Assessment of new triplet forming artificial nucleobases as RNA ligands directed towards HCV IRES IIId loop

Article abstract of DOI:10.1016/j.bmcl.2017.02.061

We report the synthesis of two new artificial nucleobase scaffolds, 1 and 2, featuring adequate hydrogen bonding donors and acceptors for the molecular recognition of U:A and C:G base pairs, respectively. The tethering of these structures to various amino acids and the assessment of these artificial nucleobase-amino acid conjugates as RNA ligands against a model of HCV IRES IIId domain are also reported. Compound 1e displayed the highest affinity (K_d twice lower than neomycin – control). Moreover, it appears that this interaction is enthalpically and entropically favored.

Full text of DOI:10.1016/j.bmcl.2017.02.061

Accepted Manuscript
Assessment of new triplet forming artificial nucleobases as RNA ligands direc-
ted towards HCV IRES IIId loop
Mauro Safir Filho, Anthony R. Martin, Rachid Benhida
PII:
S0960-894X(17)30204-4
DOI:
http://dx.doi.org/10.1016/j.bmcl.2017.02.061
BMCL 24733
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
16 December 2016
21 February 2017
23 February 2017
Please cite this article as: Filho, M.S., Martin, A.R., Benhida, R., Assessment of new triplet forming artificial
nucleobases as RNA ligands directed towards HCV IRES IIId loop, Bioorganic & Medicinal Chemistry Letters
(2017), doi: http://dx.doi.org/10.1016/j.bmcl.2017.02.061
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Graphical Abstract
Assessment of new triplet forming artificial
nucleobases as RNA ligands directed
towards HCV IRES IIId loop.
Leave this area blank for abstract info.
Mauro Safir Filho, Anthony R. Martin and Rachid Benhida*
G G
U
G
U
Kd = 4.3 µM
U
N
N
G
C
Ligand 1
N
U
G
A
U
G
C
G
A
Ac
R
N
H
H
H
N
H
O
N
G
A
A
A
G
N
N
O
H
N
U:A
C
C
G
G
G
C
N
N
5'
3'
HCV IRES
IIId loop model
For A:U base pair targeting

Assessment of new triplet forming artificial nucleobases as RNA ligands directed
towards HCV IRES IIId loop.
Mauro Safir Filhoab, Anthony R. Martina, and Rachid Benhidaa,∗
aUniversité Côte d’Azur, CNRS, Institut de Chimie de Nice UMR7272 – 06108 NICE, France; bCAPES foundation, Ministry of Education of Brazil, Brasília –
DF 70040-020, Brazil.
Abstract: We report the synthesis of two new artificial nucleobase scaffolds, 1 and 2, featuring adequate hydrogen bonding donors
and acceptors for the molecular recognition of U:A and C:G base pairs, respectively. The tethering of these structures to various amino
acids and the assessment of these artificial nucleobase-amino acid conjugates as RNA ligands against a model of HCV IRES IIId
domain are also reported. Compound 1e displayed the highest affinity (Kd twice lower than neomycin – control). Moreover, it appears
that this interaction is enthalpically and entropically favored.
Keywords: Hepatitis virus C; IRES IIId loop; RNA ligands; Inhibitors; Binding assay
Main Text:
Hepatitis C virus (HCV) affects about 170-million people worldwide hence representing a major health problem. It is a main cause
of chronic hepatitis and may lead to hepatocellular carcinoma and liver cirrhosis.1 HCV is a single-stranded positive-sense RNA virus
belonging to the flaviviridae family. Its genome is about 9,600 bases length and the several strains can be divided into 7 genotypes and
67 subtypes.2 Until recently, only few treatments against HCV were available, ribavirin and (pegylated)-interferon, while exhibiting
limited efficacy. In 2013, the FDA approval of sofosbuvir, a nucleotide inhibitor of the NS5B viral polymerase, revolutionized HCV
treatment. It yields impressive sustained virological response (SVR) and limited adverse effects.3 Nevertheless, this treatment is
contraindicated for patients suffering severe renal impairments and its efficacy on HCV genotype 3 is not satisfactory.4 Thus, new
complementary approaches to circumvent HCV infection are still needed. One of them focuses on the inhibition of HCV translation by
targeting its Internal Ribosomal Entry Site (IRES).5,6
HCV IRES is a highly conserved sequence located in the 5’-untranslated region (5’-UTR) of HCV genome and composed of three
folding domains (II-IV). It allows the recruitment of the ribosomal machinery together with a restricted number of eukaryotic initiation
factors (eIFs) to promote the translation of the viral RNA in a 5’-cap-independent fashion.7 Thus, many efforts have been devoted to the
development of new drugs able to inhibit the HCV translation by hampering its IRES functions.8 In particular, the discovery of ligands
targeting the IRES IIId loop is highly relevant as this domain is of utmost importance for the interaction with the 40S ribosomal
∗ Corresponding author. Tel.: +33-4-9207-6143; fax: +33-4-9207-6189; e-mail: benhida@unice.fr

