Development of 2-tbutyl-N-methyl pyrimidones as potent inhibitors of HIV integrase

Article abstract of DOI:10.1016/j.bmcl.2008.03.017

A series of novel 2-^tbutyl-N-methyl pyrimidone HIV-1 integrase inhibitors have been identified. Optimization of the initial lead resulted in compounds such as 9d and 14a, which showed high levels of activity in cell culture inhibiting viral rep

Full text of DOI:10.1016/j.bmcl.2008.03.017

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 2709–2713
Development of 2-tbutyl-N-methyl pyrimidones as potent
inhibitors of HIV integrase
M. Emilia Di Francesco,a,* Paola Pace,a Fabrizio Fiore,b Francesca Naimo,b
Fabio Bonelli,b Michael Rowleya and Vincenzo Summaa
aDepartment of Medicinal Chemistry, IRBM-MRL Rome, Via Pontina, Km 30.600, 00040 Pomezia, Rome, Italy
bDepartment of Drug Metabolism, IRBM-MRL Rome, Via Pontina, Km 30.600, Pomezia, Rome, Italy
Received 6 January 2008; revised 5 March 2008; accepted 6 March 2008
Available online 10 March 2008
Abstract—A series of novel 2-tbutyl-N-methyl pyrimidone HIV-1 integrase inhibitors have been identified. Optimization of the ini-
tial lead resulted in compounds such as 9d and 14a, which showed high levels of activity in cell culture inhibiting viral replication
with CIC95 of 10 nM in the presence of 50% normal human serum.
Ó 2008 Elsevier Ltd. All rights reserved.
Together with protease and reverse transcriptase, integr-
ase is one of the three virally encoded enzymes required
for the replication of the human immunodeficiency virus
(HIV-1).1 HIV integrase catalyses the insertion of the re-
verse-transcribed viral genome into the host DNA in a
multistep process. After assembling onto viral DNA,
HIV integrase cleaves the two terminal nucleotides from
each 30-end. Following nuclear entry, the viral 30-ends
are ligated into the host genome (strand transfer).2 Inhi-
bition of HIV integrase results in arrest of HIV life cycle,
and represents therefore an attractive target for the
treatment of HIV infection.3
binding to cell medium plasma proteins (analogue 2
CIC95 = 1.0 lM in 10% fetal bovine serum, FBS).6
Interestingly, further studies within this and other series
of amine analogues showed a good correlation between
an increase in the degree of substitution at the C-atom
directly attached to the pyrimidine scaffold and the
boost in potency observed in the enzyme and cellular as-
says.6,7 Indeed, the gem-dimethyl analogue 4 was
amongst the most potent compounds in the series, with
CIC95 value of 0.06 lM (10% FBS) and no shift in the
more physiological high serum conditions (CIC95
=
0.08 lM, 50% normal human serum, NHS).6
Previous studies in our group had established that 5,6-
dihydroxypyrimidine-4-carboxamides are potent inhibi-
tors of HIV integrase and efficiently suppress virus
spread in HIV-infected cells.4–6 Initial structure-activity
relationship studies in the 2-thienyl series identified 4-
fluorobenzylamide 1 as the optimal residue for enzyme
inhibition (Table 1).5 It was also established that,
although in terms of intrinsic potency a variety of resi-
dues were well tolerated as replacements of the 2-thio-
phene ring, the introduction of a basic amine within
the C-2 substituent was required to enhance activity in
cell culture by improving cell permeability and reducing
These findings prompted us to further investigate the
importance of a tetrasubstituted sp3 carbon at the C-2
position of the pyrimidine scaffold, and the tbutyl group
seemed an ideal starting point. We were pleased to find
that, despite the absence of any basic substituent,
carboxamide 5a was a potent enzyme inhibitor (IC50
=
0.01 lM) and maintained good levels of cellular activity
with a CIC95 value of 0.25 lM in 10% FBS. Despite this
encouraging result, the high shift in potency observed
for 5a in high serum conditions (CIC95 =2.5 lM)
prompted us to evaluate a series of structurally related
benzylamides on the right-hand side of the scaffold, with
the primary goal of improving cellular activity while
reducing the serum shift.
Keywords: HIV integrase; N-methyl pyrimidones; Dihydroxypyr-
imidines.
