Introduction of polar groups on the naphthalene scaffold of molecular tongs inhibiting wild-type and mutated HIV-1 protease dimerization

Article abstract of DOI:10.1039/c4md00032c

A new series of naphthalene-based molecular tongs containing polar groups at the 3-position of the naphthalene scaffold was synthesized and its anti-dimerization activity was evaluated against HIV-1 protease. The polar groups were introduced mainly via me

Full text of DOI:10.1039/c4md00032c

MedChemComm
View Article Online
View Journal | View Issue
CONCISE ARTICLE
Introduction of polar groups on the naphthalene
scaffold of molecular tongs inhibiting wild-type
and mutated HIV-1 protease dimerization†
Cite this: Med. Chem. Commun., 2014,
5, 719
R. Fanelli,‡ab A. S. Ressurreiçao,‡a L. Dufau,c J.-L. Soulier,a A. Vidu,a N. Tonali,a
˜
a
cd
a
*
*
G. Bernadat, M. Reboud-Ravaux
and S. Ongeri
A new series of naphthalene-based molecular tongs containing polar groups at the 3-position of the
naphthalene scaffold was synthesized and its anti-dimerization activity was evaluated against HIV-1
protease. The polar groups were introduced mainly via metal-catalysed cross coupling reactions such as
the Buchwald–Hartwig amination, the Sonogashira reaction and the Beller cyanation. Kinetic analyses
showed that by introducing specific polar substituents on the naphthalene scaffold the inhibitory activity
of molecular tongs against wild-type and mutated HIV-1 proteases is maintained, while increasing their
water solubility.
Received 24th January 2014
Accepted 12th March 2014
DOI: 10.1039/c4md00032c
www.rsc.org/medchemcomm
OH) from both monomers. This four-stranded region contains
the main hotspot interface residues, providing more than 50%
of the hydrogen bonds along the dimerization interface5 and
Introduction
The human immunodeciency virus type 1 (HIV-1) protease
(PR) is an important target in the ght against AIDS, since its
inhibition results in the production of an uninfectious virus.1,2
The PR inhibitors (PIs) used in highly active antiretroviral
therapy (HAART) act as transition-state analogues by targeting
the PR active site, however the emergence of several mutations
within or outside the active site3,4 can lead to PI resistance. The
mature PR is only active as a homodimer (99 residues per
monomer), in which the energy of monomer–monomer inter-
actions and consequently the energy of dimerization is not
uniformly distributed along the monomer–monomer interface.
The PR homodimer is mainly stabilised by a four-stranded
antiparallel b-sheet that involves the N-(H-Pro(1)-Gln(2)-Ile(3)-
Thr(4)) and C-termini (Cys(95)-Thr(96)-Leu(97)-Asn(98)-Phe(99)-
contributing to close to 75% of the total Gibbs free energy of PR
dimerization.6 Interestingly, this PR termini b-sheet interface
appears to be free of mutations.7 Therefore, targeting this hot-
spot could prevent the essential dimerization process or disrupt
the dimer, and as a consequence constitute an alternative to
drugs that target the active site.8,9 Several dimerization inhibi-
tors targeting the b-sheet region have been described in the
literature, namely, C- and N-terminal mimetic peptides,10–12
lipopeptides,13–15 bicyclic guanidinium and alkyl hydroxy-
benzoic acid derivatives16 and cross-linked peptides with ex-
ible17 or semi-rigid spacers.18 Our strategy for inhibiting PR
dimerization consisted of designing organised molecular tongs
displaying a rigid spacer to provide an entropic benet.19–23
Initially, we described the design of molecular tongs based on a
naphthalene or a quinoline scaffold in which two identical or
a
´
´
´
Molecules Fluorees et Chimie Medicinale, BioCIS UMR-CNRS 8076, LabEx LERMIT,
non-identical peptide strands were attached through
a
´
´
´
Universite Paris-Sud, Faculte de Pharmacie, 5 rue Jean-Baptiste Clement, 92296
carboxypropyloxy linker to enable the formation of an antipar-
allel b-sheet with the N- and C-termini of one PR monomer.19,20
Aerwards, one or two peptide fragments were replaced by
peptidomimetic ones while keeping the same array of hydrogen
bonds.21,22 Inserting a 5-acetamido-2-methoxybenzohydrazide
group into one or two strands (for example, molecular tongs 1
and 5, Fig. 1 and 2 respectively) provided efficient inhibitors of
wild-type (PR) in vitro (Kid of 0.22 and 0.28 mM for 1 and 5
respectively), with an increased metabolic stability.21,22
However, one drawback of our previously reported molecular
tongs was their lack of water solubility. More recently, our
efforts have focused on the development of more water-soluble
non-peptidic molecular tongs. For that purpose, two strategies
have been established. In a rst attempt hydrophilicity was
ˆ
Chatenay-Malabry Cedex, France. E-mail: Sandrine.Ongeri@u-psud.fr; Tel: +33
146835737
bDISMAB, Sezione di Chimica Organica “A. Marchesini”, Facolta di Farmacia,
`
`
Universita degli Studi Milano, Via Venezian 21, 20133 Milano, Italy
c
´
Sorbonne Universites, UPMC Univ Paris 06, UMR 8256, B2A, Biological Adaptation
and Ageing, Integrated Cellular Ageing and Inammation, Molecular & Functional
Enzymology, 7 Quai St Bernard, F-75005 Paris, France. E-mail: michele.reboud@
upmc.fr; Tel: +33 144275078
dCNRS, UMR 8256, B2A, Biological Adaptation and Ageing, F-75005 Paris, France
† Electronic supplementary information (ESI) available: Experimental part with
the synthesis procedures and characterisation of all the new compounds;
characteristic 1H, 13C and 2D NMR spectra of a few molecular tongs; HPLC
chromatograms of molecular tongs 3 and 8 and evolution with time of the
concentration of molecular tongs 3 and 8 incubated at 37 ꢀC in RPMI culture
medium containing 20% fetal calf serum. See DOI: 10.1039/c4md00032c
‡ These authors contributed equally to this work.
