Pyrazole NNRTIs 1: Design and initial optimisation of a novel template

Article abstract of DOI:10.1016/j.bmcl.2009.08.039

The design and synthesis of a novel series of non-nucleoside HIV reverse transcriptase inhibitors (NNRTIs) based on a pyrazole template is described. These compounds are active against wild type reverse transcriptase (RT) and retain activity against clini

Full text of DOI:10.1016/j.bmcl.2009.08.039

Bioorganic & Medicinal Chemistry Letters 19 (2009) 5599–5602
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Pyrazole NNRTIs 1: Design and initial optimisation of a novel template
a
b
b
a
a
Charles E. Mowbray a, , Catherine Burt , Romuald Corbau , Manos Perros , Isabelle Tran , Paul A. Stupple ,
*
Rob Webster c, Anthony Wood a
a Department of Discovery Chemistry, Pfizer Global Research and Development, Ramsgate Road, Sandwich, Kent CT13 9NJ, UK
b Department of Discovery Biology, Pfizer Global Research and Development, Ramsgate Road, Sandwich, Kent CT13 9NJ, UK
c Department of Pharmacokinetics, Dynamics and Metabolism Pfizer Global Research and Development, Ramsgate Road, Sandwich, Kent CT13 9NJ, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
The design and synthesis of a novel series of non-nucleoside HIV reverse transcriptase inhibitors (NNR-
TIs) based on a pyrazole template is described. These compounds are active against wild type reverse
transcriptase (RT) and retain activity against clinically important mutants.
Ó 2009 Elsevier Ltd. All rights reserved.
Received 24 June 2009
Revised 7 August 2009
Accepted 8 August 2009
Available online 13 August 2009
Keywords:
HIV
Reverse transcriptase
NNRTI
The first generation of clinically approved non-nucleoside HIV
reverse transcriptase inhibitors (NNRTIs) efavirenz, nevirapine
and delavirdine are established as key components of highly active
antiretroviral therapy (HAART).1 These drugs block the action of
the viral enzyme reverse transcriptase (RT) by binding to an allo-
steric binding site which is distinct from the catalytic site. In the
clinic combinations containing NNRTIs with other antiretroviral
agents can produce sustained reductions in viral load and slow
the rate of disease progression.2 However the first generation of
approved NNRTIs are ultimately susceptible to mutations in HIV
reverse transcriptase (RT) leading to drug resistance. Viruses with
certain key RT mutations such as K103N can become resistant to all
the first generation of approved NNRTIs.3 In contrast the experi-
mental NNRTI capravirine has been shown to retain excellent
activity against a wide variety of viruses with clinically significant
RT mutants including K103N. Furthermore two or more RT muta-
tions are required to achieve high level resistance to capravirine.4
Capravirine produced marked reductions in viral load in HIV in-
fected patients and it was hoped it would also produce reductions
in viral load in patients failing existing combination therapies.5
Following the disclosure of these pre-clinical and clinical trial re-
sults for capravirine we believed that there was an excellent
opportunity to identify still better NNRTIs which retain the broad
spectrum of activity against clinically significant HIV with muta-
tions in RT whilst reducing the required dose and pill burden, min-
imising the risk of side effects and drug–drug interactions and
hence improving the ease of compliance.6
Cl
N
O
Cl
Cl
NH2
O
N
S
CF3
N
O
NH
O
efavirenz
capravirine
As a starting point for our programme we wished to identify a
novel NNRTI series which would retain the excellent inhibition of
RT mutants demonstrated by capravirine whilst minimising the
potential for rapid in vivo clearance which would necessitate a
high clinical dose. The resilience of capravirine to a variety of
mutations to the NNRTI binding pocket has been ascribed to a com-
bination of three anchoring hydrogen bonds to the immutable RT
backbone coupled with sufficient flexibility to allow rearrange-
ment or repositioning of the lipophilic portions of the inhibitor
which contribute the majority of binding energy within the mu-
tated binding cavities.7 Therefore in a new series of NNRTIs we
planned to maximise hydrogen bonding to the RT backbone whilst
incorporating sufficient conformational flexibility. Examination of
capravirine reveals a lipophilic molecule (clog P 3.5) which would
be expected to be prone to oxidative metabolism consistent with
* Corresponding author. Tel.: +44 1304 648427; fax: +44 1304 651821.