subunit.9 Known ligands for this IRES IIId loop encompass antisense oligonucleotides10-12 and RNA aptamers13 as well as small-sized
organic molecules whose structures remain undisclosed.14
While non-coding RNA molecules are even more recognized as key players in numerous biological processes, including gene
regulation, tumorigenesis, viral translation, etc. they also appear as important druggable targets. For this purpose, several research
group, including ours, devoted much effort to develop small-sized organic molecule targeting RNAs.15-18 Our ligand design consisted in
the combination of molecular recognition elements to enhance site specificity with electrostatic interactions to strengthen the complex
stability. The preparation of these multimodal ligands was accomplished by assembling the nucleobase analog “S”, able to form triplets
through Hoogsteen interactions with U:A base pair, with basic amino acid residues (Figure 1). This strategy, based on an unprecedented
rational design of RNA ligands, was successfully applied to the discovery of new specific binders of the stem-bulge of HIV1 TAR
RNA.19,20
S
O
"S" nucleobase
N
Ac
N
N
r
H
H
Hoogsteen
H-bonding
e
c
H
a
p
s
O
N
O
H
N
N
Aminoacid
Canonical U:A
base pairing
N
N
H
N
N
Figure 1. Amino acid tethered “S” nucleobase in a “S”-U:A triplet.
In continuation of our efforts to produce innovative RNA ligands using our triplex-based rational design, we present herein the
synthesis of two new artificial nucleobase scaffolds, 1 and 2, featuring adequate hydrogen bonding donors and acceptors for the
molecular recognition of U:A and C:G base pairs, respectively. The tethering of these structures to various amino acids is also reported
(Figure 2). It is worth noting that while nucleobases 1 and 2 may involve three hydrogen bonds to increase the specificity (recognition
of U:A versus C:G), the amino acid counterparts are critical for high affinity binding (H-bonding and electrostatic ammonium-
phosphate interaction). Finally, we disclose the binding affinity of our novel structures to a model of HCV IRES IIId domain.
G G
G
U
U
O
H
N
N
U
N
N
R
G
C
1
2
N
N
H
N
H
U
G
A
U
G
C
G
A
O
R
N
Ac
H
H
H
H
O
N
N
O
H
N
N H
N
N
G
A
A
A
N
N
N
O
H
H
N
N
O
H
U:A
G
C:G
C
C
G
G
G
C
N
N
N
N
5'
3'
H
For U:A base pair targeting
For C:G base pair targeting
HCV IRES
IIId loop model
Figure 2. Structures of 1, 2 and HCV IRES IIId loop model.