This journal is © The Royal Society of Chemistry 2014
Med. Chem. Commun., 2014, 5, 719–727 | 719

View Article Online
MedChemComm
Concise Article
and an amine were respectively introduced (Fig. 1). It has been
well established that when the number of polar groups
(hydroxyl, carboxylic acid or amine groups) increases in a
molecule, water solubility is enhanced since new dipole–dipole
interactions (such as those involved in hydrogen bonding) and
thus molecule–solvent interactions are rendered possible.24 For
instance, the introduction of a N-1,2-dihydroxypropyl moiety
has been previously used to improve water solubility and thus
enhance drug-likeness properties.25 The protected precursor 2
was also evaluated in order to study the inuence of the pres-
ence or absence of a diol moiety on the PR dimerization
inhibitory activity. Starting from molecular tong 5, which is
deprived of any peptidic character,22 additional diversity was
introduced on the polar groups (Fig. 1). An aminopropanediol
and a primary amine were attached to the naphthalene scaffold
(molecules 7 and 8 respectively). The dioxolane precursor of 7
was also evaluated (molecule 6). In molecules 9 and 11, a
secondary or a tertiary amine (methylamino or piperazine) was
introduced on the scaffold via an alkylamide function. The
protected precursor 10 was also evaluated for its PR dimeriza-
tion inhibitory potency. Primary or tertiary amines were also
inserted using a propyl or methyl chain as a linker in molecules
12, 13 and 14. Finally, in molecule 15 a guanidine group, which
has been previously shown to increase water solubility,26 was
introduced. The amine and guanidine groups become charged
at biologically relevant pH values and thus should increase the
water solubility of molecules 9–15. Most of the polar groups
present on the naphthalene scaffold were introduced by metal-
catalysed cross coupling reactions.
Results and discussion
Chemistry
Fig. 1 Structure of molecular tongs 1–15.
The synthetic procedure designed to obtain the modied
naphthalene scaffolds relies on the initial introduction of a
bromine atom at the 3-position of the naphthalene-2,7-diol 16,
to allow further functionalization through metal-catalysed
reactions, such as Buchwald–Hartwig aminations and Sonoga-
shira couplings.
increased within the strands of the molecular tongs by synthe-
sizing new molecular tongs containing carbonylhydrazide
(CONHNHCO) and oligohydrazide (azatide) peptidomimetic
fragments displaying increased water solubility.23 The inhibi-
tory potency depended on the peptidomimetic strand structure
suggesting that an appropriate balance between hydrophilicity
and hydrophobicity had to be reached in order to obtain
powerful inhibitors.23 In this report, a second strategy is
described based on the introduction of hydrophilic groups on
the 3-position of the naphthalene scaffold, while keeping the
structure of the hydrophobic strands, in order to change the
physicochemical properties in the non-pharmacophore region
of the molecular tongs. Keeping the two strands of molecular
tongs 1 and 5 (those with the 5-acetamido-2-methoxybenzohy-
drazide group attached to the carboxypropyloxy linker through a
valine or the 3-amino group of a lysine residue), the main
challenge was to obtain similar or increased anti-dimerization
PR efficiency, while introducing the naphthalene moiety groups
whose steric and/or electronic effects do not destabilize the
interaction with PR monomer termini. In molecular tongs 3 and
4, (both derivatives of molecular tong 1) an aminopropanediol
Bromination of naphthalene-2,7-diol with bromine in acetic
acid gave rise to a mixture of two dibromo derivatives, the 1,3-
and 1,6-isomers,27 which by an in situ monodebromination
using Sn powder27,28 afforded the desired 3-bromonaphthalene-
2,7-diol 17, in excellent yield (94%, Scheme 1). Compound 17
was then alkylated with benzyl-4-bromobutanoate or ethyl-4-
bromobutanoate to afford the dibenzyl ester 18 (89% yield) and
the diethyl ester 19 (82% yield) respectively (Scheme 1).
Then amination reactions on the bromonaphthalene deriv-
atives 18 were performed (Scheme 2). Firstly, we investigated the
best conditions for coupling bromonaphthalene derivative 18
with a primary amine bearing a 2,2-dimethyl-1,3-dioxolane
methyl group, which would be the precursor of the polar ami-
nopropanediol group found in the target molecular tongs 3 and
7 (Fig. 1 and 2). Nucleophilic aromatic substitution of 18 was
initially investigated using a copper catalyst (CuI, 20 mol%) in
the presence of a base (Ullmann conditions),29 without (Table 1,
entry 1) or in the presence of a ligand (Table 1, entries 2–6). No
720 | Med. Chem. Commun., 2014, 5, 719–727
This journal is © The Royal Society of Chemistry 2014

View Article Online
Concise Article
MedChemComm
substitution was observed in the absence of a ligand (entry 1). benzophenone imine was engaged as a surrogate for ammonia
The introduction of L-proline30,31 as a ligand (entries 2 and 3) in the Pd-catalyzed reaction. This reagent is commercially
was not successful, resulting in complex crude mixtures. The available and allows an easy deprotection of the amino function
use of the bidentate ligand DMEDA (entries 4–6)32,33 afforded under mild conditions.40 A Buchwald–Hartwig reaction was
the desired substituted product 20, albeit in low yields. In this performed on compounds 18 and 19, using benzophenone
case, the best conditions screened were: DMEDA (10 equiv.) in imine to afford intermediates that were not isolated but
the presence of potassium carbonate (2 equiv.) in dioxane at 110 immediately hydrolysed, using methanol in the presence of
ꢀC for 24 h, which allowed the synthesis of 20, but still in sodium acetate and hydroxylamine hydrochloride, to afford
unsatisfactory yield (25%) (Table 1, entry 5). As an alternative to amines 22 (62% over two steps) and 23 (51% over two steps) in
the copper-catalyzed reaction, the Buchwald–Hartwig cross- satisfactory yields (Scheme 3).
