Towards novel S-DABOC inhibitors: Synthesis, biological investigation, and molecular modeling studies

Article abstract of DOI:10.1016/j.bmcl.2008.09.070

A small family of S-DABO cytosine analogs (S-DABOCs) has been synthesized and biologically evaluated as HIV-1 inhibitor both on wild type (wt) and drug-resistant mutants leading to the identification of an interesting compound (5d). Molecular modeling studies have been finally performed in order to rationalize the results.

Full text of DOI:10.1016/j.bmcl.2008.09.070

Bioorganic & Medicinal Chemistry Letters 18 (2008) 5777–5780
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Towards novel S-DABOC inhibitors: Synthesis, biological investigation,
and molecular modeling studies
Marco Radi a, Lucilla Angeli a, Luigi Franchi a, Lorenzo Contemori a, Giovanni Maga b, Alberta Samuele b,
Samantha Zanoli b, Mercedes Armand-Ugon c, Emmanuel Gonzalez c, Anuska Llano c, Jose A. Esté c,
Maurizio Botta a,
*
a Dipartimento Farmaco Chimico Tecnologico, Universita degli Studi di Siena, Via Alcide de Gasperi 2, I-53100 Siena, Italy
b Istituto di Genetica Molecolare, IGM-CNR, Via Abbiategrasso 207, I-27100 Pavia, Italy
c Retrovirology Laboratory irsiCaixa, Hospital Universitari Germans Trias i Pujol, Universitat Autónoma de Barcelona, E-08916 Badalona, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
A small family of S-DABO cytosine analogs (S-DABOCs) has been synthesized and biologically evaluated
as HIV-1 inhibitor both on wild type (wt) and drug-resistant mutants leading to the identification of an
interesting compound (5d). Molecular modeling studies have been finally performed in order to rational-
ize the results.
Received 16 August 2008
Revised 16 September 2008
Accepted 17 September 2008
Available online 23 September 2008
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords:
HIV-1
Reverse Transcriptase
Mutant
NNRTIs
S-DABOCs
Molecular modeling
In the fight against the AIDS plague, the selection of drug-resis-
tant viruses represents a significant obstacle to the eradication of
the HIV infection.1 To overcome the therapeutic failures connected
with the selection of mutant strains resistant to common Reverse
Transcriptase inhibitors (RTIs) or Protease inhibitors (PIs), different
molecular targets have been recently investigated. An arsenal of
anti-HIV drugs belonging to six different classes, after the recent
approval of maraviroc (entry inhibitor)2 and raltegravir (integrase
inhibitor), are currently on the market.3 Although the discovery
of the last two drugs represents an important success in the fight
against drug resistant HIV-1 viruses, a few clinical studies demon-
strated that therapies based on maraviroc and raltegravir have
comparable efficacy to regimens based on the non-nucleoside
reverse transcriptase inhibitor (NNRTI) efavirenz.4 Reverse trans-
criptase (RT) represents therefore an old but still important target
which is highly investigated for the identification of novel NNRTIs
endowed with a better activity profile against clinically relevant
mutant strains.5 A number of potent second-generation NNRTIs
are in fact currently in clinical trial (Fig. 1).6 During the course of
our long lasting studies on S-DABO RT inhibitors, we discovered
very potent compounds active on HIV-1 (wt)-infected cells at
Me
CN
O
O
CN
IDX-899
Me
Me
N
MeO P
H
N
NH2
O
N
N
H
Br
CN
NH2
Etraviridine
NC
CN
Et
Me
H
H
N
OH
O
N
N
N
N
N
Et
UK-453061
NC
Me
CN
Rilpivirine
Figure 1. Etravirine and other selected NNRTIs in clinical trial.
picomolar concentration.7 However, a pronounced loss of activity
was commonly observed when the same compounds were tested
against drug resistant HIV-1 mutants (K103N, Y181C, and
Y188L). To overcome the loss of activity against the above-
* Corresponding author. Tel.: +39 0577 234306; fax: +39 0577 234333.
E-mail address: botta@unisi.it (M. Botta).
