Effect of substitution on novel tricyclic HIV-1 integrase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2006.05.008

A series of novel tricyclic inhibitors of HIV-1 integrase enzyme was prepared. The effect of substitution at C-6 of the 9-hydroxy-6,7-dihydropyrrolo[3,4-g]quinolin-8-one compounds was studied in vitro. Inhibitors with small side chains at C-6 were general

Full text of DOI:10.1016/j.bmcl.2006.05.008

Bioorganic & Medicinal Chemistry Letters 16 (2006) 4031–4035
Effect of substitution on novel tricyclic HIV-1 integrase inhibitors
Maria Fardis,a,* Haolun Jin,a Salman Jabri,a Ruby Z. Cai,a Michael Mish,a
Manuel Tsiangb and Choung U. Kima
aDepartment of Medicinal Chemistry, Gilead Sciences, 333 Lakeside Dr., Foster City, CA 94404, USA
bDepartment of Biology, Gilead Sciences, 333 Lakeside Dr., Foster City, CA 94404, USA
Received 1 April 2006; revised 2 May 2006; accepted 2 May 2006
Available online 23 May 2006
Abstract—A series of novel tricyclic inhibitors of HIV-1 integrase enzyme was prepared. The effect of substitution at C-6 of the
9-hydroxy-6,7-dihydropyrrolo[3,4-g]quinolin-8-one compounds was studied in vitro. Inhibitors with small side chains at C-6 were
generally well tolerated by the enzyme, and the physicochemical properties of the inhibitors were improved by substitution of a small
alkyl group at this position. A second series of analogs bearing a sulfamate at the C-5 position with various C-6 substituents were
prepared to explore the interplay between the two groups. The SAR of the two classes are not parallel; modification at C-5 impacts
the effect of substitutions at C-6.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Acquired immunodeficiency syndrome (AIDS) has
become a serious epidemic since first reported in 1981.
Currently, more than 38 million people are infected
worldwide and the disease still remains a significant
problem.1 Human immunodeficiency virus (HIV), which
causes AIDS, contains three essential enzymes: reverse
transcriptase (RT), protease (PR), and integrase (IN).
Inhibitors of RT and PR are known, and various com-
binations of them are used in highly active antiretroviral
therapy (HAART). The effectiveness of HAART is lim-
ited by poor adherence to the demanding dosing regi-
men, cost, and the development of viral resistance.
Therefore, new drugs, particularly those with a novel
mode of action, are highly sought after.2
O
O
4
O
N
6
N
2
N
N
8
O
O
OH
OH
F
F
1
2
IC50: 80 nM
IC50: 60 nM
EC50: 161 nM
CC50: 5100 nM
Solubility (uM)
pH 7.3: 3
EC50: 7500 nM
CC50: 19000 nM
Solubility (uM)
pH 7.3: < 1.4
pH 2.2: < 1.4
pH 2.2: 285
Figure 1. Activity of pyrroloquinoline-6,8-dione 1 versus pyrroloquin-
oline-8-one 2.
in-house indicated that the pyridyl nitrogen, the C-9
phenol, and the C-8 carbonyl form the pharmacophore
of the compounds in binding to the enzyme. However,
the cellular antiviral activity was limited. In an effort
to improve the potency and pharmacokinetic profile of
our inhibitors, we noted that the cell-based antiviral
activity of the pyrroloquinoline-8-one 2 was substantial-
ly improved compared with the pyrroloquinoline-6,8-
dione 1.6 We hypothesized that the presence of the C-6
carbonyl in 1 yields a planar molecule with poor
solubility due to p-stacking, and that the better cellular
activity of 2 might be due to improved solubility. We
therefore decided to further elaborate that position.