E-mail address: Charles.Mowbray@Pfizer.Com (C.E. Mowbray).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.08.039

5600
C. E. Mowbray et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5599–5602
the short half life of only 7.5 min in our human liver microsome
stability assay. The imidazole nucleus is not sufficiently basic
(pKa 4.47) for ionisation of the intrinsically lipophilic molecule to
reduce the measured lipophilicity (log D >2.8) into the accepted
range for optimal oral absorption and pharmacokinetics.8 Further-
more we speculated that capravirine contains a biaryl sulphide and
sterically unencumbered 4-pyridyl group both of which could pro-
vide likely sites for oxidative metabolism. A subsequent publica-
tion has confirmed that this does occur at least in rat liver
microsomes.9 Table 1 shows published data10 and our own re-
sults11 for inhibition of RT by a range of analogues of capravirine
which demonstrate that within the series the pyridylmethyl group
in capravirine can be removed (compare capravirine to 1), albeit
with a small loss of potency. Similarly the primary carbamate
can be replaced with other moieties such as an alcohol (compare
capravirine to 1a, and 1 to 2 and 3). In addition the 3,5-dichloro
substitution on the thiophenoxy ring can be simplified to the less
lipophilic 3-fluorosubstitution (compare 1 and 4). Finally replacing
the sulphur atom with a methylene linker reduces activity signifi-
cantly (compare 1 and 5).
vulnerable sulphur atom replaced
and pyridine removed
Cl
Cl
N
OH
novel pyrazole
template
N
potential for H-bonds to
RT backbone retained
Figure 1. Design of prototype pyrazole inhibitor 6.
Cl
Cl
a
O
O
O
b
O
With these observations in mind we designed (Fig. 1) the pyra-
zole 6 as a prototype of a new series which proved to be a moder-
ately potent inhibitor of wild type RT (IC50 1.9 lM). We were
pleased with the promising activity of the pyrazole 6 and carried
out extensive profiling to determine its quality as a lead with the
potential to be optimised into a series capable of delivering a drug
candidate. The RT inhibition of the lead 6 was shown to translate
into potent antiviral activity in cell culture (EC50 33 nM, EC90
77 nM) in SupT1 cells infected with the RF strain of HIV. The com-
pound was also shown to be not cytotoxic to this cell type (CC30
>1000 nM). Although pyrazole 6 is lipophilic (clog P 4.2, log D
>4.0) it showed encouraging stability in human liver microsomes
(T1/2 50 min cf. capravirine T1/2 7.5 min). The lead 6 also displayed
excellent caco-2 flux (Papp A–B/B–A 41/9 cmSÀ1). Furthermore the
lead 6 did not have any profound polypharmacology, QTc interval
prolongation or drug–drug interaction liabilities as assessed in
wide ligand binding, HERG channel binding and P450 inhibition
studies which might limit the potential of a new series based upon
this core structure. We therefore instigated a programme to opti-
mise the potency, physicochemical and in vitro metabolic proper-
ties of this new lead.
Cl
Cl
O
Cl
Cl
c
NH
N
O
d
Cl
Cl
Cl
Cl
e
N
N
OH
N
N
CO2Et
6
We elected to optimise the 3-isopropyl and 5-methyl substitu-
ents of lead 6 first as these were introduced in the b-diketone at
the start of our synthetic route as illustrated in Scheme 1.10 We
Scheme 1. Synthesis of prototype pyrazole inhibitor 6. Reagents and conditions: (a)
3,5-dichlorobenzaldehyde, acetic acid, piperidine, toluene, reflux, 50%; (b) H2 15 psi,
Pd/BaSO4, EtOH, 80%; (c) hydrazine hydrate, EtOH, reflux, 80%; (d) ethyl bromo-
acetate, NaOEt/EtOH, 80 °C, 78%; (e) LiAlH4, Et2O, 77%.
