Benzyl amide-ketoacid inhibitors of HIV-integrase

Article abstract of DOI:10.1016/j.bmcl.2007.06.042

Integrase is one of three enzymes expressed by HIV and represents a validated target for therapy. Previous reports have demonstrated that the diketoacid-based chemotype is a useful starting point for the design of inhibitors of this enzyme. In this study, one of the ketone groups is replaced by a benzylamide resulting in a new potent chemotype. A preliminary SAR study is carried out to investigate the substitution requirements on the phenyl ring and methylene group of the benzylamide.

Full text of DOI:10.1016/j.bmcl.2007.06.042

Bioorganic & Medicinal Chemistry Letters 17 (2007) 4886–4890
Benzyl amide-ketoacid inhibitors of HIV-integrase
Michael A. Walker,a,* Timothy Johnson,a B. Narasimhulu Naidu,a Jacques Banville,c
Roger Remillard,c Serge Plamondon,c Alain Martel,c Chen Li,a Albert Torri,b
Himadri Samanta,b Zeyu Lin,b Ira Dicker,b Mark Krystalb and Nicholas A. Meanwella
aDepartment of Medicinal Chemistry, Research and Development, Bristol-Myers Squibb, The Richard L Gelb Center
for Pharmaceutical Research and Development, 5 Research Parkway, Wallingford, CT 06492, USA
bDepartment of Virology, Research and Development, Bristol-Myers Squibb, The Richard L Gelb Center for Pharmaceutical
Research and Development, 5 Research Parkway, Wallingford, CT 06492, USA
cDepartment of Medicinal Chemistry, Research and Development 100 de l’Industrie Boulevard, Que., Canada J5R1J1
Received 6 March 2007; revised 11 June 2007; accepted 12 June 2007
Available online 14 June 2007
Abstract—Integrase is one of three enzymes expressed by HIV and represents a validated target for therapy. Previous reports have
demonstrated that the diketoacid-based chemotype is a useful starting point for the design of inhibitors of this enzyme. In this study,
one of the ketone groups is replaced by a benzylamide resulting in a new potent chemotype. A preliminary SAR study is carried out
to investigate the substitution requirements on the phenyl ring and methylene group of the benzylamide.
Ó 2007 Elsevier Ltd. All rights reserved.
The ability of the viral DNA of HIV to become incorpo-
rated into the genome of the host is a hallmark of retro-
viral infection. This event is necessary1 and it ensures the
indefinite survival of the virus in the host as long as the
infected pool of cells remains alive. Integrase, which is
coded for by the virus, is responsible for this event in
the viral lifecycle. In the cytosol, intergrase trims the
30-ends of both strands of the viral DNA (vDNA) imme-
diately downstream of a conserved CA dinucleotide,
leaving a 2-base overhang at each 50-end. The enzyme
and processed vDNA are transported to the nucleus
where the two 30-ends of the DNA are inserted into
the host DNA, an event designated as strand transfer.
The subsequent steps are not well defined but include
pruning the 50-overhangs on the vDNA and repairing
the gaps created by the insertion reaction, processes
thought to be completed by host cell enzymes. HIV gen-
ome integration yields a provirus that can go on to pro-
duce new virus or lie dormant in a latent state.
false leads from authentic leads was difficult under the
early assay conditions. Second, the integrase-inhibitor
pharmacophore appears to be under-represented in
most compound collections. Recently, optimized assay
conditions have been developed which are more selective
for finding bona fide leads. This has led to the discovery
of the diketoacid derived inhibitors.2 These compounds
selectively inhibit the strand transfer reaction of the
enzyme. The mechanism of inhibition appears to involve
direct binding of the compounds to the divalent metal
(Mg2+) co-factor located at the enzyme active site.
Kinetic studies indicate that the inhibitors intercept the
enzyme-vDNA complex formed after the 30-processing
reaction and prevent the binding of this complex to host
DNA.
