Synthesis and biological evaluation of phosphonate analogues of nevirapine

Article abstract of DOI:10.1016/j.bmcl.2012.12.049

A series of nevirapine-based analogues containing the phosphonate functionality were prepared and evaluated in vitro against HIV RT. The effect of the phosphonate was evaluated against the wild type and Y181C HIV replication. An in vivo PK study was perfo

Full text of DOI:10.1016/j.bmcl.2012.12.049

Bioorganic & Medicinal Chemistry Letters 23 (2013) 1493–1497
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and biological evaluation of phosphonate analogues of
nevirapine
⇑
Jay Parrish , Leah Tong, Michael Wang, Xiaowu Chen, Eric B. Lansdon, Carina Cannizzaro, Xubin Zheng,
Manoj C. Desai, Lianhong Xu
Gilead Sciences, Inc., 333 Lakeside Drive, Foster City, CA 94404, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of nevirapine-based analogues containing the phosphonate functionality were prepared and
evaluated in vitro against HIV RT. The effect of the phosphonate was evaluated against the wild type
and Y181C HIV replication. An in vivo PK study was performed on a select analogue.
Ó 2012 Elsevier Ltd. All rights reserved.
Received 29 October 2012
Revised 8 December 2012
Accepted 13 December 2012
Available online 21 December 2012
Keywords:
HIV
Reverse transcriptase
Non-nucleoside reverse transcriptase
inhibitors
NNRTI
Tremendous advances in therapeutic options for the treatment
of human immunodeficiency virus (HIV) infection have been
achieved.1 There are now more than twenty antiretroviral agents
covering six classes marketed to treat HIV, with the non-nucleoside
reverse transcriptase inhibitors (NNRTI) remaining one of the key
components of highly active antiretroviral therapy (FDA approved
NNRTIs 1–5 shown in Fig. 1).2,3 However, rapid emergence of drug-
resistant viruses that are cross-resistant to other members of this
class limit the broad usage of NNRTI.4 Toward this end, the goal
of our project was to identify an NNRTI with activity against major
HIV-1 reverse transcriptase (RT) mutants.
hydrophobic active site channel. It was hypothesized that the ob-
served increased potency was due to a ‘solvent anchoring’ affect
provided by the phosphonate moiety.5
According to the authors, solvent anchoring enhanced the
degeneracy of the inhibitor binding states and allowed the mole-
cule to effectively adapt to the altered active site of the mutant en-
zyme, which endowed a favorable resistance profile. We set out to
improve upon existing NNRTIs by implementing this phosphonate
strategy in hopes of obtaining a desirable resistance profile. We
herein report the synthesis and evaluation of phosphonate-
containing NNRTIs based on nevirapine scaffold.
Recently in our HIV protease inhibitor (PI) project, the phospho-
nate functional group was explored as a pro-drug handle, designed
to increase half-life (t1/2) through higher intracellular loading.5a
Protease inhibitors containing a diethyl phosphonate were shown
to have favorable resistance profiles (Fig. 2). GS-8374, a darunavir
analogue, which has a diethyl phosphonate with optimal projec-
tion installed at the P2 site of the darunavir scaffold, is markedly
more potent against mutant HIV protease enzymes including most
of the clinically relevant strains.5b Interestingly, when the diethyl
phosphonate-containing inhibitors were co-crystallized with wild
type and mutant HIV protease enzymes, the phosphonate group it-
self was found to have no interactions with the active site. Instead
the group was entirely exposed to solvent at the end of a long,
Prior data from Boehringer Ingelheim detailed novel modifica-
tions of nevirapine (ViramuneÒ, 3) that improved its resistance
profile toward a key clinical mutant Y181C. This included modify-
ing the core and installing an aryl group separated from the core by
a linker of varying atom composition and length (Fig. 3).6 These
second generation nevirapine analogues, such as literature com-
pound 6, were shown to be very active against the RT enzyme as
well as the wild-type virus in cell based assays. Compared with
3, compound 6 was >200-fold more active against the Y181C mu-
tant (Table 1).6 Our modeling studies with compound 6 suggested
that a phosphonate group appended on these second generation
nevirapine compounds would potentially further enhance the
compound’s affinity for RT and improve its resistance profile
through solvent anchoring effects.
