Polyamide Amino Acids trimers as TAR RNA ligands and anti-HIV agents

Article abstract of DOI:10.1016/j.bmc.2010.09.002

Based on a split-and-mix strategy, a library of trimeric Polyamide Amino Acids (PAA) incorporating four different amino acids (Lys, Ala, Arg, and Phe) has been prepared. Screening of the batches for HIV TAR RNA binding in a fluorescent assay allowed the i

Full text of DOI:10.1016/j.bmc.2010.09.002

Bioorganic & Medicinal Chemistry 18 (2010) 7432–7438
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Polyamide Amino Acids trimers as TAR RNA ligands and anti-HIV agents
Vanessa Bonnard a, Lise Pascale a, Stéphane Azoulay a, Audrey Di Giorgio a, Christine Rogez-Kreuz b,
Karine Storck b, Pascal Clayette b, Nadia Patino a,
⇑
a Laboratoire de Chimie des Molécules Bioactives et des Arômes, UMR 6001 CNRS UNSA, Institut de Chimie de Nice, Université de Nice Sophia Antipolis, 06108 Nice Cedex, France
b Neurovirology Department, SPI-Bio, CEA, 18 route du Panorama, BP 6, 92265 Fontenay-aux-Roses Cedex, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 3 June 2010
Revised 30 August 2010
Accepted 1 September 2010
Available online 7 September 2010
Based on a split-and-mix strategy, a library of trimeric Polyamide Amino Acids (PAA) incorporating four
different amino acids (Lys, Ala, Arg, and Phe) has been prepared. Screening of the batches for HIV TAR
RNA binding in a fluorescent assay allowed the identification of several components that interact with
TAR RNA at a micromolar concentration, with a good TAR versus tRNA specificity. Some of these com-
pounds compete efficiently with the association of TAR and Tat protein. In cell cultures, these compounds
display a moderate antiviral activity, associated nevertheless with some toxicity. Overall, these results
confirm that this new family can be a basis for the design of novel RNA targeting drugs.
Ó 2010 Elsevier Ltd. All rights reserved.
Keywords:
Polyamide Amino Acids
TAR RNA ligands
Anti-HIV agents
1. Introduction
As part of our efforts to identify new types of specific RNA ligands,
wehave recentlydesigned a new family of compounds, named ‘Poly-
Small molecules that recognize and bind a three-dimensional
RNA structure with high affinity and specificity may provide new
types of drugs that exert their effects by inhibiting the biological
function of the RNA target.1,2 In this context, considerable efforts
have been directed to the finding of specific ligands of the HIV-1
TAR RNA fragment, a 59-base stem-loop structure located at the
50-ends of all nascent HIV-1 transcripts.3,4 A key step during the
HIV-1 transcription is the recognition of TAR RNA by the viral regu-
latory Tat protein, which assembles along with host cell factors
(CDK9 and cyclin T1) to form a ternary complex, resulting in a drastic
increase in the number of viral transcripts.5–8 Because the formation
of this complex is essential for HIV replication, the development of
inhibitorsof this interactioncould providethe basis for newantiviral
therapies. Several low molecular-weight compounds have been re-
ported that block in vitro the Tat/TAR interaction, among which
intercalators, aminoglycoside derivatives, peptides, and peptidomi-
metics.9 Although some of them show viral inhibition in cellular
assays, none has yet advanced to the clinic. Therefore, it is of contin-
ued interest to search for new low molecular-weight molecules that
bind with high affinity and selectivity to TAR and inhibit the forma-
tion of the ternary complex and, consequently, HIV replication.
amide Amino Acids (PAA)’, which are constituted by repetitive 2-
aminoethyl glycyl units (e.g., the PolyamideNucleic Acidsbackbone;
PNA: oligonucleotide analogs), onto which
a-amino acid residues
are grafted (Fig. 1A). Preliminary studies on eight tetra-PAA models,
based on four selected amino acids (Alanine (A), Phenylalanine (F),
Arginine (R), and Lysine (K); Fig. 1B), highlighted the potential of
these new compounds to behave as TAR RNA ligands, which bind
their target via both electrostatic and non-electrostatic interactions,
the affinity and tRNA versus TAR specificity depending on the nature
of the amino acid residues.10 Very interestingly, these tetra-PAA dis-
played also a high TAR RNA versus DNA selectivity.
In parallel to this work, and with the aim of establishing ‘struc-
ture–activity’ relationships, a library of PAA trimers, including the
four amino acids A, F, R, and K, was also prepared and the binding
to the TAR RNA fragment (Fig. 1C) was investigated. In this paper,
we report the ‘split and mix’ synthesis of this library and of se-
lected individual tri-PAA, as well as their TAR RNA binding affinity
and specificity. The ability to inhibit the TAR/tat complex and the
anti-HIV activity and cytotoxicity in PBMCs were also assessed.
