stirred for 16 h and then filtered.
The eluant was concentrated under reduced pressure and
the residuechromatographedona DEAEcolumn. The col-
umn was eluted with 0–0.4 M ammonium carbonate in
water. The fractions containing product were collected
and concentrated under reduced pressure. The residue
was dissolved in water and concentrated under reduced
pressure (repeat 4 times). The residue was then dissolved
in ethanol and concentrated under reduced pressure four
times. The crude residue was then freeze-dried to give 26
1
as a white solid (3 mg, 9%). Compound 26: H NMR
(300 MHz, D2O) d 7.54 (s, 1H), 6.25 (m, 1H), 5.27 (s,
1H), 3.79 (m, 2H), 2.28 (m, 1H), 2.05 (m, 3H), 1.82 (s,
3H). 31P NMR (121.4 MHz, D2O) d 7.9 (d, J = 25 Hz,
Pa), ꢀ5.5 (d, J = 20 Hz, Pc), ꢀ21.3 (m, Pb). HPLC
8.84 min, 97.4%.
6.3. NMR conformation determination
1
title product 23 (7.9 mg, 26%). Compound 23: H NMR
(300 MHz, CDCl3) d 8.12 (s, 1H), 7.38 (d, J = 1.2 Hz,
1H), 6.48 (m, 1H), 5.75 (m, 4H), 5.19 (d, J = 3.9 Hz,
1H), 4.93 (m, 2H), 4.07 (dd, J = 9.0, 13.8 Hz, 1H), 3.88
(dd, J = 10.2, 13.8 Hz, 1H), 2.07–2.38 (m, 4H), 1.33 (m,
15H). 31P NMR (121.4 MHz, CDCl3) d 21.2. LR
LCMS(A) (M+Na) 561.2 (2.11 min). FABHRMS
(M+Li) calc. for C20H31LiN2O13P 545.1724, found
545.1715. HPLC 25.32 min, 100%.
The JHH coupling constants were measured from one-
dimensional proton spectrum, collected with a spectral
width of 5000 Hz and 32K complex points. Two hundred
milliseconds of ROESY spectra were collected with 2048
complex points in t2, and 150 fids. The spectral width was
5000 Hz. Gaussian function was used in both dimensions.
Zero filling was used to end up with 2048 · 2048 points.
The NOE volumes were from the 200 ms ROESY data.
All data processing, including NOE volume integration,
was done by the Varian NMR software, VNMR. The
ensemble solution conformations of nucleosides and
phosphonates were derived from JHH and NOE data.
Coupling constant data were used to extract vicinal
hydrogen dihedral angles according to the Altona–Kar-
plus equation.21 H–H distance constraints falling off as
1/r6 were derived from ROESY NOE data. The molecular
species of interest were generated and manipulated using
the SYBYL computational package [SYBYL 6.92, Tripos
Inc., 1699 South Hanley Rd., St. Louis, Missouri, 63144,
USA]. Each molecule was minimized to convergence
using the MMFF94s force field, as implemented in Sybyl
6.92, applying the experimentally derived dihedral and
distance constraints. All calculations were carried out in
vacuum with a constant dielectric of e = 4.0. The resulting
molecular conformation was used to calculate ensemble-
average values for the pseudorotation angle P and the
amplitude of sugar pucker w.21
6.2.13.
(2R,5R)-[5-(5-Methyl-2,4-dioxo-3,4-dihydro-2H-
pyrimidin-1-yl)-2,5-dihydro-furan-2-yloxymethyl]-phosphon-
ic acid diisopropoxycarbonyloxymethyl ester (24). Prepared
from 8 according to the procedure described for 23. Com-
1
pound 24: H NMR (300 MHz, CDCl3) d 8.45 (s, 1H),
7.15 (s, 1H), 6.98 (s, 1H), 6.32 (d, J = 8.7 Hz, 1H), 6.1 (d,
J = 8.7 Hz, 1H), 5.6–5.8 (m, 5H), 4.95 (m, 2H), 4.0–4.12
(m, 2H), 1.92 (s, 3H), 1.35 (d, J = 7.6 Hz, 12H). 31P NMR
(121.4 MHz, CDCl3) d 20.77. LR LCMS(A) (M+Na)
559.6. FABHRMS (M+Li) calc. for C20H29LiN2O13P
543.1567, found 543.1563. HPLC 24.84 min, 99.3%.