Table 1
Wild type RT activity of truncated analogues of capravirine
z
Y
R1
N
X
R2
N
a
b
Compound
X
Y
Z
R1
—
R2
IC50 (nM)
IC50 (nM)
Efavirenz
Caparavirine
1
1a
2
3
4
5
—
S
S
S
S
S
S
CH2
—
Cl
Cl
Cl
Cl
Cl
F
—
Cl
Cl
Cl
Cl
Cl
H
—
—
8.4
93
730
110
792
NT
4-Pyridylmethyl
Me
4-Pyridylmethyl
Me
Me
Me
Me
OCONH2
OCONH2
OH
0.35
1.3
NTc
3
0.4
0.12
19
OH
CH2OH
OCONH2
OCONH2
NT
NT
Cl
Cl
a
b
c
Reported values are for inhibition of wild type RT with a poly(rA) template, (dT) 12–18 primer and dTTP as substrate.10
Inhibition of wild type RT with a poly(rA) ꢀ300 template, (dT) 16 primer and dTTP as substrate.11
Not tested.

C. E. Mowbray et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5599–5602
5601
Table 3
planned to subsequently investigate changes to the benzaldehyde
component before varying the N-substituent and then re-optimis-
ing as necessary.
IC50 fold-resistance of selected pyrazoles versus RT enzymesa bearing NNRTI
resistance mutations cf. wild type
Efavirenz
8
11
12
13
2.4
1.6
16
14
A range of pyrazoles with varied 3- and 5-substituents were
prepared and compared to the lead 6 (Table 2). Reducing the 3-iso-
propyl group to an ethyl group slightly enhanced potency but
further truncation to a methyl group was not favoured (compare
6–8). Replacing the 5-methyl group with an ethyl group also
improved potency but a larger isopropyl group was less successful
(compare 8–10). Combining the best 3- and 5-substituents gave
the 3,5-diethylpyrazole 11 as the most potent compound in this
early series. This compound was made from commercially avail-
able symmetrical 3,5-heptanedione which allowed straightforward
synthesis without the need for separation of pyrazole N-1/N-2
isomers.
To determine whether the excellent balance of activity of
capravirine versus RT with drug resistance mutations had been re-
tained within this new series of NNRTIs we tested compounds 8
and 11–14 against a panel of RT enzymes with clinically relevant12
engineered point mutations in the NNRTI binding pocket (Table 3).
As we had hoped all the compounds retained excellent activity
within twofold of their wild type activities against the clinically
common K103N and Y181C mutations. These compounds were
also active against a wider panel of RT enzymes with less clinically
K103N
Y181C
F227L
V106A
Y188C
K101E
P236L
V108I
44
1.6
2.1
16
0.8
1.1
9.0
1.8
0.6
5.1
0.2
5.5
1.0
1.3
11
1.4
1.0
9.3
3.4
0.4
5.1
0.2
5.6
2.2
0.4
1.8
0.8
3.8
2.8
1.0
2.7
1.3
5.6
0.3
7.1
2.4
0.6
5.7
0.2
6.5
2.2
0.7
5.1
0.2
7.3
a
Inhibition of wild type and mutated RT with a poly(rA) ꢀ300 template, (dT) 16
primer and dTTP as substrate.11
common NNRTI resistance mutations. Indeed all these mutant
enzymes with the exception of F227L for compounds 8, 12 and
13 remained sensitive to the compounds at concentrations within
10-fold of their wild type IC50’s. In contrast although efavirenz
retains activity against RT with the less clinically common muta-
tions it is much less active versus the K103N mutant.