2. Results and discussion
2.1. Synthesis
Abbreviations: NMP, N-methyl-2-pyrrolidinone; DIPEA, N,N-diisopropylethyl-
amine; HBTU, 2-(1-H-benzotriazol-1-yl)1,1,3,3-tetramethyluroniumhexafluorophos-
phate; Pyr, pyridine; TFA, trifluoroacetic acid; TFMSA, trifluoromethanesulfonic acid;
Ac2O, acetic anhydride; TIS, triisopropylsilane; DCM, dichloromethane.
2.1.1. Library synthesis
The tri-PAA library was prepared following an on-resin split-and-
mix strategy, starting from Boc-protected PAA (Ala) and (Phe), and
Boc-/Z-protected PAA (Lys) and (Arg) monomers (Fig. 2A). Following
⇑
Corresponding author. Tel.: +33 04 92 07 61 46; fax: +33 04 92 07 61 51.
E-mail addresses: pascal.clayette@cea.fr (P. Clayette), patino@unice.fr (N.
Patino).
0968-0896/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2010.09.002

V. Bonnard et al. / Bioorg. Med. Chem. 18 (2010) 7432–7438
7433
Figure 1. (A) Polyamide Nucleic Acids (PNA) versus Polyamide Amino Acids (PAA) structures. (B) Series of tetra-PAA previously studied.10 (C) Secondary structure of the TAR
RNA fragment used in this study.
this combinatorial strategy, the 64 compounds constituting the
library were divided into four sub-libraries of 16 tri-PAA, with
respect to the nature of the PAA located at the C-extremity
(SL(AA1–4)). Each of these sub-libraries was further split into four
batches of four tri-PAA. Each batch is characterized by the nature of
the two PAA monomers located at the N- and C-extremities (Fig. 2B).
The synthesis of each sub-library was performed as illustrated
in Figure 2C: b-alanine functionalized MBHA-LL resin was coupled
to one of the four N-protected PAA monomer, using an HBTU pro-
cedure. Unreacted amino groups on the resin were capped with a
Ac2O/pyridin/NMP solution. The resin was then split into four por-
tions and each portion was reacted with a TFA/TIS (9:1) solution
for Boc cleavage. Then, a second PAA selected among the four N-
protected PAA monomers was introduced following the same elon-
gation procedure. After capping, the four portions were mixed then
split again into four batches. Each batch was subjected to a Boc
deprotection then to the elongation procedure with the third PAA
among the four N-protected PAA monomers. After Boc cleavage
and N-acetylation of the terminal PAA residue, each batch was sub-
mitted to acidolysis (TFMSA/TFA/TIS (1:8:1) solution) for the cleav-
age of the tri-PAA from the support and the concomittant Z-
deprotection for the Arg- and/or Lys-containing batches. Precipita-
tion from cold diethyl ether and subsequent lyophylisation from
water gave the sub-library as four batches of four tri-PAA each in
good yields. These mixtures were desalted by HPLC semi-prepara-
tive and their composition was verified by ESI MS experiments.
SL(K), and SL(R)) (see data in Table 1), then on the individual tri-
PAA 1–24 and 25–28 (see data in Table 2), by monitoring the fluo-
rescence decrease of a 50-fluorescently-labelled (Alexa 488) TAR
fragment (27 mer, 5 nM) observed upon addition of increasing con-
centrations of either mixtures of four tri-PAA from a batch or single
tri-PAA (from 0.01 to 2000 lM). These experiments were per-
formed at 25 °C, in buffer A (20 mM HEPES (pH 7.4 at 25 °C),
20 mM NaCl, 140 mM KCl, and 3 mM MgCl2), which closely
approximates ionic conditions inside cells. Dissociation constants
(Kd) were then determined from the binding curves, as previously
described.10
2.2.1. Library studies
For all batches, curves fitted well a 1:1 stoichiometry model and
dissociation constants (Kd) were determined, these values indicat-
ing the ‘global’ binding activity of the corresponding batches (Ta-
ble 1). A good discrimination between the different batches was
obtained with Kd values ranging from 1.5 to 150
brary containing at its C-extremity the Ala residue (i.e., SL(A)) glob-
ally showed the lowest binding affinity (25 M < Kd < 150 M),
comparatively to SL(K) (7 M < Kd < 65 M), SL(R) (4 M < Kd <
48 M), and SL(F) (1.5 M < Kd < 24 M), the later sub-library dis-
lM. The sub-li-
l
l
l
l
l
l
l
l
playing the strongest affinity. In addition, within each sub-library
and with respect to the nature of the PAA at the N-extremity of
the trimer, the Kd value is decreasing along the Ala ꢁ Lys >
Arg > Phe sequence. These results indicate that, unlike the Ala res-
idue, the aromatic Phe monomer, the cationic Arg monomer and, to
a lesser extent, the cationic Lys one, contribute to enhance the TAR
binding affinity.
In order to identify tri-PAA hits, the 24 tri-PAA (1–24) constitut-
ing six of the most interesting batches (FXF), (RXF), (FXR), (FXK),
(RXR), and (RXK) were synthesized individually (see above). More-
over, we also prepared as controls the four tri-PAA (25–28) consti-
tuting the batch (KXA), which exhibited only a poor global binding
2.1.2. Individual tri-PAA synthesis
Global interaction studies of the 16 batches led us to synthesize
individually 28 tri-PAA (1–28). Their synthesis was performed as
described above for the preparation of the sub-libraries (Fig. 2C).
