Synthesis and anti-HIV activity of 2′-fluorine modified nucleoside phosphonates: Analogs of GS-9148

Article abstract of DOI:10.1016/j.bmcl.2007.11.126

Modified purine analogs of GS-9148 [5-(6-amino-purin-9-yl)-4-fluoro-2,5-dihydro-furan-2-yloxymethyl]-phosphonic acid (2′-Fd4AP) were synthesized and their anti-HIV potency evaluated. The antiviral activity of guanosine analog (2′-Fd4GP) was comparable that of to 2′-Fd4AP in MT-2 cells, but selectivity was reduced.

Full text of DOI:10.1016/j.bmcl.2007.11.126

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 1116–1119
Synthesis and anti-HIV activity of 20-fluorine modified nucleoside
phosphonates: Analogs of GS-9148
Richard L. Mackman,* Kuei-Ying Lin, Constantine G. Boojamra, Hon Hui,
Janet Douglas, Deborah Grant, Oleg Petrakovsky, Vidya Prasad,
Adrian S. Ray and Tomas Cihlar
Gilead Sciences, Inc., 333 Lakeside Drive, Foster City, CA 94404, USA
Received 17 November 2007; accepted 30 November 2007
Available online 5 December 2007
Abstract—Modified purine analogs of GS-9148 [5-(6-amino-purin-9-yl)-4-fluoro-2,5-dihydro-furan-2-yloxymethyl]-phosphonic acid
(20-Fd4AP) were synthesized and their anti-HIV potency evaluated. The antiviral activity of guanosine analog (20-Fd4GP) was com-
parable that of to 20-Fd4AP in MT-2 cells, but selectivity was reduced.
Ó 2007 Elsevier Ltd. All rights reserved.
NH2
N
N
O
There are currently eight approved nucleos(t)ide HIV re-
verse transcriptase (RT) inhibitors (N(t)RTI), but due to
cross resistance and long term toxicities there remains a
clinical need to develop novel n(t)RTI that can expand
the current treatment options, especially for HIV pa-
tients harboring resistant virus.1 In the preceding paper,
we described the synthesis and anti-HIV evaluation of
GS-9148 (1, [5-(6-amino-purin-9-yl)-4-fluoro-2,5-dihy-
dro-furan-2-yloxymethyl]-phosphonic acid, 20-Fd4AP),
a novel nucleoside phosphonate RT inhibitor (Fig. 1).2
GS-9148 has a promising antiviral resistance profile,
retaining potency toward HIV RT resistant virus con-
taining M184V, multiple thymidine analog mutations
(TAMs), and K65R resistance mutations.3 GS-9131
(2), an ethylalaninyl phosphonoamidate prodrug of
GS-9148, demonstrated excellent potency toward multi-
ple subtypes of HIV-1 clinical isolates (mean
EC50 = 37 nM), several fold better than 30-azido-20,30-
dideoxythymidine (3, AZT).3 The amidate prodrug is
readily hydrolyzed inside lymphoid cells to parent 1
and subsequently anabolized to the active diphosphate
metabolite.4 The diphosphate metabolite has demon-
strated a long intracellular half-life in vitro (activated
CD4+ cells of 19 h) and in vivo (circulating dog PBMCs
of >24 h), a property that would support once-daily dos-
ing.3,4 Based on these data we have continued to explore
NH2
N
N
N
N
O
O
P
N
N
O
HN
O
O
O
P
O
O
HO
OH
F
F
1. GS-9148 (2'-Fd4AP)
2. GS-9131
O
NH2
N
N
NH
N
N
O
P
OH
N
O
O
O
O
HO
HO
N3
3. AZT
4. d4AP
Figure 1. HIV RT inhibitors.
the SAR around GS-9148 with the aim of further
improving the potency and resistance profile.
The 20-fluorine group of 1 was rationally designed into
d4AP (4) to address undesirable activity toward host
mitochondrial DNA c polymerase.2,3,5 The success of
this modification demanded that it be retained in future
analogs. Hence the next generation analogs of 1 focused
on modifications of the purine base and reducing the
saturated ring. The saturated beta-fluorine analog 6
was easily prepared by radical deselanation of 5, an
intermediate in the synthesis of 1, followed by hydrolysis
of the ester and pivaloyl groups (Scheme 1).2,6 The syn-
thetic route that yielded 1 also formed the basis for the
preparation of additional purine analogs 19, 24, and 26
Keywords: HIV; Phosphonate; Antiviral.
*
Corresponding author. Tel.: +1 650 522 5258; fax: +1 650 522
5899; e-mail: rmackman@gilead.com
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.11.126