We investigated phosponate-containing nevirapine analogues
by examining SAR with two- or three-atom linkers (containing C,
O, or N) between the nevirapine core and the aryl group. As such,
⇑
Corresponding author.
E-mail address: Jay.Parrish@gilead.com (J. Parrish).
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2012.12.049

1494
J. Parrish et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1493–1497
the synthesis for the two-atom C-linked series is detailed in
Scheme 1. Starting from known compound 7, Sonogashira coupling
to the acetylene adduct, followed by a two-step reduction provided
phenol 8 in moderate yield.7 Alkylation of 8 installed the desired
diethyl phosphonate moiety to give 9 in 54%, which was then
hydrolyzed using TMSBr buffered with 2,6-lutidine to reveal diacid
10 in nearly quantitative yield. The highly polar diacids were iso-
lated via HPLC (0–100% MeCN–H2O gradient, 5% TFA modifier).
The synthesis for the two-atom N-linked series is detailed in
Scheme 2. Diethyl hydroxymethylphosphonate was triflated at
low temperature in high yield. The resulting triflate was reacted
with 4-hydroxybenzaldehyde in the presence of Cs2CO3 to afford
A. Next, reduction of nitro group in compound 11 (prepared
according to literature procedures),7 followed by a one-pot reduc-
tive amination of the resulting amine gave diethyl phosphonate 12
in good yield. The phosphonate was unmasked with buffered
TMSBr to afford diacid 13 in 95%.
The inhibitory activity against both the RT enzyme and HIV rep-
lication of the two-atom linker (C- and N-linked) series is shown in
Table 1. Non-phosphonate, second generation compounds 8 and 14
(synthesis not shown), prepared as references, showed cell-based
activity on par with nevirapine (3) against the WT virus. Yet both
compounds demonstrated a significant improvement in resistance
profile against the Y181C mutant relative to compound 3 (16- and
13-fold reduction comparing with >100-fold decrease). Incorporat-
ing the diethyl phosphonate moiety gave 9, which had a further
improvement in activity against WT (1.5- to 3-fold) and Y181C
(1.5- to 2-fold) when compared with 8 and 14. In addition, it is
>40-fold more active against the Y181C mutant than 3, and resis-
tance was improved to 7-fold from >100-fold.
Figure 1. Marketed HIV NNRTI’s.
Figure 2. Darunavir and GS–8374.
Figure 3. Second generation nevirapine compounds.
Table 1
Two-atom linker SAR for C- and N-linked nevirapine analogues
Compound
X
R
IC50 (nM)
EC50 (nM)
Fold change
CC50 (lM)
WT
Y181C
Efavirenz (1)
Nevirapine (3)
6
8
14
9
10
12
13
—
—
—
CH2
CH2
CH2
CH2
NH
NH
—
—
—
H
30
309
69
—
—
100
1810
70
0.55
74
8
43
100
33
>5000
10
>5000
1.6
>10,000
53
670
1273
234
>5000
94
>5000
3.0
>100
7
16
13
7
n/a
9
n/a
—
—
—
—
—
—
71
—
—
CH3
CH2P(O)(OEt)2
CH2P(O)(OH)2
CH2P(O)(OEt)2
CH2P(O)(OH)2
3800

J. Parrish et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1493–1497
1495
Scheme 1. Synthesis of two-atom linker, C-linked phosphonates. Reagents and conditions: (a) 4-benzyloxyphenyl acetylene, PdCl2(PPh3)2, TEA, DMF, 90 °C, 16 h, 60%; (b) H2
(150 psi), PtO2, EtOAc, 60 °C, 2 h, 99%; (c) H2 (1 atm), 10% Pd/C, EtOH, rt, 3 h, 52%; (d) TfOCH2P(O)(OEt)2, Cs2CO3, THF, rt, 54%; (e) TMSBr, 2,6-lutidine, MeCN, rt, 1 h, 99%.