Evolution of this chemotype has yielded a number of
compounds which have been examined in the clinic
(see Fig. 1), including S-1360 (Shionogi, GlaxoSmithK-
line),3 L-870810, and MK-0518 (Merck).4,5 Figure 2
shows a generic diketoacid and the minimal pharmaco-
phore-elements needed to bind to integrase. The dike-
toacid portion of the template is believed to bind to
the Mg2+ co-factor located in the active site, while
the aryl group participates in a specific interaction with
an adjacent hydrophobic pocket or surface. SAR stud-
ies on the aryl group indicate that the structural ele-
ments of the enzyme which form this hydrophobic
The discovery of clinically relevant inhibitors of HIV-
integrase for antiviral therapy has proven to be a chal-
lenging task for two major reasons. First, distinguishing
Keywords: Integrase; HIV; Antiviral; AIDS; Diketoacid; Amide-
ketoacid.
*
Corresponding author. E-mail: michael.a.walker@bms.com
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.06.042

M. A. Walker et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4886–4890
4887
O
OH
O
OH
Since it is our contention that 1 and 2 possess different
integrase-binding configurations it was of interest to
see what effect the analogous modification would have
on 2. In contrast to the results obtained on going from
1 to 3, the diketoacid derivative 5 retained activity, in
the absence of the enol. Since it is logical to assume that
the 2-carbon spacer group of 5 is associated with even a
larger entropic penalty for binding compared to the sin-
gle carbon linker of 3, the difference in activity between
these homologues suggests that the longer linker allows
the phenyl group to reach a hydrophobic domain on the
enzyme that is not accessible by the 1 carbon linker.
Taking advantage of the improved activity associated
with introducing a conformational lock into the ring
via an amide bond the same approach was applied to
5. The 1000-fold increase in activity for compound 6 rel-
ative to 5 provides support for this hypothesis.10
H
N
O
N
N
H
N
N
N
F
N
O
N
O2S
L-870810
S-1360
F
F
O
OH
F
O
O
Cl
N
OH
H
N
N
O
O
OH
O
N
H
N
N
MK-0518
GS-9137
Figure 1. Clinically studied integrase inhibitors.
domain are well defined. This being the case, the linker
group which couples the aryl group to the ketoacid is a
critical feature in controlling activity by altering the
position of the aryl group with respect to the Mg2+
-
With this new chemotype in hand, a survey of the phenyl
ring SAR was carried out to probe the nature of the
hydrophobic pocket believed to be spatially distinct
from that of the anilide series. While a report by Daeya-
ert et al.11 reported the use of the benzylamide ketoacid
template to develop a QSAR model for integrase, based
on molecular properties such as sterics, polarizability,
charge, H-bond acceptors, and donors, we were more
interested in examining the SAR with a higher degree
of precision, optimizing activity, and to relate this to
the diketoacid chemotype represented by compounds 1
and 2. Substituents were chosen in order to examine
the effects of electron donation/withdrawal and sterics.
It should be noted that substitution of the phenyl ring
also has the potential of changing the electronics of
the adjacent amide bond. This in turn will alter the
ability of amide-carbonyl to bind to Mg2+ in the active
site. As shown in Table 1, ortho-substitution reduces
activity relative to the unsubstituted compound 6. The
SAR of the meta-position appears to be driven primarily
by electron induction, in that inductive electron with-
drawal is unfavorable. In contrast, electron withdrawal
at the para-position leads to improved activity (cf. 9c),
but this effect is offset somewhat by the size of the group.
binding portion of the template.6 Very recently another
apparently new chemotype has emerged represented by
the clinical compound GS-9137 (Fig. 1, Japan Tobacco
and Gilead).7
The diketoacid based inhibitors have been under inves-
tigation in our laboratories, in particular the SAR asso-
ciated with the linker group.8 Our efforts in this area
began with the characterization of diketoacid 1 and tri-
ketoacid 2. Using these two compounds, we have previ-
ously shown that there are probably two different aryl
binding sites located adjacent to the Mg2+ binding
Asp, Asp, Glu (DDE)-triad. Access to these binding
sites is controlled by the structure of the linker group,
in the present case the pyrrole ring of 1 and the enol
moiety of 2. Of interest, deannulation of 1 to provide
3 leads to a complete loss in activity, but the anilide
ketoacid derivative 4 displays very good inhibitory
activity.9 Based on this result it is tempting to speculate
that an H-bond donating group is required for activity;
however, our results suggest that the primary contribu-
tion of the amide bond is to serve as a conformational
lock.