Design of the platforms 1 and 2 was achieved by analogy with the existing structures of artificial nucleobases known to establish
Hoogsteen base pairing in triplex-forming oligonucleotides.21 Thus, the molecular structure of 1 is an analog of the artificial nucleobase
“S”, already reported by our group, where the thiazole ring is replaced by a pyrimidine moiety. It is worth noting that the
aminopyrimidine system should account for more H-bonding ability and π-stacking interactions along the RNA helix, hence modulating
the binding properties. Moreover, this replacement will also affect the possible stacking interactions. Similarly to 1, the structure of 2 is
analogous to a benzimidazolylglycine compound (BIG) that was reported by Sasaki and Maeda22 and further studied by our group in
aminoglycoside conjugates.23 Notably, the BIG moiety have not been linked to amino acids. Thus, we herein complement this study and
will use these data for comparative purpose. This should allow us to identify which base pair is best suited for IRES IIId loop targeting
using our multimodal ligands.
The first step of our study consists of the synthesis of platforms 1 and 2 and their derivatization with various amino acids. Thus, we
first undertook the preparation of 1, following a linear 4-step strategy (Scheme 1, steps i-iv) starting from m-nitroacetophenone (3).
Firstly, 3 was homologated into the corresponding enaminone 4 using the N,N-dimethylamine dimethylacetal reagent. Next, the
construction of the 2-aminopyrimidine ring was performed by cyclization of 4 with guanidine in nBuOH to afford the biaryl compound
5. Then, the free amino function of 5 was acylated using acetyl chloride in anhydrous pyridine. This acylation provided 6 in moderate
yields (27%) probably owing to the low nucleophilicity of the 2-aminopyrimidine moiety. Finally, compound 1 was obtained after
reduction of the nitro function into an amino group that will serve as anchoring point for various amino acids tethering (Scheme 2).
NMe2
N
(i)
(ii)
N
O
O
O2N
NH2
O2N
O2N
N
3
4 (89%)
5 (92%)
N
(iii)
(iv)
N
N
H2N
NHAc
O2N
NHAc
1 (71%)
6 (27%)
Scheme 1. Synthetic pathway for the preparation of 1. Reagents and conditions (i) N,N-dimethylformamide dimethylacetal, xylene, 140°C, 12 h; (ii) guanidine
hydrochloride, nBuOH, 100°C, 24 h; (iii) AcCl, pyridine, 0-50°C, 12 h; (iv) NaBH4, Pd/C, MeOH, t.a., 3 h.
With compound 1 in hands, we proceeded to its connection and unmasking to various amino acids through their carboxylic acid
function. These structures, 1a-1j, were obtained in decent to good yields (48-77%) and are summarized on Scheme 2. The variety of
connected amino acids was selected to evaluate the structure-binding relationship in terms of (1) chain length between the carboxylic
acid and the amine, (2) number and pKa of amino functions, (3) presence of an α-amino group or an aromatic ring.

N
O
N
O
N
(i)
R1
N
OH
NH
NHAc
(ii)
H2N
NHAc
R1
R2
1
R2
1a-1j
N
N
NHAc
O
N
O
N
NH
NHAc
NH
H2N
H2N
3
1a, 71%
1b, 59%
N
N
O
N
O
N
Me
NH
NHAc
NH
NHAc
N
H2N
3
5
1c, 62%
1d, 77%
Me
N
N
O
N
O
N
NH
NHAc
NH
NHAc
H2N
H2N
H2N
4
1e, 53%
1f, 67%
H2N
H
N
N
N
O
N
H2N
H2N
O
O
N
N
NH
NHAc
NH
NHAc
HN
5
NH2
1g, 64%
1h, 48%
N
N
N
N
N
N
N
NH
1i, 70%
NHAc
H2N
NH
NHAc
O
O
1j, 76%
Scheme 2. Synthesis and structures of ligands 1a-1j. Reagents and conditions (i) N-Boc protected amino acid, HBTU, HOBt, DIEA, DMF, 40°C, 24 h; (ii)
TFA, CH2Cl2, r.t., overnight. Yields are isolated overall yields over the 2 steps.
Following the preparation of the first family of ligands, we proceeded to the synthesis of the second series (2a-2c) designed for C:G
base pair targeting (2). As aforementioned, this second set of compounds was prepared for comparative purpose with the first family
(1a-1j). The synthesis of 2a-2c, was performed in a 4-step sequence. Starting from the N-Fmoc-glycine, after the formation of its
corresponding acid chloride (Scheme 3, step i), the amide bond was established with the 2-aminobenzimidazole and the N-Fmoc was
subsequently unmasked, thus affording 7 in a 57% overall yield (3 steps). Next, the free amine was engaged in a peptide coupling with
three different N-Boc protected amino acids to yield 8a-8c. Finally, acidic treatment prompted the N-Boc deprotection and furnished 2a-
2c in 79-86% yields over two steps (Scheme 3).
O
N
O
H
N
H
N
CO2H
(i)-(iii)
(iv)
(v)
N
H
NH NH2
N
H
NHR
n
NHFmoc
N
O
7 (57%, 3 steps)
8a; n = 1, R = Boc; 86%
8b; n = 3, R = Boc; 81%
8c; n = 5, R = Boc; 79%
2a; n = 1, R = H; quant.
2b; n = 3, R = H; quant.
2c; n = 5, R = H; quant.
Scheme 3. Synthetic pathway for the preparation of 1a-1j. Reagents and conditions (i) thionyl chloride, THF/CH2Cl2, 60°C, 0.5 h; (ii) 2-aminobenzimidazole,
DMAP, DIEA, DMF, 100°C, 24 h; (iii) piperidine, DMF, r.t., 2 h; (iv) N-Boc protected amino acid, HBTU, HOBt, DIEA, DMF, 40°C, 24 h; (v) TFA, CH2Cl2,
r.t., overnight.