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.09.070

5778
M. Radi et al. / Bioorg. Med. Chem. Lett. 18 (2008) 5777–5780
reported by us,8a the key intermediate 2 was reacted with five dif-
O
N
ferent Grignard reagents to give the alcohols 3a–e which were then
oxidized to the corresponding C6-keto derivatives 4a–e by reaction
with Dess–Martin periodinane (Scheme 1). Finally, selective C4
chlorination in refluxing POCl3 followed by nucleophilic substitu-
tion with methanolic ammonia gave the S-DABOC derivatives
5a–e in which the X and Z moieties were kept fixed (and equal
to S and CO, respectively) while introducing different R1 groups
on C60 (see Table 1).
MeO
HN
S
Cl
Cl
S
-DABO (1)
R2
R2
K103 P236
N
On the other hand, starting from the simpler derivative 5a, R1
was kept fixed (and equal to Ph) while additional functionalization
was introduced in C60 and C2 in order to select the most interesting
substituents X and Z (for the antiviral activity) to be subsequently
coupled with the best R1 group. A series of more flexible deriva-
tives was initially synthesized by reduction of the C60 keto moiety
with NaBH4 to give the corresponding alcohol 6 which was further
converted into the C6-chloromethyl phenyl and C6-fluoromethyl
phenyl derivatives (7a and 7b) by reaction with DAST and PCl5,
respectively (Scheme 2).
R3
L100
K102
Y318
N
NH2
R1
Y
Me
S
N
N
Me
R1
X
N
I
Z
R4
S-DABOCs
Figure 2. S-DABO precursor (1), S-DABOCs and new S-DABOC analogs with generic
structure I.
mentioned mutants, we decided to synthesize simplified analogues
of the S-DABO precursor 1 (Fig. 2) in order to identify a smaller
molecule able to give specific interactions with residues of the allo-
steric site not associated with any drug-resistant mutation.
Accordingly, the S-DABO cytosine analog family (S-DABOC) has
been recently identified.8 Based on the encouraging preliminary
results on HIV inhibition, this study has been focused on the devel-
opment of novel S-DABOC analogs with generic structure I in order
to get additional information on the SAR for this new family of
inhibitors (Fig. 2).
O
O
Me
Me
OH
i.
HN
MeS
HN
MeS
N
CHO
N
3a-e
R1
2
ii.
As starting point for this work, we decided to maintain invariate
the functional groups which were previously found important for
the RT:S-DABOCs interaction and to introduce additional variabil-
ity at C2 and C6 positions (X, Z, and R1): the target compounds
(generic structure I) should therefore bear a C4 amino group
(responsible for key hydrogen bond interaction with the backbone
of Lys103 and Pro236) and a C5 methyl group (which seems to be
accommodated in a hydrophobic region defined by Leu100, Lys102,
and Tyr 318).8b Following a versatile synthetic approach previously
O
NH2
N
iii.
Me
Me
O
HN
MeS
N
O
N
MeS
R1
R1
4a-e
5a-e
Scheme 1. Reagents and conditions: (i) RMgBr, THF, rt, 1 h; (ii) Dess–Martin
periodinane, CH2Cl2, rt, 1 h; (iii) a—POCl3, reflux, 1 h; b—NH3(g)/MeOH, steal bomb,
100 °C, 2 h.
Table 1
Anti HIV-1 activity and cytotoxicity of S-DABOC derivatives
NH2
Me
R1
N
Me
X
N
I
Z
a,b
a,c
d
Entry
Compound
X
Z
R1
ID50
wt
(lM)
ED50
(lM)
CC50
NL4-3 wt
K103N
Y181C
1.71 (28)
>25
>25
>25
>25
1
2
3
4
5
6
7
8
5a
6
7a
7b
8
S
S
S
S
CO
Ph
Ph
Ph
Ph
Ph
Ph
Ph
1.9
7.7
>20
0.06
>25
11.98
>25
10.45
0.44
3.4
>25
>25
0.034
>19.37
0.007
0.04
1.67 (28)e
>25
>25
>25
>25
>25
>25
>25
>25
>25
>25
>25
CHOH
CHCl
CHF
CNOH
CO
CO
CO
CO
CO
>25
>25
>25
>20
>20
16
>20
>20
>20
S
9
SO2
NH
S
S
S
19.12 (43)
10.37 (23)
10
5b
5c
5d
5e
1
>25
>25
>25
>25
>25
>25
CH2Ph
CH2CH2Ph
3-F–Ph
Naphthyl
9
10
11
12
13
14
0.2
2.7
0.004
0.695 (20)
>19.37
4.7 (671)
0.68 (17)
0.133 (133)
0.779 (23)
S
CO
>19.37
>19.37
>25
>2
ndf
NVP
EFV
—
—
—
—
—
—
—
—
>2
0.008 (8)
0.001
>1
a
Data represent mean values of at least two experiments.
b
c
d
e
f
ID50: Inhibiting dose 50 or needed dose to inhibit 50% of the enzyme.