The analogues 5a–f were prepared by the displacement
of ester 7 with a variety of benzylamines, as described
in Scheme 1. The synthesis of 5,6-dihydroxypyrimidine-
*
Corresponding author. Tel.: +39 06 910 93 563; fax: +39 06 910 93
654; e-mail: mariaemilia_difrancesco@merck.com
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.03.017

2710
M. E. Di Francesco et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2709–2713
Table 1. Development of the 2-tbutyl-dihydroxypyrimidine scaffold
1
2
3
Me, Me
4
5a
R, R0 = H, H
a
Inhibition of strand transfer IC50
0.01
0.2
0.01
0.005
0.01
b
Antiviral activity in cell culture CIC95
10% FBS
50% NHS
>10
ndc
1.0
>1.0
0.12
0.50
0.06
0.08
0.25
2.5
a Assays were performed with recombinant HIV-1 integrase preassembled on immobilized oligonucleotides. Inhibitors were added after assembly and
washings, and IC50 is the concentration of inhibitor that reduces HIV integrase activity by 50%. Results are the mean of at least three independent
experiments. SD was always 8% of the value. For details see Ref. 9 (lM).
b Spread assay: 95% cell culture inhibitory concentrations for inhibition of the spread of HIV-1 infection in cell culture, using HIV-1IIIb and MT-4 T-
lymphoid cells, in a medium containing 10% heat-inactivated fetal bovine serum. Results are the mean of at least 3 independent experiments, SD
was always < 10% of the value. For details see Ref. 10 (lM).
c Not determined.
after chromatographic purification). Treatment of 8
with a choice of amines gave the desired N-methyl pyr-
imidone benzylamides 9a–f.
Interestingly, 4-fluorobenzylamide 9a (Table 2) showed
a comparable level of potency in low serum conditions
with the corresponding dihydroxypyrimidine analogue
5a (CIC95 = 0.12 lM vs CIC95 = 0.25 lM, respectively)
but 9a had only a 3-fold shift in 50% NHS instead of
the 10-fold shift previously observed for 5a
(CIC95 = 0.37 lM vs CIC95 = 2.5 lM, respectively).
This first result was further supported by other N-
methyl pyrimidone derivatives, which were found to be
consistently more potent than their corresponding dihy-
Scheme 1. Preparation of compounds 5a–f and 9a–f. Reagents and
droxypyrimidine analogues and with lower serum shift
conditions: (a)NH2OH, MeOH, 55 °C (60%); (b) Dimethylacetylen-
edicarboxylate, CHCl3, 60 °C; then xylene, reflux (27%); (c) RNH2,
MeOH, 70 °C (43–76% after preparative HPLC purification); (d)
Bz2O, Pyr, DCM (81%); (e) LiH, Me2SO4, dioxane, 60 °C (77%).
(Table 2, analogues 9b, 9d–f). In particular N-methyl
amide 9d and sulfone 9f showed potency in the low
nanomolar range and virtually no shift between the en-
zyme and cellular assays, even in high serum conditions
(9d CIC95 = 0.01/0.01 lM; 9f CIC95 = 0.01/0.02 lM in
10% FBS/50% NHS).
4-carboxylate core has been previously reported5,6 and
features the conversion of nitrile 6 into the correspond-
ing amidoxime, followed by reaction with dimethyl-
acetylenedicarboxylate and cyclisation to the desired
pyrimidine methyl ester 7. Amongst the various ana-
logues synthesised both N-methyl amide 5d and sulfon-
amide 5f (Table 2) retained high levels of inhibition of
viral spread in infected cells, with low shift between
10% FBS and 50% NHS (CIC95 = 0.06/0.12 lM for
5d and CIC95 = 0.06/0.06 lM for 5f).8
Prompted by these findings, further evaluation of the
two most promising N-methyl pyrimidone analogues
9d and 9f was continued in vitro. Metabolic stability
studies of the two analogues in rat, dog and human li-
ver microsomes showed minimal turnover in the pres-
ence of NADPH, suggesting a marginal involvement
of oxidative metabolism in the clearance of both com-
pounds. In contrast, high turnover was observed in the
presence of UDPGA, with sulfone 9f having a higher
glucuronidation rate than 9d in all species (ranking or-
der: rat > human > dog; rat intrinsic clearance 63 lL/
min/mg for 9f vs 16 lL/min/mg for 9d).11 The pharma-
