Fragment-based drug design targeting syntenin PDZ2 domain involved in exosomal release and tumour spread

Article abstract of DOI:10.1016/j.ejmech.2021.113601

Syntenin stimulates exosome production and its expression is upregulated in many cancers and implicated in the spread of metastatic tumor. These effects are supported by syntenin PDZ domains interacting with syndecans. We therefore aimed to develop, throu

Full text of DOI:10.1016/j.ejmech.2021.113601

European Journal of Medicinal Chemistry 223 (2021) 113601
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Fragment-based drug design targeting syntenin PDZ2 domain
involved in exosomal release and tumour spread
a
a
b
a
€
Manon Garcia , Laurent Hoffer , Raphael Leblanc , Fatiha Benmansour ,
Mikael Feracci a, Carine Derviaux a, Antonio Luis Egea-Jimenez b, Philippe Roche a,
Pascale Zimmermann b c, Xavier Morelli a, Karine Barral a
,
, *
a
ꢀ
ꢀ
Centre de Recherche en Cancerologie de Marseille (CRCM), Integrative Structural & Chemical Biology, Aix-Marseille Universite, Inserm 1068, CNRS 7258,
Institut Paoli Calmettes, 13009, Marseille, France
b
ꢀ
ꢀ
ꢀ
Equipe Labellisee Ligue 2018, Centre de Recherche en Cancerologie de Marseille, Aix-Marseille Universite, Inserm1068, CNRS7258, Institut Paoli-Calmettes,
13009 Marseille, France
c Department of Human Genetics, K. U. Leuven, B-3000, Leuven, Belgium
a r t i c l e i n f o
a b s t r a c t
Article history:
Syntenin stimulates exosome production and its expression is upregulated in many cancers and impli-
cated in the spread of metastatic tumor. These effects are supported by syntenin PDZ domains interacting
with syndecans. We therefore aimed to develop, through a fragment-based drug design approach, novel
inhibitors targeting syntenin-syndecan interactions. We describe here the optimization of a fragment,
‘hit’ C58, identified by in vitro screening of a PDZ-focused fragment library, which binds specifically to the
syntenin-PDZ2 domain at the same binding site as the syndecan-2 peptide. X-ray crystallographic
structures and computational docking were used to guide our optimization process and lead to com-
Received 22 March 2021
Received in revised form
28 May 2021
Accepted 30 May 2021
Available online 5 June 2021
Keywords:
pounds 45 and 57 (IC50 ¼ 33
mM and 47 mM; respectively), two representatives of syntenin-syndecan
Fragment-based drug design
Small molecule PPI inhibitor
Syntenin-syndecan interaction inhibitors
Exosomes biogenesis
interactions inhibitors, that selectively affect the syntenin-exosome release. These findings demon-
strate that it is possible to identify small molecules inhibiting syntenin-syndecan interaction and exo-
some release that may be useful for cancer therapy.
Cancer therapy
© 2021 Elsevier Masson SAS. All rights reserved.
1. Introduction
32 kDa PDZ-tandem (PDZ1 and PDZ2) intracellular protein, has a
large variety and diversity of interaction partners, including syn-
Exosomes are lipid-bilayer nanovesicles, released from most cell
types, which appear to be important during cancer progression,
with a role in metastatic niche formation and tumor immune
escape [1,2]. ‘Cancer exosomes’ derived from tumor cells are thus
gaining considerable attention as oncotargets [3]. Syntenin, a
decans [4]. Syndecans (SDCs) are transmembrane proteins carrying
heparin sulfate chains involved in numerous cellular processes,
including cellecell and cellematrix adhesion, signal transduction,
and vesicular trafficking [5e7]. P. Zimmermann and coworkers
have identified a molecular pathway, supported by the syndecan
heparan sulfate proteoglycans and their adaptor protein syntenin,
that binds to the cytosolic tail of syndecans, which regulates the
biogenesis and the uptake of a specific subclass of exosome [8].
Recent studies established that syntenin-exosomes might be
implicated in the processes of malignant cell invasion and the
establishment of niches that promote tumor metastasis [9]. Indeed,
syntenin expression is upregulated in many cancer cell lines (car-
cinomas and others cancers) suggesting that it may be an effective
target to intervene in metastasis since gain of syntenin expression
in tumors cells has been associated with the invasion and the
metastatic potential of various solid cancers [10].
Abbreviations: FBDD, fragment-based drug design; PPI, protein-protein inter-
action; PDB, protein data bank; LE, ligand effipciency; LLE, lipophilic Ligand Effi-
ciency; BB, building block; IC50, inhibitory concentration needed to decrease by 50%
the enzyme activity; EC50, effective concentration inducing 50% of the maximum
effect on infected cells; CC50, cytotoxic concentration that reduces the infected cell
number by 50%; SD, standard deviation; HTRF, Homogeneous Time Resolved
Fluorescence; LC/MS, liquid chromatography / mass spectroscopy; NMR, nuclear
magnetic
resonance;
rt,
room
temperature;
EDCI,
1-ethyl-3-(3-
dimethylaminopropyl)carbodiimide; THF, tetrahydrofuran; FA, formic acid; VS,
virtual screening; SAR, Structure-activity relationship.
* Corresponding author.
As knowledge has grown regarding cellular protein interaction
E-mail address: karine.barral@inserm.fr (K. Barral).
https://doi.org/10.1016/j.ejmech.2021.113601
0223-5234/© 2021 Elsevier Masson SAS. All rights reserved.

M. Garcia, L. Hoffer, R. Leblanc et al.
European Journal of Medicinal Chemistry 223 (2021) 113601
networks and their role in numerous cell disorders, including
cancer, targeting protein-protein interactions (PPI) has emerged as
an encouraging therapeutic approach for the treatment of solid
tumors or hematologic malignancies [11e13]. Promising results
have been obtained with PPI inhibitors such as p53, BclxL or more
recently bromodomain inhibitors [13e18]. We therefore aim to
apply a Fragment-Based Drug Design (FBDD) strategy for the rapid
identification of starting hits and their subsequent optimization
into compounds that can selectively disrupt syntenin-syndecan
PDZ-domain mediated interactions showing known implications
in cancer development and dissemination.
targeted its first PDZ domain and had a Kd value of 21 mM
(LE ¼ 0.16) [23]. In this study, we performed SAR analysis and
fragment hit optimization aiming to improve the affinity of C58 for
the second PDZ domain of Syntenin. X-ray crystallographic struc-
tures and molecular modeling were used to guide the fragment
optimization process in order to generate new analogues by adding
chemical groups that might establish strong interactions with
syntenin-PDZ2 domain.
2. Results and discussion
In a previous study, Leblanc et al. screened a fragment library,
containing 139 potential PDZ-inhibitory fragments, on syntenin-
syndecan-2 complex by HTRF. Five fragments were identified and
tested for their selectivity in two other PDZ-mediated interaction
complexes implicated in oncology, namely GRASP/Jam-C and Erbin/
P0071 complexes [19]. The most potent and selective fragment, C58
(Fig. 1), was confirmed in a secondary dose-response screening to
be a reliable inhibitor of syntenin-syndecan interaction with an IC50
2.1. Fragment expansion and SAR around analogues of C58
A hit identified from screening traditionally require expansion
to provide evidence SAR based upon chemical similarity. Therefore,
fragment C58 was used as starting point to select the best fragment
hit involved in the subsequent optimization process. For this pur-
pose, a rapid fragment hit expansion has been considered by syn-
thetizing nearest analogues around fragment C58 (Fig. 1).
X-ray crystal structure of C58 bound to syntenin was resolved
(PDB code 6R9H) and revealed that C58 is located within the
syntenin-PDZ2 domain on the same binding pocket as syndecan-2
peptide as shown in Fig. 1A (a superimposition view of binding
mode of compounds C58 and C-terminal end peptide of Syndecan is
given Figure S1A in Supp. Data). The phenyl ring of fragment C58
of 350
mM (LE ¼ 0.25). Additional experiment on MCF7 breast
carcinoma cells established that C58 decreases MCF7 migration and
colony formation, alters mammosphere formation and exosomal
release [19]. Taken together, these data clearly indicated that
fragment C58 is a selective inhibitor of the syntenin-exosomal
pathway.
It should be noted that PDZ domains belong to protein-protein
interaction (PPI) networks, and were considered as poorly drug-
gable targets for the development of potent small organic probes.
This was confirmed by the screening of large compound libraries in
which no potent hits were identified [20]. However, several studies
have reported that small organic compounds were able to disrupt
PPIs between PDZ domains and their endogenous protein partners
[21]. Some of these studies also included intensive
structureeactivity relationship (SAR) studies in which dozens an-
alogues were synthesized and tested. As expected, due to the poor
druggable nature of PDZ domains, most non-peptide compounds
exhibits a strong
p-stacking with Phe213, its nitrogen atom in-
teracts with the backbone carbonyl of Phe211 through hydrogen
bonds and its carboxylic acid forms the expected canonical
hydrogen-bonding interactions with backbone amide group of
Val209 and Gly210 [19]. Based on these structural data, we explored
four different modifications to improve the affinity of C58 while
maintaining its canonical binding mode (Fig. 1C).
The main modification involved the terminal hydrophobic
moiety of fragment C58. Indeed, PDZ domains are known to
recognize C-terminal hydrophobic moieties from their partners.
The last residue from syntenin endogenous ligands (syndecan
proteins) is an alanine that contains a very small hydrophobic
sidechain. Surprisingly, structural data highlighted that PDZ
domain from syntenin was also able to bind peptides with large
hydrophobic terminal residue (PDB 1W9E). However, binding of
the phenylalanine required an induced fit to accommodate the
large benzyl group. Thus, 3 different terminal modifications (Ala-
exhibited moderate inhibition, i.e. IC50 in the 10
mM range. One
notable exception was recently reported with the development of
high affinity compounds (IC50 value below 100 nM) for the Pick1
PDZ domain [22]. To our knowledge, this is the most potent non-
peptide PDZ inhibitor discovered to date. Regarding the syntenin
protein, the most potent organic compound reported to date
Fig. 1. A. Crystal structure of syntenin PDZ2 domain in complex with compound C58 (PDBID: 6R9H). Compound C58 is depicted in cylinder representation dark green carbon atom.
Nitrogen, oxygen, chlorine and sulphur are colored in blue, red, green and yellow, respectively. Small red spheres represent water molecules around the binding sites. Hydrogen
bond interactions are displayed as black dotted lines. Amino acid residues highlighted are those establishing Van der Waals interactions with compound C58. B. 2D structure of hit
fragment C58. C. 2D structures of putatively interesting explored modifications around fragment C58.
2