EC50: Effective concentration 50 or needed concentration to inhibit 50% HIV-induced cell death, evaluated with MTT method in MT-4 cells.
CC50: Cytotoxic concentration 50 or needed concentration to induce 50% death of noninfected cells evaluated with the MTT method in MT-4 cells.
Fold-resistance is reported in parenthesis and expresses the ratio of EC50 value against drug-resistant strain and EC50 of the wild type NL4-3 strain.
Nd, not determined.

M. Radi et al. / Bioorg. Med. Chem. Lett. 18 (2008) 5777–5780
5779
keto moiety (entries 8–11), the best substituent was found to be
the 3-fluorophenyl and the corresponding compound 5d repre-
sents the most active derivative of the series both against HIV-1
wild type and resistant mutants (Table 1). As a general consider-
ation, the increase of the C6 substituent flexibility either by intro-
ducing of a spacer between the C60 keto moiety and the phenyl ring
(5b and 5c) or by converting 5a into the C6-hydroxymethy phenyl,
C6-chloromethy phenyl and C6-fluoromethyl phenyl derivatives
(6, 7a, and 7b respectively), always led to a loss of activity (Table
1). In addition, the introduction of a polar moiety in C2 (X = SO2,
compound 9 and X = NH, compound 10) always led to a loss of
activity. The most important requirements for a good antiviral
activity in the S-DABOC series seems to be the presence of a rigid
structure characterized by a small substituted benzoyl group in
C6: compound 5d showed an activity profile better than that of
both the S-DABO precursor 1 and nevirapine, exhibiting fold resis-
tance values comparable to this commercial drug, EC50 values in
the high-nanomolar range against all the studied mutants and
low cytotoxicity. (see Fig. 3).
In order to have an idea of the druglikeness of our S-DABOC
inhibitors, significant physicochemical properties were predicted
using QikProp v2.59 and the results obtained were compared with
that of common anti-HIV drugs on the market (Table 2). In a
recently published review,6 Sweeney and Klumpp reported the
importance of pharmacokinetic factors in the success of first-line
anti-HIV therapies based on NNRTIs. A head to head comparison
between nevirapine and efavirenz, highlighted that despite the lat-
ter possesses higher in vitro antiviral activity, its poor pharmacoki-
netic profile make it clinically comparable to the less active
nevirapine. Efavirenz has in fact shown low solubility (LogS),
extensive binding to human serum proteins (LogK) and high inci-
dence of CNS side effects (LogBB) that can be easily predicted with
QikProp (Table 2). It was interesting to note that the predicted
physicochemical properties for the most active S-DABOC 5d seems
to be similar to that of nevirapine thus making it a good candidate
for further studies.
NH2
Me
4
N
2
6
6'
O
MeS
N
iv.
i.
NH2
5a
NH2
Me
OH
Me
N
N
iii.
O
MeS
N
MeS
O
N
O
6
9
NH2
Me
NOH
N
ii.
v.
MeS
N
NH2
NH2
8
Me
Me
N
N
O
X
MeHN
N
MeS
N
10
7a; X = F
7b; X = Cl
Scheme 2. Reagents and conditions: (i) NaBH4, MeOH, rt, 10 min; (ii) DAST, CH2Cl2,
0 °C, 1 h (for 7a); SOCl2, CH2Cl2, rt, 24 h (for 7b); (iii) NH2OH?HCl, AcONa, EtOH,
reflux, 24 h; (iv) OxoneÒ, MeOH/H2O 1:1, rt, 4 h; v. CH3NH2, MW, 170 °C, 30 min.