With these fully characterized new compounds in hand (1a-1j and 2a-2c),24 we proceeded to their evaluation as ligands of the RNA
model of HCV IRES IIId loop presented on figure 2. We used for this study a classical fluorescence-based assay for quantification of
RNA-ligand interactions (IRES fragment labelled at its 5’-end with a fluorescent reporter Alexa488®). Indeed, the dissociation constant
(Kd) of each ligand was determined by plotting the fluorescence intensity as a function of the ligand concentrations. Neomycin, a well-
known non-specific RNA binder20 was used as a positive control. Based on these results concerning compounds 1a-1j (Table 1), some
structure-binding relationship (SBR) could be drawn. We observed that the nature of the linked amino acid dramatically influences the
20°CKd with values ranging from 4.3 to 863.4 µM; neomycin exhibiting a Kd of 8.5 µM. Interestingly, the nucleobase 1 with no appended
amino acids, or the lysine amino acid alone are unable to bind IRES RNA (Table 1, entry 15 & 16), demonstrating therefore that both
domains (nucleoabse and amino acid) cooperate in binding to RNA. The first trend of this SBR concerns the length of the hydrocarbon
chain between carbonyl and terminal amine of the amino acid moiety. Its lengthening from one to five methylene units correlates with a
73-fold decrease in Kd (Table 1, entries 2-4). Additionally, switching from a primary terminal amine to a more basic tertiary
dimethylamino function is also beneficial in terms of binding (Table 1, entry 3 vs entry 5). The number of amino group and H-bond
donor/acceptor is also highly important as their increase favors the binding (entries 6-8). Nevertheless, this last point must be in balance
with an appropriate chain length for an optimal effect. The use of histidine or other aromatic amino acids (entries 9-11) displayed
limited efficacy. Finally, concerning the three compounds belonging to the 2-aminobenzimidazole series (2a-2c), their binding were too
weak to be measured. In the present study, the ligand exhibiting the highest affinity for IRES IIId loop is 1e with a Kd of 4.3 µM, two-
fold more potent than neomycin. While it obviously does not compare favorably with oligonucleotide ligands,10-12 this Kd value falls in
the same range of other small molecule ligands of IRES IIa loop.8
Table 1. Kd values and thermodynamic parameters for ligand/IRES IIId interaction.
Number of 20°CKd
∆G°
∆H°
T∆S°
Entry Ligand
aminesa
(µM)b
(kJ/mol)
(kJ/mol)
(kJ/mol)
1
2
3
4
5
6
7
8
9
Neomycin
6
1
1
1
1
2
2
2
2
8.5
-28.1
n.a.c
-39.8 4.4
n.a.c
-11.7 4.4
n.a.c
1a
1b
1c
1d
1e
1f
863.4
220.5
11.8
36.9
4.3
n.a.c
n.a.c
n.a.c
-27.7
-25.2
-30.0
-27.8
-26.5
-25.9
-3.1 0.4
-8.0 1.1
-25.7 2.6
-40.3 3.1
-33.8 2.4
-48.5 5.1
+24.6 0.4
+17.2 1.1
+4.3 2.7
-12.5 3.2
-7.3 2.4
-22.6 5.2
10.1
16.8
19.7
1g
1h