generic chemotype
Mg+2
Mg+2
O
O
O
aryl
X
O
linker
diketoacid
Cl
Cl
O
OH
O
OH
O
OH
OH
OH
OH
N
O
O
N
H
O
H
1
3
4
IC50 = 0.07 μM
IC50 > 200 μM
IC50 = 0.7 μM
O
OH
O
OH
O
OH
OH
OH
N
OH
H
HO
O
O
2
O
5
6
IC50 = 34 μM
IC50 = 30 μM
IC50 = 0.3 μM
Figure 2. Diketoacid pharmacophore and the corresponding activity of related diketoacid and amide-ketoacid inhibitors.

4888
M. A. Walker et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4886–4890
Table 1. In vitro SAR of phenyl ring substitution12
Table 3. SAR of multiple substituents
O
OH
O
OH
R
OH
OH
N
Ar
N
H
H
O
O
Compound
∗
IC50 (lM)
Compound
R
IC50 (lM)
Ar
Cl
∗
N
17
0.7
R
R
∗
∗
∗
R
OMe
c
b
∗
18
19
7.5
0.3
a
MeO
O
6
–H
–Me
0.3
1.3
7a,b,c
8a,b,c
9a,b,c
10a,b,c
11
0.3
1.4
4.2
0.2
1.1
2.1
9
∗
–OMe 0.5
2.4
1.2
0.6
—
–F
–Cl
1.8
1.7
—
Cl
∗
–NO2
–CF3
–t-Bu
12
13
—
—
—
—
20
21
22
0.2
1.7
0.2
180
F
F
∗
F
Table 2. In vitro SAR of benzylic methylene
Cl
Compound
R
IC50 (lM)
Cl
∗
∗
∗
14
>300
present on the molecules as well as the presence of a
highly acidic carboxylate the antiviral activity was mod-
est. As shown in Figure 3, compound 9c, which pos-
sesses potent in vitro activity, exhibited an
EC50 = 5.9 lM. Assuming that the H-bond donating
groups were adversely affecting the penetration of the
compound into the cell these groups were blocked in a
systematic fashion. There was only a slight improvement
in activity when 9c was delivered as the corresponding
dioxolane pro-drug derivative 23 probably due to pre-
mature hydrolysis of the acetonide-ester prior to enter-
ing the cell. On the other hand, methylation of the
amide to yield 24 provides a 10-fold increase in cell cul-
ture activity while having only a marginal effect on
in vitro activity, apparently due to increased cellular
uptake. It is interesting to point out the necessity of
the methyl group to achieve improved activity for this
chemotype compared to the aza-naphthyridine and
pyrimidine chemotypes (cf. L-870810 and MK-0518)
where the opposite effect is seen.
15
16
24
3
∗
F
Apparently, the depth of the pocket which surrounds
the benzyl group appears to be limited, therefore only
small groups can be accommodated. Altogether, these
results suggest a well-organized hydrophobic pocket
which is sensitive to the electronic nature of the p-sur-
face and size of the phenyl group.
Apart from the phenyl group, substitution of the ben-
zylic methylene was examined. As illustrated by com-
pounds 14, 15, and 16 methylation results in
attenuation of activity indicating that there is restricted
tolerance for substitution at this site. Interestingly, this
site is sensitive to the enantiomeric configuration of
the chiral carbon atom. The R-isomer has much lower
activity than the S-isomer.
O
OH
O
O
O
OH
As shown in Table 3, the SAR of multiply functional-
ized aryl groups is not straightforward in that the substi-
tuent effects are not additive. When combined, the
groups appear to be acting synergistically to reduce or
increase activity beyond what is expected based on the
SAR shown in Table 1.
N
H
N
H
O
O
F
F
9c
23
EC50 = 5.9 μM
EC50 = 1.2 μM
O
OH
OH
N
O
In order to evaluate the antiviral activity of the amide-
ketoacid chemotype the compounds were tested in cell
culture against a luciferase containing non-infectious
HIV-reporter virus.13 As might be expected based on
the number of H-bond donor and acceptor functionality
F
24
IC50 = 0.11 μM
EC50 = 0.6 μM
Figure 3. Cell culture activity of amide-ketoacid derivatives of 9c.