M. Garcia, L. Hoffer, R. Leblanc et al.
European Journal of Medicinal Chemistry 223 (2021) 113601
like, Val-like and Phe-like) were investigated in the SAR studies. A
thioether function was also present in the original fragment hit.
This sub-optimal moiety, from a medicinal chemistry point of view,
was replaced by 1) its similar ether function and 2) with one
methylene spacer in the SAR studies. Finally, the influence of the
chlorine atom was also assessed by synthesizing the chlorine-free
version of each compound.
In a first stage, we aimed to synthesize analogues with a spacer
length of two atoms (SeCH2 or OeCH2) between the phenyl ring
and the amide bond. A general synthesis of amide derivatives is
depicted in Scheme 1. To proceed, we directly condensed the
commercial acetyl chloride derivatives (1e4) with the corre-
20 exhibits an inhibition of 16%, and (iii) chlorine as para substit-
uent because the corresponding non-substituted analogue 14 dis-
plays an inhibition of 20%.
Among the analogues 41e53 with a spacer length of three
atoms, we identified compound 45 as the most potent inhibitor of
syntenin-syndecan-2 interaction, with an inhibition of 94%. As
previously, these data revealed the importance of (i) valine as
amino acid residue in Cter versus alanine or phenylalanine, (ii)
sulphur atom versus oxygen or methylene in the spacer, and (iii)
chlorine as para substituent. Adding one supplementary methylene
group between the sulphur atom and the amide bond resulted in a
more than 10-fold improvement of the affinity (IC50 ¼ 33
350
M for compounds 45 and C58, LE ¼ 0.31/LLE ¼ 1.31 and
LE ¼ 0.25/LLE ¼ 0.74, respectively). Affinity of compound 45 was
validated in the presence of 0.01% NP40 detergent, (IC50 ¼ 45
and 864 M for compounds 45 and C58, respectively), demon-
strating that its activity is not due to aggregation effect (Fig. 2). As
reported by Leblanc et al. [19] secondary dose-responses
mM and
sponding
L-aminoacid methyl ester hydrochloride (Alanine or
m
Valine) in a mixture dichloromethane/trimethylamine at room
temperature [24]. Methyl ester intermediates 5e12 obtained with
good to excellent yields were then hydrolysed with LiOH to give
derivatives 13e20 [25].
In a second stage, we performed the synthesis of analogues with
a spacer length of three atoms (SeCH2eCH2, OeCH2eCH2 or
(CH2)3) between the phenyl ring and the amide bond (Scheme 2).
Indeed, molecular modeling studies, relying on the X-ray structure
of C58 (16), suggested that the addition of one carbon spacer in the
compound structure could maintain both the PDZ canonical bind-
ing mode with the carboxylate group and p-stacking with Phe213,
while optimizing the torsion angles from the flexible carbohydrate
spacer.
Thus, thiophenol and 4-chlorothiophenol were converted to 3-
(phenylthio)propanoic acid 23 and 3-(4-chlorophenylthio)prop-
anoic acid 24 by S-alkylation with 3-chloropropionic acid under
alkaline conditions [26,27]. Then, derivatives 23 and 24, as well as
commercial propionic acid derivatives 25e27, were condensed
mM
m
a
screening confirms the selectivity of analogue 45 for syntenin,
with no effect on GRASP/Jam-C and Erbin/P0071 complexes (two
other PDZ domains implicated in oncology). Compound 45 was also
validated using SPR in an ‘orthogonal’ validation with KD values of
130
shown).
To validate and characterize the binding mode of compound 45,
mM for syntenin-PDZ tandem with C-terminal (data not
we soaked it with syntenin crystals. The X-ray crystal structure of
45 bound to syntenin was resolved (PDB code 6RLC) and its binding
mode into syntenin PDZ2 domain was confirmed with the expected
low strain energy in the carbohydrate spacer (Fig. 3). The phenyl
ring of 45 exhibits a strong
p-stacking with Phe213, its nitrogen
atom makes a hydrogen-bonding interaction with the backbone
carbonyl of Phe211 and its carboxylic acid forms the expected ca-
nonical hydrogen-bonding interactions with backbone amide
group of Val209 and Gly210 [19].
Therefore, optimized fragment 45 maintains canonical binding
mode and presents a better biological evaluation than the previ-
ously identified fragment C58 (a superimposition view of binding
mode of compounds C58 and 45 with syntenin is given Figure S1B
in Supp. Data). For these reasons, we performed the subsequent
growing optimization process starting from fragment 45.
with the corresponding L-aminoacid methyl ester hydrochloride
(Alanine, Valine or Phenylalanine) in a mixture dichloromethane/
trimethylamine at room temperature in the presence of coupling
reagent EDCI [28,29]. As previously, methyl ester intermediates
28e40 obtained with good yields were then hydrolysed with LiOH
to give derivatives 41e53 [25].
A primary screening was performed by HTRF with analogues
13e20 and 41e53, at a concentration of 400 mM, against Syntenin-
Syndecan2 complex (GST-S1FL/SDC2) (Table 1). None of the ana-
logues 13e20 with a spacer length of two atoms displayed a better
affinity than fragment C58 (analogue 16, inhibition of 56%, IC50 of
2.2. Optimization of fragment 45 by fragment growing
350 mM). These data revealed the importance of (i) valine as ter-
minal amino acid residue because the corresponding alanine
analogue 15 shows an inhibition of 25%, (ii) sulphur atom versus
oxygen in the spacer because the corresponding phenoxy analogue
Based on X-ray crystallographic structure of syntenin in com-
plex with fragment 45, our recently published in silico chemistry
driven strategy DOTS was applied to the design of optimized
Scheme 1. Synthesis of analogues 13e20.
3

M. Garcia, L. Hoffer, R. Leblanc et al.
European Journal of Medicinal Chemistry 223 (2021) 113601
Scheme 2. Synthesis of analogues 41e53 and their ester derivatives 28e40.
Table 1
Inhibition of Syntenin-Syndecan2 complex interaction in the presence of analogues 13e20 and 41e53 at 400
m
M confirmed in counter screening experiment using HTRF assay.
inhibitors [30]. The in silico pipeline can be briefly described as
follows (see experimental section for details): a focused virtual
chemical library was first designed around fragment 45. The
growing strategy consists in exploring the chemical space around
this core starting from the logical p-chlorophenyl optimization
vector that was naturally included in its structure. Then, the
resulting chemical library was virtually screened using the S4MPLE
molecular modeling tool [30e32] with constraints on the shared
substructure. Constraints were added in the first stage of the
simulation to maintain the original binding mode of the fragment,
according to the generic growing paradigm in fragment-based drug
discovery. In other words, atoms from fragment 45 were
4

M. Garcia, L. Hoffer, R. Leblanc et al.
European Journal of Medicinal Chemistry 223 (2021) 113601
Fig. 2. Selectivity profiles and IC50 of compounds C58 (red) and 45 (blue) using HTRF assay without and with 0.01% NP40.
Fig. 3. Crystal structure of syntenin PDZ2 domain in complex with compound 45 (PDBID: 6RLC). Compound 45 is displayed in cylinder representation orange carbon atom. Ni-
trogen, oxygen, chlorine and sulphur are colored in blue, red, green and yellow, respectively. Small red spheres represent water molecules around the binding sites. Hydrogen bond
interactions are depicted as black dotted lines. Amino acid residues indicated are those establishing Van der Waals interactions with compound 45.
constrained while newly added atoms from virtual synthesis were
free to explore putative adjacent sub-pockets without any con-
straints. The goal was to either catch additional interactions in the
vicinity of compound 45 or to extend this fragment towards K214,
N215, K250 and D251 residues (see illustration in Figure S2 in Supp.
Data), because these residues were known to interact with PIP2
ligand in previously published “Syntenin-PDZ-domain/peptide/
PIP2” ternary complexes (PDB 4Z33) [33]. However, it is important
to mention that these polar and flexible solvent exposed residues
do not define a clear druggable sub-pocket. This was confirmed by
the FTMap web-server [34], that identifies binding hot spots of
macromolecules using in silico mapping of small organic probes. It
has been proven that regions that can bind clusters of multiple
probes correlate with known binding hot spots of macromolecules.
Using the Syntenin 4Z33 PDB structure as input, the computational
mapping identified 2 main regions within the PDZ2-domain that
are predicted to be potential binding hot spots. The first concerns
the carboxylate binding loop (V209, G210) and the second is located
near Phe213. It should be noted that both regions perfectly match
the interaction residues with fragment 45 and bound peptide.
However, no particular binding hot spots regions were predicted
for K214, N215, K250 and D251 residues. The FTMap simulation
confirmed that this PIP2-binding area was not a well-defined
druggable sub-pocket (see Figure S2 in Supp. Data).
The final selection of compounds was achieved after visual in-
spection of the top100 best poses with a focus on compounds that
were predicted to interact with previously listed residues. In the
end, ten compounds were selected using Suzuki-Miyaura coupling
reaction (Table 2).
At this stage, we focused on the synthesis of biaryl compounds
reported in Table 2, and we decided to explore a strategy using a
Suzuki-Miyaura coupling reaction that would allow more flexibility
5

M. Garcia, L. Hoffer, R. Leblanc et al.
European Journal of Medicinal Chemistry 223 (2021) 113601
Table 2
2D structure of selected compounds by the virtual growing strategy. Building blocks used to build the compounds are depicted with reactive groups (boronic acid or ester)
leading to aromatic-carbon bonds highlighted in red.
in introducing a large diversity of functionalities. To improve the
coupling, we set up a general procedure starting with a 4-
bromophenyl intermediate (more reactive substrate than 4-
chlorophenyl derivatives depicted in Scheme 1) to generate the
selected library. As previously described, 4-bromothiophenol was
converted to 3-(4-bromophenylthio)propanoic acid 55 by S-alkyl-
ation with 3-chloropropionic acid under alkaline conditions with
excellent yield (Fig. 4A) [26,27]. Compound 55 was condensed with
for inhibitory activity on Syntenin-Syndecan2 complex through
HTRF assay. Compounds 57 and 62 displayed the lowest IC50 values,
47 mM and 7.5 mM respectively (Fig. 4B). Analogue 57 corresponds to
the brominated version of 45 (Br vs Cl, as para substituent) and
analogue 62 contains a phenyl ring substituted in para position
with a methyl mesylate instead of chlorine.
Compounds 45, 57 and 62 were investigated using cellular as-
says. We observed an unusual MCF7 exosomes secretion profile
after treatment with compound 62 suggesting a tendency to
aggregate (see Figure S3 in Supp. Data). For these reasons, IC50
values for compounds 57 and 62 were also evaluated in presence of
0.01% NP40 detergent to eliminate potential false positive due to
aggregation effects in bioassays. IC50 value was not significantly
L
-valine methyl ester hydrochloride in the presence of EDCI yielded
the desired 4-bromophenyl intermediate 56 [28,29]. To compare
the effect of halogens (Br vs Cl, analogue 45) as para substituent,
intermediate 56 was hydrolysed with LiOH to give derivative 57
[25]. Then, we turned to substitution of the bromo group at the C4-
position under standard Suzuki-Miyaura conditions in the presence
of Pd(PPh3)4 under aqueous basic conditions [35]. After few reac-
tion optimization such as performing the reaction with phenyl-
boronic acid and 56 using different bases (sodium, potassium or
cesium carbonate) in various solvents such toluene, THF, 1,4-
dioxane or ethanol, we were able to conclude that use of
Pd(PPh3)4 (5 mol%) and aqueous Na2CO3 in a mixture toluene/
ethanol (4/1) at reflux temperature achieved optimum yield (re-
actions monitored by LC/MS analysis). Furthermore, under these
conditions, we noticed that the cross-coupled product is hydro-
lysed to the corresponding carboxylic acid derivative, probably due
to methyl ester sensitivity to basic hydrolysis. Therefore, interme-
diate 56 reacted readily with boronic acids VS02, VS04e07 and VS-
09 as well as boronic esters VS01, VS03 and VS08 in the presence of
Pd(PPh3)4 and K2CO3 in refluxing aqueous toluene/ethanol mixture
for 18h providing the corresponding carboxylic acid derivatives
58e66 with good to moderate yields (Fig. 4A).
affected by the detergent for compound 57 (51
without and with detergent, respectively). However, IC50 value for
compound 62 increased to 204 M in the presence of NP40
mM and 47 mM,
m
(Fig. 4B). The goal of our growing strategy was to target residues
K214, N215, K250, D251 and S252 residues, because they were
either residues in the nearby environment of original fragment or
residues involved in binding of the PIP2 ligand in 4Z33 ternary
complex. However, the growing stage failed to identify any com-
pound with an improved affinity compared to compound 45
(Fig. 4B). Since no better compound was identified using the first
small growing round and that targeted residues were not predicted
to represent some well-defined pocket, the growing approach was
stopped at this stage. In the end, compounds 45 and 57 are the two
most potent inhibitors in our study.
Subsequently, exosomal fractions released by MCF7 breast
cancer cells treated with compounds 45 and 57 (and their corre-
sponding methyl esters 32 and 56 used as controls) were then
analyzed by western blot, investigating their effect on both
syntenin-dependent and -independent exosomal markers (Fig. 5A).
The amounts of exosomal syntenin were significantly decreased
Despite several attempts in different reaction conditions, inter-
mediate 56 failed to react with the commercial pyrazole boronic
ester VS10 (see Table 2). The nitrogen-rich heterocycle here could
act as a ligand causing side reactions.
when cells were treated with with 100
mM of compound 45, but not
The newly synthetized derivatives 57e66 were then evaluated
with 100 M of C58 [19] or the inactive compound 32 (ester
m
6