Recently, attention has been focused on the develop-
ment of inhibitors targeted at HIV IN. Lack of a close
analog of this enzyme in mammals provides further
incentive toward development of a nontoxic antiviral
therapeutic. A large number of inhibitors of IN have
been identified in vitro, but compounds that have proven
effective in vivo remain scarce.3 We and others have pre-
viously identified tricyclic compounds such as 1 that
demonstrate good potency against the enzyme
(Fig. 1).4,5 An extensive series of analogs prepared
Keywords: HIV; Integrase; AIDS; Dihydropyrrolo[3,4-g]quinolin-
8-one; Acquired immunodeficiency syndrome.
*
Our initial goal was to investigate the effect of
substitution at the C-6 position of 5-methoxy-pyrrolo-
Corresponding author. Tel.: +1 650 522 5194; fax: +1 650 522
5899; e-mail: mfardis@gilead.com
0960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.05.008

4032
M. Fardis et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4031–4035
quinolin-8-ones on enzymatic and cell-based antiviral
activity. In addition, since C-6 substituents have the
potential for additional interactions with the enzyme,
we wanted to explore the relationship between substitu-
tions at the C-5 and the C-6 positions. To address the
above questions, an initial series of C-6 analogs carrying
5-methoxy was prepared. Subsequently, we also
explored a series of sulfamates at C-5.
O
O
a
N
N
N
N
O
O
F
F
OTIPS
OH
9
10
Scheme 3. Reagents and conditions: (a) i—Et2Zn, CH2Cl2, 0 ꢁC; ii—
TFA, 0 ꢁC, 15 min; iii—CH2I2, CH2Cl2; iv—Compound 9, CH2Cl2,
0 ꢁC to rt, overnight.
The common intermediate 3, which was utilized in the
preparation of all these analogs, was synthesized as
shown in Scheme 1. The pyrroloquinoline-6,8-dione 1
was prepared as described in the literature.7 The C-9
phenol can be preferentially protected under thermody-
namic conditions by heating the reaction. The effect of
various protecting groups on the chemoselectivity of
subsequent reactions for the C-6 position was explored.
TIPS proved to be optimal, presumably for steric
reasons.
O
O
a
N
N
b
N
N
O
O
OTIPS
OH
F
F
11
12
Scheme 4. Reagents and condition: (a) LiHMDS, THF, 0 ꢁC, MeI
(58%); (b) TFA, CH2Cl2.
A series of small alkyl groups at C-6 were prepared by
selective nucleophilic addition of alkyl anions to the car-
bonyl group (Scheme 2). Specifically, addition of methyl
Grignard to the appropriately protected compound 4 led
to the rapid and clean production of 5.8 Quenching the
reaction with 1 N aqueous HCl conveniently gave 8 in
one step. Reduction of 5 using triethylsilane with
BF3ÆOEt2 in dichloromethane yielded 6 in near quantita-
tive yield. Removal of the TIPS group was achieved
using TFA in dichloromethane to provide 7. Repeating
this reaction sequence, it was found that extended expo-
sure of compound 5 to BF3ÆOEt2 results in the cleavage
of the protecting group, providing 7 in one step.
To complete the SAR of this series of compounds, we
also prepared the C-6 gem-dimethyl compound 12. This
analog was prepared by direct alkylation of the
protected analog 11 under strongly basic conditions at
low temperature (Scheme 4).
The tolerance of the enzyme for solubility-enhancing
substituents at C-6 was explored next. We chose to
attach different substituents onto the C-6 position
via nucleophilic addition to N-acyliminium ions pre-
pared from the protected 6-hydroxy-pyrroloquinolin-
8-one 13 (Scheme 5). Reduction of the imide 4 was
accomplished using an excess of NaBH4 in MeOH10
to provide 13 as the only regioisomer. Deprotection
provided compound 14. Exposure of 13 to methanol
in TFA provided methyl ether 15. We concluded that
the reactive acyliminium ion can be generated under
mild conditions from the 6-OH compound 13. The
formation of the thioether 16 required forcing condi-
tions leading us to believe that stability of 16 is im-
proved compared with that of 15.