M. A. Walker et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4886–4890
4889
The synthesis of the benzylic amide-ketoacid com-
pounds is readily accomplished using either of the meth-
ods shown in Scheme 1.14 Acylation of benzylamine A
provides amide B which can be readily condensed with
dimethyloxylate forming the amide-ketoester C. Like-
wise, the benzylamide can be amidated with F15 to yield
D. The desired benzyl amide-ketoacid is easily generated
following saponification of C and D. The simplicity of
these methods is readily apparent and affords a conve-
nient platform for the parallel synthesis of small
libraries.
hydrophobic pocket on integrase that can discriminate
groups based on their size and electronic properties.
Returning briefly to Figure 2, one can see where the
diketoacid inhibitor 1 derives its activity. As illustrated
in Figure 4, in analogy to the amide group of 6, the pyr-
role-ring locks the attached phenyl group into what is
likely the preferred conformation via resonance and also
pushes electron density back toward the diketoacid
through conjugation, increasing its affinity toward
Mg2+. This is similar to the proposed function of the
amide group of the amide-ketoacid inhibitors. However,
the amide group is less structurally complicated and
therefore a more efficient functional group.
In summary, the results reported above demonstrate
that benzylic amide-ketoacids are potent inhibitors of
HIV-integrase. Moreover, the amide-based chemotype
yields a dramatic improvement in activity over the cor-
responding diketoacid based template represented by
compound 5. This is believed to be the result of the
reduced rotational freedom of an amide C–N bond com-
pared to an acyclic C–C bond. Thus, as demonstrated in
previous publications, the relative position of the aryl
group with respect to the Mg2+-binding portion of the
molecule is important for activity.
The amide-ketoacid chemotype provides a very versatile
platform for investigating the pharmacophore require-
ments for the design of integrase inhibitors. The rela-
tively straightforward synthesis provides rapid
assembly of compounds for SAR studies. As mentioned
above, both the ester and acid forms16 of the amide-
ketoacid chemotype have been exploited by others, for
example, Daeyaert et al. for the development of a QSAR
model. While this model proved useful for predicting the
binding of amide-ketoacid esters and acids to integrase
it did not provide information regarding the binding
configuration to the enzyme.
In general, the unsubstituted benzylic group of com-
pound 6 appears to be fairly well optimized for binding
to integrase. The SAR associated with the benzylic ele-
ment appears to be straightforward when substituted
by a single group, as illustrated in Tables 1 and 2. How-
ever, the SAR of multiple substituents attached to the
phenyl ring is more complicated. Nonetheless, the
SAR of the benzylic group points to a well-defined
A feature worth mentioning is the high ‘binding effi-
ciency’ of this chemotype where P2/3’s of the atoms
of the chemotype are directly interacting with the
enzyme. Another interesting feature of this chemotype
is the large response in cell culture activity achieved by
methylation of the benzylamide nitrogen. This provides
an additional vector for compound modification and
leaves open the possibility for improving activity even
further by attaching additional functional groups. More
complex structural variations have been studied in our
laboratories and will reported in due course.
O
O
O
Cl
F
O
O
O
OH
R
R
R
O
a
b
NH2
N
N
H
H
O
References and notes
A
B
C
c
1. LaFemina, R. L.; Schneider, C. L.; Robbins, H. L.;
Callahan, P. L.; LeGrow, K.; Roth, E.; Schleif, W. A.;
Emini, E. A. J. Virol. 1992, 66, 7414.
2. (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) Goldgur, Y.; Craigie, R.; Cohen, G. H.; Fujiwara, T.;
Yoshinaga, T.; Fujishita, T.; Sugimoto, H.; Endo, T.;
Murai, H.; Davies, D. R. Proc. Natl. Acad. Sci. U.S.A.
1999, 96, 13040.
d
O
OH
O
O
R
OH
c
O
N
N
H
H
O
O
D
E
Scheme 1. Reagents: (a) CH3COCl, i-Pr2NEt, CH2Cl2; (b) dimethy-
loxalate, LiHMDS, THF; (c) NaOH; (d) F, i-Pr2NEt, CH2Cl2.