M. Garcia, L. Hoffer, R. Leblanc et al.
European Journal of Medicinal Chemistry 223 (2021) 113601
Fig. 4. A. General scheme for synthesis of derivatives 57 and 58e66 via Suzuki-Miyaura coupling. B. Structure, IC50 and LE values of analogues 57e66 on Syntenin-Syndecan2
complex HTRF assay. (a) in parenthesis IC50 and LE values with NP40 0.01%. Additional physicochemical efficiency metrics are shown Table S1 in Supp. Data.
analogue of 45 used as control, Fig. 5A & B). On the other hand,
compound 57 exhibits a tendency to decrease the amounts of
exosomal syntenin, but not its negative control 56 (Fig. 5A & B).
Consistent with our previous study [19], CD63 and the syntenin-
independent exosomal marker Flotillin-1 remained unaffected
regardless of the treatment (Fig. 5A & B). As expected, the different
treatments do not alter the cellular levels (Fig. 5A & B).
4. Experimental section
4.1. Chemistry
All commercial reagents and solvents were purchased from
Sigma Aldrich, Fisher Scientific or Fluorochem companies.
Furthermore, all dry solvents were obtained via Sigma Aldrich with
Sure/Seal™ system and regular solvents were obtained via Sigma
Aldrich at technical grade. Analytical thin layer chromatographiy
(TLC) of the reactions was performed on silica gel 60F 254
aluminium plates (Merck) of 0.2 mm thickness with appropriate
3. Conclusion
solvents. The spots were examined with UV light (
l
¼ 254 nm) and
We established a FBDD process against syntenin-syndecan in-
teractions based on a structure-guided optimization strategy. Initial
fragment hit C58 identified by HTRF screening was an attractive
starting point for inhibitory design. X-ray crystallographic struc-
tures and computational docking were used to guide the fragment
growing and allowed the synthesis of the first selective syntenin-
PDZ2 inhibitors. Investigations into their capacity to control exo-
somal pathways validate syntenin-PDZ2 as a promising therapeutic
target for inhibiting oncogenic processes and could pave the way
for effective therapies for primary tumors and metastasis in the
future.
Ninhydrin Spray. Preparative flash column chromatographies were
performed using silica gel (Merck) G60 230e240 under compressed
air. The 1H NMR and 13C NMR spectra were determined with a
BRUCKER AMX 250 MHz or BRUKER Avance III nanobay 400 MHz.
The chemical shifts are reported in ppm and coupling constants (J)
are reported in hertz. Reaction monitoring and purity of com-
pounds were recorded by using analytical Agilent Infinity high
performance liquid chromatography (Column Zorbax SB-C18
1.8
m
M (2.1 ꢀ 50 mm); Mobile phase (A: 0.1% FA H2O, B: 0.1% FA
CH3CN, Time/%B: 0/10, 4/90, 7/90, 9/10, 10/10); Flow rate 0.3 mL/
7

M. Garcia, L. Hoffer, R. Leblanc et al.
European Journal of Medicinal Chemistry 223 (2021) 113601
Fig. 5. MCF-7 cells were treated with vehicle (Ctrl) or the indicated compound (32, 45, 56 & 57; 100
mM) in medium containing exosome-depleted FCS (10%) for 16h. A. Exosomes
and corresponding total cell lysates were analyzed by western blot, tracing different markers, as indicated. B. Histograms represent mean signal intensities ( SEM) in exosomes,
relative to controls (Ctrl). Data were obtained from N independent experiments, as indicated. Statistical analysis was performed using the one-way analysis of variance (ANOVA)
with a Tukey posttest.
min with DAD at 254 nM. All tested compounds yielded data
consistent with a purity of ꢁ95%. Low-resolution mass spectra were
obtained with Agilent SQ G6120B mass spectrometer in positive
and/or negative electrospray modes. Resolution Mass Spectra
5 mmol) in dry dichloromethane was added dropwise the corre-
sponding acetyl chloride derivative (1e4) (1 eq, 1.0 mmol) at 4 ꢂC.
The mixture was stirred under argon atmosphere at 4 ꢂC during 1 h
then 1e3h at room temperature. After completion, the resulting
mixture was washed twice with 1 N HCl solution, once with water,
once with brine, dried over Na2SO4, and evaporated in vacuo. Crude
was purified by column chromatography eluting with
dichloromethane-methanol (100e98:2).
(HRMS) were obtained on
spectrometer.
a SYNAPT G2-S WATERS mass
4.1.1. General procedure for the synthesis of methyl propanoate and
butanoate intermediates 5e12
To a stirred solution of
(Alanine or Valine) (4 eq, 4.0 mmol) and triethylamine (5 eq,
L
-aminoacid methyl ester hydrochloride
4.1.1.1. (S)-methyl
2-(2-(phenylthio)acetamido)propanoate
5.
1H NMR (CDCl3, 250 MHz):
d
¼ 7.29e7.09 (m, 6H), 4.48 (p,
8