6.2.14.
(2R,5R)-[5-(5-Methyl-2,4-dioxo-3,4-dihydro-2H-
pyrimidin-1-yl)-2,5-dihydro-furan-2-yloxymethyl]-phosphon-
ic acid diphosphate (25). Prepared from 8 according to the
procedure described for 26. The product was purified using
TEA bicarbonate buffer, instead of ammonium carbonate,
1
to isolate the bis-triethylamine salt of compound 25: H
NMR (300 MHz, D2O) d 7.21 (s, 1H), 6.66 (s, 1H), 6.27
(d, J = 6.0 Hz, 1H), 6.02 (d, J = 6.3 Hz, 1H), 5.76 (s, 1H),
3.77 (m, 2H), 2.98 (m, 12H), 1.98 (s, 3H), 1.12 (m, 18H).
31P NMR (121.4 MHz, D2O) d 6.8 (d, J = 22 Hz, Pa),
ꢀ12.6 (d, J = 22 Hz, Pc) ꢀ25.3 (m, Pb). LR LCMS(A)
(M+H) 464. HPLC 9.05 min, 100%.
6.4. Biological assays
HIV RT inhibition assay wild-type p66/p51 heterodimer
HIV-1 reverse transcriptase was cloned, expressed, and
isolated as described previously.6 Enzymatic RT reac-
tions contained the following components: 50 mM
Tris–HCl, pH 7.8, 60 mM KCl, 5 mM MgCl2, 3 mM
DTT, 150 lg/mL activated calf thymus DNA, 500 lg/mL
bovine serum albumin, 5% glycerol, cognate [33P]dNTP
at the Km concentration (0.15–0.25 lM depending on
dNTP), 25 lM each of the remaining three dNTPs,
and serial dilutions of the tested nucleotide triphosphate
analog. Each inhibitor concentration was set up in
duplicate and pre-incubated at 37 ꢁC. Reactions were
6.2.15. (2R,5R)-[5-(5-Methyl-2,4-dioxo-3,4-dihydro-2H-
pyrimidin-1-yl)-tetrahydro-furan-2-yloxymethyl]-phosphon-
ic acid diphosphate (26). The ddTP disodium salt 7 was
treated as described in the preparation of 23 to give the free
acid of 7. The acid (24 mg, 79 lmol) was dissolved in DMF
(2 mL) and then CDI (65 mg, 0.40 mmol), dissolved in
DMF (1 mL), was added. The mixture was stirred for
16 h and then quenched with methanol (24 lL). The mix-

R. L. Mackman et al. / Bioorg. Med. Chem. 15 (2007) 5519–5528
5527
3. http://AIDSinfo.nih.gov. Guidelines for the Use of Anti-
retroviral Agents in HIV-1 Infected Adults and
Adolescents.
4. Miller, M. D.; Margot, N.; Lu, B.; Zhong, L.; Chen, S.-S.;
Cheng, A.; Wulfsohn, M. J. Infect. Diseases 2004, 189,
837.
5. Robbins, B. L.; Srinivas, R. V.; Kim, C.; Bischofberger,
N.; Fridland, A. Antimicrob. Agents Chemother. 1998, 42,
612.
started by adding RT enzyme to a final concentration of
5 nM. Following the incubation for 6 and 12 min, reac-
tion aliquotes were transferred into EDTA and spotted
onto Unifilter 96-well CE81 DEAE plate. Filter plate
was washed with 125 mM Na2HPO4 and the radioactiv-
ity on filter was counted after adding liquid scintillant.