3. (a) Fujiwara, T. XIV International AIDS Conference,
Barcelona, July 2002 (TuPeB443); (b) Billich, A. Curr.
Opin. Invest. Drugs 2003, 4, 206; (c) Kawasuji, T.;
Yoshinaga, T.; Sato, A.; Yodo, M.; Fujiwara, T.; Kiyama,
R. Bioorg. Med. Chem. 2006, 14, 8430.
4. (a) Little, S.; Drusano, G.; Schooley, R., et al. In 12th
Conference on Retroviruses and Opportunistic Infections,
Boston, Massachusetts, February 2005; (b) Hazuda, D. J.;
Anthony, N. J.; Gomez, R. P.; Jolly, S. M.; Wai, J. S.;
Zhuang, L.; Fisher, T. E.; Embrey, M.; Guare, J. P.;
Egbertson, M. S.; Vacca, J. P.; Huff, J. R.; Felock, P. J.;
Witmer, M. V.; Stillmock, K. A.; Danovich, R.; Grobler,
Cl
Cl
O
O
OH
O- OH
OH
OH
OH
OH
N
H
N+
H
O
O
1
6
OH
O- OH
N
N+
H
H
O
O
Figure 4. Resonance structures of compounds 1 and 6.

4890
M. A. Walker et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4886–4890
J.; Miller, M. D.; Espeseth, A. S.; Jin, L.; Chen, I. W.; Lin,
11. Daeyaert, F. F. D.; Vinkers, H. M.; d Jonge, M. R.;
Heeres, J.; Koymans, L. M. H.; Lewi, P. J.; Janssen, P. A.
Internet Electron. J. Mol. Des. 2004, 3, 528,
www.biochempress.com.
J. H.; Kassahun, K.; Ellis, J. D.; Wong, B. K.; Xu, W.;
Pearson, P. G.; Schleif, W. A.; Cortese, R.; Emini, E.;
Summa, V.; Holloway, M. K.; Young, S. D. Proc. Natl.
Acad. Sci. U.S.A. 2004, 101, 11233; (c) Zhuang, L.; Wai, J.
S.; Embrey, M. W.; Fisher, T. E.; Egbertson, M. S.; Payne,
L. S.; Guare, J. P., Jr.; Vacca, J. P.; Hazuda, D. J.; Felock,
P. J.; Wolfe, A. L.; Stillmock, K. A.; Witmer, M. V.;
Moyer, G.; Schleif, W. A.; Gabryelski, L. J.;
Leonard, Y. M.; Lynch, J. J., Jr.; Michelson, S. R.;
Young, S. D. J. Med. Chem. 2003, 46, 453.
12. To evaluate in vitro activity against HIV-integrase, 5 pmol
of biotin labeled substrate DNA was bound to 100 lg of
Streptavidin coated PVT SPA beads (Amersham Pharma-
cia Biotech). Recombinant integrase (0.26 ng) was incu-
bated with the beads for 90 min at 37 °C. Unbound
enzyme was removed by washing the complex followed by
addition of inhibitors and 0.1 fmol of P33 labeled target
DNA. The reaction was stopped by adding EDTA to a
final concentration of 10 mM. Samples were counted in
TopCountNXT (Packard) and the CPM was used as a
measure of integration. The reaction condition was as
described in Engelman, A.; Craigie, R. J. Virol. 1995, 69,
5908–5911. The sequences of substrate and target DNA
were described in Nucleic Acid Research 1994 22, 1121–
1122.