M. Garcia, L. Hoffer, R. Leblanc et al.
European Journal of Medicinal Chemistry 223 (2021) 113601
J ¼ 7.2 Hz,1H), 3.63 (s, 3H), 3.56 (d, J ¼ 0.9 Hz, 2H),1.25 (d, J ¼ 7.2 Hz,
3H). LC/MS (ESI): 252.5 [M ꢃ H]-, 254.4 [MþH]þ; light brown oil.
Yield ¼ 92%.
4.1.2. General procedure for the synthesis of amino acid derivatives
13e20, 41e53 and 57
To a stirred solution of ester intermediate derivative (5e12,
28e40 or 56) (1eq, 0.5 mmol) in THF at 0 ꢂC, aqueous solution of
LiOH (10 eq, 5.0 mmol) was added dropwise. The resulting mixture
was stirred 1 h at room temperature. After completion, THF was
removed in vacuo. Then, 1 N HCl solution was added dropwise until
pH 2e3 and the aqueous layer was extracted with ethyl acetate. The
organic layer was washed once with water, once with brine, dried
over Na2SO4, and evaporated in vacuo to afford either a crude
residue which does not requires further purification or a crude
which was purified by column chromatography eluting with
dichloromethane-methanol (95:5).
4.1.1.2. (S)-methyl 3-methyl-2-(2-(phenylthio)acetamido)butanoate
6. 1H NMR (DMSO‑d6, 400 MHz):
d
¼ 8.39 (bd, J ¼ 8.3 Hz, 1H),
7.39e7.35 (m, 2H), 7.34e7.26 (m, 2H), 7.23e7.15 (m, 1H), 4.49 (dd,
J ¼ 8.3 Hz and 6.2 Hz, 1H), 3.80 (d, J ¼ 14.5 Hz, 1H), 3.70 (d,
J ¼ 14.5 Hz, 1H), 3.63 (s, 3H), 2.02 (m, 1H), 0.85 (d, J ¼ 6.8 Hz, 3H),
0.82 (d, J ¼ 6.8 Hz, 3H). 13C NMR (DMSO‑d6, 101 MHz):
d
¼ 172.24,
168.76, 136.54, 129.36, 128.30, 126.25, 58.01, 52.21, 36.23, 30.48,
19.33, 18.47; LC/MS (ESI): 280.5 [M ꢃ H]-, 282.4 [MþH]þ; light
yellow oil. Yield ¼ 98%.
4.1.2.1. (S)-2-(2-(phenylthio)acetamido)propanoic acid 13. 1H NMR
4.1.1.3. (S)-methyl
2-(2-(4-
chlorophenylthio)acetaamido)prop-
¼ 8.58 (bd, J ¼ 7.0 Hz, 1H),
(DMSO‑d6, 250 MHz):
d
¼ 12.69 (bs, 1H), 8.49 (d, J ¼ 7.2 Hz, 1H),
amoate 7. 1H NMR (CDCl3, 400 MHz):
d
7.42e7.26 (m, 4H), 7.19 (t, J ¼ 7.0 Hz,1H), 4.24 (p, J ¼ 7.1 Hz,1H), 3.72
(s, 2H), 1.28 (d, J ¼ 7.3 Hz, 3H). 13C NMR (DMSO‑d6, 63 MHz):
7.38 (bs, 4H), 4.27 (p, J ¼ 7.2 Hz,1H), 3.69 (s, 2H), 3.61 (s, 3H),1.26 (d,
J ¼ 7.3 Hz, 3H). 13C NMR (CDCl3, 101 MHz):
d
¼ 173.21, 168.07,
d
¼ 173.88, 167.63, 136.15, 128.86, 127.82, 125.73, 47.76, 36.02, 17.19;
LC/MS (ESI): 238.6 [M
ꢃ
H]-, 240.4 [MþH]þ; beige solid.
135.65, 131.02, 130.22, 129.25, 52.37, 48.30, 36.46, 17.42; LC/MS
(ESI): 286.5 [M ꢃ H]-, 288.1 [MþH]þ; white solid. Yield ¼ 69%.
Yield ¼ 77%.
4.1.2.2. (S)-3-methyl-2-(2-(phenylthio)acetamido)butanoic acid 14.
1H NMR (DMSO‑d6, 300 MHz):
4.1.1.4. (S)-methyl 2-(2-(4-chlorophenylthio)acetamido)-3-
methylbutanoate 8. 1H NMR (DMSO‑d6, 400 MHz):
d
¼ 7.21 (bs,
d
¼ 12.66 (bs, 1H), 8.19 (d, J ¼ 8.6 Hz,
1H), 7.39e7.35 (m, 2H), 7.3e7.25 (m, 2H), 7.22e7.14 (m, 1H), 4.16
(dd, J ¼ 8.6 Hz and 5.5 Hz, 1H), 3.81 (d, J ¼ 14.5 Hz, 1H), 3.71 (d,
J ¼ 14.5 Hz, 1H), 2.03 (m, 1H), 0.84 (d, J ¼ 6.4 Hz, 3H), 0.82 (d,
4H), 7.09 (bd, J ¼ 8.4 Hz, 1H), 4.43 (dd, J ¼ 8.9 Hz and 4.8 Hz, 1H),
3.65 (d, J ¼ 16.8 Hz, 1H), 3.63 (s, 3H), 3.55 (d, J ¼ 16.8 Hz, 1H),
2.11e1.99 (m, 1H), 0.74 (d, J ¼ 6.8 Hz, 3H), 0.73 (d, J ¼ 6.9 Hz, 3H). 13
C
J ¼ 6.4 Hz, 3H). 13C NMR (DMSO‑d6, 75 MHz):
d
¼ 173.20, 168.52,
NMR (DMSO‑d6, 101 MHz):
d
¼ 170.87, 166.40, 131.98, 131.94,
136.62, 129.34, 128.27, 126.18, 57.82, 36.37, 30.47, 19.47, 18.23; LC/
128.90, 128.38, 56.33, 51.18, 36.68, 30.15, 17.79, 16.53; LC/MS (ESI):
MS (ESI): 266.6 [M ꢃ H]-, 268.3 [MþH]þ; beige solid. Yield ¼ 88%.
314.6 [M ꢃ H]-, 316.2 [MþH]þ; colorless oil. Yield ¼ 90%.
4.1.2.3. (S)-2-(2-(4- chlorophenylthio)acetaamido)propamoic acid
4.1.1.5. (S)-methyl 2-(2-phenoxyacetamido)propanoate 9. 1H NMR
15. 1H NMR (DMSO‑d6, 300 MHz):
d
¼ 12.56 (bs, 1H), 8.43 (d,
(DMSO‑d6, 400 MHz):
d
¼ 8.50 (bd, J ¼ 7.3 Hz, 1H), 7.34e7.26 (m,
J ¼ 7.2 Hz, 1H), 7.39 (d, J ¼ 8.9 Hz, 2H), 7.34 (d, J ¼ 8.8 Hz, 2H), 4.20
(p, J ¼ 7.3 Hz, 1H), 3.69 (s, 3H), 1.25 (d, J ¼ 7.3 Hz, 3H). 13C NMR
2H), 7.00e6.94 (m, 3H), 4.54 (d, J ¼ 14.8 Hz, 1H), 4.50 (d, J ¼ 14.8 Hz,
1H), 4.39 (p, J ¼ 7.3 Hz, 1H), 3.64 (s, 3H), 1.33 (d, J ¼ 7.3 Hz, 3H). 13
C
(DMSO‑d6, 75 MHz):
d
¼ 174.24, 167.85, 135.75, 130.96, 130.20,
129.22, 48.28, 36.59, 17.70; LC/MS (ESI): 272.5 [M ꢃ H]-, 274.1
NMR (DMSO‑d6, 101 MHz):
d
¼ 173.18, 168.13, 158.21, 129.91, 121.64,
115.19, 67.08, 52.40, 47.76, 17.32; LC/MS (ESI): 236.6 [M ꢃ H]-, 238.5
[MþH]þ; colorless oil. Yield ¼ 74%.
[MþH]þ; white solid. Yield ¼ 78%.
4.1.2.4. (S)-2-(2-(4-chlorophenylthio)acetamido)-3-methylbutanoic
acid 16. 1H NMR (DMSO‑d6, 300 MHz):
d
¼ 12.04 (bs, 1H), 8.06 (d,
4.1.1.6. (S)-methyl 3-methyl-2-(2-phenoxyacetamido)butanoate 10.
1H NMR (DMSO‑d6, 400 MHz):
d
¼ 8.28 (bd, J ¼ 8.3 Hz, 1H),
J ¼ 8.5 Hz, 1H), 7.21 (d, J ¼ 8.6 Hz, 2H), 7.15 (d, J ¼ 8.7 Hz, 2H), 3.96
(dd, J ¼ 8.5 Hz and 5.5 Hz, 1H), 3.64 (d, J ¼ 14.5 Hz, 1H), 3.53 (d,
J ¼ 14.5 Hz, 1H), 1.98e1.70 (m, 1H), 0.85e0.53 (m, 6H). 13C NMR
7.35e7.25 (m, 2H), 6.95 (m, 3H), 4.62 (d, J ¼ 14.7 Hz, 1H), 4.58 (d,
J ¼ 14.7 Hz, 1H), 4.43 (dd, J ¼ 8.3 Hz and 6.4 Hz, 1H), 3.65 (s, 3H),
2.16e1.98 (m, 1H), 0.88 (d, J ¼ 6.8 Hz, 3H), 0.86 (d, J ¼ 6.8 Hz, 3H).
(DMSO‑d6, 75 MHz):
d
¼ 172.63, 167.91, 135.26, 130.39, 129.58,
128.71, 57.31, 35.85, 29.92, 18.96, 17.74; LC/MS (ESI): 300.1 [M ꢃ H]-,
302.1 [MþH]þ; HRMS (TOF, ESI-) cald for C13H15NO3SCl [M ꢃ H]-
300.0467, found 360.0462; LC/MS (ESI): 300.1 [M ꢃ H]-, 302.1
[MþH]þ; white solid. Yield ¼ 96%.
13C NMR (DMSO‑d6, 101 MHz):
d
¼ 172.22, 168.50, 158.27, 129.90,
121.54, 115.03, 66.80, 57.60, 52.25, 30.30, 19.42, 18.67; LC/MS (ESI):
264.7 [M ꢃ H]-, 266.5 [MþH]þ; colorless oil. Yield ¼ 63%.
4.1.1.7. (S)-methyl 2-(2-(4-chlorophenoxy)acetamido)propanoate 11.
4.1.2.5. (S)-2-(2-phenoxyacetamido)propanoic acid 17. 1H NMR
1H NMR (CDCl3, 250 MHz):
d
¼ 7.20 (d, J ¼ 9.1 Hz, 2H), 6.99 (bd,
(DMSO‑d6, 300 MHz):
d
¼ 12.62 (bs, 1H), 8.27 (d, J ¼ 7.4 Hz, 1H),
J ¼ 6.7 Hz,1H), 6.80 (d, J ¼ 9.1 Hz, 2H), 4.50 (m,1H), 4.40 (s, 2H), 3.69
(s, 3H), 1.38 (d, J ¼ 7.2 Hz, 3H). LC/MS (ESI): 270.5 [M ꢃ H]-, 272.2
[MþH]þ; colorless oil. Yield ¼ 89%.
7.34e7.26 (m, 2H), 7.02e6.93 (m, 3H), 4.48 (s, 2H), 4.31 (p,
J ¼ 7.3 Hz, 1H), 1.31 (d, J ¼ 7.3 Hz, 3H). 13C NMR (DMSO‑d6, 75 MHz):
d
¼ 174.21, 167.88, 158.23, 129.90, 121.63, 115.22, 67.17, 47.72, 17.61;
LC/MS (ESI): 222.8 [M
ꢃ
H]-, 224.4 [MþH]þ; white solid.
4.1.1.8. (S)-methyl 2-(2-(4-chlorophenoxy)acetamido)-3-
Yield ¼ 78%.
methylbutanoate 12. 1H NMR (DMSO‑d6, 400 MHz):
d
¼ 8.33 (bd,
J ¼ 8.2 Hz, 1H), 7.34 (d, J ¼ 9.1 Hz, 2H), 6.96 (d, J ¼ 9.1 Hz, 2H), 4.64
(d, J ¼ 14.7 Hz, 1H), 4.59 (d, J ¼ 14.7 Hz, 1H), 4.23 (dd, J ¼ 8.2 Hz and
6.4 Hz, 1H), 3.65 (s, 3H), 2.14e1.98 (m, 1H), 0.88 (d, J ¼ 6.8 Hz, 3H),
4.1.2.6. (S)-3-methyl-2-(2-phenoxyacetamido)butanoic acid 18.
1H NMR (DMSO‑d6, 300 MHz):
d
¼ 12.68 (bs, 1H), 7.99 (d, J ¼ 8.4 Hz,
1H), 7.35e7.24 (m, 2H), 7.00e6.91 (m, 3H), 4.59 (s, 2H), 4.22 (dd,
J ¼ 8.5 Hz and 5.7 Hz, 1H), 2.10 (m, 1H), 0.88 (d, J ¼ 6.7 Hz, 3H), 0.86
0.86 (d, J ¼ 6.8 Hz, 3H). 13C NMR (DMSO‑d6, 101 MHz):
d
¼ 172.19,
168.21, 157.19, 129.63, 125.24, 116.86, 67.07, 57.65, 52.25, 30.27,
19.42, 18.70; LC/MS (ESI): 298.5 [M ꢃ H]-, 300.1 [MþH]þ; colorless
oil. Yield ¼ 93%.
(d, J ¼ 6.7 Hz, 3H). 13C NMR (DMSO‑d6, 75 MHz):
d
¼ 173.11, 168.25,
158.26, 129.91, 121.57, 115.08, 66.95, 57.33, 30.38, 19.54, 18.37; LC/
MS (ESI): 250.7 [M ꢃ H]-, 252.4 [MþH]þ; white solid. Yield ¼ 45%.
9