Scheme 2. Synthesis of two-atom linker, N-linked phosphonates. Reagents and conditions: (a) SnCl2Á2H2O, concd HCl, AcOH, rt, 16 h, 82%; (b) A, AcOH, MeOH, rt, 2 h, then
NaBH3CN, rt, 5.5 h, 79%; (c) TMSBr, 2,6-lutidine, MeCN, rt, 3 h, 95%; (d) Tf2O, py, DCM, À78 °C, 3 h, >90%; (e) 4-hydroxybenzaldehyde, Cs2CO3, THF, rt, 3 h, 87%.
In order to improve our understanding of these chemical mod-
ifications, an X-ray co-crystal structure of HIV-1 RT with 9 was
determined at 2.5 Å resolution. As expected, the tri-cyclic core
binds in a similar fashion as nevirapine (Fig. 4). The 2 carbon-
linked phenyl phosphonate extends through a solvent channel
formed by the b13–b14 hairpin loop and sheet b6. Compared to
nevirapine, significant movement of b13–b14 sheets is observed
to accommodate 9. Through the solvent channel, the phenyl moi-
ety makes close contacts to Pro236 and Val106. The diethyl phos-
phonate is able to extend to the protein surface and makes
additional interactions with Pro225 and Ser105. The side-chain of
Ser105 is positioned between the diethyl moieties. Furthermore,
one ethyl is in close proximity with the side-chain of His198,
Arg199, and Thr107.
With the X-ray data in hand, we felt validated to continue our
search for additional modifications. For example, improvements
were seen when the nitrogen linker was employed, where diethyl
phosponate compound 12, with WT and Y181C EC50 = 10 and
94 nM, respectively, was the most potent example with the com-
parable resistance profile in the two-atom linker series. As an addi-
tional benefit, the N-linker series was generally more soluble than
the C-linker series (i.e., 9 vs 12, solubility >2.5-fold higher at pH
1.5). In order to evaluate the role of the diethyl group, phosphonate
diacids 10 and 13 were evaluated for both the carbon- and nitro-
gen-containing linkers. In both cases, the compounds were 20- to
50-fold less potent than the diethyl phosphonate against the en-
zyme and devoid of activity in the cell-based assays due to
permeability.
To further expand upon the SAR, we sought to investigate the
three-atom linker, O-linked series. The synthesis is detailed in
Scheme 3. Based on SAR past precedence, for potency reasons we
decided to incorporate a methyl group at the position meta to
the phosphonate.6 As such, compound 7 was subjected to a two-
carbon homologation under Stille conditions, followed by hydrob-
oration and oxidation to yield hydroxyethyl 15 in moderate yield.
Next, we prepared the coupling partner starting with the TBS pro-
tection of 4-iodo-2-methylphenol in 32% yield. This protected io-
dide was then treated with diethyl phosphite under standard
palladium coupling conditions to afford the phosphonate in mod-
erate yield, which was subsequently deprotected to afford diethyl
4-hydroxy-3-methylphenylphosphonate (B). Alcohol 15 was then
coupled with B under Mitsunobu conditions to afford diethyl phos-
phonate 16 in 18% yield. The diethyl phosphonate was unmasked
with buffered TMSBr to afford diacid 17 in 95%.
Figure 4. Co-crystal structure of 9 and HIV-1 RT (yellow) aligned with nevirapine
(3) co-crystal structure (green), PDB code 1VRT. The alignment was created from
superimposing residues 91–118, 156–215, and 317–319 of the p66 subunit and
137–139 of the p51 subunit. Residues from the b12–b14 sheets were excluded in
the alignment due the large movements observed.
The activity for the three-atom linker series is shown in Table 2,
where the phosphonate is directly connected to the aryl ring and
via a three atom oxygen-containing linker. As a comparison, liter-
ature compound 6 was ninefold more potent against the WT and

1496
J. Parrish et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1493–1497
Scheme 3. Synthesis of three-atom linker, O-linked phosphonates. Reagents and conditions: (a) vinylSnBu3, PdCl2(PPh3)2, TEA, DMF, 90 °C, 5 h, 84%; (b) 9-BBN, THF, rt, 2 h,
then alkaline H2O2, rt, 1 h, 30%; (c) 1 N HCl, THF, rt, 2 h, 89%; (d) B, DIAD, PPh3, THF, rt, 4 h, 18% (e) TMSBr, 2,6-lutidine, MeCN, rt, 3 h, 95%; (f) TBSCl, imid, DCM, rt, 1 h, 32%; (g)
(EtO)2P(O)H, Pd(PPh3)4, TEA, 90 °C, 16 h, 67%; (h) TBAF, THF, rt, 2 h, 89%.