The metabolic stability of several members of this series was
examined in human liver microsomes (Table 2). Decreasing the
size of the 3-substituent reduced the half life and if this was
accompanied by an increase in lipophilicity the effect was magni-
fied (compare 6, 8 and 10). If the overall lipophilicity was kept con-
stant but the 3-substituent reduced in size the half life was also
reduced (compare 6 and 11). We therefore speculated that the
principal route of oxidative metabolism of these compounds was
via a stabilized benzylic type radical at the pyrazole 3-position
which was subsequently converted to a hydroxylated product. For-
mation of this radical by a P450 enzyme would be expected to be
sensitive to the steric environment of the benzylic position and the
overall lipophilicity of the molecule explaining the observed
trends. We therefore prepared the 3-trifluoromethyl analogue 12
which should not be susceptible to this route of oxidation and
we were rewarded with an increased half life, however the mole-
cule was still metabolised demonstrating that alternative sites
for oxidation exist in this template. That oxidation at more than
one site was probably contributing to the metabolism of these
molecules was further indicated by the increased stability of a 5-
trifluoromethyl compound compared with its 5-ethyl analogue
(compare 11 and 14). The most stable member of this series, com-
pound 12 was progressed to an isolated perfused rat liver prepara-
tion where extremely rapid glucuronidation of the alcohol moiety
was detected giving an estimated unbound clearance value of
approximately 10,000 mL/min/kg. This showed that in addition to
removing or blocking oxidatively vulnerable groups we would
need to reduce glucuronidation.
Table 2
Variation of pyrazole 3- and 5-substituents
Cl
Cl
R2
N
OH
N
R1
a
b
Compound
R1
R2
IC50
(lM)
T1/2 (min)
Efavirenz
6
7
8
—
—
0.0084
1.9
0.77
2.1
1.0
6.5
0.66
0.92
0.63
0.54
—
Pri
Et
Me
Me
Me
Et
50
NTc
39
NT
4
18
79
NT
32
Me
Me
Me
Et
CF3
CF3
Et
9
10
11
12
13
14
Pri
Et
Me
Et
CF3
a
Inhibition of wild type RT with a poly(rA) ꢀ300 template, (dT) 16 primer and
dTTP as substrate.11
b
Half life of compound in a human liver microsome preparation.
Not tested.
c
Figure 2. Model showing possible hydrogen bonding to wild type RT by alcohol 11 (shown in yellow in the left hand image) compared to the carbamate of capravirine
(shown in green in the right hand image). Ligand–protein distances (Å) are shown in black and indicated with dashed lines.

5602
C. E. Mowbray et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5599–5602
S.; Jacobsen, D. M.; Montaner, J. S. G.; Richman, D. D.; Yeni, P. G.; Volberding, P.
From these initial investigations of RT inhibition SAR and met-
A. J. Am. Med. Assoc. 2008, 300, 555. For further information see also:
www.who.int/hiv/en/.
2. Abdel-Malak, M.; Gallati, C.; Mousa, S. A. Drugs Future 2008, 33, 691.
3. Bachelor, L. T. Drug Resist. Updat. 1999, 2, 56.
4. Fujiwara, T.; Sato, A.; El-Farrash, M.; Miki, S.; Abe, K.; Isaka, Y.; Kodama, M.;
Wu, Y.; Chen, L. B.; Harada, H.; Sugimoto, H.; Hatanaka, M.; Hinuma, Y.
Antimicrob. Agents Chemother. 1998, 42, 1340.
5. Brown, W. M. Curr. Opin. Anti-infect. invest. Drugs 2000, 2, 286.
6. Subsequent to this analysis capravirine failed to meet its primary outcome in
two Phase 2 clinical trials, and no further development of CPV is planned. This
vindicated our decision to look for compounds with improved properties
compared to capravirine. (a) Sato, A.; Hammond, J.; Alexander, T. N.; Graham, J.