M. Garcia, L. Hoffer, R. Leblanc et al.
European Journal of Medicinal Chemistry 223 (2021) 113601
4.1.2.7. (S)-2-(2-(4-chlorophenoxy)acetamido)propanoic acid 19.
[MþH]þ 316.0774, found 316.0776; white solid. Yield ¼ 89%.
1H NMR (DMSO‑d6, 250 MHz):
d
¼ 12.72 (bs, 1H), 8.40 (d, J ¼ 7.3 Hz,
1H), 7.35 (d, J ¼ 8.7 Hz, 2H), 7.01 (d, J ¼ 8.8 Hz, 2H), 4.42 (s, 2H), 4.32
4.1.2.14. (S)-2-(3-(4-cholorophenylthio)propanamido)-3-
(p, J ¼ 7.0 Hz, 1H), 1.34 (d, J ¼ 7.2 Hz, 3H). 13C NMR (DMSO‑d6,
phenylpropamoic acid 46. 1H NMR (DMSO‑d6, 250 MHz):
d
¼ 12.82
63 MHz):
d
¼ 171.33, 164.71, 154.09, 126.68, 122.38, 113.99, 64.34,
(bs, 1H), 8.34 (d, J ¼ 8.1 Hz, 1H), 7.43 (d, J ¼ 8.5 Hz, 2H), 7.35 (d,
J ¼ 8.7 Hz, 2H), 7.33e7.23 (m, 5H), 4.48 (m, 1H), 3.09 (m, 3H), 2.88
(dd, J ¼ 8.4 Hz and 6.0 Hz, 1H), 2.44 (t, J ¼ 7.2 Hz, 2H); 13C NMR
44.79, 14.56; LC/MS (ESI): 256.5 [M ꢃ H]-, 258.3 [MþH]þ; white
solid. Yield ¼ 95%.
(DMSO‑d6, 63 MHz):
d
¼ 172.66, 169.87, 137.60, 135.18, 130.58,
4.1.2.8. (S)-2-(2-(4-chlorophenoxy)acetamido)-3-methylbutanoic
130.19,129.05,128.88,128.06,126.30, 53.38, 36.98, 34.85, 28.59; LC/
acid 20. 1H NMR (CDCl3, 300 MHz):
d
¼ 7.21 (d, J ¼ 9.0 Hz, 2H), 6.92
MS (ESI): 362.6 [M ꢃ H]-, 364.1 [MþH]þ; HRMS (TOF, ESIþ) cald for
C
18H19NO3SCl [MþH]þ 364.0774, found 364.0776; white solid.
(bd, J ¼ 8.8 Hz, 1H), 6.82 (d, J ¼ 8.9 Hz, 2H), 4.23 (dd, J ¼ 8.8 Hz and
4.7 Hz, 1H), 4.47 (s, 1H), 2.27e2.13 (m, 1H), 0.91 (d, J ¼ 6.9 Hz, 3H),
Yield ¼ 97%.
0.87 (d, J ¼ 6.9 Hz, 3H). 13C NMR (CDCl3, 101 MHz):
d
¼ 174.28,
167.45, 154.65, 128.71, 126.37, 115.14, 66.53, 55.61, 29.97, 17.99,
16.54; LC/MS (ESI): 284.5 [M ꢃ H]-, 285.9 [MþH]þ; white solid.
Yield ¼ 85%.
4.1.2.15. (S)-2-(3-(phenoxypropanamido)propanoic acid 47. 1H NMR
(DMSO‑d6, 250 MHz):
d
¼ 12.63 (bs, 1H), 8.35 (d, J ¼ 7.3 Hz, 1H),
7.38e7.17 (m, 2H), 7.02e6.84 (m, 3H), 4.27 (p, J ¼ 5.8 Hz, 1H), 4.18
(td, J ¼ 6.2 Hz and 2.8 Hz, 2H), 2.62 (td, J ¼ 6.2 Hz and 1.2 Hz, 2H),
4.1.2.9. (S)-2-(3-phenylthio)propanamido)propamoic
1H NMR (DMSO‑d6, 250 MHz):
acid
41.
1.30 (d, J ¼ 7.3 Hz, 3H). 13C NMR (DMSO‑d6, 63 MHz):
d
¼
13C NMR
d
¼ 12.59 (bs, 1H), 8.31 (d, J ¼ 7.2 Hz,
(DMSO‑d6, 63 MHz)
d
¼ 173.92, 169.26, 158.48, 130.20, 129.39,
1H), 7.39 (bd, J ¼ 4.3 Hz, 4H), 7.26 (dd, J ¼ 8.7 Hz and J ¼ 4.3 Hz, 1H),
120.57, 116.00, 114.65, 110.85, 64.01, 47.52, 35.18, 17.35; LC/MS (ESI):
4.25 (p, J ¼ 7.3 Hz, 1H), 3.18 (t, J ¼ 7.3 Hz, 2H), 2.50 (t, J ¼ 7.3 Hz, 2H),
236.6 [M ꢃ H]-, 238.7 [MþH]þ; HRMS (TOF, ESIþ) cald for
1.30 (d, J ¼ 7.3 Hz, 3H); 13C NMR (DMSO‑d6, 101 MHz):
d
¼ 174.59,
C
12H16NO4 [MþH]þ 238.1074, found 238.1079; colorless oil.
170.34, 136.45, 129.55, 128.64, 126.20, 47.94, 35.18, 28.44, 17.69; LC/
MS (ESI): 252.6 [M ꢃ H]-, 254.3 [MþH]þ; HRMS (TOF, ESIþ) cald for
Yield ¼ 90%.
C
12H16NO3S [MþH]þ 254.0851, found 254.0851; white powder.
Yield ¼ 93%.
4.1.2.16. (S)-3-methyl-2-(3-(phenoxypropanamido)butanoic acid 48.
1H NMR (DMSO‑d6, 250 MHz):
d
¼ 12.70 (bs, 1H), 8.19 (d, J ¼ 8.6 Hz,
4.1.2.10. (S)-3-methyl-2-(3-phenylthio)propanamido)butanoic acid
1H), 7.38e7.20 (m, 2H), 7.00e6.84 (m, 3H), 4.21 (m, 1H), 4.18 (t,
J ¼ 5.8 Hz, 2H), 2.82e2.58 (m, 2H), 2.16e1.99 (m, 1H), 0.91 (d,
42. 1H NMR (DMSO‑d6, 250 MHz):
d
¼ 12.67 (bs, 1H), 8.17 (d,
J ¼ 8.6 Hz, 1H), 7.39 (bd, J ¼ 4.5 Hz, 4H), 7.26 (dd, J ¼ 8.6 Hz and
J ¼ 4.4 Hz,1H), 4.23 (dd, J ¼ 8.5 and J ¼ 5.9 Hz,1H), 3.19 (t, J ¼ 7.2 Hz,
2H), 2.60 (t, J ¼ 7.2 Hz, 2H), 2.16e2.01 (m, 1H), 0.93 (bd, J ¼ 5.7 Hz,
J ¼ 6.7 Hz, 6H). 13C NMR (DMSO‑d6, 63 MHz):
d
¼
13C NMR (63 MHz,
DMSO)
d
¼ 172.82, 169.80, 158.49, 129.39, 120.55, 114.60, 64.16,
57.18, 35.23, 29.93, 19.04, 18.00; LC/MS (ESI): 264.4 [M ꢃ H]-, 266.3
[MþH]þ; HRMS (TOF, ESIþ) cald for C14H20NO4 [MþH]þ 266.1387,
found 266.1395; colorless oil. Yield ¼ 76%.
6H); 13C NMR (DMSO‑d6, 63 MHz):
d
¼ 172.75, 170.32, 136.16,
128.97,128.56,125.78, 57.28, 34.97, 29.88, 28.67,19.04,18.05; LC/MS
(ESI): 280.6 [M ꢃ H]-, 282.3 [MþH]þ; HRMS (TOF, ESIþ) cald for
C
14H20NO3S [MþH]þ 282.1164, found 282.1165; white solid.
4.1.2.17. (S)-2-(3-(phenoxypropanamido)-3-phenylpropanoic
acid
¼ 12.83 (bs, 1H), 8.37 (d,
49. 1H NMR (DMSO‑d6, 250 MHz):
d
Yield ¼ 96%.
J ¼ 8.0 Hz, 1H), 7.36e7.18 (m, 7H), 7.08e6.83 (m, 3H), 4.48 (m, 1H),
4.10 (t, J ¼ 6.2 Hz, 2H), 3.09 (dd, J ¼ 13.8 Hz and 4.9 Hz,1H), 2.90 (dd,
J ¼ 13.8 Hz and 9.3 Hz, 1H), 2.58 (t, J ¼ 6.1 Hz, 2H). 13C NMR
4.1.2.11. (S)-3-phenyl-2-(3-phenylthio)propanamido)propamoic acid
43. 1H NMR (DMSO‑d6, 250 MHz):
d
¼ 12.72 (bs, 1H), 8.24 (d,
(DMSO‑d6, 63 MHz):
d
¼
13C NMR (DMSO‑d6, 63 MHz)
d
¼ 172.73,
J ¼ 8.1 Hz, 1H), 7.31e7.10 (m, 10H), 4.37 (td, J ¼ 9.2 Hz and J ¼ 5.0 Hz,
1H), 3.04e2.93 (m, 3H), 2.78 (dd, J ¼ 13.7 Hz and J ¼ 9.4 Hz, 1H),
169.45, 158.42, 137.61, 129.37, 129.06, 128.06, 126.30, 120.56, 114.60,
63.97, 53.39, 36.92, 35.26; LC/MS (ESI): 312.5 [M ꢃ H]-, 314.2
[MþH]þ; HRMS (TOF, ESIþ) cald for C18H20N2O4 [MþH]þ 314.1387,
found 314.1390; colorless oil. Yield ¼ 75%.
2.34 (t, J ¼ 7.3 Hz, 2H); 13C NMR (DMSO‑d6, 63 MHz):
d
¼ 172.67,
169.98,137.62,136.04,129.05,128.97,128.53,128.05,126.29,125.79,
53.39, 36.96, 35.04, 28.46; LC/MS (ESI): 328.5 [M ꢃ H]-, 330.3
[MþH]þ; HRMS (TOF, ESIþ) cald for C18H20NO3S [MþH]þ 330.1164,
found 330.1165; white solid. Yield ¼ 49%.
4.1.2.18. (S)-2-(3-(4-chlorophenoxy)propanamido)propanoic acid 50.
1H NMR (DMSO‑d6, 250 MHz):
d
¼ 8.07 (d, J ¼ 7.2 Hz, 1H), 7.13 (d,
4.1.2.12. 1 2 12. (S)-2-(3-(4-cholorophenylthio)propanamido)prop-
amoic acid 44. 1H NMR (CDCl3, 250 MHz):
d
¼ 7.28e7.14 (m, 4H),
J ¼ 9.0 Hz, 2H), 6.76 (d, J ¼ 9.0 Hz, 2H), 4.08e3.92 (m, 3H),
4.00e2.40 (t, J ¼ 6.1 Hz, 2H), 1.07 (d, J ¼ 7.3 Hz, 3H); 13C NMR
6.17 (bd, J ¼ 7.0 Hz,1H), 4.51 (p J ¼ 7.1 Hz,1H), 3.12 (t, J ¼ 7.2 Hz, 2H),
2.45 (t, J ¼ 7.2 Hz, 2H), 1.38 (d, J ¼ 7.2 Hz, 3H); 13C NMR (CDCl3,
(DMSO‑d6, 63 MHz):
d
¼ 174.19, 169.01, 157.36, 129.15, 124.36,
116.44, 64.56, 47.79, 35.10, 17.55; LC/MS (ESI): 270.5 [M ꢃ H]-, 272.3
[MþH]þ; HRMS (TOF, ESIþ) cald for C12H15NO4Cl [MþH]þ
272.0690, found 272.0689; white solid. Yield ¼ 52%.
63 MHz):
d
¼ 207.45, 170.05, 133.02, 131.95, 130.52, 128.30, 47.29,
35.12, 28.86, 17.02; LC/MS (ESI): 286.5 [M ꢃ H]-, 288.1 [MþH]þ;
HRMS (TOF, ESIþ) cald for C12H15NO3SCl [MþH]þ 288.0461, found
288.0461; white solid. Yield ¼ 85%.
4.1.2.19. (S)-2-(3-(4-chlorophenoxy)propanamido)-3-
4.1.2.13. (S)-2-(3-(4-chlorophenylthio)propanamido)-3-
methylbutanoic acid 51. 1H NMR (CDCl3, 250 MHz):
d
¼ 7.16 (d,
methylbutanoic acid 45. 1H NMR (DMSO‑d6, 300 MHz):
d
¼ 12.51
J ¼ 7.6 Hz, 2H), 6.77 (d, J ¼ 7.6 Hz, 2H), 6.40 (d, J ¼ 8.5 Hz, 1H), 4.55
(dd, J ¼ 8.6 Hz and J ¼ 4.7 Hz, 1H), 4.17 (t, J ¼ 5.6 Hz, 2H), 2.67 (t,
J ¼ 5.7 Hz, 2H), 2.19 (m,1H), 0.94e0.85 (m, 6H); 13C NMR (DMSO‑d6,
(bs, 2H), 8.04 (d, J ¼ 8.5 Hz, 1H), 7.38 (d, J ¼ 9.2 Hz, 2H), 7.34 (d,
J ¼ 9.3 Hz, 2H), 4.16 (dd, J ¼ 8.4 Hz and 6.0 Hz, 1H), 3.14 (t, J ¼ 7.1 Hz,
2H), 2.54 (t, J ¼ 7.0 Hz, 2H), 2.03 (m, 1H), 0.87 (bd, J ¼ 6.6 Hz, 6H);
63 MHz):
d
¼ 172.81, 169.68, 157.34, 129.15, 124.34, 116.37, 64.66,
13C NMR (DMSO‑d6, 75 MHz)
d
173.41,170.71,135.69,130.86,130.42,
57.17, 35.07, 29.91, 19.03, 17.99; LC/MS (ESI): 298,5 [M ꢃ H]-, 300,1
[MþH]þ; HRMS (TOF, ESIþ) cald for C14H19NO4Cl [MþH]þ 300.1003,
found 300.1003; white solid. Yield ¼ 74%.
129.40, 57.62, 34.99, 30.34, 28.91, 19.58, 18.48. LC/MS (ESI): 314.6
[M ꢃ H]-, 316.2 [MþH]þ; HRMS (TOF, ESIþ) cald for C14H19NO3SCl
10