It should be noted that tri-PAA incorporating two and three Arg
residues were obtained in lower yields than the other ones (HPLC
yields from 58% to 85%). Mono, di, and tri acetylated derivatives
were identified as by-products by MS experiments, likely demon-
strating that under the acid conditions of Boc cleavage, partial Z-
deprotection of guanidinium moieties occurred, leading to their
acetylation during capping steps. All the 28 tri-PAA were easily
purified by semi-preparative reverse-phase HPLC and unambigu-
ously identified by HRMS.
activity (Kd = 107 lM).
2.2.2. Individual tri-PAA
The tri-PAA 25–28 exhibited
a
low affinity for TAR
(Kd P 46 lM; Table 2), demonstrating that global Kd values likely
reflect the binding potential of the four individual tri-PAA consti-
tuting batches. Concerning compounds 1–24, Kd values ranged
2.2. TAR RNA binding affinity
from 2 to 80 lM. Within all batches, a similar tendency was
observed, the binding affinity increasing along the sequence:
XAX0 < XKX0 6 XRX0 6 XFX0. These results confirm that the Ala
residue disfavors the interaction with TAR, whatever its position
The TAR binding affinity was first investigated on the 16
batches (4 ꢀ 4) constituting the four sub-libraries (SL(A), SL(F),

7434
V. Bonnard et al. / Bioorg. Med. Chem. 18 (2010) 7432–7438
Figure 2. (A) Structure of the four Boc/Z protected PAA monomers. (B) Structure of the tri-PAA library. (C) Sub-libraries synthesis: (i) PAA (AA1), HBTU, DIEA, NMP; (ii) Ac2O/
Py/NMP (15/10/75); (iii) TFA/TIS (90/10); (iv) TFMSA/TIS/TFA (10/10/80).
Table 1
Global TAR RNA binding activities (Kd) of tri-PAA sub-librariesa
Sub-library SL (PAA)
SL (A)
Batches (PAAN-X-PAAC)
Kd (M)
Sub-library SL (PAA)
SL (K)
Batches (PAAN-X-PAAC)
Kd
(
l
M)
(AXA)
(FXA)
(KXA)
(RXA)
(AXF)
(FXF)
(KXF)
(RXF)
150 50
25 10
(AXK)
(FXK)
(KXK)
(RXK)
(AXR)
(FXR)
(KXR)
(RXR)
65
7
37
10
48 10
4.2 0.6
8
1
4
5
107
29 10
24
1.5 0.7
9
SL (F)
6
SL (R)
8
2
14
7
6
2
3.8 0.9
a
All standard fluorescence measurements were performed in buffer A (20 mM HEPES (pH 7.4 at 25 °C), 20 mM NaCl, 140 mM KCl, and 3 mM MgCl2). Each batch (PAAN-X-
PAAC) contains four compounds, in which the nature of the internal X residue varies (X = PAA(A) or (F) or (K) or (R)).

V. Bonnard et al. / Bioorg. Med. Chem. 18 (2010) 7432–7438
7435
‘anagram’ compounds FRR 8 and RRF 16 (K00 /Kd = 5.7 and 13.6,
respectively).
Table 2
d
Dissociation constants (Kd) for individual tri-PAA-TAR RNA interactionsa
Batches
(FXF)
Tri-PAA
Kd
(
l
M)
Batches
(RXF)
Tri-PAA
Kd
(lM)
2.3. FRET study of Tat/TAR complex inhibition
FAF 1
FFF 2
FKF 3
FRF 4
ndb
ndb
8
RAF 13
RFF 14
RKF 15
RRF 16
RAR 17
RFR 18
RKR 19
RRR 20
RAK 21
RFK 22
RKK 23
RRK 24
34
3
7
2
80
8
12
14
76 20
6
10
10
7
2
2
1
9
2
1
4
3
1
5
1
1
The extent to which tri-PAA could disrupt the TAR/tat complex
was investigated with trimers 4, 6, 8, and 14–16, which are among
the best ligands. This study was performed using a FRET-based
competition assay based on the Alexa (A488) 50-labelled-TAR
RNA fragment as the fluorescent donor and a Tat (47–58) fragment
labeled with rhodamine as the fluorescent acceptor (Rh-Tat).12
Association of Rh-Tat with the A488-TAR RNA resulted in a 1.4-fold
decrease in fluorescence, probably by driving the peptide in an ex-
tended conformation, which allows the energy transfer between
the two dyes to occur. Titrating A488-TAR with Rh-Tat established
the 1:1 complex stoichiometry. In our conditions, a Kd value of
25 nM could be calculated from these data, which is in the same
range than reported values.12,13 We performed single-point mea-
surements of the inhibition potency of the candidate molecules
3
(FXR)
FAR 5
FFR 6
FKR 7
FRR 8
FAK 9
FFK 10
FKK 11
FRK 12
KAA 25
KFA 26
KKA 27
KRA 28
22
2
(RXR)
(RXK)
4
2.2 0.7
68
3
(FXK)
8
1
3
2
1
3
4
9
7
(KXA) control
>800
46 12
109 20
76 14
a
All standard fluorescence measurements were performed in buffer A.