Preparation of the spirocyclopropyl 10 was accom-
plished from compound 9 using the (iodomethyl)zinc
species activated with TFA (Scheme 3).9
We were particularly interested in the antiviral activity
of the charged analogs. We utilized compound 16 to
rapidly provide access to both carboxylic acid 17 and
amine 18 (Scheme 6).
Scheme 1. Reagents: (a) TIPSCl, Im, DMF; (b) MeI, K2CO3, ACN
(72% over two steps).
O
O
O
O
OH
a
b
N
N
N
N
TIPS
N
N
O
O
O
F
O
F
O
F
O
TIPS
TIPS
6
4
5
d
O
O
c
N
N
N
N
O
O
F
OH
F
OH
7
8
Scheme 2. Reagents: (a) i—MeMgBr, THF, rt; ii—satd aq NH4Cl, EtOAc (99%); (b) BF3ÆOEt2, TES, CH2Cl2, (99%); (c) and (d) TFA, CH2Cl2.

M. Fardis et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4031–4035
HO
4033
O
b
N
14
15
N
N
O
F
OH
O
O
O
HO
N
O
N
O
O
N
c
a
N
N
O
O
F
OTIPS
F
MeO
OH
O
F
OTIPS
4
13
O
S
N
O
d
16
N
F
OH
Scheme 5. Reagents and conditions: (a) NaBH4, MeOH, CH2Cl2, overnight; (b) TBAF, THF, (100%); (c) TFA, MeOH, 1 h (100%);
(d) HS(CH2)2CO2Me, TFA, CH2Cl2, 1 day.
DMAP in good yield. Removal of the C-6 carbonyl
was performed using a two-step procedure: reduction
using NaBH4 to form the aminal, followed by a second
reductive elimination of the aminal using triethylsilane
and TFA to provide the dihydrosulfamate 24. Prepara-
tion of compounds 21–23 was accomplished following
the procedures described for 7, 8, and 10.
HO
O
S
O
a
N
N
MeO
O
O
F
OH
O
S
N
17
N
TFA.HN
a, b. c
O
F
N
OH
O
S
16
O
All compounds were evaluated in the IN strand transfer
assay, as well as a cell-based antiviral assay using the
MT4 cell line.11 The data are summarized in Table 1.
N
N
O
F
OH
18
Comparison of compound 1, our initial lead, with the
rest of the analogs demonstrates the benefit of modifica-
tion at C-6 on antiviral activity. Compound 8 which
maintained the planar conformation showed weaker
activity among 2–10. Compounds 2 and 7 had very sim-
ilar enzymatic activities, demonstrating the tolerance of
the enzyme toward substitution with small alkyl groups.
The spirocyclopropyl variant 10 had excellent activity
against the enzyme. The solubility of the spirocyclic ana-
log 10 similar to 2 and 7 was improved over our initial
lead 1. Further substitution at C-6 was not tolerated
by the enzyme as the activity of the gem-dimethyl
compound 12 dropped by over an order of magnitude
Scheme 6. Reagents and conditions: (a) LiOH, THF, MeOH, H2O
(58% over three-steps—see Scheme 5); (b) 1-BOC-piperazine, EDC,
CH2Cl2, TEA; (c) TFA, CH2Cl2,, 1 h.
Of various C-5 substitutions, dimethyl sulfamate (24)
gave the best cell-based antiviral activity (2:
EC50 = 89 nM, 24: EC50 = 7 nM). Therefore, a series
of compounds carrying this substituent at C-5 with
small alkyl groups at C-6 was prepared (Scheme 7).