The oxime 8 was easily obtained reacting compound 5a with
hydroxylamine hydrochloride, while the functionalizations in C2
was accomplished by converting 5a to the corresponding sulfone
9 and final nucleophilic substitution with methylamine under
microwave assisted condition to give compound 10. All the synthe-
sized compounds were evaluated in enzymatic tests for their abil-
ity to inhibit wild type (wt) as well as on MT-4 cells for cytotoxicity
and anti-HIV activity, in comparison with nevirapine, efavirenz and
the S-DABO 1 used as reference drugs. In particular, the mutants
K103N and Y181C were used for tests on cell lines. The biological
results reported in Table 1 showed that among the Z substituents
(entries 1–5), the best one was represented by the keto moiety (en-
try 1), while the introduction of a methylsulfonyl or a methylamino
group on C2 in place of the methylthio determined a loss of activity
(entries 6–7). Among the R1 groups to be combined with the C6
Docking simulations were finally conducted in order to get fur-
ther insight on the SAR for the synthesized inhibitors starting from
the crystal structure of the TNK-651:RT (wt) complex. A detailed
analysis of the binding site was performed in order to identify
the features responsible for the better activity of 5d with respect
Figure 3. Graphical representation of the binding mode of the S-DABOCs 5a and 5d within the allosteric site of the TNK-651:RT complex (PDB code: 1RT2). Regions of
minimum energy as derived from molecular interaction field calculations are also displayed for probes F (cyan spheres), CH3 group (yellow spheres), CH aromatic or vinyl (red
spheres), and neutral flat NH2 (blue spheres). (A) Docked conformation of 5a (sticks, green carbon atoms) and (B) docked conformation of 5d (sticks, purple carbon atoms).

5780
M. Radi et al. / Bioorg. Med. Chem. Lett. 18 (2008) 5777–5780
Table 2
Phe227 and Trp229 which could account for the better activity
profile.
Predicted physiochemical associated properties for synthesized S-DABOCs and
commercial anti-HIV drugs
In conclusion,
a SAR study on the synthesized S-DABOC
Compound
QP-
QP-
Number of
QP-
QP-
QP-
analogues led to the identification of an interesting inhibitor (5d)
endowed with nanomolar activity and predicted pharmacokinetic
properties similar to that of anti-HIV drugs on the market. Molec-
ular modeling studies suggested the important features responsi-
ble for the RT:S-DABOCs interaction and will be of help in the
synthesis of more active analogues.
LogPa
LogSb
metabolitesc
LogKd
LogBBe
Cacof
5a
6
7a
7b
8
5b
9
10
5c
5d
5e
1
2.28
2.24
3.57
3.26
1.89
2.74
0.27
1.47
3.05
2.54
3.45
5.5
ꢀ3.43
ꢀ3.21
ꢀ4.28
ꢀ3.79
ꢀ3.05
ꢀ3.9
1
3
3
3
1
2
1
1
3
1
1
4
6
2
4
2
ꢀ0.12
ꢀ0.17
0.23
ꢀ0.51
ꢀ0.57
ꢀ0.06
ꢀ0.12
ꢀ0.9
1039.53
978.04
2334.41
2144.49
509.9
1098.97
136.37
539.46
1138.96
1042.52
997.79
1725.36
1917.57
1315.89
243.63
55.64
0.15
ꢀ0.25
0.01
ꢀ0.57
ꢀ1.44
ꢀ0.88
ꢀ0.65
ꢀ0.41
ꢀ0.6
ꢀ2.48
ꢀ2.91
ꢀ4.23
ꢀ3.87
ꢀ4.95
ꢀ7.18
ꢀ3.67
ꢀ5.1
ꢀ0.6
ꢀ0.28
0.09
Acknowledgments
This study was supported by grants from the European THiNC
Consortium. The University of Siena and the Spanish FIS
PI060624 are also acknowledged for financial support. S.Z. is the
recipient of a Buzzati-Traverso Fellowship.
ꢀ0.07
0.31
0.89
0.01
0.29
0.15
ꢀ0.36
ꢀ0.1
Nevirapine
Efavirenz
Delavirdine 2.61
2.32
3.51
ꢀ0.02
ꢀ1.6
ꢀ5.75
ꢀ5.99
Supplementary data
Etravirine
2.67
0.25
ꢀ2.1
a
Predicted octanol/water partition coefficient; range of recommended values
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2008.09.070.
(ꢀ2.0)–(+6.5).
b
c
Predicted aqueous solubility; range of recommended values: (ꢀ6.5)–(+0.5).
Number of likely metabolic reactions; range of recommended values: 1–8.
Prediction of binding to human serum albumin; range of recommended values:
d
References and notes
(ꢀ1.5)–(+1.5).
e
Predicted brain/blood partition coefficient; range of recommended values:
(ꢀ3.0)–(+1.2).