Not determined.
b
at four fixed concentrations (1, 10, 100, and 400 lM). In all cases,
within the trimer sequence, contrary to the aromatic Phe and cat-
ionic Arg residues and, to a lesser extend, Lys ones. The presence of
at least one Phe residue was found to favor affinity. Indeed, 19, 20,
23, and 24 trimers, which contain only Lys/Arg residues, display a
an increase of the fluorescence emission was observed, demon-
strating that tri-PAA compete with the Tat fragment for the same
binding site on TAR. All tested compounds were able to completely
displace Tat at the highest concentration, the most potent com-
pounds, RFF 14, RKF 15, and RRF 16 displaying an IC50 value around
lower affinity (14 P Kd P 10 lM) than any trimer deriving from a
combination of at least one Phe with the cationic Lys/Arg residues
(e.g., 3, 4, 6–8, 10–12, and 14–16). These results are in line with
those observed in the case of tetra-PAA, confirming that TAR affin-
ity does not result exclusively from non-specific electrostatic inter-
actions with Lys and Arg units. The best tri-PAA ligands (Kd =
1 lM (Fig. 4).
2.4. Antiviral potency of tri-PAA compounds
To have an insight into the potential of PAA to inhibit the HIV-1
replication, the antiviral activity of the six tri-PAA 4, 6, 8, and 14–
16, was evaluated on phytohemaglutinin-P (PHA-P)-activated
PBMC experimentally infected with the X4-tropic HIV-1-LAI strain,
which possesses the same TAR sequence than the mini-TAR used in
this study for acellular assays. Before the infection step, cells were
2–3
2 Phe/1 Arg (e.g., 4, 6, and 14) or 2 Arg/1 Phe (e.g., 8 and 16),
excepting trimer RFR 18 (Kd = 8 M). Among the Phe/Lys series,
the sequence FFK 10 (Kd = 3 M) is the only one displaying an affin-
lM) result from a combination of Phe and Arg residues, either
l
l
ity for TAR comparable to that of the Phe/Arg ones.
Furthermore, in a given series, TAR affinity does not seem to de-
pend on the trimer sequence, as tri-PAA containing the same three
residues located at different positions within the trimer display
comparable Kd values (see, for example, 4, 6, and 14 in the Phe/
Arg trimer series and 7, 12, 15, and 22, in the Phe/Arg/Lys series).
This observation could be explained by the inherent flexibility of
the PAA structure. Indeed, upon TAR binding, various conforma-
tions could be adopted, as illustrated in Figure 3, which represents
models of complexes formed between TAR and three palindromic
tri-PAA (FRF 4, FFR 6, and RFF 14), obtained by molecular docking
experiments (data not shown).11 Following these models, these
three compounds bind, as in the case of Tat protein and Tat derived
analogues, to the major groove of TAR, 4 and 6 in a folded confor-
mation and 14 in an extended one.
pre-treated for 30 min with the PAA, at a concentration of 100 lM.
At the same time, cytotoxicity of the sample was evaluated in unin-
fected PHA-P-activated PBMC, by the classical MTT assay. As
shown in Table 4, all tested PAA decreased at 100 lM the HIV rep-
lication (around 50–60%) and the cell viability (around 30%). No
clear difference could be seen between these six compounds. Fur-
ther experiments are required to elucidate the origine of this anti-
viral activity and toxicity. Concerning the later, it may be related to
the low TAR versus DNA specificity observed in the case of these
tri-PAA.
3. Conclusions
Starting from a small library of 64 tri-PAA generated from only
4 different amino acids, we have identified several compounds able
to bind TAR RNA as tightly as tetra-PAA, that is, with a micromolar
affinity, but with a lower TAR RNA versus DNA specificity. These
compounds also prevent the Tat/TAR association at a micromolar
level. Even if they display only a moderate antiviral activity and
some cytotoxicity, altogether these results are encouraging as this
library of short length PAA explores only a very small part of the
combinatorial diversity potentially achievable for this family of
compounds. Thus, one can expect that introducing other amino
acid residues than the four taken as models in this study could lead
to TAR tri-PAA ligands with both an improved affinity and specific-
ity for the target. Moreover, constraining the conformational free-
dom of these highly flexible trimeric structures by cyclization
could also impact on these two parameters. Overall, PAA represent
interesting structures for the development of more potent HIV
TAR/tat inhibitors.