coupling reaction was attempted,34,35 in which C(sp2)–N bond
Derivatives 22 and 23 were then converted into halides 24
formation is palladium-catalysed, in the presence of the (80% yield) and 25 (75% yield) by reaction with chloroacetyl
bidentate ligand, 9,9-dimethyl-4,5-bis(diphenylphosphino)- chloride in dry dichloromethane (Scheme 3). Treatment of
xanthene (xantphos)36,37 and a base (Table 2). The best results compound 24 with methylamine in the presence of triethyl-
(Table 2, entries 2 and 3) were obtained using palladium(II) amine and a catalytic amount of potassium iodide in DMF
acetate as a catalyst and potassium carbonate as a base, in dry afforded amine 26 (60% yield). Following the same procedure,
dioxane. The reaction was complete and 20 was obtained in intermediate 25 was coupled to the mono protected Cbz-
satisfactory yield (70%), only when a large excess of base was piperazine to afford compound 28 in good yield (86%). In this
used (20 equiv., Table 2, entry 3). The same observation was case, we used the naphthalene scaffold bearing the carboxy-
already reported in the literature for the amination of chloro- propyl linker protected as an ethyl ester in order to obtain an
pyridines and chloro-pyridazinones.38,39 On the other hand, the orthogonal protection with respect to the Cbz protected piper-
use of stronger bases (e.g., sodium-tert-butoxide, Table 2, entry azine ring (Scheme 3). Finally, cleavage of the benzyl ester of 26
5) mainly led to degradation products and only traces of the by catalytic hydrogenation, using Pd/C as a catalyst, afforded 27
desired substituted product 20 were recovered. The catalyst (84% yield), and basic hydrolysis of the ethyl ester of 28, by
tris(dibenzylideneacetone)dipalladium was less efficient than aqueous NaOH 2 N in methanol, gave 29 (86% yield).
palladium(II) acetate (Table 2, entry 6).
The second goal of this chemistry section was to introduce
Finally, cleavage of the benzyl ester by catalytic hydrogena- alkylamino chains into the naphthalene scaffold. The Sonoga-
tion using Pd/C as catalyst afforded the desired compound 21 shira coupling, consisting of a palladium catalysed sp2–sp
(Scheme 2). The Buchwald–Hartwig amination reaction was coupling reaction between an aryl halide and a terminal alkyne,
then applied under the conditions previously described (palla- in the presence of an amine playing the role of a base, has
dium acetate 10 mol%, xantphos 20 mol%, K2CO3 20 equiv., in become the most important method to prepare aryl alkynes,41
ꢀ
dry dioxane at 110 C) for the introduction of the amino (NH2) used here as precursors of arylalkanes. Sonogashira also
group on the naphthalene scaffold. For that purpose, reported that addition of a catalytic amount of copper(I) iodide
greatly accelerates the reaction.42 The Sonogashira reaction is
usually performed using
a palladium-phosphane ligand
complex (e.g. Pd(PPh3)4 or PdCl2(PPh3)2) as a catalyst in the
presence of a catalytic amount of copper(I) salt and an amine
(as a solvent or in large excess) under homogeneous conditions.
Starting from the bromo derivative 18, the coupling with the
N,N-dimethylpropargylamine in the presence of Pd(PPh3)4 as a
catalyst (5%) and copper iodide (10%) as a co-catalyst, triethy-
lamine and using DMF as a solvent, at 80 ꢀC afforded
compound 30 in satisfactory yield (60%, Scheme 4). The use of
freshly prepared PdCl2(PPh3)2 (from PdCl2 and PPh3) or longer
reaction times did not lead to higher yields. Compound 30 was
then treated with H2 and Pd/C in a (1 : 1) mixture of MeOH and
EtOAc in order to reduce the triple bond, and concomitantly
cleave the benzyl ester by hydrogenolysis, affording compound
31 in excellent yield (98%). The Sonogashira coupling of
compound 19 with N,N-di(benzyloxycarbonyl)propargylamine,
using the same conditions as for 18, afforded compound 32 in
moderate yield (41%, Scheme 4). Compound 32 was then
treated with aqueous NaOH 2 N in methanol to afford the cor-
responding acid 33 aer treatment with aqueous HCl 1 N (61%,
Scheme 4). It is worth noting the loss of one Cbz moiety during
the basic hydrolysis of the methyl ester.
Scheme 1 Reagents and conditions: (i) Br2, CH3COOH, 1.5 h, then Sn
powder, H2O, 80 ꢀC, 24 h. (ii) Br(CH2)3COOCH2Ph or Br(CH2)3COOEt,
K2CO3, dry DMF, 50 ꢀC, overnight.
Scheme 2 Reagents and conditions: (i) (2,2-dimethyl-[1,3]-dioxolan-
4-yl)-methylamine, amination reaction. (ii) H2, 10% Pd/C, CH3OH–
EtOAc (1 : 1, v/v), r.t., 24 h.
This journal is © The Royal Society of Chemistry 2014
Med. Chem. Commun., 2014, 5, 719–727 | 721

View Article Online
MedChemComm
Concise Article
Table 1 Ullmann-type conditions used in the amination reaction of 18 with CuI as the copper source
Entry
Ligand (mol%)
Base (equiv.)
Solvent
Temp
Duration
Result
1
2
3
4
5
6
—
K2CO3 (3)
K2CO3 (3)
K2CO3 (2)
K2CO3 (2)
K2CO3 (2)
K2CO3 (2)
DMSO
DMSO
DMSO
Dioxane
Dioxane
Dioxane
135 ꢀC
135 ꢀC
110 ꢀC
75 ꢀC
48 h
48 h
24 h
24 h
24 h
24 h
SM
DP
DP
SM
Proline (40)
Proline (20)
DMEDA (10)
DMEDA (10)
DMEDA (20)
110 ꢀC
125 ꢀC
25%a (+SM)
8%a (SM + DP)
a
Yield of a pure isolated compound. Abbreviations used in the table: SM ¼ starting material, DP ¼ degradation products.
Table 2 Buchwald–Hartwig conditions used in the amination reaction
In order to enlarge the scope of the substitution pattern of
the main bromonaphthalene scaffold, a cyano group was
introduced, and subsequently reduced to an aminomethyl
group. Aromatic nitriles have been prepared in numerous ways.
These methods include the Rosemund-von Braun reaction of
aryl halides43 and the diazotization of anilines with subsequent
Sandmeyer reaction.44 More recently, the transition metal-cat-
alysed cyanation of aryl halides has been widely employed.45–47
In a rst attempt, a Stille coupling, which is a versatile C–C
bond forming reaction between a stannane and an aryl halide,
was performed. Therefore, compound 19 was treated with
Bu3SnCN (2 equiv.) in the presence of tetrakis palladium
of 18 (in dioxane, at 110 ꢀC, for 15 h)
Pd source (10
mol%)
Entry
Ligand (mol%)
Base (equiv.)
Resulta
1
2
3
4
5
6
Pd(OAc)2
Pd(OAc)2
Pd(OAc)2
Pd(OAc)2
Pd(OAc)2
Pd2(dba)3
Xantphos (20)
Xantphos (20)
Xantphos (20)
Xantphos (20)
Xantphos (20)
Xantphos (20)
K2CO3 (5)
27%
66%
70%
51%
4%b
50%
K2CO3 (10)
K2CO3 (20)
Cs2CO3 (20)
NaO-t-Bu (2)
K2CO3 (20)
a
b
Yield of a pure isolated compound. The reaction mainly led to
degradation products.