M. Garcia, L. Hoffer, R. Leblanc et al.
European Journal of Medicinal Chemistry 223 (2021) 113601
4.1.2.20. (S)-2-(3-(4-chlorophenoxy)propanamido)-3-
1.0
mmol)
in
dry
dichloromethane,
hydrochloride
N-Ethyl-N0-(3-
(2 eq,
phenylpropanoic acid 52. 1H NMR (DMSO‑d6, 250 MHz):
d
¼ 12.85
dimethylaminopropyl)carbodiimide
(bs, 1H), 8.41 (d, J ¼ 8.0 Hz, 1H), 7.40e7.26 (m, 7H), 6.94 (d,
J ¼ 9.0 Hz, 2H), 4.51 (m, 1H), 4.13 (t, J ¼ 6.0 Hz, 2H), 3.12 (dd,
J ¼ 13.7 Hz and J ¼ 4.8 Hz, 1H), 2.93 (dd, J ¼ 13.7 Hz and J ¼ 9.3 Hz,
1H), 2.60 (t, J ¼ 6.0 Hz, 2H); 13C NMR (DMSO‑d6, 63 MHz):
2.0 mmol) and triethylamine (2 eq, 2 mmol) were added. The
mixture was stirred under argon atmosphere during 5e16 h at
room temperature. After completion, the resulting mixture was
washed twice with 1 N HCl solution, once with water, once with
brine, dried over Na2SO4, and evaporated in vacuo. Crude was pu-
rified by column chromatography eluting with dichloromethane-
methanol (100e98:2).
d
¼ 172.72, 169.33, 157.28, 137.61, 129.13, 129.05, 128.06, 126.30,
124.35, 116.38, 64.49, 53.40, 36.91, 35.11; LC/MS (ESI): 346,6
[M ꢃ H]-, 348,3 [MþH]þ; HRMS (TOF, ESIþ) cald for C18H19NO4Cl
[MþH]þ 348.1003, found 348.1006; white solid. Yield ¼ 94%.
4.1.4.1. (S)-methyl 2-(3-(phenylthio)propanamido)propanoate 28.
1H NMR (DMSO‑d6, 300 MHz):
d
¼ 8.32 (bd, J ¼ 6.8 Hz, 1H), 7.33 (d,
4.1.2.21. (S)-3-methyl-2-(4-(phenylbutanamido)butanoic acid 53.
1H NMR (DMSO‑d6, 250 MHz):
d
¼ 12.65 (bs, 1H), 8.03 (d, J ¼ 8.3 Hz,
J ¼ 4.3 Hz, 4H), 7.25e7.14 (m, 1H), 4.27 (p, J ¼ 7.2 Hz, 1H), 3.62 (s,
3H), 3.13 (t, J ¼ 7.3 Hz, 2H), 2.45 (t, J ¼ 7.3 Hz, 2H), 1.26 (d, J ¼ 7.3 Hz,
1H), 7.36e7.26 (m, 2H), 7.28e7.18 (m, 3H), 4.20 (dd, J ¼ 8.1 Hz and
J ¼ 6.0 Hz,1H), 2.61 (t, J ¼ 7.5 Hz, 2H), 2.25 (t, J ¼ 7.3 Hz, 2H), 2.11 (m,
1H), 1.92e1.74 (m, 2H), 0.93 (bd, J ¼ 6.7 Hz, 6H); 13C NMR
3H). 13C NMR (DMSO‑d6, 75 MHz):
d 173.52, 170.50, 136.41, 129.54,
128.76, 126.24, 52.26, 48.02, 35.16, 28.51, 17.45; LC/MS (ESI): 266.5
[M ꢃ H]-, 268.4 [MþH]þ; colorless oil. Yield ¼ 45%.
(DMSO‑d6,63 MHz):
d
¼
d 173.01, 172.18, 141.87, 128.24, 128.19,
125.64, 57.20, 34.66, 34.60, 29.72, 27.14, 19.10, 18.10; LC/MS (ESI):
262.6 [M
C
Yield ¼ 92%.
ꢃ
H]-, 264.4[MþH]þ; HRMS (TOF, ESIþ) cald for
4.1.4.2. (S)-methyl 3-methyl-2-(3-(phenylthio)propanamido)buta-
15H22NO3 [MþH]þ 264.1594, found 264.1590; white solid.
noate 29. 1H NMR (DMSO‑d6, 400 MHz):
d
¼ 8.24 (bd, J ¼ 8.2 Hz,
1H), 7.33 (d, J ¼ 4.3 Hz, 4H), 7.24e7.17 (m, 1H), 4.27 (dd, J ¼ 8.2 Hz
and 6.4 Hz,1H), 3.63 (s, 3H), 3.13 (t, J ¼ 7.2 Hz, 2H), 2.46 (t, J ¼ 7.3 Hz,
2H), 2.07e1.90 (m, 1H), 0.88 (d, J ¼ 6.8 Hz, 3H), 0.86 (d, J ¼ 6.8 Hz,
4.1.2.22. (S)-2-(3-(4-bromophenylthio)propanamido)-3-
methylbutanoic acid 57. 1H NMR (DMSO‑d6, 250 MHz):
3H). 13C NMR (DMSO‑d6, 101 MHz):
d¼172.54, 170.99, 136.47,
d
¼ 12.67
(bs, 2H), 8.16 (d, J ¼ 8.6 Hz, 1H), 7.56 (d, J ¼ 8.6 Hz, 2H), 7.33 (d,
J ¼ 8.6 Hz, 2H), 4.21 (dd, J ¼ 8.6 Hz and 5.8 Hz,1H), 3.20 (t, J ¼ 7.1 Hz,
2H), 2.59 (t, J ¼ 7.0 Hz, 2H), 2.07 (m,1H), 0.92 (d, J ¼ 6.8 Hz, 3H), 0.93
129.55, 128.70, 126.22, 57.86, 52.10, 35.03, 30.38, 29.32, 28.62,
19.43, 18.73; LC/MS (ESI): 294.6 [M ꢃ H]-, 296.1 [MþH]þ; colorless
oil. Yield ¼ 55%.
(d, J ¼ 6.8 Hz, 3H). 13C NMR (DMSO‑d6, 63 MHz):
¼ 172.96, 170.16,
d
135.71,131.74, 129.96, 118.46, 57.04, 34.30, 29.80, 28.05, 19.07,17.92;
LC/MS (ESI): 358.8 [M ꢃ H]- and isotopic peak 360.8, 359.8 [MþH]þ
and isotopic peak 361.9; HRMS (TOF, ESIþ) cald for C14H19NO3SBr
[MþH]þ 360.0269, found 360.0272; white solid. Yield ¼ 96%.
4.1.4.3. (S)-methyl
3-phenyl-2-(3-(phenylthio)propanamido)prop-
¼ 8.38 (bd, J ¼ 7.7 Hz,
anoate 30. 1H NMR (DMSO‑d6, 300 MHz):
d
1H), 7.35e7.18 (m, 10H), 4.49 (m, 1H), 3.60 (s, 3H), 3.06e2.99 (m,
3H), 2.88 (dd, J ¼ 13.7 Hz and 9.1 Hz, 1H), 2.41 (t, J ¼ 7.3 Hz, 2H). 13
C
NMR (DMSO‑d6, 75 MHz):
d
¼ 172.41, 170.65, 137.62, 136.37, 129.53,
4.1.3. General procedure for the synthesis of propanoic acid
derivatives 23, 24 and 55
128.73, 128.67, 126.98, 126.23, 54.01, 52.26, 37.22, 35.16, 28.51; LC/
MS (ESI): 342.5 [M ꢃ H]-, 344.2 [MþH]þ; white solid. Yield ¼ 71%.
To a stirred solution of commercial thiophenol derivative (1 eq,
5.0 mmol) and sodium hydroxide (2.4 eq, 12.0 mmol) in water,
aqueous solution of 3-chloropropionic acid was added dropwise
(1.7 eq, 8.5 mmol). The mixture was refluxed during 2e16 h. After
completion, 1 N HCl solution was added dropwise at 4 ꢂC until pH 1,
to afford a crude precipitate which was filtered and dried. The crude
residue does not require further purification.
4.1.4.4. (S)-methyl 2-(3-(4- chlorophenylthio)propanamido)prop-
amoate 31. 1H NMR (DMSO‑d6, 300 MHz):
d
¼ 8.32 (bd, J ¼ 6.8 Hz,
1H), 7.39 (d, J ¼ 8.9 Hz, 2H), 7.35 (d, J ¼ 9.0 Hz, 2H), 4.26 (p,
J ¼ 7.3 Hz, 1H), 3.62 (s, 3H), 3.13 (t, J ¼ 7.2 Hz, 2H), 2.44 (t, J ¼ 7.2 Hz,
2H), 1.25 (d, J ¼ 7.3 Hz, 3H). 13C NMR (DMSO‑d6, 75 MHz):
d
¼ 173.49, 170.38, 135.57, 130.90, 130.44, 129.42, 52.26, 48.01,
34.97, 28.63, 17.46; LC/MS (ESI): 301.1 [MþH]þ; white solid.
Yield ¼ 41%.
4.1.3.1. 3-(phenylthio)propanoic acid 23. 1H NMR (CDCl3, 250 MHz):
d
7.39e7.07 (m, 5H), 3.09 (t, J ¼ 7.3 Hz, 2H), 2.61 (t, J ¼ 7.3 Hz, 2H).
LC/MS (ESI): 181.6 [M ꢃ H]-, 183.6 [MþH]þ; white powder.
Yield ¼ 94%.
4.1.4.5. (S)-methyl 2-(3-(4-chlorophenylthio)propanamido)-3-
methylbutanoate 32. 1H NMR (DMSO‑d6, 300 MHz):
d
¼ 8.19 (bd,
4.1.3.2. 3-(4-chlorophenylthio)propanoic acid 24. 1H NMR (CDCl3,
J ¼ 8.1 Hz, 1H), 7.38 (d, J ¼ 8.9 Hz, 2H), 7.35 (d, J ¼ 9.0 Hz, 2H),
4.26e4.12 (m, 1H), 3.63 (s, 3H), 3.14 (t, J ¼ 7.1 Hz, 2H), 2.53 (t,
J ¼ 7.1 Hz, 2H), 2.00 (m, 1H), 0.87 (d, J ¼ 6.6 Hz, 3H), 0.86 (d,
250 MHz):
d
7.35e7.08 (m, 4H), 3.07 (t, J ¼ 7.2 Hz, 2H), 2.59 (t,
J ¼ 7.2 Hz, 2H). 13C NMR (CDCl3, 63 MHz):
d 175.58, 132.74, 132.16,
J ¼ 6.7 Hz, 3H). 13C NMR (DMSO‑d6, 75 MHz):
d
¼ 172.47, 170.86,
130.80, 128.18, 33.13, 28.39. LC/MS (ESI): 215.7 [M ꢃ H]-; white
powder. Yield ¼ 75%.
135.62, 130.89, 130.45, 129.41, 57.87, 52.06, 34.91, 30.39, 28.83,
19.40, 18.72; LC/MS (ESI): 328.63 [M ꢃ H]-, 330.1 [MþH]þ; white
solid. Yield ¼ 62%.
4.1.3.3. 3-(4-bromophenylthio)propanoic acid 55. 1H NMR (CDCl3,
250 MHz):
d
7.35 (d, J ¼ 8.4 Hz, 2H), 7.17 (d, J ¼ 8.4 Hz, 2H), 3.07 (t,
J ¼ 7.2 Hz, 2H), 2.60 (t, J ¼ 7.2 Hz, 2H). 13C NMR (CDCl3, 63 MHz):
4.1.4.6. (S)-methyl 2-(3-(4-chlorophenylthio)propanamido)-3-
d
178.34, 134.72, 132.79, 132.51, 121.44, 34.68, 29.49. LC/MS (ESI):
phenylpropanoate 33. 1H NMR (DMSO‑d6, 300 MHz):
d
¼ 8.38 (bd,
259.3 [M ꢃ H]- and isotopic peak 361.4; white solid. Yield ¼ 97%.
J ¼ 7.8 Hz, 1H), 7.38 (d, J ¼ 8.6 Hz, 2H), 7.31 (d, J ¼ 8.5 Hz, 2H),
7.28e7.17 (m, 5H), 4.48 (m,1H), 3.60 (s, 3H), 3.09e2.96 (m, 3H), 2.88
(dd, J ¼ 13.7 Hz and 9.2 Hz, 1H), 2.40 (t, J ¼ 7.2 Hz, 2H); 13C NMR
4.1.4. General procedure for the synthesis of amino acid derivatives
28e40 and 56
(DMSO‑d6, 75 MHz):
d
¼ 172.38, 170.54, 137.59, 135.52, 130.89,
To a stirred solution of the corresponding propionic acid de-
130.42, 129.52, 129.41, 128.68, 126.99, 54.00, 52.26, 37.23, 34.97,
28.64; LC/MS (ESI): 376.5 [M ꢃ H]-, 378.1 [MþH]þ; white powder.
Yield ¼ 54%.
rivative (23e27 or 55) (1 eq, 1.0 mmol) and
L-aminoacid methyl
ester hydrochloride (Alanine, Valine or Phenylalanine) (1 eq,
11