TAR RNA specificity of the eight tri-PAA displaying the highest
affinities was evaluated in the presence of tRNA and dsDNA as
competitors. Binding affinities K0d and Kd00 (collected in Table 3)
were measured in the presence of a 100-fold nucleotide excess of
either a mixture of natural tRNA or a 15-mer DNA duplex, respec-
tively. For all the tri-PAA tested, only a slight TAR affinity decrease
was observed in presence of tRNA (maximum threefold for FKR 7
and RFF 14), demonstrating a high TAR RNA versus tRNA specific-
ity, as observed for longer PAA.10 On the other hand, and unlike tet-
ra-PAA, which were shown to have a high TAR RNA versus DNA
specificity even in the case of highly charged derivatives, tri-PAA
exhibit a 5–14-fold lower TAR affinity in the presence of DNA com-
petitor. This decrease of specificity is not correlated to the number
of cationic charges on tri-PAA (see, for example, K00/Kd values for
d
tetra-cationic FFK 10 (11.9) and penta-cationic RKF 15 (4.4)) but
seems to be rather sequence-dependent, as demonstrated by com-
paring, for example, the specificity ratio K00/Kd for tetra-cationic
d

7436
V. Bonnard et al. / Bioorg. Med. Chem. 18 (2010) 7432–7438
Figure 3. Models illustrating the binding of three palindromic tri-PAA with HIV-1 TAR RNA. (A) FRF (4); (B) FFR (6); (C) RFF (14).11
Table 3
Dissociation constants for tri-PAA-TAR RNA interactions in the absence (Kd) and in the presence of tRNA (K0 ) and DNA (K00 ) competitorsa
d
d
K0d (TAR) (
l
M) with tRNA mixb
K0d/Kd
K0d0 (TAR) ( M) with dsDNAc
l
K0d0 /Kd
Tri-PAA
Kd (TAR) (lM) without competitor
FRF 4
FFR 6
3
2
4
1
1
1
1.1 0.4
4.3 1.8
0.4
2.1
3
1
2.1
3
17.5 4.3
28.4 4.5
33.6 6.2
12.5 2.8
33.3 4.2
14.9 4.2
31.1 6.5
27.3 5.9
5.8
14.2
8.4
5.7
11.9
4.9
FKR 7
FRR 8
FFK 10
RFF 14
RKF 15
RRF 16
12
8
2.2 0.7
2.8 1.0
3
7
2
2.2 1.1
5.9 1.4
9.1 3.7
4.4 0.8
2.7 0.6
2
2
1
0.6
1.3
4.4
13.6
a
b
c
All standard fluorescence measurements were performed in buffer A (20 mM HEPES (pH 7.4 at 25 °C), 20 mM NaCl, 140 mM KCl, and 3 mM MgCl2).
Measured in the presence of a 100-fold nucleotide excess of a mixture of natural tRNA (tRNAmix).
Measured in the presence of a 100-fold nucleotide excess of a 15-mer duplex DNA.
Table 4
Anti-HIV activity and cellular toxicity of selected tri-PAA
Tri-PAA
% of HIV-1 inhibition at 100
53
60 17
l
M
% cellular toxicity at 100 lM
FRF 4
FFR 6
5
26
27
29
28
26
33
33
29
1
1
2
5
4
6
6
5
FKR 7
FRR 8
FFK 10
RFF 14
RKF 15
RRF 16
46
56
51
6
4
5
53 11
54
58 13
6
array detector. 0.1% TFA/water (Solvent A) and 0.1% TFA/acetoni-
trile (Solvent B) were used as elution solvents. A gradient of A/B
(100/0 for 7 min then 100/0 to 50/50 for 23 min) was employed
at a flow rate of 1 mL/min. Semi-preparative HPLC were performed
on a Waters system (600E system controller, 2487 dual wave-
Figure 4. TAR/tat complex inhibitor effect of tri PAA.
length absorbance detector) with
a Thermo RP18 column
4. Experimental section
(10 ꢀ 250 mm, 5 m, 300 Å) at a flow rate of 3 mL/min, with the
l
same solvent gradient as for analytical separations. Lyophilization
was performed on a Flexy-Dry FTS system. ESI HRMS analysis was
carried out on an LTQ Orbitrap hybrid mass spectrometer with an
electrospray ionization probe (Thermo Scientific, San Jose, CA) by
direct infusion from a pump syringue, to confirm correct molar
mass and high purity of the compounds. 1H and COSY spectra were
recorded with a Brucker AV 500 spectrometer, using deuterated
D2O purchased from euristop. For all tri-PAA, the presence of a
mixture of several isomers cause signals multiplication, giving
4.1. Material and methods
Solvents and reagents were obtained from commercial sources
and used without further purification. N-protected PAA monomers
were synthesized as described previously.10 HPLC analyses were
performed at room temperature on a Thermo RP18 column
(3.2 ꢀ 250 mm, 5
lm, 300 Å) using a Waters apparatus (St Quentin
en Yvelines, France) including HPLC Alliance 2695, 996 photodiode

V. Bonnard et al. / Bioorg. Med. Chem. 18 (2010) 7432–7438
7437
complex 1H spectra. Thus, the reported data are those of the major
conformers.
cleaved (and totally deprotected) using a TFMSA/TFA/TIS (1:8:1,
v:v:v) solution (3 mL) for 4 h. The resulting solution combined with
1 mL of TFA (used to wash the resin) was added to cold anhydrous
diethyl ether (20 mL) and the mixture was centrifugated (10 mn,
3000 tr/mn). The supernatant was removed, the crude solid tri-
PAA was washed with 30 mL of diethyl ether and the resulting
solution was centrifugated. This step was repeated three times.