ꢀ
(20 mol%) as a catalyst in 1,2-dichloroethane at 70 C, but no
reaction occurred. The general problem with metal-catalysed
cyanation is the high affinity of cyanide species for the metal
center, which results in the fast deactivation of the catalytic
system by formation of stable cyanide complexes. Beller et al.
described copper-catalysed cyanation with potassium ferrocya-
nate (K4[Fe(CN)6]), which has the advantage of being essentially
the least toxic cyanide source conceivable.48 This method fore-
sees the use of imidazole ligands to control the stability and
selectivity of the copper catalyst. Using this methodology,
compound 19 was treated with K4[Fe(CN)6] in the presence of
CuI (10 mol%) as a catalyst and 1-butyl-1H-imidazole as a ligand
in dry toluene at reux affording compound 34 in satisfactory
Scheme
3 Reagents and conditions: (i) benzophenone imine,
Pd(OAc)2 (10 mol%), xantphos (20 mol%), K2CO3 (20 equiv.), dry Scheme 4 Reagents and conditions: (i) 18, N,N-dimethylpropargyl-
dioxane, 110 ꢀC, 15 h, then NH2OH$HCl, NaOAc$3H2O, CH3OH, r.t., 6 amine, CuI (10 mol%), Pd(PPh3)4 (5 mol%), Et3N, DMF, 80 ꢀC, 6.5 h. (ii)
h. (ii) ClCH2COCl, Et3N, CH2Cl2, 0 ꢀC, 10 min, then r.t., 2 h. (iii) 24, 19, N,N-di(benzyloxycarbonyl)propargylamine, CuI (10 mol%),
CH3NH2$HCl, Et3N, DMF, r.t., 48 h. (iv) H2, 10% Pd/C, CH3OH, r.t., 18 h. Pd(PPh3)4 (5 mol%), Et3N, DMF, 80 ꢀC, 15 h. (iii) H2, 10% Pd/C CH3OH–
(v) 25, N-Cbz-piperazine, Et3N, DMF, r.t., 24 h. (vi) NaOHaq 2 N, EtOAc 1 : 1 (v/v), r.t., 14 h. (iv) NaOHaq 2 N, CH3OH–THF 1 : 1 (v/v), r.t.,
CH3OH, r.t., 35 h, then HClaq 1 N.
5.5 h, then HClaq 1 N.
722 | Med. Chem. Commun., 2014, 5, 719–727
This journal is © The Royal Society of Chemistry 2014

View Article Online
Concise Article
MedChemComm
yield (70%, Scheme 5). The cyano group was then reduced into
the aminomethyl derivative by treatment of compound 34 with
Ni-RANEY® in a hydrogen atmosphere (60 psi) in ethanol, and
in the presence of ammonia, affording derivative 35 in 98%
yield (Scheme 5). Compound 35 was further treated with CbzCl
in the presence of triethylamine in dry dichloromethane
affording the corresponding N-protected amine 36 (83% yield,
Scheme 5). Compound 36 was then converted into the free
carboxylic acid derivative 37 by treatment with aqueous NaOH 2
N in methanol (96% yield, Scheme 5). Amine 35 was reacted
with 1,3-bis(tert-butoxycarbonyl)-2-methyl-2-thiopseudourea in
the presence of HgCl2 and triethylamine in dry dichloro-
methane to give compound 38 in good yield (62%, Scheme 5).
Finally, compound 38 was treated with aqueous NaOH 2 N in
MeOH–THF and, subsequently, with HCl 4 N in dioxane, to
hydrolyse the esters and cleave the Boc moieties respectively,
affording compound 39 in quantitative yield (Scheme 5).
Molecular tongs 2 and 6 were synthesised from scaffold 21
and peptidomimetics 40 (ref. 21) and 41 (ref. 22) respectively, in
satisfactory yields (50% and 59%) using a standard coupling
protocol (using HBTU and HOBt, in the presence of DIPEA, in
dry DMF) (Scheme 6). The 2,2-dimethyl-1,3-dioxolane group of
molecule 2 was then cleaved by HCl in EtOH–THF, affording
molecular tong 3, in satisfactory yield (60%). Since molecule 6
contained acid labile protecting groups (Na-Boc) in its strands,
the diol deprotection step was performed with Amberlyst 15
resin in methanol,49 affording molecular tong 7 in low yield
(30%). An alternative to the selective dimethyl-1,3-dioxolane
cleavage was tested in the presence of p-toluenesulfonic acid in
methanol,50 affording molecular tong 7 with a similar yield.
The synthesis of molecular tongs 4 and 8–11 is described in
Scheme 7. Scaffold 22 was deprotected by catalytic hydro-
genolysis with hydrogen and Pd/C in MeOH–EtOAc affording
the corresponding carboxylic acid 42 in quantitative yield.
Compound 42 was then coupled to peptidomimetic arms 40
and 41 (Scheme 6) using the same standard coupling protocol
described above (HBTU–HOBt, DIPEA, in dry DMF), to afford
molecular tongs 4 and 8 in satisfactory yields (68% and 72%,
respectively). In the same manner, scaffolds 27 and 29 were
coupled to peptidomimetic arm 41 to afford molecular tongs 9
Scheme 6 Synthesis of molecular tongs 2–3 and 6–7. Reagents and
conditions: (i) peptidomimetic 40 or 41, HBTU, HOBt, DIPEA, DMF, r.t.,
48 h. (ii) 2, HClaq 1 N, EtOH–THF, 55 ꢀC, 1 h, or 6, Amberlyst™ 15,
MeOH, r.t., 4 h.
and 10 in good yields (93% and 61% respectively). The Cbz
protecting group of the piperazinyl ring of molecule 10 was
removed by hydrogenolysis using Pd/C in methanol, to give
molecular tong 11 in quantitative yield.
The synthesis of molecular tongs 12 and 13 is described in
Scheme 8. Scaffolds 31 and 33 were coupled to peptidomimetic
arm 41 (using HBTU–HOBt, DIPEA, in dry DMF) to afford
molecular tongs 13 and 43 in very good yields (86% for both
compounds). Treatment of molecule 43 in a hydrogen atmo-
sphere using 10% Pd/C in methanol afforded molecular tong 12
in excellent yield (92%).