P.; Binford, S.; Sugita, K.; Sugimoto, H.; Fujiwara, T.; Patick, A. K. Antiviral Res.
2006, 70, 66; (b) Pesano, R.; Piraino, S.; Hawley, P.; Hammond, J.; Tressler, R.L.;
Ryan, R.J.; Nickens, D.; Ruiz, R. 2005, Abstracts of Papers, 555, Proceedings of the
Twelfth Conference on Retroviruses and Opportunistic Infections, Boston.; (c)
Hawley, P.; Hammond, J.; Ryan, R.J.; Tressler, R.L.; Raber, S.R.; Hodges, M. 2006,
Abstracts of Papers, J-130, Proceedings of the 13th Conference on Retroviruses
and Opportunistic Infections, Colorado.
abolic stability it was clear that we had identified a potent series
of inhibitors but to achieve further improvements in stability to-
wards metabolism would need a reduction in lipophilicity and/or
additional modification of the remaining vulnerable sites whilst
further optimising potency. Further studies toward these goals
are described in the following papers in this series.
Finally we suggest that the alcohol present in inhibitors such as
11 is able to make hydrogen bonds to proline-236 and lysine-103
of the RT main chain as shown in Figure 2. This is similar to that
observed for the carbamate of capravirine in the published crystal
structure with wild type RT.7
Acknowledgements
This Letter describes the work of a number of people in addition
to the authors. We would especially like to thank Dave Beal and
Chris Carr for compound synthesis, Lesley Fishburn for RT testing,
Julie Mori for antiviral testing and Gill Allan for metabolism
studies.
7. Ren, J.; Nichols, C.; Bird, L. E.; Fujiwara, T.; Sugimoto, H.; Stuart, D. I.; Stammers,
D. K. J. Biol. Chem. 2000, 275, 14316.
8. (a) Van de Waterbeemd, H.; Smith, D. A.; Beaumont, K.; Walker, D. J. Med. Chem.
2001, 44, 1313; (b) Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J.
Adv. Drug Delivery Rev. 2001, 46, 3.
9. Ohkawa, T.; Goto, S.; Miki, S.; Sato, A.; Kuroda, T.; Iwatani, K.; Takeuchi, M.;
Nakano, M. Xenobiotica 1998, 28, 877.
10. Sugimoto, H.; Fujiwara, T. European Patent Application EP 0786455A1.
11. Corbau, R. G.; Mowbray, C. E.; Perros, M.; Stupple, P. A.; Wood, A. World Patent
Application WO 200204424. The large absolute difference in IC50s in these two
assays is due to the different experimental conditions used. The rank order of
potencies within the series is the same in both assay formats.
12. (a) Corbau, R., et al. Poster Presentation, 47th Interscience Conference on
Antimicrobial Agents and Chemotherapy, September 17–20, 2007, Chicago.;
(b) Corbau, R.; Mori, J.; Phillips, C.; Fishburn, L.; Martin, A.; Mowbray, C.; Panton,
W.; Smith-Burchnell, C.; Thornberry, A.; Ringrose, H.; Knöechel, T.; Irving, S.;
Westby, M.; Wood, A.; Perros, M. Antimicrob. Agents Chemother. 2009, in press.
References and notes
1. (a) Panel on Antiretroviral Guidelines for Adults and Adolescents. Guidelines
for the use of antiretroviral agents in HIV-1-infected adults and adolescents.
Department of Health and Human Services. November 3, 2008; 1–139.
Available at http://www.aidsinfo.nih.gov.; (b) Antiretroviral Treatment of
Adult HIV Infection: 2008 Recommendations of the International AIDS
Society-USA Panel: Hammer, S. M.; Eron, J. J., Jr.; Reiss, P.; Schooley, R. T.;
Thompson, M. A.; Walmsley, S.; Cahn, P.; Fischl, M. A.; Gatell, J. M.; Hirsch, M.