The monoprotected compound 3 was transformed to
sulfamate 19 using dimethyl sulfamoyl chloride and
O
N
O
O
O N
O
O
S
S
OH
O
O
a
b
N
N
N
N
N
N
O
O
O
F
F
F
OTIPS
OTIPS
OTIPS
3
19
20
d
e
c
f
O
O
O
N
O
O
O
N
O
O
O
O
S
S
S
S
O
N
O
N
N
N
N
N
N
N
N
N
O
O
O
O
F
OH
F
OH
F
OH
F
OH
24
23
22
21
Scheme 7. Reagents and conditions: (a) ClSO2N(Me)2, CH2Cl2, DMAP, TEA (75%); (b) i—MeMgBr, THF, rt; ii—aq 1 N HCl, CH2Cl2 (87%); (c)
BF3ÆOEt2, TES, CH2Cl2 (75%); (d) TFA, CH2Cl2 (75%); (e) i—Et2Zn, CH2Cl2, 0 ꢁC; ii—TFA, 0 ꢁC, 15 min; iii—CH2I2, CH2Cl2; iv—Compound 20,
CH2Cl2, 0 ꢁC to rt, overnight; (f) i—NaBH4, MeOH, THF (85%); ii—TES, TFA, CH2Cl2 (95%).

4034
M. Fardis et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4031–4035
Table 1. Strand transfer inhibition, antiviral activity, cytotoxicity, and
selected physicochemical properties of C-6 pyrrolloquinoline-8-ones
pyrroloquinoline-6,8-diones was improved by removing
the C-6 carbonyl group. Among the substitutions placed
at the C-6, small alkyl groups such as the methyl analog
7 and the spirocyclopropyl compound 10 had the most
attractive in vitro profiles. Additionally, the solubility of
C-6 substituted analogs was improved compared to the
parent imide 1. This class of compounds has favorable
drug-like properties and is being further pursued.
Compound
IC50
(nM)
EC50
(nM)
CC50
(nM)
Solubility (lM)
pH: 7.3/2.2
1
2
80
60
7500
89
19,000
5100
1600
6100
3300
427
<1.4/<1.4
3/285
3/193
7
88
77
8
150
7
375
124
1110
256
415
430
3400
685
8
10
12
14
15
16
17
18
21
22
23
24
26/49
210
36
9740
11,400
2800
24,000
5250
755
Acknowledgments
64
310
295
59
The authors are grateful to Dr. Will Watkins for his crit-
ical reading of the manuscript. The authors also thank
James Chen for developing a working model of the en-
zyme, Fang Yu, Xiaoping Qi, Ameneh Zeynalzadegan,
and Gregg Jones for carrying out the biological screens.
815
420
357
255
2/11
52
1650
14,200
1400
<1/<1
2.1/11
<2.3/7
443
7
Supplementary data
compared to the spirocyclopropyl analog 10. Therefore,
subsequent analogs were prepared using a linear side
chain to avoid a steric clash with the enzyme.
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmcl.2006.05.008.
In an attempt to install polar groups at the C-6 position,
compounds 14, 17, and 18 were prepared. Chemical sta-
bility studies indicated that compound 16 was stable un-
der the assay conditions. Thus, while 16 lacked the
necessary antiviral activity, we used it as a precursor
for preparation of acidic analog 17 and basic compound
18. The compounds with polar substitutions had poor
potency in the antiviral assay and were not further
pursued.
References and notes
1. Jaffe, H. Science 2004, 305, 1243.
2. Pommier, Y.; Johnson, A. A.; Marchand, C. Nat. Rev.
Drug Disc. 2005, 4, 236.
3. First report on discovery of diketo acid inhibitors of IN:
(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) A more current report: Little, S.; Drusano, D.;
Schooley, R.; Haas, D.; Kumar, P.; Hammer, S.; McMa-
hon, D.; Squires, K.; Asfour, R.; Richman, D.; Chen, J.;
Saah, A.; Leavitt, R.; Hazuda, D. J.; Nguyen, B. Y.;
Protocol 004 Study Team, Abstracts of Papers, 161, 12th
Conference on Retroviruses and Opportunistic Infections,
Boston, MA, February 22–25, 2005; Foundation of
Retrovirology and Human Health, Alexandria, VA, 2005.
4. Chen, J. M.; Chen, X.; Fardis, M.; Jin, H.; Kim, C. U.;
Schacherer, L. N. WO 04035576; Chem. Abstr. 2004, 140,
375156.