R. L. Mackman et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1116–1119
1117
adenosine deaminase type II.9 Attempts to oxidize 12
to the 50-carboxylic acid without protection of the gua-
nosine base proved difficult. Therefore, the 2-amino
group was protected once again as the 2-N-isobutyryl
amide 13. Acetylation of the ribose hydroxyl groups,
followed by Mitsunobu mediated installation of the 6-
O benzyl ether, and subsequent removal of the acetyl
groups yielded the base protected intermediate 14.11 14
was converted to acid 15, using platinum-catalyzed oxi-
dation in 37% yield.12 Decarboxylative dehydration of
acid 15 by heating in the presence of DMF-dineopentyl-
alcohol formed the intermediate glycal 16, which, with-
out purification, was activated with PhSeCl and
converted to phosphonate 17.13 The instability of the
glycal intermediate was also a challenge in this synthetic
route and care had to be used to avoid extensive depu-
rination and furan formation. Oxidation of selenium
by O3 followed by elimination provided olefin derivative
18. Finally, 18 was treated with TMSI in DMF and
2,4,6-collidine at 45 °C to simultaneously remove both
the phosphonate ester groups and the 6-O benzyl ether.
Treatment with concentrated ammonium hydroxide to
remove the isobutyryl group then yielded the ammo-
nium salt of guanosine phosphonate 19.
NHpiv
NH2
N
N
N
N
N
N
N
(i)-(iii)
O
Et2O3P
O
O
O
N
H2O3P
F
PhSe
F
5
6, 2'-FddAP
Scheme 1. Reagents and conditions: (i) n-Bu3SnH, AIBN, benzene; (ii)
NaOMe, MeOH, 100%; (iii) TMSBr, DMF, lutidine, 50 °C, then
NH4OH.
(Scheme 2 and 3). The focus was directed specifically to
purine analogs since earlier pyrimidine analogs of 4
demonstrated poor resistance profiles toward key HIV
mutations.7 Guanosine analog 19 was prepared from
the fluoro sugar 7 according to Scheme 2. Initially, con-
version of 7 to the anomeric bromide, by treatment with
HBr,2 was followed by the attempted addition of 2-N-
isobutyryl amide, 6-O diphenylcarbamoyl protected
guanine to provide 8. Unfortunately, the addition was
low-yielding and also resulted in significant amounts
of the N-7 substituted product that made efficient isola-
tion of the desired N-9 product problematic. Explora-
tion of alternative guanine precursors identified 2,6-
dichloropurine as the optimal precursor, which coupled
with the anomeric bromide to yield intermediate 9 in
64% yield.8 Treatment of 9 with sodium azide in reflux-
ing EtOH resulted in 10 which was then reduced and
debenzoylated using NaBH4 in ethanol to provide diam-
inopurine nucleoside 11.9,10 Diaminopurine 11 was then
enzymatically converted to guanosine analog 12, by
A logical approach to the 2,6-diaminopurine phospho-
nate analog 24 was to utilize the 2,6-diaminopurine
nucleoside 11, but attempts to form the corresponding
glycal intermediate, after oxidiation to the 50-carboxylic
acid, were unsuccessful. Therefore, an alternative route
from 2,6-dichloro intermediate 9 was pursued (Scheme
R
NH2
N
O
N
N
N
N
N
N
N
NH
R
O
OBz
i, ii
iii
O
O
N
R1
N
NH2
v
BzO
BzO
iv
O
N
BzO
BzO
HO
HO
F
F
HO
F
HO
12 R = NH2
13 R = NHCOiPr
F
7
11
8
9
R = OCON(Ph)2, R1 = NHCOiBu
R = R1 = Cl
10 R = R1 = N3
vi
vii-ix
OBn
N
OBn
N
OBn
N
OBn
N
N
N
O
N
N
N
N
N
O
O
O
O
xi
O
N
N
H
x
N
O
F
N
N
O
N
N
HO
O
N
N
Et2O3P
O
HO
H
H
H
HO
F
HO
15
F
PhSe
F
14
17
16
xii
O
N
OBn
N
N
N
N
NH
NH2
O
xiii-xiv
O
O
N
H2O3P
O
O
N
N
H
Et2O3P
F
F
19, 2'-Fd4GP
18
Scheme 2. Reagents and conditions: (i) HBr, AcOH; (ii) NaH, CH3CN, 2,6-dichloropurine, 64%; (iii) NaN3, EtOH, reflux, 1 h, 100%; (iv) NaBH4,
EtOH , dioxane; (v) adenosine deaminase type II, rt, 16 h, 100%; (vi) TMSCl, pyridine, (iPrCO)2O, concd NH4OH, 70%; (vii) Ac2O, Py, 71%; (viii)
t
Ph3P, BnOH, DIAD, dioxane; (ix) NaOMe, MeOH, rt, 44% (two steps); (x) O2, 10% Pt/C, H2O, 60 °C, 24 h, 37%; (xi) a—Me2NCH(OCH2 Bu)2,
CH3SO3H, 100 °C; b—PhSeCl, HOCH2P(O)(OEt)2, AgClO4; (xii) O3, CH2Cl2, 0–45 °C; (xiii) TMSI, 2,4,6-collidine, DMF, 45 °C; (xiv) concd
NH4OH, 45 °C, 5 h.