13. A recombinant NL-Rluc virus was constructed in which
a section of the nef gene from NL4-3 was replaced with
the Renilla Luciferase gene. The NL-RLuc virus was
prepared by co-transfection of two plasmids, pNLRLuc
and pVSVenv. The pNLRLuc contains the NL-Rluc
DNA cloned into pUC18 at the PvuII site, while the
pVSVenv contains the gene for VSV G protein linked to
an LTR promoter. Transfections were performed at a
1:3 ratio of pNLRLuc to pVSVenv on 293T cells using
the LipofectAMINE PLUS kit from Invitrogen (Carls-
bad, CA) according to manufacturer’s instruction, and
the pseudotype virus generated was titered in MT-2
cells. Susceptibility of viruses to compounds was deter-
mined by incubation in the presence of serial dilutions
of the compound. The 50% effective concentration
(EC50) was calculated by using the exponential form
of the median effect equation where (Fa) = 1/[1 + (ED50/
drug concn)m] (Johnson, V. A,; Byington, R.T. Infec-
tivity Assay. In Techniques in HIV Research, Aldovini,
A., Walker, B. D., Eds.; Stockton Press: New York,
1990, pp 71–76).
5. (a) Morales-Ramirez, J. O.; Teppler, H.; Kovacs, C., et al.
In 10th European AIDS Conference, Dublin, Ireland,
November 17–20, 2005; (b) Markowitz, M.; Morales-
Ramirez, J. O.; Nguyen, B.-Y.; Kovacs, C. M.; Steigbigel,
R. T.; Cooper, D. A.; Liporace, R.; Schwartz, R.; Isaacs,
R.; Gilde, L. R.; Wenning, L.; Zhao, J.; Teppler, H.
JAIDS, J. Acquir. Immune Defic. Syndr. 2006, 43, 509; (c)
Grinsztejn, B.; Nguyen, B. Y.; Katlama, C.; Gatell, J. M.;
Lazzarin, A.; Vittecoq, D.; Gonzalez, C. J.; Chen, J.;
Harvey, C. M.; Isaacs, R. D. Lancet 2007, 369, 1261,
Protocol 005 Team.
6. Wai, J. S.; Egbertson, M. S.; Payne, L. S.; Fisher, T. E.;
Embrey, M. W.; Tran, L. O.; Melamed, J. Y.; Langford,
H. M.; Guare, J. P., Jr.; Zhuang, L.; Grey, V. E.; Vacca, J.
P.; Holloway, M. K.; Naylor-Olsen, A. M.; Hazuda, D. J.;
Felock, P. J.; Wolfe, A. L.; Stillmock, K. A.; Schleif, W.
A.; Gabryelski, L. J.; Young, S. D. J. Med. Chem. 2000,
43, 4923.
7. (a) DeJesus, E. In 13th Conference on Retroviruses and
Opportunistic Infections, Denver CO, February 5–8, 2006;
(b) Ramanathan, S.; Shen, G.; Cheng, A.; Kearney, B. P.
J. Acquir. Immune Defic. Syndr. 2007 ahead of print.
8. Walker, M. A.; Johnson, T.; Ma, Z.; Banville, J.;
Remillard, R.; Kim, O.; Zhang, Y.; Staab, A.; Wong,
H.; Torri, A.; Samanta, H.; Lin, Z.; Deminie, C.; Terry,
B.; Krystal, M.; Meanwell, N. Bioorg. Med. Chem. Lett.
2006, 16, 2920.
9. Walker, M. A.; Johnson, T.; Ma, Z.; Zhang, Y.; Banville,
J.; Remillard, R.; Plamondon, S.; Pendri, A.; Wong, H.;
Smith, D.; Torri, A.; Samanta, H.; Lin, Z.; Deminie, C.;
Terry, B.; Krystal, M.; Meanwell, N. Bioorg. Med. Chem.
Lett. 2006, 16, 5818.
14. Walker, M. A.; Johnson, T. A.; Meanwell, N. A.; Banville,
J. U.S. Patent Appl. US-20030027847, June 18th, 2001.
15. Zhu, K.; Simpson, J. H.; Delaney, E. J.; Nugent, W. A.
J. Org. Chem. 2007, 72, 3949.
16. In our own studies, we have found that the acid form is
preferred for binding to integrase. For example, the ethyl
ester of compound 6 had IC50 = 10.5 lM.
10. (a) Walker, M. A.; Johnson, T. D.; Meanwell, N. A.;
Banville, J. WO-2001096283, December 20th, 2001; (b)
Katoh, S.; Miyazaki, S.; Habuka, N. WO-2003016266,
February 27th, 2003.