10
11
12
13
14
15
16
1i
1
1
1
1
1
0
2
659.3
232.3
n.b.d
n.b.d
n.b.d
n.b.d
n.b.d
n.a.c
n.a.c
n.a.c
n.a.c
n.a.c
n.a.c
n.a.c
n.a.c
n.a.c
n.a.c
n.a.c
n.a.c
n.a.c
n.a.c
n.a.c
n.a.c
n.a.c
n.a.c
n.a.c
n.a.c
n.a.c
1j
2a
2b
2c
1
Lysine
aNumber of amino functions, i.e. sites of protonation on the amino acid side chain; bKd measurements were performed at 20°C in a 20 mM HEPES buffer (pH =
7.4) containing 20 mM of NaCl, 140 mM of KCl and 3 mM of MgCl2 – Kd values are reported with an uncertainty of 10%.20 cn.a.: not accessible; bn.b.: no
binding at >30 °C.
Following the determination of the Kd of each ligand for IRES IIId at 20°C, we complemented these datasets by measuring the
equilibrium constants at various temperature ranging from 5 to 36°C; thus we could access the corresponding thermodynamic
parameters (Table 1).24 It appears that the Gibbs free energy variations (∆G°) fall in the same range for all the ligands (∆∆G°max = 4.1
kJ/mol); although the enthalpic and entropic contributions may vary drastically. In fact, compounds exhibiting the longest flexible
alkylamino side chains tend to bind their RNA target in an entropically driven fashion (i.e. 1c-1e). This could be due to some degrees of
residual motility of the formed complex and/or desolvation contribution.25 Counterintuitively, 1g binding seems to be enthalpically
driven. This could be explained by the presence of two amines and one amide function at the end of 1g side chain, which might lock the
conformation of the ligand at its binding site, hence minimizing the entropic contribution. Finally, ligands displaying short side chain
bearing H-bond donors/acceptors (1f & 1h) feature the largest enthalpic contribution.
Having evaluated the affinity of these new ligands for IRES IIId RNA, we selected the best ligand (1e) and its related congener (1c)
to study their specificity towards the unique U:A base pair located at the stem-bulge junction. For this purpose, we used a mutated
sequence of the IRES IIId (Mut-IIId) in which the U:A base pair is replaced by a C:G (Table 2). Importantly, we found that the binding
of 1e to Mut-IIId was drastically decreased with more than 100-fold compared with the wild-type sequence (Kd values 4.3 vs 554.4 µM,
Table 2, entry 1), proving therefore the high selectivity for U:A over C:G. The same behavior was observed using the ligand 1c since
the corresponding Kd was too high to be determined (Table 2, entry 2). These pieces of evidence confirmed again that the ligands
derived from 1 effectively target the U:A base pair of IRES IIId. To further support this result, we assessed the ligand 2c against Mut-
IIId. While 2c does not bind at all to the wild-type sequence, it displayed a Kd value of 41.4 µM against Mut-IIId (Table 2, entry 3).
Since several C:G base pairs are present in Mut-IIId and the wild-type IRES IIId sequences, this last observation suggests that our
ligands feature some degree of site specificity and are tailored to target the stem-bulge junction where the U:A base pair is located.
Finally, we evaluated the ability of our ligands 1e and 1c to bind the IRES IIId loop sequence in the presence of natural tRNA
competitors (Table 2). To do so, we measured the Kd value of 1e and 1c againt IRES IIId in the presence of a 50-fold excess of a