1118
R. L. Mackman et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1116–1119
Cl
N
N
OMe
N
OMe
N
N
N
N
N
N
N
O
ii
i
O
9
Cl
O
N
Cl
O
N
Cl
BzO
BzO
HO
HO
F
HO
F
HO
F
20
21
iii-iv
NH2
N
NH2
N
OMe
N
N R
N
N
N
vi
N
v
O
P
OH
O
P
OH
O
N
NH2
N
O
N
O
N
NH2
O
O
O
Et2O3P
EtO
HO
F
F
F
23
22 R = Cl
25 R = H
24, 2'-Fd4DAPP
vii
O
N
N
NH
O
O
N
H2O3P
F
26, 2'-Fd4IP
t
Scheme 3. Reagents and conditions: (i) NaOMe, MeOH, 85%; (ii) O2, 10% Pt/C, H2O, NaHCO3, 70 °C, 24 h, 57%; (iii) a—Me2NCH(OCH2 Bu)2,
CH3SO3H, 110 °C; b—PhSeCl, HOCH2P(O)(OEt)2, AgClO4; (iv) O3, CH2Cl2, 0–45 °C; (v) concd NH4OH–EtOH (5:1 ratio), 140 °C; (vi) TMSBr,
2,4,6-collidine, DMF, 45 °C, 5 h; (vii) TMSBr, DMF, 50 °C; then NH4OH, 45%.
3). Treatment of 9 with NaOMe in methanol generated
nucleoside 20. Platinum-catalyzed oxidation provided
the 50-carboxylic acid 21, and conversion to the phos-
phonate 22 via the glycal proceeded smoothly. Ammo-
nolysis in a steel bomb at 140 °C overnight displaced
the chlorine, methoxy group, and one phosphonate
ethyl group to afford the 2,6 diaminopurine phosphonic
acid monoester 23. Subsequent treatment with TMSBr
then generated the target 2,6 diaminopurine phospho-
nate 24. Inosine nucleoside phosphonate 26 was easily
obtained from 2514 in 45% yield by initial deprotection
of the phosphonate ester groups using TMSBr, followed
by treatment with ammonium hydroxide.
cytotoxicity assay there was a clear difference in selectiv-
ity, in favor of 1. It is interesting to note that the corre-
sponding nucleoside analogs of 1 and 19 (i.e., 20-Fd4A
and 20-Fd4G) have been reported and the guanosine
nucleoside analog was also marginally less potent than
the corresponding adenine analog (5-fold).15 The hypo-
xanthine and 2,6-diaminopurine nucleoside phospho-
nate analogs, 26 and 24, respectively, were inactive up
to 300 lM. This is in contrast to the corresponding
nucleoside hypoxanthine and 2,6-diaminopurine ana-
logs, 20-Fd4I and 20-Fd4DAP, respectively, which were
both reported to be quite potent (EC50 < 5 lM).15 These
results suggest that these base modified nucleoside phos-
phonates may be limited in their intracellular metabolic
pathways to generate the active diphospho-phospho-
nates (triphosphate equivalent). The reduced 20-F-20,30-
dideoxy adenine phosphonate analog 6 demonstrated
weak antiviral activity. Previous studies on non-fluori-
nated 20,30-dideoxy phosphonate analogs have shown
that the saturated ribose ring is inferior in potency to-
ward HIV RT compared to the unsaturated 20,30-dide-
oxy-20,30-didehydro ribose ring, potentially due to the
saturated ribose ring adopting a more southern confor-
mation.7 The 20-beta-fluorine group would be expected
to also favor a southern conformation. Therefore, the
diminished antiviral activity of 6 is likely due, in part,
to reduced potency toward HIV RT.
Guanosine phosphonate, 19 (20-Fd4GP), displayed anti-
viral activity (EC50 = 17.3 lM) in a similar range to that
of adenosine analog, 1 (20-Fd4AP), and ꢀ5-fold less po-
tent than tenofovir (Table 1). However, in the MT-2 cell
Table 1. Anti-HIV activity and cytotoxicty
Compound
Code
WT HIVa
MT-2b
EC50 (lM)
CC50 (lM)
Tenofovir
20-Fd4AP
20-Fd4GP
20-Fd4IP
20-Fd4DAPP
20-FddAP
3.6 (1.5)
12.3 (3.4)
17.4 (5.3)
>300
>1000
>1000
700
1
19
26
24
6
—
—
>300
283
>1000
Synthesis of several purine and ribose modified analogs
of the 20-fluoro modified nucleoside phosphonate GS-
9148, 1, identified the guanosine analog, 19, as a potent
but less selective nucleoside phosphonate inhibitor of
HIV.
a Antiviral activity in MT-2 cells infected with HIV-1 IIIb. Values are
the results of at least 2 experiments, standard deviation given in
parentheses.
b Cytotoxicity in uninfected MT-2 cells.