M. Garcia, L. Hoffer, R. Leblanc et al.
European Journal of Medicinal Chemistry 223 (2021) 113601
4.1.4.7. (S)-methyl
2-(3-(phenoxypropanamido)propanoate
34.
J ¼ 8.5 Hz, 2H), 7.15 (d, J ¼ 8.5 Hz, 2H), 5.98 (bd, J ¼ 8.6 Hz, 1H), 4.50
(dd, J ¼ 8.8 Hz and 4.6 Hz, 1H), 3.66 (s, 3H), 3.12 (t, J ¼ 7.2 Hz, 2H),
2.45 (td, J ¼ 7.2 Hz and 1.6 Hz, 2H), 2.07 (m, 1H), 0.67 (d, J ¼ 6.9 Hz,
1H NMR (DMSO‑d6, 250 MHz):
d
¼ 8.49 (bd, J ¼ 7.0 Hz, 1H),
7.36e7.25 (m, 2H), 6.99e6.90 (m, 3H), 4.33 (p, J ¼ 7.3 Hz, 1H), 4.19
(td, J ¼ 6.1 Hz and 2.2 Hz, 2H), 3.65 (s, 3H), 2.62 (t, J ¼ 6.2 Hz, 2H),
3H), 0.82 (d, J ¼ 6.9 Hz, 3H). 13C NMR (CDCl3, 63 MHz):
d
¼ 170.97,
1.31 (d, J ¼ 7.3 Hz, 3H). 13C NMR (DMSO‑d6, 63 MHz):
d
¼ 172.97,
168.97, 133.05, 130.58, 129.75, 118.83, 55.50, 50.72, 34.40, 29.83,
27.91, 17.42, 16.29; LC/MS (ESI): 372.3 [M ꢃ H]- and isotopic peak
374,4, 374.1 [MþH]þ and isotopic peak 376.1; white solid.
Yield ¼ 72%.
169.45, 158.46, 129.39, 120.58, 114.62, 63.93, 51.69, 47.54, 35.11,
17.05; LC/MS (ESI): 252.5 [MþH]þ; colorless oil. Yield ¼ 67%.
4.1.4.8. (S)-methyl 3-methyl-2-(3-(phenoxypropanamido)butanoate
35. 1H NMR (DMSO‑d6, 250 MHz):
d
¼ 8.35 (bd, J ¼ 8.3 Hz, 1H),
4.1.5. General procedure for Suzuki synthesis of valine derivatives
58e66
7.35e7.24 (m, 2H), 7.00e6.88 (m, 3H), 4.26 (dd, J ¼ 8.2 Hz and
6.5 Hz, 1H), 4.18 (t, J ¼ 6.1 Hz, 2H), 3.66 (s, 3H), 2.78e2.59 (m, 2H),
To a solution of compound 56 (1eq, 0.2 mmol), boronic acid or
ester derivative (1.5 eq, 0.3 mmol) and Pd[P(Ph3)]4 (5 mol%) in a
mixture toluene/ethanol (2/0.5 mL) was added an aqueous solution
of Na2CO3 (2 eq, 0.4 mmol, 0.5 mL). The resulting mixture was
refluxed under argon atmosphere during 18 h, cooled and then
concentrated. To the resulting residue was added dropwise 1 N HCl
solution 4 ꢂC until pH 1, the aqueous layer was extracted twice with
ethyl acetate. The combined organic layers were washed once with
water, once with brine, dried over Na2SO4, and evaporated under
vacuum. Crude was purified by column chromatography eluting
with dichloromethane-methanol (98:2 to 80:20).
2.06 (m, 1H), 0.93 (d, J ¼ 6.6 Hz, 3H), 0.89 (d, J ¼ 6.6 Hz, 3H). 13
C
NMR (DMSO‑d6, 63 MHz):
d
¼
13C NMR (63 MHz, DMSO)
d 171.95,
169.96, 158.47, 129.39, 120.56, 114.58, 64.08, 57.40, 51.48, 35.14,
29.96, 22.50, 18.87, 18.20, 15.61; LC/MS (ESI): 280.4 [MþH]þ;
colorless oil. Yield ¼ 82%.
4.1.4.9. (S)-methyl
2-(3-(phenoxypropanamido)-3-phenyl-prop-
anoate 36. 1H NMR (DMSO‑d6, 250 MHz):
d
¼ 8.55 (bd, J ¼ 7.8 Hz,
1H), 7.35e7.20 (m, 7H), 6.99e6.84 (m, 3H), 4.53 (m, 1H), 4.11 (t,
J ¼ 6.7 Hz, 2H), 3.63 (s, 3H), 3.12e2.86 (m, 2H), 2.56 (t, J ¼ 6.3 Hz,
2H). 13C NMR (DMSO‑d6, 63 MHz):
d
¼ 171.86, 169.63, 158.41, 137.20,
129.38, 129.00, 128.16, 126.45, 120.58, 114.59, 63.90, 53.51, 51.70,
4.1.5.1. (2S)-3-methyl-2-(3-{[4-(3-sulfamoylphenyl)phenyl]sulfanyl}
36.85, 35.18; LC/MS (ESI): 328.3 [MþH]þ; colorless oil. Yield ¼ 91%.
propanamido)butanoic acid 58. 1H NMR (CD3OD, 250 MHz):
d
¼ 8.04 (t, J ¼ 1.6 Hz, 1H), 7.75 (m, 2H), 7.52 (d, J ¼ 8.6 Hz, 2H), 7.50
4.1.4.10. (S)-methyl 2-(3-(4- chlorophenoxy)propanamido)prop-
(m, 1H), 7.37 (d, J ¼ 8.6 Hz, 2H), 4.23 (d, J ¼ 5.6 Hz, 1H), 3.15 (t, J ¼
7.2 Hz, 2H), 2.54 (t, J ¼ 7.2 Hz, 2H), 2.15e1.96 (m, 1H), 0.88 (d, J ¼
6.7 Hz, 3H), 0.85 (d, J ¼ 6.7 Hz, 3H).13C NMR (CD3OD, 63 MHz
amoate 37. 1H NMR (DMSO‑d6, 250 MHz):
d
¼ 8.49 (d, J ¼ 7.0 Hz,
1H), 7.35 (d, J ¼ 9.1 Hz, 2H), 6.97 (d, J ¼ 9.1 Hz, 2H), 4.32 (p,
J ¼ 7.2 Hz, 1H), 4.18 (td, J ¼ 6.1 Hz and J ¼ 2.3 Hz, 2H), 3.64 (s, 3H),
2.61 (t, J ¼ 6.1 Hz, 2H), 1.30 (d, J ¼ 7.3 Hz, 3H); 13C NMR (DMSO‑d6,
d
¼ 175.26, 173.91, 145.76, 142.62, 138.53, 137.81, 131.31, 131.11,
130.72, 128.64, 125.83, 125.21, 59.30, 36.41, 31.75, 30.03, 19.70,
18.38; LC/MS (ESI): 434.5 [M ꢃ H]-, 436.5 [MþH]þ; HRMS (TOF,
ESIþ) cald for C20H25N2O5S2 [MþH]þ 437.1205, found 437.1208;
purity ¼ 99%; pale yellow solid. Yield ¼ 63%.
63 MHz):
d
¼ 172.95, 169.32, 157.33, 129.16, 124.38, 116.42, 64.46,
51.70, 47.54, 34.96, 17.05; LC/MS (ESI): 286.2 [MþH]þ; white pow-
der. Yield ¼ 73%.
4.1.4.11. (S)-methyl 2-(3-(4-chlorophenoxy)propanamido)-3-
4.1.5.2. (2S)-3-methyl-2-[3-({4-[4-(piperidine-1-sulfonyl)phenyl]
methylbutanoate 38. 1H NMR (DMSO‑d6, 250 MHz):
d
¼ 8.12 (d,
phenyl}sulfanyl)propanamido]butanoic acid 59. 1H NMR (DMSO‑d6,
J ¼ 8.3 Hz, 1H), 7.11 (d, J ¼ 9.0 Hz, 2H), 6.72 (d, J ¼ 9.0 Hz, 2H), 4.02
(dd, J ¼ 8.3 Hz and J ¼ 6.5 Hz, 1H), 3.95 (t, J ¼ 6.1 Hz, 2H), 3.43 (s,
3H), 2.56e2.33 (m, 2H), 1.82 (m, 1H), 0.68 (d, J ¼ 6.7 Hz, 3H), 0.65 (d,
400 MHz):
d
¼ 8.02 (bd, J ¼ 8.4 Hz, 1H), 7.93 (d, J ¼ 8.5 Hz, 2H), 7.79
(d, J ¼ 8.5 Hz, 2H), 7.73 (d, J ¼ 8.5 Hz, 2H), 7.45 (d, J ¼ 8.5 Hz, 2H),
4.16 (dd, J ¼ 8.6 Hz and 5.7 Hz, 1H), 3.48e3.39 (m, 2H), 3.22 (t,
J ¼ 7.2 Hz, 2H), 2.92 (t, J ¼ 5.4 Hz, 4H), 2.10e1.99 (m, 1H), 1.61e1.50
(m, 4H), 1.42e1.34 (m, 2H), 0.88 (d, J ¼ 6.8 Hz, 3H), 0.87 (d,
J ¼ 6.7 Hz, 3H); 13C NMR (DMSO‑d6, 63 MHz):
d
¼ 171.92, 169.83,
157.33, 129.17, 124.36, 116.37, 64.59, 57.40, 51.49, 34.98, 29.95, 18.87,
18.19; LC/MS (ESI): 314.2 [MþH]þ; colorless oil. Yield ¼ 68%.
J ¼ 6.8 Hz, 3H).13C NMR (DMSO‑d6, 101 MHz):
d
¼ 185.40, 170.63,
144.14, 137.81, 135.83, 134.70, 128.68, 128.60, 128.09, 127.56, 57.93,
47.05, 35.12, 30.50, 28.35, 25.19, 23.29, 19.71, 18.53; LC/MS (ESI):
502.7 [M ꢃ H]-, 504.6 [MþH]þ; HRMS (TOF, ESIþ) cald for
4.1.4.12. (S)-methyl 2-(3-(4-chlorophenoxy)propanamido)-3-phenyl-
propanoate 39. 1H NMR (DMSO‑d6, 250 MHz):
d
¼ 8.55 (d,
J ¼ 7.8 Hz, 1H), 7.38e7.22 (m, 7H), 6.91 (d, J ¼ 9.0 Hz, 2H), 4.53 (m,
1H), 4.11 (t, 6.1 Hz, 2H), 3.63 (s, 3H), 3.06 (dd,
C
25H33N2O5S2 [MþH]þ 505.1831, found 505.1833; purity ¼ 98%;
J
¼
light brown solid. Yield ¼ 48%.
J ¼ 13.7 Hz J ¼ 5.5 Hz, 1H), 2.93 (dd, J ¼ 13.7 Hz and J ¼ 9.2 Hz, 1H),
2.58 (t, J ¼ 6.0 Hz, 2H); 13C NMR (DMSO‑d6, 63 MHz):
d
¼ 171.84,
4.1.5.3. (2S)-2-(3-{[4-(6-acetamidopyridin-3-yl)phenyl]sulfanyl}
169.50, 157.27, 137.18, 129.15, 128.99, 128.16, 126.46, 124.37, 116.37,
64.43, 53.50, 51.71, 36.84, 35.03; LC/MS (ESI): 362.2 [MþH]þ; %;
white powder. Yield ¼ 77%.
propanamido)-3- methylbutanoic acid 60. 1H NMR (DMSO‑d6,
250 MHz):
d
¼ 10.62 (s, 1H), 8.65 (d, J ¼ 1.6 Hz, 1H), 8.17 (d,
J ¼ 8.8 Hz, 1H), 8.09 (dd, J ¼ 8.7 Hz and 2.4 Hz, 1H), 7.86 (bd,
J ¼ 8.5 Hz, 1H), 7.69 (d, J ¼ 8.4 Hz, 2H), 7.43 (d, J ¼ 8.4 Hz, 2H), 4.12
(dd, J ¼ 8.6 Hz and 5.4 Hz, 1H), 3.20 (t, J ¼ 7.1 Hz, 2H), 2.63e2.55 (m,
2H), 2.13 (s, 3H), 2.10e2.00 (m, 1H), 1.42e1.34 (m, 2H), 0.87 (d,
J ¼ 6.8 Hz, 3H), 0.86 (d, J ¼ 6.7 Hz, 3H).13C NMR (DMSO‑d6, 63 MHz):
4.1.4.13. (S)-methyl
3-methyl-2-(4-(phenylbutanamido)butanoate
40. 1H NMR (DMSO‑d6, 250 MHz):
d
¼ 8.15 (d, J ¼ 8.0 Hz, 1H),
7.36e7.27 (m, 2H), 7.24e7.17 (m, 3H), 4.19 (dd, J ¼ 8.0 Hz and
J ¼ 6.5 Hz, 1H), 3.65 (s, 3H), 2.59 (t, J ¼ 7.6 Hz, 2H), 2.21 (t, J ¼ 7.3 Hz,
2H), 2.04 (m, 1H), 1.88e1.75 (m, 2H), 0.91 (d, J ¼ 6.8 Hz, 3H), 0.89 (d,
d
¼ 174.45, 169.86, 169.28, 151.30, 145.43, 135.77, 135.73, 134.09,
130.23, 128.60, 126.82, 113.16, 57.99, 34.83, 30.66, 28.27, 23.89,
J ¼ 6.7 Hz, 3H); 13C NMR (DMSO‑d6,63 MHz):
d
¼ 172.30, 172.12,
19.43, 18.04; LC/MS (ESI): 413.6 [M ꢃ H]-, 415.7 [MþH]þ; HRMS
141.81,128.24,128.20,125.66, 57.42, 51.41, 34.61, 34.44, 29.77, 27.02,
(TOF, ESIþ) cald for
C
21H26N3O4S [MþH]þ 416.1644, found
18.92, 18.30; LC/MS (ESI): 278.4 [MþH]þ; colorless oil. Yield ¼ 35%.
416.1645; purity ¼ 98%; light orange solid. Yield ¼ 58%.
4.1.4.14. (S)-methyl 2-(3-(4-bromophenylthio)propanamido)-3-
4.1.5.4. (2S)-2-[3-({4-[4-({[(tert-butoxy)carbonyl]amino}methyl)
phenyl]phenyl}sulfanyl)propanamido]-3-methylbutanoic acid 61.
methylbutanoate 56. 1H NMR (CDCl3, 250 MHz):
d
¼ 7.34 (d,
12