mixture of natural tRNA. This competition experiment revealed that 1e and 1c are able to selectively target the IIId loop of IRES since
only a marginal increase of the corresponding Kd values was observed (2.4- and 1.4-fold, respectively).
Table 2. Kd determination to probe the site-selectivity and the specificity of the ligands.
Entry Ligand
20°CKd (µM)a
IRES IIId Mut-IIIdb IRES IIId + tRNAc
1
2
3
1e
1c
2c
4.3
11.8
n.b. d
554.4
n.b.d
41.4
10.5
17.1
—
aKd measurements were performed at 20°C in a 20 mM HEPES buffer (pH = 7.4) containing 20 mM of NaCl, 140 mM of KCl and 3 mM of MgCl2 – Kd values
are reported with an uncertainty of 10%.20 bSequence of Mut-IIId: 5’-GCCGAGGAGUGUUGGGUCGCGCAAGGC-3’. cMeasured in the presence of a 50-
fold excess of a mixture of natural tRNA dn.b.: no binding.
In conclusion, we reported herein two new series of rationally design multimodal RNA ligands featuring new artificial nucleobases.
These compounds proved easily accessible in only 4 to 6 steps. We presented their ability to bind to a IIId loop model of HCV IRES
and compared their potency with neomycin. It is noteworthy that this study discloses the first structures of small-molecule ligands of
IRES IIId loop. Low micromolar dissociation constant could be obtained with a two-fold lower value for our best ligand (1e) compared
with neomycin. Moreover, it seems that its binding is both enthalpically and entropically favored (T∆S° = +4.3 kJ/mol). This study
validates our rational design strategy and could now be used to design new series of multimodal RNA ligands and paves the way
towards more potent compounds.
Acknowledgments
This work was supported by CNRS, University of Nice Sophia Antipolis and CAPES (fellowship to MSF, process number
99999.001495/2015-01).
References and notes
1. Webster, D. P.; Klenerman, P.; Dusheiko, G. M. Lancet 2015, 385, 1124–1135.
2. Smith, D. B.; Bukh, J.; Kuiken, C.; Muerhoff, A. S.; Rice, C. M.; Stapleton, J. T.; Simmonds, P. Hepatology 2014, 59, 318–327.
3. Gentile, I.; Maraolo, A. E.; Buonomo, A. R.; Zappulo, E.; Borgia, G. Expert Opin. Drug Discovery 2015, 10, 1363–1377.
4. Nakamura, M.; Kanda, T.; Haga, Y.; Sasaki, R.; Wu, S.; Nakamoto, S.; Yasui, S.; Arai, M.; Imazeki, F.; Yokosuka, O. World J. Hepatol. 2016, 8, 183–
190.
5. Gallego, J.; Varani, G. Biochem. Soc. Trans. 2002, 30, 140–145.
6. Davis, D. R.; Seth, P. P. Antiviral Chem. Chemother. 2011, 21, 117–128.
7. Lukavsky, P. J. Virus Res. 2009, 139, 166–171.
8. Dibrov, S. M.; Parsons, J.; Carnevali, M.; Zhou, S.; Rynearson, K. D.; Ding, K.; Garcia Sega, E.; Brunn, N. D.; Boerneke, M. A.; Castaldi, M. P.;
Hermann, T. J. Med. Chem. 2014, 57, 1694–1707.
9. Angulo, J.; Ulryck, N.; Deforges, J.; Chamond, N.; Lopez-Lastra, M.; Masquida, B.; Sargueil, B. Nucleic Acids Res. 2015, gkv1325–17.
10. Jubin, R.; Vantuno, N. E.; Kieft, J. S.; Murray, M. G.; Doudna, J. A.; Lau, J. Y.; Baroudy, B. M. J. Virol. 2000, 74, 10430–10437.

11. Hanecak, R.; Brown-Driver, V.; Fox, M. C.; Azad, R. F.; Furusako, S.; Nozaki, C.; Ford, C.; Sasmor, H.; Anderson, K. P. J. Virol. 1996, 70, 5203–5212.
12. Tallet-Lopez, B.; Aldaz-Carroll, L.; Chabas, S.; Dausse, E.; Staedel, C.; Toulmé, J.-J. Nucleic Acids Res. 2003, 31, 734–742.
13. Kikuchi, K.; Umehara, T.; Fukuda, K.; Kuno, A.; Hasegawa, T.; Nishikawa, S. Nucleic Acids Res. 2005, 33, 683–692.
14. Baugh, C.; Shaohui Wang; Bin Li; Appleman, J. R.; Thompson, P. A. J. Biomol. Screening 2009, 14, 219–229.
15. Thomas, J. R.; Hergenrother, P. J. Chem. Rev. 2008, 108, 1171–1224.
16. Guan, L.; Disney, M. D. ACS Chem. Biol. 2012, 7, 73–86.
17. Hermann, T. WIREs RNA 2016, 7, 726–743.
18. Di Giorgio, A.; Tran, T. P. A.; Duca, M. Future Med. Chem. 2016, 8, 803–816.
19. Duca, M.; Malnuit, V.; Barbault, F.; Benhida, R. Chem. Commun. 2010, 46, 6162.
20. Joly, J.-P.; Mata, G.; Eldin, P.; Briant, L.; Fontaine-Vive, F.; Duca, M.; Benhida, R. Chem. Eur. J. 2014, 20, 2071–2079.
21. Malnuit, V.; Duca, M.; Benhida, R. Org. Biomol. Chem. 2011, 9, 326–336.
22. Sasaki, S.; Nakashima, S.; Nagatsugi, F.; Tanaka, Y.; Hisatome, M.; Maeda, M. Tetrahedron Lett. 1995, 36, 9521–9524.
23. Vo, D. D.; Staedel, C.; Zehnacker, L.; Benhida, R.; Darfeuille, F.; Duca, M. ACS Chem. Biol. 2014, 9, 711–721.
24. See Electronic Supplementary Information (ESI) for more details.
25. Klebe, G. Nat Rev Drug Discov 2015, 14, 95–110.
Supplementary Material
Supplementary material includes the synthetic procedures for the preparation of all reported compounds, their characterization, copies of 1H,
13C and HRMS spectra; experimental procedures for the determination of Kd and thermodynamic parameters.