Tri-PAA were then dissolved in water and purified by semi-pre-
parative HPLC (A/B: 100/0 for 7 min then 100/0 to 50/50 for
23 min, flow rate of 1 mL/min). After lyophilization, their structure
were confirmed by HRMS experiments.
4.2. Synthesis
The tri-PAA library as well as individual tri-PAA were prepared
on MBHA resin LL (100-200 mesh, 0.67 mmol/g, Novabiochem)
starting from the four protected PAA monomers Boc-PAA (AlaZ),
Boc-PAA (PheZ), Boc-PAA (LysZ), and Boc-PAA (ArgZ), whose syn-
thesis has been previously published.10 Syntheses were performed
in a 20 mL glass peptide vessel fitted with a polyethylene filter
disk. Solvents and soluble reagents were removed by filtration un-
der Argon. All syntheses and washes were done at 25 °C.
4.3. Fluorescence binding assays
4.2.1. Tri-PAA library
Unless otherwise stated, all reagents and solvents were of bio-
molecular grade and from Sigma (St Louis, USA). HEPES [4-(2-
hydroxyethyl)-1-piperazineethanesulfonic acid] and all inorganic
salts for buffers were purchased from Calbiochem (molecular biol-
ogy grade). TAR RNA and DNA oligonucleotides were purchased
from IBA GmbH and used without further purification. A mixture
of pre- and mature yeast tRNAs (containing over 30 different spe-
cies from baker’s yeast Saccharomyces cerevisiae) was purchased
from Sigma (type X-SA). Stocks of tRNAmix can be quantified in
its native form (without base hydrolysis) using an extinction coef-
ficient of 9640 cmꢂ1 Mꢂ1 per base.14
The tri-PAA library was elaborated following a split-and-mix
strategy. Sub-libraries SL (F), (K), and (R) were prepared as de-
scribed below for the synthesis of the sub-library SL (A), given as
an example:(a) A preactivated (3 min) mixture of Boc-b-alanine
(28.3 mg, 0.15 mmol), DIEA (105 lL, 0.6 mmol), and HBTU
(57 mg, 0.15 mmol) in 1.5 mL of NMP was added to 450 mg
(0.3 mmol) of MBHA resin. After 2 h of stirring, the resin was
washed with NMP until pH 7. (b) The remaining amino groups
on the resin were capped using a Ac2O/Pyr/NMP (15:15:70,
v:v:v) solution (2 ꢀ 4 mL; 2 ꢀ 10 min), then the resin was washed
with NMP (2 ꢀ 10 mL) and DCM (3 ꢀ 10 mL). (c) The Boc group
All standard fluorescence measurements were performed in buf-
fer A (20 mM HEPES (pH 7.4 at 25 °C), 20 mM NaCl, 140 mM KCl, and
3 mM MgCl2). Buffers were filtered through 0.22-lm Millipore fil-
was cleaved with
a
TFA/TIS (9:1, v:v) solution (2 ꢀ 4 mL;
2 ꢀ 15 mn) then the resin was washed with DCM (3 ꢀ 10 mL)
and NMP (2 ꢀ 4 mL). (d) A preactivated (3 min) mixture of Boc-
ters (GP ExpressPLUS membrane). A small aliquot (50–100 mL)
was first filtered and then discarded to avoid any contaminants that
might be leached from the filter. The solutions to be used in the fluo-
rescence experiments were prepared by diluting the concentrated
stocks in Milli-Q water and filtered again as described above.
For competitive experiments in the presence of a dsDNA, a 15-
mer sequence (50-CGTTTTTATTTTTGC-30) and its complement, an-
nealed beforehand, were added to buffer A to obtain a 100-fold
nucleotide excess regarding TAR RNA (900 nM duplex; 5 nM RNA).