cokinetic profile of both analogues was then evaluated
in Sprague–Dawley rat (3 mpk iv and po, Table 4),
where 9f showed a higher level of plasma clearance
than 9d (37 ml/min/Kg vs 24 ml/min/Kg, respectively),
thus substantiating the findings from the in vitro met-
Contemporarily, a similar approach was undertaken
with a series of analogues based on the structurally re-
lated N-methyl pyrimidone scaffold, the previous work
of which in our group showed to have similar or im-
proved activity profile to the dihydroxypyrimidine one
(Scheme 1, 9a–f).7 The required ester 8 was prepared
from 7 via regioselective 5-benzoylation followed by
LiH-promoted N-alkylation (N- vs O-alkylation 4:1

M. E. Di Francesco et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2709–2713
Table 2. Benzylamide SAR in the dihydroxypyrimidone and N-methylpyrimidone series
2711
a
a
a
a
R
Compound
IC50 (lM)
CIC95 (lM)
Compound
IC50 (lM)
CIC95 (lM)
10% FBS 50% NHS
10%FBS 50% NHS
5a
5b
5c
0.01
0.01
0.01
0.25
1.0
2.5
9a
9b
9c
0.01
0.02
0.02
0.12
0.12
1.0
0.37
0.5
>1.0
>1.0
>1.0
>1.0
5d
5e
5f
0.02
0.02
0.06
0.06
0.12
0.06
0.12
>1.0
0.06
9d
9e
9f
0.02
0.02
0.02
0.01
0.02
0.01
0.01
0.12
0.02
a For footnote details see Table 1.
abolic stability experiments. Moreover, while both
compounds had a comparable behaviour in terms of
terminal plasma half lives and acceptable bioavailabil-
ity (T1/2iv = 1.6 h; F = 18% for 9 d; T1/2iv = 1.8 h; F =
17% for 9f), exposure after oral administration for sul-
fone 9f was less than half the exposure of amide 9d
(AUC 0.6 lM h vs 1.4 lM h, respectively).
The results for analogues 14a–i are reported in Table 3.
The unsubstituted amide 14a retained both the intrinsic
potency against integrase (strand transfer IC50
=
0.02 lM) and the antiviral activity in cell culture
(CIC95 = 0.01 lM in 50% NHS) in the same range of
N-methyl amide analogue 9d. Increasing the length of
the amide substituent, branching at the a-position and
introducing a tertiary amide all proved to be detrimen-
tal in terms of intrinsic potency. Simple replacement of
the N-methyl residue of 9d with the N-ethyl or N-ipro-
pyl as in 14c and 14e resulted in a 3- and 4-fold loss
in potency respectively (strand transfer IC50 = 0.07 lM
for 14c and IC50 = 0.09 lM for 14e vs IC50 = 0.02 lM
for 9d), while with the N-dimethyl amino group of
14g a more dramatic 9-fold loss was observed (strand
transfer IC50 = 0.18 lM). Diminished activity in cell
culture in the presence of 50% NHS was observed for
the most lipophilic analogues (14f: CIC95 = 0.12 lM
and 14e: CIC95 = 0.06 lM, 1% and 2% fraction un-
bound, respectively). Conversely, a lower protein bind-
ing can explain the re-established cellular activity for
compounds such as 14g, only moderately active in the
strand assay but amongst the most potent of the series
in high serum conditions (14g: 18% fraction unbound,
IC50 = 0.18 lM; CIC95 = 0.03 lM 10% FBS and 50%
NHS).
Given the overall more favourable profile of 9d over 9f,
we set out to explore if possible replacements of the 2-N-
methyl carboxamide could offer an improvement to the
pharmacokinetic behaviour of inhibitors in this series
while retaining the same levels of cellular activity ob-
served in high serum conditions for 9d.
To allow maximum flexibility in the preparation of the
2-carboxamide analogues, we settled on the synthesis
of 2-carboxy methyl ester 12, a stable intermediate
which, upon conversion to the corresponding carboxylic
acid 13, could give a variety of 2-amides analogues
(Scheme 2). Thus, acid 11 was pre-activated with
EDCI/HOBt and treated with benzylamine 1013 and a
base to give ester 12. Hydrolysis in standard conditions
afforded the required acid 13 (89% yield over three steps)
which could then be progressed to amides 14a–i either
via conversion to the corresponding acid chloride or
pre-activation with EDCI/HOBt.