Percent inhibition and IC50 values were determined
using GraphPad Prizm program.
6. White, K. L.; Margot, N. A.; Ly, J. K.; Chen, J. M.; Ray,
A. S.; Pavelko, M.; Wang, R.; McDermott, M.; Swami-
nathan, S.; Miller, M. D. AIDS 2005, 19, 1751.
7. Parikh, U. M.; Koontz, D. L.; Chu, C. K.; Schinazi, R. F.;
Mellors, J. W. Antimicrob. Agents Chemother. 2005, 49,
1139.
8. Kim, C. U.; Bronson, J. J.; Ferrare, L. M.; Martin, J. C.
Bioorg. Med. Chem. Lett. 1992, 2, 367.
9. Secrist, J. A., III; Riggs, R. M.; Comber, R. N.;
Montgomery, J. A. Nucleosides Nucleotides Nucleic Acids
1992, 11, 947.
10. Freeman, G. A.; Rideout, J. L.; Miller, W. H.; Reardon, J.
E. J. Med. Chem. 1992, 35, 3192.
11. Rajwanshi, V. K.; Prhvac, M.; Fagan, P.; Brooks, J. L.;
Hurd, T.; Cook, D. P.; Wang, G. Nucleosides Nucleotides
Nucleic Acids 2005, 24, 179.
12. (a) Teng, K.; Bharadwaj, R.; Cook, D. P. Synlett 1996,
346; (b) Teng, K.; Cook, P. D. J. Org. Chem. 1994, 59,
278.
13. Avino, A.; Garcia, R. G.; Diaz, A.; Albericio, F.;
Eritja, R. Nucleosides Nucleotides Nucleic Acids 1996,
15, 1871.
14. Epp, J. B.; Widlanski, T. S. J. Org. Chem. 1999, 64, 293.
15. Sato, Y.; Tateno, G.; Seio, K.; Sekine, M. Tetrahedron
Lett. 2002, 43, 3251.
16. O’Dell, C. A.; Shealy, Y. F. Nucleosides Nucleotides
Nucleic Acids 1994, 13, 1929.
17. Kim, C. U.; Luh, B. Y.; Martin, J. C. J. Org. Chem. 1991,
56, 2642.
18. Arimilli, M. N.; Kim, C. U.; Dougherty, J.; Mulato, A.;
Oliyai, R.; Shaw, J. P.; Cundy, K. C.; Bischofberger, N.
Antiviral Chem. Chemother. 1997, 8, 557.
19. Huang, H.; Chopra, R.; Verdine, G. L.; Harrison, S. C.
Science 1998, 282, 1669.
20. Reardon, J. E. Biochemistry 1992, 31, 4473.
21. (a) Altona, A.; Sundaralingam, M. J. Am. Chem. Soc.
1973, 95, 2333; (b) Altona, C.; Sundaralingam, M. J. Am.
Chem. Soc. 1972, 94, 8205.
22. (a) Plavec, J.; Koole, L. H.; Chattopadhaya, J. J.
Biochem. Biophys. Methods 1992, 25, 253; (b) Plavec, J.;
Tong, W.; Chattopadhyaya, J. J. Am. Chem. Soc. 1993,
115, 9734.
23. Using the methods described herein the conformational
parameters for 20,30-dideoxyuridine were determined to be
P = 48ꢁ, w = 27ꢁ and for 30 deoxythymidine to be P = 42ꢁ,
w = 27ꢁ, respectively.