The synthesis of molecular tongs 14–15 is described in
Scheme 9. Scaffolds 37 and 39 were coupled to peptidomimetic
Scheme 7 Synthesis of molecular tongs 4 and 8–11. Reagents and
conditions: (i) H2, 10% Pd/C, MeOH–EtOAc, r.t., 24 h. (ii) Peptidomi-
metics 40 or 41, HBTU, HOBt, DIPEA, DMF, r.t., 48 h. (iii) Peptidomi-
metic 41, HBTU, HOBt, DIPEA, DMF, r.t., 48 h. (iv) H2, 10% Pd/C, MeOH,
r.t., 19 h.
Scheme 5 Reagents and conditions: (i) K4[Fe(CN)6], CuI, 1-butyl-1H-
imidazole, toluene, reflux, 46 h. (ii) H2, RANEY® Ni, NH3 1 N in EtOH,
DMF, 4 bar, r.t., 2 days. (iii) CbzCl, Et3N, CH2Cl2, 0 ꢀC, 30 min, then r.t.
overnight
or
1,3-bis(tert-butoxycarbonyl)-2-methyl-2-thio-
pseudourea, HgCl2, NEt3, CH2Cl2, r.t., 17 h. (iv) NaOHaq 2 N, CH3OH, r.t. Scheme 8 Synthesis of molecular tongs 12 and 13. Reagents and
2 h, then HClaq 1 N. (v) NaOHaq 2 N, CH3OH–THF 1 : 1 (v/v), r.t., 4 h, conditions: (i) peptidomimetic 41, HBTU, HOBt, DIPEA, DMF, r.t., 48 h.
HClaq 1 N then HCl 4 N in dioxane, r.t., overnight.
(ii) H2, 10% Pd/C, MeOH, r.t., 48 h.
This journal is © The Royal Society of Chemistry 2014
Med. Chem. Commun., 2014, 5, 719–727 | 723

View Article Online
MedChemComm
Concise Article
axis intercept were obtained (Fig. 2b). Compounds 10 and 13
were not inhibitors at the highest tested concentration (28 mM),
whereas compounds 9, 11 and 12 inhibited the enzyme poorly
(18, 18 and 35% inhibition at 28 mM, respectively). The most
efficient inhibitors 3, 4, 8 and 14 were also assayed against the
multidrug-resistant mutated protease ANAM-11 (which bears
11 mutations: Leu10Ile/Met36Ile/Ser37Asp/Met46Ile/Arg57-
Lys/Leu63Pro/Ala71Val/Gly73Ser/Ile84Val/Leu90Met/Ile93Leu,
Fig. 2c).51 This protease is analogous to that found in multi-
resistant viruses.
Biological evaluation of the new molecular tongs showed
that the introduction of a small amino group directly attached
to the naphthalene scaffold in molecules 4 and 8 almost
preserved the dimerization inhibitory efficiency (0.40 mM for 4
versus 0.22 mM for 1 (ref. 22) and 0.55 mM for 8 versus 0.28 mM for
5 (ref. 22)). The aminopropanediol group decreased the
dimerization inhibitory efficiency only slightly in the case of
molecule 3 (0.60 mM for 3 relative to 0.22 mM for 1 (ref. 22)) and
noticeably in the case of 7 (2.75 mM for 7 relative to 0.28 mM for 5
(ref. 22)). It is noteworthy that both dioxolane precursors
(compounds 2 and 6) exhibited dramatically decreased inhibi-
tion (2.70 mM for 2 versus 0.60 mM for the aminopropanediol 3
and 5.90 mM for 6 versus 2.75 mM for the aminopropanediol 7).
This decreased inhibitory activity of molecular tongs 2 and 6
suggests a possible alteration of the hypothesised binding mode
due to the presence of the 2,2-dimethyl-1,3-dioxolane methyl
group. Even if introduced in the non-pharmacophore region,
such a group might weaken the interaction between the pepti-
domimetic strands and the PR-termini. The methylamino group
preserved the dimerization inhibitory potency of molecular
tong 14 (0.30 mM for 4 versus to 0.22 mM for 1 (ref. 22)).
Remarkably, the four new molecular tongs 3, 4, 8 and 14
behaved as antidimers against the multi-mutated protease
ANAM-11 (Table 3). However, when the amino group was
attached through a longer alkyl or alkylamide chain the inhib-
itory activity is dramatically decreased (molecules 9, 11 and 12)
or even suppressed (molecules 10 and 13). These results might
be attributed to the length and exibility of the alkyl or
Scheme
9 Synthesis of molecular tongs 14–15. Reagents and
conditions: (i) peptidomimetic 41, HBTU, HOBt, DIPEA, DMF, r.t., 48 h.
(ii) H2 (10 bar), 10% Pd/C, MeOH, 50 ꢀC, 3 min.
arm 41 (using HBTU–HOBt, DIPEA, in dry DMF) to give
molecular tongs 44 and 15 in satisfactory yields (65% for both
compounds). Surprisingly, the classical deprotection conditions
used previously (H2, 10% Pd/C, MeOH) were inefficient for the
cleavage of the Cbz moiety present on molecule 44. Therefore,
more drastic conditions were used, and treatment of molecule
44 in a hydrogen atmosphere under pressure (10 bar) using 10%
Pd/C in hot methanol (50 ꢀC) afforded molecular tong 14 in
modest yield (32%).
Biology
The ability of compounds 2–4 and 6–15 to inhibit PR
(recombinant wild-type protease) was evaluated in a uori-
metric assay at the optimum pH (4.7) and 30 ꢀC.19–22 Zhang–
Poorman kinetic analyses were performed to identify the
mechanism of inhibition. Plots of [E]0/Ovi versus Ovi were con-
structed, where vi is the initial rate. The results are summarised
in Table 3. Compounds 2–4, 6–8 and 14 efficiently inhibited PR.
Parallel lines were obtained for these compounds, demon-
strating that they act as pure dimerization inhibitors (Fig. 2a).