2712
M. E. Di Francesco et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2709–2713
i
Scheme 2. Preparation of compounds 14a–i. Reagents and conditions: (a) NaOH, THF/H2O; (b) EDCIÆHCl, HOBt, Pr2EtN, DCM; (c) NaOH,
i
THF/H2O (89% from 8); (d) EDCIÆHCl, HOBt, Pr2EtN, DCM, R0R00NH or (COCl)2, cat. DMF, DCM, 0 °C 1 h then R0R00NH, DCM 0 °C to rt
(24–36% after preparative HPLC purification); (e) CF3CO2H, DCM.
Table 3. SAR of 2-carboxamide analogues 14a–i
Compound
NR0R00
Inhibition of strand
a
transfer IC50 (lM)
Antiviral activity in cell
b
culture CIC95 (lM)
Protein Binding fu (%)c
10% FBS
50% NHS
14a
9d
NH2
NHCH3
0.02
0.02
0.07
0.07
0.09
0.09
0.18
0.01
0.01
0.02
0.03
0.06
0.03
0.03
0.01
0.01
0.02
0.06
0.06
0.12
0.03
6
7
14c
14d
14e
14f
14g
NHCH2CH3
NH(CH2)2CH3
NHCH(CH3)2
NHCH2CF3
N(CH3)2
3
2
1
1
18
14h
0.23
0.12
0.12
ndd
34
14i
0.49
0.12
0.12
a For footnote details see Table 1.
b For footnote details see Table 1.
c Fraction unbound to human plasma proteins as determined by ultrafiltration method. For details see Ref. 12.
d Not determined.
The pharmacokinetic profile of the most promising
amide analogues was determined in Sprague–Dawley
rat (Table 4). Carboxamide 14a showed a marginal
improvement in bioavailability but suffered from a sub-
stantial increase in plasma clearance and reduction in
exposure over N-methyl amide 9d (F% 23 vs 18; Cl 55 vs
24 ml/min/kg for 9d; AUC 0.6 lM.hr for 14a vs
1.4 lM h for 9d). On the other hand the mono-substituted
amides 14c and 14e had comparable levels of clearance to
9d, and showed a promising increase in bioavailability
(F = 30% and 40% for 14c and 14e, respectively).
Table 4. Pharmacokinetic profile of 9f, 9d and further 2-carboxamide
analogues in rata
Compound Fb (%) T1/2 (h) Cld (mL/min/kg) AUCe (lM h)
c
9f
9d
17
18
23
30
44
1.8
1.6
3.6
2.6
1.6
37
24
55
21
35
0.6
1.4
0.6
1.8
3.5
14a
14c
14e
a 3 mg/Kg (n = 3); vehicle iv 20% DMSO/60% PEG400/20% Water, po
1%methylcellulose (suspension).
b Oral bioavailability.
c Terminal phase plasma half life following iv administration.
d Plasma clearance.
In summary, optimization of 5a resulted in the identifi-
cation of a series of potent HIV integrase inhibitors such
e Area under the curve following po administration.

M. E. Di Francesco et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2709–2713
2713
7. Gardelli, C.; Nizi, E.; Muraglia, E.; Crescenzi, B.; Ferrara,
M.; Orvieto, F.; Pace, P.; Pescatore, G.; Poma, M.; Rico
Ferreira, M.; Scarpelli, R.; Homnick, C. F.; Ikemoto, N.;
Alfieri, A.; Verdirame, M.; Bonelli, F.; Gonzalez Paz, O.;
Taliani, M.; Monteagudo, E.; Pesci, S.; Laupher, R.;
Felock, P.; Stillmock, K.; Hazuda, D. J.; Rowley, M.;
Summa, V. J. Med. Chem. 2007, 50, 4953.
as sulfone 9f and carboxamides 14a and 9d characterized
by a minimal scaffold and excellent cellular activity in
the presence of 50% NHS (CIC95 = 0.01 lM for 9d
and 14a). Studies in this series are ongoing to identify
the optimal combination of potency, activity in cell cul-
ture and pharmacokinetic properties.
8. For the synthesis of 4-fluoro-2-(N-methylbenzamide)-
benzylamine see: (a) Egbertson, M. S.; Moritz, H. M.;
Melamed, J. Y.; Wei, H.; Perlow, D. S.; Kuo, M. S.;
Embrey, M.; Vacca, J. P.; Zrada, M. M.; Cortes, A. R.;
Wallace, A.; Leonard, Y.; Hazuda, D. J.; Miller, M. D.;
Felock, P.; Stillmock, K.; Witmer, M. V.; Schleif, W.;
Gabryelski, L.; Moyer, G.; Ellis, J. D.; Jin, L.; Xu, W.;
Braun, M. P.; Kassahun, K.; Tsou, N. N.; Young, S.