6.5. Antiviral activity and cytotoxicity
Appropriate compound dilutions were prepared in tripli-
cate in 96-well plates. MT-2 cells were infected in bulk
with HIV-1 IIIb for 3 h at 37 ꢁC and added to plates at
a density of 20,000/well in a final volume of 200 lL. After
a 5-day incubation at 37 ꢁC, the virus-induced cytopathic
effect was determined using a cell viability assay. 100 lL
of media was removed from each well and replaced
with 100 lL of phosphate-buffered saline containing
1.7 mg/mL XTT [2,3,-bis(methoxy-4-nitro-5-sulfophe-
nyl)-2H-tetrazolium-5-carboxanilide] and 5 lg/mL phen-
azine methosulfate. Following a 1-h incubation at 37 ꢁC,
20 lL/well of 2% Triton X-100 was added and absorbance
was read at 450 nm. The data were plotted as cell death
versus drug concentration. Cell death was expressed as a
percentage of the signal from samples with fully sup-
pressed virus replication following the subtraction of sig-
nal from untreated infected control. The concentration of
each drug that inhibited the virus-induced cytopathic ef-
fect by 50% (EC50) was calculated using GraphPad Prizm
program. Cytotoxicity in MT-2 cells was determined un-
der identical conditions except that cells were uninfected
and higher drug concentrations were tested.
6.6. Resistance profiling
All HIV-1 variants containing drug resistance mutations
in reverse transcriptase were recombinant strains and
were kindly provided by Nicolas Margot of Gilead Sci-
ences Inc. Molecular construction and characterization
of viruses with K65R or M184V has been previously de-
scribed.33 HIV-1 strain designated 6TAMs contains the
reverse transcriptase gene isolated from an NRTI-experi-
enced patient with M41L/D67N/K70R/L210W/T215Y/
K219Q mutations. The virus has been produced by
homologous recombination as described for the other
two mutants.33 All viruses were propagated in MT-2 cells,
titered, and used in parallel with wild-type HIV-1 HXB2
strain to determine the fold resistance of tested inhibitors
using the above described XTT-based cytopathicity
assay.
24. Marquez, V. E.; Ezzitouni, A.; Russ, P.; Siddiqui, M. A.;
Ford, H., Jr.; Feldman, R. J.; Mitsuya, H.; George, C.;
Barchi, J. J., Jr. J. Am. Chem. Soc. 1998, 120, 2780.
25. Harte, W. E., Jr.; Starrett, J. E.; Martin, J. C.; Mansuri,
M. M. Biochem. Biophys. Res. Commun. 1991, 175, 298.
26. Lavie, A.; Konrad, M. Mini-Rev. Med. Chem. 2004, 4,
351.
Acknowledgment
The authors thank M. Desai for helpful discussions
during the preparation of this manuscript.
27. White, K. L.; Chen, J. C.; Feng, J. Y.; Margot, N. A.; Ly,
J. K.; Ray, A. S.; MacArthur, H. L.; McDermott, M. J.;
Swaminathan, S.; Miller, M. D. Antiviral Ther. 2006, 11,
155.
References and notes
1. De Clercq, E. J. Clin. Virol. 2004, 30, 115.
2. Chapman, T. M.; McGavin, J. K.; Noble, S. Drugs 2003,
63, 1597.
28. Sluis-Cramer, N.; Arion, D.; Kaushik, N.; Lim, H.;
Parniak, M. A. Biochem. J. 2000, 348, 77.
29. Engel, R. Chem. Rev. 1977, 77, 349.

5528
R. L. Mackman et al. / Bioorg. Med. Chem. 15 (2007) 5519–5528
30. (a) Arion, D.; Kaushik, N.; McCormick, S.; Borkow, G.;
Parniak, M. A. Biochemistry 1998, 37, 15908; (b) Meyer, P. R.;
Matsuura,S.E.;So,A.G.;Scott,W.A.PNAS 1998,95, 13471.
31. Boyer, P. L.; Sarafianos, S. G.; Arnold, E.; Hughes, S. H.
J. Virology 2001, 75, 4832.
32. Stanford HIV drug resistance database http://hivdb.
stanford.edu/, 2006.
33. White, K. L.; Margot, N. A.; Wrin, T.; Petropoulos, C. J.;
Miller, M. D.; Naeger, L. K. Antimicrob. Agents Chemo-
ther. 2002, 46, 3437.