Compound 15 was able to inhibit PR, however it acted as a
competitive inhibitor since altered slopes and the unaltered y-
Fig. 2 Zhang kinetic analyses of PR by compounds 3 (a) and 15 (b), and ANAM-11 by compound 8 (c), at pH 4.7 and 30 ꢀC. PR inhibition: [PR] ¼
5.38 ꢁ 18 nM; [S]0 ¼ 5.2 mM; [3] ¼ 17 mM (-) and 11 mM (:); [15] ¼ 22 mM 2 (-) and 11 mM 2 (:). ANAM-11 inhibition: [PR] ¼ 5.33 ꢁ 18 nM; [S]0 ¼ 5.2
mM; [8] ¼11 mM (-) and 5.6 mM 2. Enzyme activity was also always measured without inhibitor (C). Each experiment was at least performed in
triplicate.
724 | Med. Chem. Commun., 2014, 5, 719–727
This journal is © The Royal Society of Chemistry 2014

View Article Online
Concise Article
MedChemComm
Table 3 Inhibition of PR and mutated HIV-1 protease ANAM-11 by compounds 1–15 (30 ꢀC and pH 4.7)
Kida (mM)
Hydrophobic surface
area (A )
Hydrophilic surface
area (A )
2
˚
2
Comp
PR
ANAM-11
C log P
Retention timec (min)
˚
1
2
3
4
5
6
7
8
0.22 (ref. 22)
2.70
0.60
0.40
0.28 (ref. 22)
5.90
2.75
0.55
0.30
nd
nd
0.94
0.94
2.00
nd
nd
0.83
2.45
6.5
7.3
4.8
5.6
8.7
9.5
6.9
7.7
7.4
1305
1455
1316
1261
1681
1844
1713
1635
1664
359
368
456
402
489
519
600
536
558
13.8
14.2
11.3
10.6
14.8
15.2
12.7
12.2
11.1
14
Hydrophobic surface
area (A )
Hydrophilic surface
area (A )
Kicb (mM)
C log P
7.2
Retention timec (min)
11.2
2
2
˚
˚
Comp
15
15
1647
632
% Inhibition
at 28 mM (PR)
Hydrophobic surface
area (A )
Hydrophilic surface
area (A )
2
˚
2
˚
Comp
C log P
Retention timec (min)
9
18
0
18
35
0
7.1
9.6
6.2
8.0
9.2
1755
2032
1724
1719
1833
561
599
522
569
510
11.1
13.8
11.2
11.4
11.4
10
11
12
13
a
b
Dimerization inhibition constant (dissociation constant of monomer–inhibitor complexes). Competitive inhibition constant (dissociation
constant of a dimer–inhibitor complex). Standard errors of initial rates are less than 5%. nd: not determined. C log P and solvent accessible
surface areas are indicated. Determined on a column Sunre 2.1 ꢂ 100 mm 3.5 mm, ow rate 1 mL minꢁ1; detection at 254 nm; solvent (A)
c
H2O + 0.2% formic acid and (B) CH3CN, mixture A/B: from 95/5 to 0/100 in 20 min.
alkylamide chain allowing the amino groups to contract aminopropanediol (factor of 1.35 between molecular tongs 1
hydrogen bonds with the peptidomimetic arms, thus and 3 and factor of 1.26 between molecular tongs 5 and 7). The
precluding interactions with the PR termini. The introduction amine in molecules 4 and 8, the methylamine in molecules 14
of a guanidinium group at the 3-position of the naphthalene and 9, and the methylguanidine in molecule 15 have similar
scaffold (molecule 15) did not lead to PR dimerization inhibi- effects on the C log P (factors around 1.12–1.20). The amino-
tion. Conversely to that reasonably expected, the positively propyl group in 12 and the dimethylaminopropyl substituent in
charged guanidinium of 15 led to a shi from dimerization 13 have a negligible effect on the C log P (factor of 1.08 and 1.05
inhibition towards active-site competitive inhibition. The respectively). The hydrophilic surface provided by the polar
decrease or suppression of the dimerization inhibitory activity groups is increased in all cases, while the hydrophobic surface
of molecules 9–13 and 15 might be assigned to a destabilization provided by the strands is almost always preserved. The esti-
of the hypothesised binding mode with PR termini due to steric mated C log P values were compared with the experimentally
and/or electronic effects. In particular, the positive charge determined chromatographic retention times, because the
resulting from protonation of the amine (molecules 9–13 and reverse phase liquid chromatography has been previously
14) and guanidine (molecule 15) groups might form a salt reported as a model to estimate aqueous solubility.57 In the rst
bridge with the C-terminal carboxylate of Phe-99. It was previ- series of molecules 1–4, retention time differences were in
ously reported that bicyclic guanidinium spacers introduce accordance with their C log P. Molecules 3 and 4 were less
additional electrostatic hydrogen bonding interactions with the retained on the column than molecules 1 and 2 (Table 3). In the
C-terminal Phe-99 carboxylate leading to dimerization second series of molecules 5–15, retention time differences
inhibitors.16
were in accordance with their C log P for the neutral molecules
Three physical properties (estimated log P, referred to as 5–8 while ionisable compounds 9 and 11–15 have similar
C log P, hydrophobic surface area and hydrophilic surface area) retention times (Table 3).57c Experimental solubility in water was
of molecular tongs 1–15 were calculated by molecular modeling determined for the more active and the more polar compounds
(see ESI†52–56) in order to correlate the nature of the 3-substit- 3 and 4 (z0.3 mg mLꢁ1). These data show that submicromolar
uent on the naphthalene scaffold with the hydropathic prop- inhibitory activity can be conserved (Kid ¼ 0.60 and 0.40 mM for
erties of the molecules (Table 3). The more favorable groups molecules 3 and 4 respectively compared to 0.22 mM for
capable of decreasing C log P are the piperazine ring (factor of molecular tong 1) while solubility is increased.
1.40 between molecular tongs
5
and 11) and the
This journal is © The Royal Society of Chemistry 2014
Med. Chem. Commun., 2014, 5, 719–727 | 725

View Article Online
MedChemComm
Concise Article
Although highlighting the difficulty involved, these data
show the possibility of introducing polar groups on the naph-
thalene scaffold and nding a compromise between hydrophi-
licity provided by the polar groups and inhibitory activity
essentially supported by the peptidomimetic strands.
Finally, we determined the stability of molecular tongs 3 and
8 in RPMI culture medium containing 20% fetal calf serum. We
previously reported that the introduction of the 5-acetamido-2-
methoxybenzohydrazide group in a single strand increased the
metabolic stability of the molecular tongs.21 Remarkably,
compounds 3 and 8 were not signicantly degraded aer
incubation during 48 h (see ESI†).