5. Verschueren, W. G.; Dierynck, I.; Amssoms, K. I. E.; Hu,
L.; Boonants, P. M. J. G.; Pille, G. M. E.; Daeyaert, F. F.
D.; Hertogs, K.; Surleraux, D. L. N. G.; Wigerinck, P. B.
T. P. J. Med. Chem. 2005, 48, 1930.
6. Jin, H. et al., Bioorg. Med. Chem. Lett., in press.
7. Murray, W. M.; Semple, J. E. Synthesis 1996, 10, 1180.
8. Deniau, E.; Enders, D. Tetrahedron Lett. 2000, 41, 2347.
9. Yang, Z.; Lorenz, J. C.; Shi, Y. Tetrahedron Lett. 1998, 39,
8621.
The enzymatic activity of compounds bearing a sulfa-
mate at C-5 was generally poorer compared with
analogs with a methyl group at this position. However,
the methyl sulfamate 21 as well as the dihydrosulfamate
24 both had impressive antiviral activities. This class of
compounds is less soluble than the C-5 methyl series and
did not offer an improvement in the physicochemical
profile of the compounds.
In general, the SAR of C-5 sulfamates does not parallel
that of C-5 methyl series. This may be due to the steric
or electronic factors. Sterically, the presence of substitu-
ents at C-6 may alter the preferred conformation of C-5
side chain compared with the dihydro C-6 analogs. Elec-
tronically, the presence of an electron-withdrawing
group at C-5 might affect the pKa of the C-9 phenol
which in turn changes the binding properties of the
pharmacophore region. Further studies are necessary
to clarify the importance of each factor affecting the
enzymatic activity.
10. Decroix, B.; Netchita¨ılo, P. J. Heterocycl. Chem. 2000, 37,
827.
11. (a) Strand transfer assay was modified from a previous
report (Hazuda, et al. Nucleic. Acids Res. 1994, 22 1121).
Biotinylated donor DNA was bound to reacti-bind high
binding capacity streptavidin-coated white plates. DIG-
tagged target DNA with anti-DIG antibody-conjugated
horseradish peroxidase detection was used.; (b) For
antiviral assay, 50 ll of 2X test concentration of 5-fold
serially diluted drug in culture medium was added to each
well of a 96-well plate (9 concentrations) in triplicate.
In summary, a new class of inhibitors of IN with favorable
in vitro potency is reported herein. The pharmacophore
region of the molecule, the 1-nitrogen, the 8-carbonyl,
and the 9-phenol, was left intact. A number of analogs
with various substitutions at C-6 were evaluated in order
to understand the relationship between C-6 and C-5
groups. In general, the antiviral activity of the

M. Fardis et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4031–4035
4035
MT-2 or MT-4 cells were infected with HIV-1 IIIB at an
moi of 0.01 for 3 h. Fifty microliters of infected cell
suspension in culture medium (ꢀ1.5 · 104 cells) were then
added to each well containing the drug dilutions. The
plates were incubated at 37 ꢁC for 5 days. One hundred
microliters of CellTiter-GloTM Reagent (catalog # G7571,
Promega Biosciences, Inc., Madison, WI) was then added
to each well. Cell lysis was allowed to complete by
incubating at room temperature for 10 min. Chemilumi-
nescence was then read. For the cytotoxicity assay, the
protocol is identical to that of the antiviral assay, except
that uninfected cells and a 3-fold serial dilution of drugs
were used.; (c) The effect of compounds binding to serum
protein components was evaluated by determining the
antiviral EC50 in MT-2 cells in 10% FBS in the presence or
absence of serum concentrations of HSA (35 mg/ml) or a1-
AGP (1.5 mg/ml). From the EC50 data in the presence of
each individual protein, the EC50 resulting from the
combined effect of both proteins (as in serum) can be
calculated. The derivation of the appropriate equation for
this calculation can be made through competitive binding
assumptions.