1. (a) Yeni, P. J. Hepatol. 2006, 44, S100–S103; (b) Crespan, E.; Locatelli, G. A.;
Cancio, R.; Hubscher, U.; Spadari, S.; Maga, G. Antimicrob. Agents Chemother.
2005, 49, 4546–4554.
f
Predicted apparent Caco-2 cell permeability in nm/s; range of recommended
values: <25 poor; >500 great.
2. Este, J. A.; Telenti, A. Lancet 2007, 370, 81.
3. (a) Deeks, S. G.; Kar, S.; Gubernick, S. I.; Kirkpatrick, P. Nat. Rev. Drug Discov.
2008, 7, 117; (b) Oversteegen, L.; Shah, M.; Rovini, H. Nat. Rev. Drug Discov.
2007, 6, 951–952.
4. (a) Markowitz, M. et al J. Acquir. Immune Defic. Syndr. 2007, 46, 125–133; b Saag,
M., et al. Fourth International AIDS Society Conference on HIV Pathogenesis,
Treatment and Prevention, Sydney, Australia, 2007; WESS104.
to 5a. For this purpose, molecular interaction fields (MIFs) were
calculated for the binding site using the software Grid10 (for fur-
ther details see Supplementary data). Details derived from both
docking studies and Grid analysis allowed to rationalize the biolog-
ical results. In summary, two different binding modes were found
for the most active inhibitors: compound 5a showed a binding
mode similar to that previously found8b and characterized by key
hydrogen bond interaction with the backbone of Lys103 and
Pro236. This kind of interactions have been suggested to be very
important for maintaining the activity against the RT mutants
since are less likely to be disrupted as the result of a mutation.11
Alternatively, the presence of a meta-fluoro substituent on the
C6-benzoyl ring determine a 180° flip of the molecule along the
CO axis in order to locate the fluoro atom and other important fea-
tures on the corresponding minima calculated by GRID. In this ori-
entation, compound 5d gives two hydrogen bond contacts with the
5. Ren, J.; Stammers, D. K. Virus Res. 2008, 134, 157–170.
6. Sweeney, Z. K.; Klumpp, K. Curr. Opin. Drug Discov. Dev. 2008, 11, 458–470.
7. (a) Manetti, F.; Esté, J. A.; Clotet-Codina, I.; Armand-Ugon, M.; Maga, G.;
Crespan, E.; Cancio, R.; Mugnaini, C.; Bernardini, C.; Togninelli, A.; Carmi, C.;
Alongi, M.; Petricci, E.; Massa, S.; Corelli, F.; Botta, M. J. Med. Chem. 2005, 48,
8000–8008; (b) Mugnaini, C.; Alongi, M.; Togninelli, A.; Gevariya, H.; Brizzi,
A.; Manetti, F.; Bernardini, C.; Angeli, L.; Tafi, A.; Bellucci, L.; Corelli, F.;
Massa, S.; Maga, G.; Samuele, A.; Facchini, M.; Clotet-Codina, I.; Armand-
Ugón, M.; Esté, J. A.; Botta, M. J. Med. Chem. 2007, 50, 6580–6595; c Botta,
M.; Corelli, F.; Petricci, E.; Radi, M.; Maga, G.; Esté, J. A.; Mai, A. WO 2007/
043094 A2.
8. (a) Radi, M.; Contemori, L.; Castagnolo, D.; Spinosa, R.; Esté, J. A.; Massa, S.;
Botta, M. Org. Lett. 2007, 9, 3157–3160; (b) Radi, M.; Falciani, C.; Contemori, L.;
Petricci, E.; Maga, G.; Samuele, A.; Zanoli, S.; Terrazas, M.; Castria, M.;
Togninelli, A.; Esté, J. A.; Clotet-Codina, I.; Armand-Ugòn, M.; Botta, M.
ChemMedChem 2008, 3, 573–579.
9. QikProp, version 2.5, Schrödinger, LLC, New York, NY, 2005.
10. Grid 22, Molecular Discovery Ltd., Pinner, Middlesex, UK.
11. Ren, J.; Nichols, C. E.; Bird, L. E.; Fujiwara, T.; Sugimoto, H.; Stuart, D. I.;
Stammers, D. K. J. Biol. Chem. 2000, 275, 14316–14320.
backbone of Lys103 and Tyr188, p–p interaction with Tyr318 and
important hydrophobic interactions between the small C2-methyl-
thio group and the hydrophobic pocket defined by Tyr181, Tyr188,