For competitive experiments in the presence of a tRNA, the mix-
ture of pre- and mature yeast tRNAs was added to buffer A to ob-
tain a 100-fold nucleotide excess regarding TAR RNA. Stock
solutions of tRNA were prepared in water and quantified using
an extinction coefficient of 9640 cmꢂ1 Mꢂ1 per base.14
PAA (AlaZ) monomer (0.225 mmol), DIEA (183
lL, 1.05 mmol),
and HBTU (80 mg, 0.21 mmol) in 2 mL of NMP was added to the re-
sin. After 1 h of stirring, the resin was washed with NMP
(2 ꢀ 10 mL) and DCM (3 ꢀ 10 mL). This coupling step was moni-
tored by a Kaiser test and repeated twice. After capping of unre-
acted amino groups (step b), the resin was dried then split into
four portions. On each portion, after cleavage of the Boc group
and resin washing (step c), one preactivated Boc-PAA protected
monomer among the four was condensed, using the same proce-
dure than above (step d). After a capping step (step b), the four por-
tions were mixed then subjected to a common Boc deprotection
(step c) before being dried then split into four batches. To each
batch was condensed one Boc-PAA protected monomer among
the four (step d). After Boc cleavage (step c) and N-acetylation of
the last PAA residue (step b), each batch was submitted to acidol-
ysis using a 1:8:1 (v:v:v) TFMSA/TFA/TIS solution (4 mL, 4 h). For
each batch, the resulting solution combined with 1 mL of TFA (used
to wash the resin) was then added to cold diethyl ether (30 mL)
and the solution was centrifugated (10 mn, 3000 tr/mn). The
supernatant was removed, the residual solid was washed with
30 mL of diethyl ether and the resulting solution was centrifugat-
ed. This step was repeated three times. The solid was then dis-
solved in water and the mixture, containing four tri-PAA, was
basically purified by HPLC semi-preparative (100% A for 7 min then
100% A to 50% A/50% B for 23 min; flow rate: 3 mL/min). After
lyophilization, the content of the four batches constituting the SL
(A) sub-library was analysed by ESI MS experiments.
Ligand solutions were prepared as serial dilutions by an epMotion
automated pipetting system (eppendorf) in buffer A at a concentra-
tion two times higher than the desired final concentration to allow
for the subsequent dilution during the addition of the RNA solution.
The appropriate ligand solution (50 lL) was then added to a well of a
non-treated black 96-well plate (Nunc 237105), in triplicate. Refold-
ing of the RNA was performed using a thermocycler (ThermoStat
Plus Eppendorf) as follows: the RNA, diluted in 1 mL of buffer A,
was first denatured by heating to 90 °C for 2 min then cooled to
4 °C for 10 min followed by incubation at 20 °C for 15 min. After
refolding, the RNA was diluted to a working concentration of
10 nM through addition of the appropriate amount of buffer A. The
tube was mixed and 50 lL of the RNA solution was added to each
well containing ligand. This subsequent dilution lowered the final
RNA concentration to 5 nM. The fluorescence was measured on a
GeniosPro (Tecan) with an excitation filter of 485 10 nm and an
emission filter of 535 15 nm. Each point was measured 10 times
with a 500 ls integration time and averaged. Binding was allowed
to proceed overnight at 5 °C to achieve equilibrium.
4.2.2. Individual tri-PAA
Individual tri-PAA were elongated on b-alanine functionalized
MBHA resin (300 mg), following the standard procedure: (i) cou-
pling conditions: couplings were performed with a NMP solution
(1.5 mL) containing a preactivated (3 min) mixture of Boc-PAA pro-
tected monomer (1.5 equiv), DIEA (7 equiv), and HBTU (1.4 equiv).
Each coupling step was monitored by a Kaiser test and repeated
twice. (ii) Boc cleavage: TFA/TIS (9:1, v:v), 2 ꢀ 15 min, DCM wash,
NMP wash. (iii) Cleavage from the resin: the compounds were
4.4. Tat/TAR complex inhibition study by FRET assay
Determination of FRET assay fluorescence-based binding assays
were performed in 96-well microplates (nunc) at 25 °C and the

7438
V. Bonnard et al. / Bioorg. Med. Chem. 18 (2010) 7432–7438
fluorescence was measured on a GeniosPro (Tecan) with an excita-
tion filter of 485 10 nm and an emission filter of 535 15 nm.
TAR and Tat49–57 peptide (sequence: Rhodamine-YGRKKRRQRRRP,
EzBiolab, USA) were both used at concentrations of 10 nM in buffer
B (50 mM Tris–HCl, 20 mM KCl, 0.02% Tween 20, pH 7.4). Prior to
analyses, the RNA was treated as described above.
The emission of the background (TK buffer only), pure peptide
and of the Tat–TAR complex was established first. Single-point
measurements of potential inhibitors were then carried out in trip-
licates at varying concentrations.
Acknowledgements
We thank SIDACTION, the Agence Nationale de Recherche sur le
SIDA and the Caisse Primaire d’Assurance Maladie des Professions
Libérales for financial support. V. Bonnard and L. Pascale are recip-
ients of a MENRT Ph.D. fellowship. We gratefully acknowledge Dr.
Christophe Guilbert (Department of Pharmaceutical Chemistry,
University of California San Francisco) for molecular docking
experiments.
Supplementary data
4.4.1. Data analysis
Binding data were analyzed using the non-linear least-squares
numerical solver-based binding data global analysis program BIO-
EQS, in which the calculated binding surface is obtained using a
numerical constrained optimization chemical equilibrium solver.15
Unless otherwise stated, binding profiles were well modeled using
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmc.2010.09.002.
References and notes
1. Thomas, J. R.; Hergenrother, P. J. Chem. Rev. 2008, 108, 1171.
2. Chow, C. S.; Bogdan, F. M. Chem. Rev. 1997, 97, 1489.
3. Karn, J. J. Mol. Biol. 1999, 293, 235.
a simple model assuming the one to one stoichiometry.