2 N. E. Kohl, E. A. Emini, W. A. Schleif, L. J. Davis,
J. C. Heimbach, R. A. Dixon, E. M. Scolnick and I. S. Sigal,
Proc. Natl. Acad. Sci. U. S. A., 1988, 85, 4686–4690.
3 A. Carr, Nat. Rev. Drug Discovery, 2003, 2, 624–634.
4 (a) R. T. D'Aquila, J. M. Schapiro, F. Brun-Vezinet, B. Clotet,
B. Conway, L. M. Demeter, R. M. Grant, V. A. Johnson,
D. R. Kuritzkes, C. Loveday, R. W. Shafer and
D. D. Richman, Top. HIV Med., 2002, 10, 21–25; (b) E. De
Clercq, Nat. Rev. Microbiol., 2004, 2, 704–720.
5 R. Ishima, R. Ghirlando, J. Tozser, A. M. Gronenborn,
D. A. Torchia and J. M. Louis, J. Biol. Chem., 2001, 276,
49110–49116.
6 M. J. Todd, N. Semo and E. Freire, J. Mol. Biol., 1998, 283,
475–488.
7 A. Gustchina and I. T. Weber, Proteins, 1991, 10, 325–329.
8 (a) N. Boggetto and M. Reboud-Ravaux, Biol. Chem., 2002,
383, 1321–1324; (b) L. Bannwarth and M. Reboud-Ravaux,
Biochem. Soc. Trans., 2007, 35, 551–554.
Conclusions
A new series of naphthalene-based molecular tongs containing
polar groups at the 3-position of the naphthalene scaffold was
designed and synthesised. The polar groups were introduced
mainly via metal-catalysed cross coupling reactions on bromo-
naphthalene derivatives, in very satisfying yields. The ability of
the new molecules to disrupt the PR termini b-sheet interface in
order to inhibit PR was analysed. The best of our molecular
tongs inhibited PR dimerization with an inhibition constant Kid
of 0.3 mM. Furthermore, most of the polar groups decreased the
C log P parameter of the molecules and thus their water solu-
bility. We have demonstrated that the inhibitory efficiency is
strictly related to the molecule structure, indicating that it is
essential to have an appropriate balance between the length
and/or bulkiness of the substituent and its hydrophilicity, in
order to maintain effective anti-dimerization activity. As shown
by inhibitors 3 and 4, introducing an aminopropanediol moiety
or an amino group on the naphthalene scaffold increases water
solubility without compromising inhibitory efficiency. The
ability of these proteolysis-resistant molecules to inhibit multi-
mutated ANAM-11 suggests that they have the potential to
successfully overcome the resistance presently encountered
with classical PI protease inhibitors.
9 K. Hamacher, Gene, 2008, 422, 30–36.
10 H. J. Schramm, H. Nakashima, W. Schramm, H. Wakayama
and N. Yamamoto, Biochem. Biophys. Res. Commun., 1991,
179, 847–851.
11 Z. Y. Zhang, R. A. Poorman, L. L. Maggiora, R. L. Heinrikson
and F. J. Kezdy, J. Biol. Chem., 1991, 266, 15591–15594.
12 J. Franciskovich, K. Houseman, R. Mueller and
J. Chiemlevski, Bioorg. Med. Chem. Lett., 1993, 3, 765–768.
13 H. J. Schramm, E. de Rosny, M. Reboud-Ravaux, J. Buttner,
A. Dick and W. Schramm, Biol. Chem., 1999, 380, 593–596.
14 J. Dumond, N. Boggetto, H. J. Schramm, W. Schramm,
M. Takahashi and M. Reboud-Ravaux, Biochem. Pharmacol.,
2003, 65, 1097–1102.
15 L. Bannwarth, T. Rose, L. Dufau, R. Vanderesse, J. Dumond,
B. Jamart-Gregoire, C. Pannecouque, E. De Clercq and
M. Reboud-Ravaux, Biochemistry, 2009, 48, 379–387.
16 (a) P. Breccia, N. Boggetto, R. Perez-Fernandez, M. Van Gool,
M. Takahashi, L. Rene, P. Prados, B. Badet, M. Reboud-
Ravaux and J. de Mendoza, J. Med. Chem., 2003, 46, 5196–
5207; (b) O. A. Flausino Jr, L. Dufau, L. O. Regasini,
ˆ
M. S. Petronio, D. H. S. Silvab, T. Rosec, V. S. Bolzani and
Acknowledgements
M. Reboud-Ravaux, Curr. Med. Chem., 2012, 19, 4534–4540.
17 R. Zutshi, J. Franciskovich, M. Slultz, B. Schwetizer,
P. Bishop, M. Wilson and J. Chmielewski, J. Am. Chem.
Soc., 1997, 119, 4841–4845.
18 L. G. Ulysse and J. Chiemlevski, Bioorg. Med. Chem. Lett.,
1998, 8, 3281–3286.
19 A. Bouras, N. Boggetto, Z. Benatalah, E. de Rosny, S. Sicsic
and M. Reboud-Ravaux, J. Med. Chem., 1999, 42, 957–962.
20 N. Merabet, J. Dumond, B. Collinet, L. Van Baelinghem,
N. Boggetto, S. Ongeri, F. Ressad, M. Reboud-Ravaux and
S. Sicsic, J. Med. Chem., 2004, 47, 6392–6400.
21 L. Bannwarth, A. Kessler, S. Pethe, B. Collinet, N. Merabet,
N. Boggetto, S. Sicsic, M. Reboud-Ravaux and S. Ongeri,
J. Med. Chem., 2006, 49, 4657–4664.
We thank Claire Troufflard (BioCIS, UMR 8076) for NMR
experiments. The Italian Research Ministry provided nances
for RF. Financial support for ASR and AV (Marie Curie Early
Stage training Fellowship of the European Community's Sixth
Framework Programme: contract MESTCT-2004-515968) was
`
provided by the European Community. The Ministere de la
Recherche et des Technologies (MRT) provided nancial
support for LD. We thank Prof. Ernesto Freire of the Depart-
ment of Biology and Biocalorimetry Center, Johns Hopkins
University, Baltimore, MD, for the generous gi of the plasmid
encoding the ANAM-11 mutant.
22 A. Vidu, L. Dufau, L. Bannwarth, J. L. Soulier, S. Sicsic,
U. Piarulli, M. Reboud-Ravaux and S. Ongeri,
ChemMedChem, 2010, 11, 1899–1906.