Table 2
Three-atom linker SAR for nevirapine analogues
Compound
R
IC50 (nM)
EC50 (nM)
Fold change
CC50 (lM)
WT
Y181C
Efavirenz (1)
Nevirapine (3)
6
16
17
—
—
30
0.55
74
8.2
3.1
1050
1.6
>10,000
53
3
>100
7
3
>5
—
—
40
50
62
309
69
48
29
CO2H
P(O)(OEt)2
P(O)(OH)2
9
>5000
showed a >200-fold improvement against the Y181C mutant when
compared to 3. Changing the carboxylic acid in 6 to a diethyl phos-
phonate group gave the most potent compound (16) in the series,
which was ꢀ2.6-fold more potent against the WT and more than
>twofold improvement in resistance profile against Y181C than 6
and >20-fold more active than nevirapine against WT RT, all while
maintaining a wide therapeutic index (SI >10,000). Compound 16
had activity that was on par with efavirenz, the most widely used
NNRTI in clinical situations.8 The diacid (17) showed an excellent
IC50 (29 nM), but the compound showed limited activity in cell-
based assays, which we hypothesize is due to permeability issues.
Modeling of compounds 6 and 16 into a more open form of the
NNRTI binding pocket revealed an excellent superposition and
overlap of the pair, indicating that in this case the phosphonate
is an effective isostere replacement of carboxylic acid within HIV
RT (Fig. 5). Solvent anchoring allows the compound to adapt to
changes in the binding pocket more readily. In the case of the
NNRTI binding pocket, the addition of the phosphonate allows
for new interactions to the protein surface. The loss of aromatic
contacts due to the Y181C mutation is mitigated by the solvent
anchoring effect and increasing interactions with the protein.
Due to its excellent potency, compound 16 was selected for fur-
ther profiling. The compound was a crystalline solid (mp = 180 °C)
and displayed modest solubility at pH 1.5 (29
lM) and pH 7.4.
(4 M). Caco-2 assays were run on compounds in this class and
l
no P-gp issues were found. An iv PK study was performed in male
beagle dogs, with PK properties measured in both plasma and
peripheral blood mononuclear cells (PBMC’s) and analyzed by LC/
MS/MS. The performance of 16 in dog was poor: the compound
suffered from rapid clearance (CL = 2.34 L/h/kg) and a short half-
life (t1/2 = 0.48 h) in plasma. Although not investigated in detail
(no metabolite ID studies were performed), we hypothesized that
metabolism issues were responsible for the high clearance.
In conclusion, several phosphonate containing NNRTI com-
pounds based on the second generation nevirapine scaffold have
been prepared and tested in vitro. Utilizing the phosphonate group
provided compounds with excellent anti-viral activities against the
WT and mutant Y181C HIV-1 RT. By using the phosphonate, we
were able to prepare a molecule with potency on par with
Figure 5. Superposition of 6 onto phosphonate 16.

J. Parrish et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1493–1497
1497
Lee, W. A. J. Mol. Biol. 2006, 363, 635; (b) He, G.-X.; Yang, Z.-Y.; Williams, M.;
Callebaut, C.; Cihlar, T.; Murray, B. P.; Yang, C.; Mitchell, M. L.; Liu, H.; Wang, J.;
Arimilli, M.; Eisenberg, E.; Stray, K. M.; Tsai, L. K.; Hatada, M.; Chen, X.; Chen, J.