DG° values
D
Gꢃ=RTÞ
were converted to Kd values as Kd ¼ eꢂð
.
4. Aboul-ela, F.; Karn, J.; Varani, G. Nucleic Acids Res. 1996, 24, 3974.
5. Frankel, A. D. Curr. Opin. Genet. Dev. 1992, 2, 293.
6. Wei, P.; Garber, M. E.; Fang, S. M.; Fischer, W. H.; Jones, K. A. Cell 1998, 92, 451.
7. Garber, M. E.; Wei, P.; Kewal Ramani, V. N.; Mayall, T. P.; Herrmann, C. H.; Rice,
A. P.; Littman, D. R.; Jones, K. A. Genes Dev. 1998, 12, 3512.
4.5. Antiviral activity of tri-PAA compounds
Antiviral activity of the six selectioned tri-PAA was evaluated on
phytohemaglutinin-P (PHA-P)-activated PBMC experimentally in-
fected with the X4-tropic HIV-1-LAI strain.16 These cells pre-trea-
ted for 30 min by six concentrations of each drug (1:10 dilutions
8. Richter, S. N.; Cao, H.; Rana, T. M. Biochemistry 2002, 41, 6391.
9. Dinesh, C. U.; Rana, T. M.. In Small Molecule DNA and RNA Binders; Demeunynck,
M., Bailly, C., Wilson, W. D., Eds.; Wiley-VCH: Weinheim, 2003; Vol. 1, pp 58–
87. Chapter 4; Luedtke, N. W.; Tor, Y.. In Small Molecule DNA and RNA Binders;
Demeunynck, M., Bailly, C., Wilson, W. D., Eds.; Wiley-VCH: Weinheim, 2003;
Vol. 1, pp 18–40. Chapter 2; Richter, S. N.; Palù, G. Curr. Med. Chem. 2006, 13,
1305; Yang, M. Curr. Drug Targets Infect. Disord. 2005, 5, 433; Krebs, A.; Ludwig,
V.; Boden, O.; Gobel, M. W. ChemBioChem 2003, 4, 972; Froeyen, M.; Herdewijn,
P. Curr. Top. Med. Chem. 2002, 2, 1123.
10. Bonnard, V.; Azoulay, S.; Di Giorgio, A.; Patino, N. Chem. Commun. 2009, 2302.
11. Docking experiments were carried out using the virtual docking program
MORDOR (Molecular Recognition with a Driven Dynamics optimizer). For
details on MORDOR, see: Guilbert, C.; James, T. L. J. Chem. Inf. Model. 2008, 48,
1257. Pinto, I. G.; Guilbert, C.; Ulyanov, N. B.; Stearns, J.; James, T. L. J. Med.
Chem. 2008, 51, 7205.
12. Cao, H.; Tamilarasu, N.; Rana, T. M. Bioconjug. Chem. 2006, 17, 352.
13. Murchie, A. I. H.; Davis, B.; Isel, C.; Afshar, M.; Drysdale, M. J.; Bower, J.; Potter,
A. J.; Starkey, I. D.; Swarbrick, T. M.; Mirza, S.; Prescott, C. D.; Vaglio, P.; Aboul-
ela, F.; Karn, J. J. Mol. Biol. 2004, 336, 625.
14. Luedtke, N. W.; Liu, Q.; Tor, Y. Biochemistry 2003, 42, 11391.
15. Royer, C. A. Anal. Biochem. 1993, 210, 91.
16. Barré-Sinoussi, F.; Chermann, J. C.; Rey, F.; Nugeyre, M. T.; Chamaret, S.; Gruest,
J.; Dauguet, C.; Axler-Blin, C.; Vezinet-Brun, F.; Rouzioux, C.; Rozenbaum, W.;
Montagnier, L. Science 1983, 220, 868.
17. Baleux, F.; Loureiro-Morais, L.; Hersant, Y.; Clayette, P.; Arenzana-Seisdedos, F.;
Bonnaffé, D.; Lortat-Jacob, H. Nat. Chem. Biol. 2009, 5, 743.
between 100 lM and 1 nM) and infected with one hundred 50%
tissue culture infectious doses (TCID50) of the HIV-1-LAI strain.
This virus was amplified in vitro on PHA-P-activated PBMC and
the viral stock titrated using PHA-P-activated PBMC and the Kär-
ber’s formula. In our antiviral assay, molecules were maintained
throughout the culture, and cell supernatants were collected at
day 7 post-infection and stored at ꢂ20 °C. Azidothymidine (AZT)
at 10 nM was used in these experiments as internal control; this
AZT concentration corresponds to the ED99 in our experimental
model.17 Viral replication was measured by quantifying reverse
transcriptase (RT) activity in cell culture supernatants using the
RetroSys HIV RT kit (Innovagen). In parallel, cytotoxicity of the
samples was evaluated in uninfected PHA-P-activated PBMC by
methyltetrazolium salt (MTT) assay on day 7. Experiments were
performed in triplicate and percents of inhibition were calculated
to quantify the cell viability and the anti-HIV activity of each
PAA compound.