References
1 S. Orozlan and R. B. Luig, Curr. Top. Microbiol. Immunol.,
1990, 157, 153–185.
726 | Med. Chem. Commun., 2014, 5, 719–727
This journal is © The Royal Society of Chemistry 2014

View Article Online
Concise Article
MedChemComm
˜
`
23 L. Dufau, A. S. M. Ressurreiçao, R. Fanelli, N. Kihal, A. Vidu, 41 R. Chinchilla and C. Nagera, Chem Rev., 2007, 107, 874–922.
T. Milcent, J.-L. Soulier, J. Rodrigo, A. Desvergne, K. Leblanc, 42 K. Sonogashira, Y. Tohda and N. Hagihara, Tetrahedron Lett.,
B. Crousse, M. Reboud-Ravaux and S. Ongeri, J. Med. Chem.,
2012, 55, 6762–6775.
24 N. Sharin, Int. J. Pharm. Life Sci., 2013, 1, 1–9.
25 D. B. Horton, K. B. Siripurapu, G. Zheng, P. A. Crooks and
1975, 16, 4467–4470.
43 J. Lindley, Tetrahedron, 1984, 40, 1433–1456.
44 I. P. Beletskaya, A. S. Sigeev, A. S. Peregudov and
P. V. Petrovskii, J. Organomet. Chem., 2004, 689, 3810–3812.
L. P. Dwoskin, J. Pharmacol. Exp. Ther., 2011, 339, 286–297. 45 M. Sundermeier, A. Zopf and M. Beller, Angew. Chem., Int.
26 I. A. Sogias, V. V. Khutoryanskiy and A. C. Williams,
Macromol. Chem. Phys., 2010, 211, 426–433.
27 R. G. Cooke, B. L. Johnson and W. R. Owen, Aust. J. Chem.,
1960, 13, 256–260.
Ed., 2003, 42, 1661–1664.
46 T. Schareina, A. Zapf and M. Beller, J. Organomet. Chem.,
2004, 689, 4576–4583.
47 B. Wang, R. Zhao, B. Chen and B. Balasubramanian,
ARKIVOC, 2010, 47–52.
28 E. Martinborough, T. M. Denti, P. P. Castro, T. B. Wyman,
´
¨
C. B. Knobler and F. Diederich, Helv. Chim. Acta, 1995, 78, 48 T. Schareina, A. Zapf, A. Cotte, N. Muller and M. Beller,
1037–1066.
Synthesis, 2008, 20, 3351–3355.
29 F. Ullmann, Ber. Dtsch. Chem. Ges., 1903, 36, 2382–2384.
30 D. Ma, Q. Cai and H. Zhang, Org. Lett., 2003, 5, 2453–2455.
49 J. Chun, L. He, H.-S. Byun and R. Bittman, J. Org. Chem.,
2000, 65, 7634–7640.
31 H. Zhang, Q. Cai and D. Ma, J. Org. Chem., 2005, 70, 5164– 50 T. K. Chakraborty, A. K. Hussain and S. P. Joshi, Chem. Lett.,
5173. 1992, 2385–2388.
32 A. Klapars, J. C. Antilla, X. Huang and S. L. Buchwald, J. Am. 51 S. Muzammil, P. Ross and E. Freire, Biochemistry, 2003, 42,
Chem. Soc., 2001, 123, 7727–7729. 631–638.
33 Y. Liu, Y. Bai, J. Zhang, Y. Li, J. Jiao and X. Qi, Eur. J. Org. 52 (a) J. Sadowski, J. Gasteiger and G. Klebe, J. Chem. Inf.
Chem., 2007, 6084–6088.
34 J. Louie and J. F. Hartwig, Tetrahedron Lett., 1995, 36, 3609–
Comput. Sci., 1994, 34, 1000–1008; (b) The 3D structure
generator CORINA is available from Molecular Networks
GmbH, Erlangen, Germany.
3612.
35 A. S. Guram, R. A. Rennels and S. L. Buchwald, Angew. Chem., 53 The OPLS_2005 parameters are described in, J. L. Banks,
Int. Ed., 1995, 34, 1348–1350.
36 J. Yin, M. M. Zhao, M. A. Huffman and J. M. McNamara, Org.
Lett., 2002, 4, 3481–3484.
37 E. Garnier, J. Audoux, E. Pasquinet, F. Suzenet, D. Poullain,
B. Lebret and G. Guillaumet, J. Org. Chem., 2004, 69, 7809–
7815.
H. S. Beard, Y. Cao, A. E. Cho, W. Damm, R. Farid,
A. K. Felts, T. A. Halgren, D. T. Mainz, J. R. Maple,
R. Murphy, D. M. Philipp, M. P. Repasky, L. Y. Zhang,
B. J. Berne, R. A. Friesner, E. Gallichio and R. M. Levy,
J. Comput. Chem., 2005, 26, 1752–1780.
¨
54 MacroModel, version 9.9, Schrodinger, LLC, New York, NY,
`
38 T. H. M. Jonckers, B. U. W. Maes, G. L. F. Lemiere and
2012.
¨
55 LigPrep, version 2.5, Schrodinger, LLC, New York, NY, 2012.
R. Dommisse, Tetrahedron, 2001, 57, 7027–7034.
ˇ
`
¨
39 J. Kosmrlj, B. U. W. Maes, G. L. F. Lemiere and A. Haemers, 56 QikProp, version 3.5, Schrodinger, LLC, New York, NY, 2012.
Synlett, 2000, 1581–1584.
57 (a) Y.-P. Chin, W. J. Weber JRt and T. C. Voice, Water Res.,
1986, 20, 1443–1450; (b) L.-G. Danielsson and Y.-H. Zhang,
Trends Anal. Chem., 1996, 15, 188–196; (c) S. K. Poole and
C. F. Poole, J. Chromatogr. B: Anal. Technol. Biomed. Life
Sci., 2003, 797, 3–19.
40 M. L. Gelmi, D. Pocar, G. Pontremoli, S. Pellegrino,
E. Bombardelli, G. Fontana, A. Riva, W. Balduini,
S. Carloni, M. Cimino and F. Johnson, J. Med. Chem., 2006,
49, 5571–5577.
This journal is © The Royal Society of Chemistry 2014
Med. Chem. Commun., 2014, 5, 719–727 | 727