M.; Wang, Y.; Lee, M. S.; Strickley, R. G.; Iwata, Q.; Zheng, X.; Kim, C. U.;
Swaminathan, S.; Desai, M. C.; Lee, W. A.; Xu, L. MedChemCommun 2011, 2, 1093;
(c) Callebaut, C.; Stray, K.; Tsai, L.; Williams, M.; Yang, Z.-Y.; Cannizzaro, C.;
Leavitt, S. A.; Liu, X.; Wang, K.; Murray, B. P.; Mulato, A.; Hatada, M.; Priskich, T.;
Parkin, N.; Swaminathan, S.; Lee, W.; He, G.-X.; Xu, L.; Cihlar, T. Antimicrob.
Agents Chemother. 2011, 55, 1366.
efavirenz. While the PK results for this compound were less than
desirable, we continue to investigate the phosphonate approach
for the nevirapine series as well as applying it to other structural
classes of NNRTIs. Results will be reported in due course.
Acknowledgment
6. (a) Klunder, J. M.; Hoermann, M.; Cywin, C. L.; David, E.; Brickwood, J. R.;
Schwartz, R.; Barringer, K. J.; Pauletti, D.; Shih, C.-K.; Erickson, D. A.; Sorge, C. L.;
Joseph, D. P.; Hattox, S. E.; Grob, P. M. J. Med. Chem. 1998, 41, 2960; (b) Cywin, C.
L.; Klunder, J. M.; Hoermann, M.; Brickwood, J. R.; David, E.; Grob, P. M.;
Schwartz, R.; Pauletti, D.; Barringer, K. J.; Shih, C.-K.; Sorge, C. L.; Erickson, D. A.;
Joseph, D. P.; Hattox, S. E. J. Med. Chem. 1998, 41, 2972; (c) Yoakim, C.; Bonneau,
P. R.; Déziel, R.; Doyon, L.; Duan, J.; Guse, I.; Landry, S.; Malenfant, E.; Naud, J.;
Ogilvie, W. W.; O’Meara, J. A.; Plante, R.; Simoneau, B.; Thavonekham, B.; Bös,
M.; Cordingley, M. G. Bioorg. Med. Chem. Lett. 2004, 14, 739–742; (d) O’Meara, J.
A.; Yoakim, C.; Bonneau, P. R.; Bös, M.; Cordingley, M. G.; Déziel, R.; Doyon, L.;
Duan, J.; Garneau, M.; Guse, I.; Landry, S.; Malenfant, E.; Naud, J.; Ogilvie, W. W.;
Thavonekham, B.; Simoneau, B. J. Med. Chem. 2005, 48, 5580.
The authors thank would like to thank Drs. Thorsten Kirschberg,
Dean Boojamra, and Richard Mackman for helpful discussions.
References and notes
1. (a) Dunne, M. AIDS 2007, 21, S73; (b) Flexner, C. Nat. Rev. Drug Discovery 2007, 6,
959; (c) De Clercq, E. Nat. Rev. Drug Discovery 2007, 6, 1001.
2. ‘Drugs Used in the Treatment of HIV Infection’ HIV/AIDS Therapies, HIV and
AIDS, FDA: Rockville, http://www.fda.gov/oashi/aids/virals.html.
3. (a) De Clercq, E. Antiviral Res. 1998, 38, 153; (b) Balzarini, J. Curr. Top. Med. Chem.
2004, 4, 921; (c) De Clercq, E. Chem. Biodivers. 2004, 1, 44.
4. (a) Phillips, A. Nat. Med. 2001, 7, 993; (b) De Clercq, E. Expert Opin. Emerg. Drugs
2005, 10, 241.
5. (a) Cihlar, T.; He, G.-X.; Liu, X.; Chen, J. M.; Hatada, M.; Swaminathan, S.;
McDermott, M. J.; Yang, Z.-Y.; Mulato, A. S.; Chen, X.; Leavitt, S. A.; Stray, K. M.;
7. (a) For preparation of iodide 7 and nitro 11, see: (a) Cywin, C. L.; Hoermann, M.;
Klunder, J. M. EP 1997/767172 A1.; (b) Cywin, C. L.; Hoermann, M.; Klunder, J. M.
US 1998/5705499 A and Ref. 6a,b.
8. Oversteegen, L.; Shah, M.; Rovini, H. Nat. Rev. Drug Discovery 2007, 6, 951.