Biorg. Med. Chem. Lett. 2007, 17, 1392; for the synthesis
of 4-fluoro-2-(methylsulfonyl)-benzylamine see: (b) Per-
low, D. S.; Kuo, M. S.; Moritz, H. M.; Wai, J. S.;
Egbertson, M. S. Syn. Comm. 2007, 37, 1887.
Acknowledgment
We thank Massimiliano Fonsi for microsome stability
studies; Kara A. Stillmock, Peter J. Felock and William
A. Schlief for HIV biological testing.
References and notes
9. Hazuda, D. J.; Felock, P.; Hastings, J. C.; Pramanik, B.;
Wolfe, A. J. Virol. 1997, 71, 7005.
1. (a) Esposito, D.; Craigie, R. Adv. Virus Res. 1999, 52, 351;
(b) Asante-Appiah, E.; Skalka, A. M. Adv. Virus Res.
1999, 52, 351.
2. (a) Chiu, T. K.; Davies, D. R. Curr. Top. Med. Chem.
2004, 4, 965; (b) Pommier, Y.; Johnson, A. A.; Marchand,
C. Nat. Rev. Drug Disc. 2005, 4(3), 236.
3. (a) Hazuda, D. J.; Felock, P.; Witmer, M.; Wolfe, A.;
Stillmock, K.; Grobler, J. A.; Espeseth, A.; Gabryelski, L.;
Schleif, W.; Blau, C.; Miller, M. D. Science 2000, 287, 646;
(b) Gordon, C. P.; Griffith, R.; Keller, P. A. Med. Chem.
2007, 3, 199; (c) Anthony, N. J. Curr. Top. Med. Chem.
2004, 4, 979.
4. Summa, V.; Petrocchi, A.; Matassa, V. G.; Gardelli, C.;
Muraglia, E.; Rowley, M.; Gonzalez Paz, O.; Laupher, R.;
Monteagudo, E.; Pace, P. J. Med. Chem. 2006, 49, 6646.
5. Petrocchi, A.; Koch, U.; Matassa, V. G.; Pacini, B.;
Stillmock, K.; Summa, V. Biorg. Med. Chem. Lett. 2007,
17, 350.
10. Vacca, J. P.; Dorsey, B. D.; Schleif, W. A.; Levin, R. B.;
McDaniels, S. L.; Darke, P. L.; Zugay, J.; Quintero, J.;
Blahy, O. M.; Roth, E.; Sardana, V. V.; Schlabach, A. J.;
Graham, P. I.; Condra, J. H.; Gotlib, L.; Holloway, M.
K.; Lin, J.; Chen, I.; Vastag, K.; Ostovic, D.; Anderson, P.
S.; Emini, E. A.; Huff, J. R. Proc. Natl. Acad. Sci. U.S.A.
1994, 91, 4096.
11. The concentration of compounds 9d and 9f was 1 lM and
microsomes 1 mg/ml. The degradation of the substrates
was measured by LC/MS/MS during a 90 min incubation.
12. The test compound is dissolved in DMSO and incubated
with human plasma at 37 °C for 1 h. The mixture is then
spun in an ultrafiltration device and the filtrate is assayed
for free drug by analytical LC/MS/MS detection. For
more details see also Ref. 8a.
13. Lyle, F .R. U.S. Patent 6,973,257, 1995; Egbertson, M. S.;
Langford, H.; Melamed, J. Y.; Wai, J. S.; Wei, H.; Perlow,
D. S.; Embrey, M; Young, S. D. Preparation of N-
(substituted benzyl)-8-hydroxy-1,6-naphthyridine-7-car-
boxamides useful as HIV integrase inhibitors for
treatment of HIV infection, WO 2003077857.
6. Pace, P.; Di Francesco, M. E.; Gardelli, C.; Harper, S.;
Muraglia, E.; Nizi, E.; Orvieto, F.; Petrocchi, A.; Poma,
M.; Rowley, M.; Scarpelli, R.; Laupher, R.; Gonzalez Paz,
O.; Monteagudo, E.; Bonelli, F.; Hazuda, D. J.; Stillmock,
K.; Summa, V. J. Med. Chem. 2007, 50, 2225.