R. L. Mackman et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1116–1119
1119
7. 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.
8. Shortnacy-Fowler, A. T.; Tiwari, K. N. Nucleosides
Nucleotides 2001, 20, 1583.
References and notes
1. Gallant, J. E.; Gerondelis, P. Z.; Wainberg, M. A.;
Shulman, N. S.; Haubrich, R. H.; St Clair, M.; Lanier,
E. R.; Hellmann, N. S.; Richman, D. D. Antiviral Ther.
2003, 8, 489.
9. Chavis, C.; Dumont, F.; Wightman, R. H.; Ziegler, J. C.;
Imbach, J. L. J. Org. Chem. 1982, 47, 202.
10. Gaffner, B. L.; Ti, G. S.; Jones, R. A. J. Am. Chem. Soc.
1982, 104, 1316.
11. Khazanchi, R.; Yu, P.-L.; Johnson, F. J. Org. Chem. 1993,
58, 2552.
2. Boojamra, C. G. et al. preceding paper.
3. Cihlar, T.; Ray, A. S.; Boojamra, C.; 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.; Carey, A.; Pakdaman, R.; Mackman, R. Antimicrob.
Agents Chemother. accepted for publication.
12. Moss, G. P.; Reese, C. B.; Schofield, K.; Shapiro, R.;
Todd, A. R. J. Chem. Soc. 1963, 1149.
13. Kim, C. U.; Luh, B. Y.; Martin, J. C. J. Org. Chem. 1991,
56, 2642.
14. Boojamra, C. G.; Lin, K.-Y.; Mackman, R. L.; Markev-
itch, D. Y.; Petrakovsky, O. V.; Ray, A. S.; Zhang, L.
PCT International Patent Application WO 06/110157
2006.
4. Ray. A. S.; Vela, J. E.; Boojamra, C. G.; Zhang, L.; Hui,
H.; Callebaut, C.; Stray, K.; Mackman, R. L.; Cihlar, T.
Antimicrob. Agents Chemother. accepted for publication.
5. 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.; Markev-
itch, D.; Vela, J.; Ray, A.; Cihlar, T. Bioorg. Med. Chem.
Lett. 2007, 17, 6785.
15. Lee, K.; Choi, Y.; Gumina, G.; Zhou, W.; Schinazi, R.;
Chu, C. K. J. Med. Chem. 2002, 45, 1313.
6. Bhalchandra, J.; Reese, C. B.; Varaprasad, C. V. N. S.
Nucleosides Nucleotides 1995, 14, 209.