M. Garcia, L. Hoffer, R. Leblanc et al.
European Journal of Medicinal Chemistry 223 (2021) 113601
1H NMR (CDCl3, 250 MHz):
d
¼ 7.47 (d, J ¼ 8.2 Hz, 2H), 7.44 (d,
4.1.5.9. (2S)-2-(3-{[4-(3-hydroxyphenyl)phenyl]sulfanyl}prop-
J ¼ 8.2 Hz, 2H), 7.35 (d, J ¼ 8.4 Hz, 2H), 7.28 (d, J ¼ 8.2 Hz, 2H), 6.03
(bd, J ¼ 8.5 Hz,1H), 4.83 (bs,1H), 4.52 (dd, J ¼ 8.8 Hz and 4.8 Hz,1H),
4.28 (bd, J ¼ 3.5 Hz, 2H), 3.18 (t, J ¼ 7.2 Hz, 2H), 2.49 (t, J ¼ 7.2 Hz,
2H), 2.09 (m, 1H), 1.40 (s, 9H), 0.88 (d, J ¼ 6.9 Hz, 3H), 0.84 (d,
anamido)-3-methylbutanoic acid 66. 1H NMR (CD3OD, 250 MHz):
d
¼ 7.43 (d, J ¼ 8.6 Hz, 2H), 7.33 (d, J ¼ 8.6 Hz, 2H), 7.14 (t, J ¼ 7.9 Hz,
1H), 6.95 (dm, J ¼ 7.7 Hz, 1H), 6.91 (bt, J ¼ 2.1 Hz, 1H), 6.66 (ddd,
J ¼ 8.1 Hz, 2.1 Hz and 1.0 Hz 1H), 4.18 (d, J ¼ 5.5 Hz, 1H), 3.12 (t,
J ¼ 7.2 Hz, 2H), 2.52 (t, J ¼ 7.2 Hz, 2H), 2.14e2.00 (m, 1H), 0.88 (d,
J ¼ 7.1 Hz, 3H), 0.85 (d, J ¼ 7.1 Hz, 3H). 13C NMR (CD3OD, 101 MHz):
J ¼ 6.9 Hz, 3H). 13C NMR (DMSO‑d6, 101 MHz):
d
¼ 170.51, 160.99,
156.29, 139.94, 138.28, 137.82, 135.70, 129.15, 128.03, 127.57, 126.73,
78.27, 58.21, 43.54, 35.33, 30.64, 28.84, 28.73, 19.84, 18.55; LC/MS
(ESI): 484.4 [M ꢃ H]-, 430.6 [M-tBu]þ and 508.6 [MþNa]þ; HRMS
(TOF, ESIþ) cald for C26H34N2O5SNa [MþH]þ 509.2086, found
509.2088; purity ¼ 97%; light brown solid. Yield ¼ 21%.
d
¼ 175.48, 172.44, 157.57, 141.67, 139.15, 134.84, 129.59, 129.52,
127.11, 117.67, 113.99, 113.16, 58.75, 35.25, 30.44, 29.06, 18.46, 17.06;
LC/MS (ESI): 371.6 [M ꢃ H]-, 373.7 [MþH]þ; HRMS (TOF, ESI-) cald
for C20H22NO4S [M ꢃ H]- 372.1270, found 372.1269; purity ¼ 99%;
light brown solid. Yield ¼ 34%.
4.1.5.5. (2S)-2-[3-({4-[4-(methanesulfonyloxy)phenyl]phenyl}sul-
4.2. HTRF GST-S1FL/SDC2 assay
fanyl)propanamido]-3-methylbutanoic acid 62. 1H NMR (DMSO‑d6,
250 MHz):
d
¼ 7.82 (bs, 1H), 7.77 (d, J ¼ 8.5 Hz, 2H), 7.65 (d,
4.2.1. Production of GST-syntenin full length for HTRF experiments
Competent E.coli (ER2566 strain) cells were transformed with
the human syntenin1 pGEX-5X expression vector (GE Healthcare).
Expression of N-terminally GST-tagged syntenin-1 full length was
induced overnight at 30 ꢂC and by the addition of 0.4 mM IPTG
(Fisher Scientific). Protein was purified using GSTrap4B columns
28-4017-45 (GE Healthcare).
J ¼ 8.2 Hz, 2H), 7.43 (d, J ¼ 8.5 Hz, 2H), 7.42 (d, J ¼ 8.2 Hz, 2H), 4.09
(dd, J ¼ 8.1 Hz and 5.3 Hz, 1H), 3.42 (s, 3H), 3.19 (t, J ¼ 7.1 Hz, 2H),
2.36 (t, J ¼ 6.9 Hz, 2H), 2.06 (m, 1H), 0.86 (d, J ¼ 6.4 Hz, 3H), 0.84 (d,
J ¼ 6.3 Hz, 3H). 13C NMR (DMSO‑d6, 101 MHz):
d
¼ 176.84, 170.21,
148.97, 139.05, 136.63, 136.56, 128.88, 128.52, 127.82, 123.20, 58.69,
37.91, 35.43, 30.91, 28.72, 20.00, 18.58; LC/MS (ESI): 449.6 [M ꢃ H]-,
451.5 [MþH]þ; HRMS (TOF, ESIþ) cald for C21H26NO6S2 [MþH]þ
452.1202, found 452.1205; purity ¼ 97%; light orange solid.
Yield ¼ 42%.
4.2.2. HTRF screen
HTRF assays were performed in white 96Well Small Vol-
ume™HiBase Polystyrene Microplates (Greiner) with
working volume of 100 L, as described previously [17]. Briefly,
compounds were firstly dispensed into the wells at 400 M for the
a total
4.1.5.6. (2S)-2-[3-({4-[4-(2-carboxyethyl)phenyl]phenyl}sulfanyl)
m
propanamido]-3-methylbutanoic acid 63. 1H NMR (CD3OD,
m
250 MHz):
d
¼ 7.46 (d, J ¼ 8.1 Hz, 2H), 7.43 (d, J ¼ 7.8 Hz, 2H), 7.34 (d,
primary screen or with serial DMSO dilutions for IC50 measure-
ment assays. Primary screening assays have been performed in
monoplicate while IC50 measurements were performed in tripli-
cates. All HTRF reagents were purchased from CisBio Bioassays and
reconstituted according to the supplier protocols. The plates were
spin for 3 min at 500g and then incubated at 4 ꢂC for 16 h. The HTRF
signals were recorded on a POLAR star Omega plate reader (BMG
Labtech) with an excitation filter at 337 nm and fluorescence
wavelength measurement at 620 and 665 nm, an integration delay
J ¼ 8.3 Hz, 2H), 7.20 (d, J ¼ 8.0 Hz, 2H), 4.21 (bd, J ¼ 5.5 Hz, 1H), 3.12
(t, J ¼ 7.1 Hz, 2H), 2.85 (t, J ¼ 7.7 Hz, 2H), 2.52 (bt, J ¼ 7.4 Hz, 4H),
2.06 (m, 1H), 0.88 (d, J ¼ 6.7 Hz, 3H), 0.85 (d, J ¼ 6.7 Hz, 3H). 13
C
NMR (DMSO‑d6, 101 MHz):
d
¼ 173.97, 170.61, 140.75, 137.93, 137.62,
135.55, 134.65, 129.34, 129.18, 127.52, 126.77, 58.00, 35.72, 35.32,
30.55, 30.49, 28.85, 19.76, 18.54; LC/MS (ESI): 427.6 [M ꢃ H]-, 429.6
[MþH]þ; HRMS (TOF, ESIþ) cald for C23H28NO5S [MþH]þ 430.1688,
found 430.1690; purity ¼ 95%; white solid. Yield ¼ 30%.
of 60ms and an integration time of 400ms. Results were analyzed
with a two-wavelength signal ratio: [intensity (665 nm)/intensity
4.1.5.7. (2S)-2-[3-({4-[3-(hydroxymethyl)phenyl]phenyl}sulfanyl)
(620 nm)]*104. Percentage of inhibition was calculated using the
propanamido]-3-methylbutanoic acid 64. 1H NMR (CD3OD,
following equation:
%
inhibition
¼
[(compound signal)-(min
250 MHz):
d
¼ 7.15 (m, 1H), 7.48 (d, J ¼ 8.6 Hz, 2H), 7.44e7.38 (m,
signal)]/[(max signal)-(min signal)]*100, where ‘max signal’ is the
signal ratio with the compound vehicle alone (DMSO) and ‘min
signal’ the signal ratio without protein B. For IC50 measurements,
values were normalized and fitted with Prism (GraphPad software)
using the following equation: Y ¼ 100/(1þ((X/IC50)∧Hill slope)).
1H), 7.35 (d, J ¼ 8.6 Hz, 2H), 7.29 (d, J ¼ 7.5 Hz, 1H), 7.26e7.19 (m,
1H), 4.57 (s, 2H), 4.20 (bd, J ¼ 5.5 Hz, 1H), 3.13 (t, J ¼ 7.2 Hz, 2H, 2.53
(t, J ¼ 7.2 Hz, 2H), 2.05 (m, 1H), 0.88 (d, J ¼ 6.8 Hz, 3H), 0.85 (d,
J ¼ 6.8 Hz, 3H). 13C NMR (CD3OD, 63 MHz):
d
¼ 175.10, 173.11,
142.66, 140.98, 139.67, 135.61, 130.24, 129.18, 127.80, 126.27, 125.91,
125.55, 64.40, 58.97, 35.79, 31.02, 29.63, 19.00, 17.62; LC/MS (ESI):
385.6 [M ꢃ H]-, 387.7 [MþH]þ; HRMS (TOF, ESI-) cald for
4.3. Molecular modeling
C
21H24NO4S [M ꢃ H]- 386.1426, found 386.1429; purity ¼ 99%;
4.3.1. Binding site preparation
yellow solid. Yield ¼ 56%.
MOE version 2016 (Chemical Computing Group Inc., Montreal,
QC, Canada) was used to prepare the binding sites starting from the
X-ray structure of the syntenin PDZ domain in complex with
compound 45 (6RLC). All residues, with at least one atom within
14 Å radius from reference fragment were selected to define the
binding site. A large binding site was defined because the docking
engine (S4MPLE) relies on a FF-based energy function and the goal
is to explore adjacent sub-pockets using a growing strategy. Two
different binding site files were prepared because both potential
orientations of the N215 sidechain (flip of terminal amide) were
considered.
4.1.5.8. (2S)-2-[3-({4-[6-(2-methoxyethoxy)pyridin-3-yl]phenyl}sul-
fanyl)propanamido]-3-methylbutanoic acid 65. 1H NMR (DMSO‑d6,
400 MHz):
d
¼ 8.46 (d, J ¼ 2.3 Hz, 1H), 8.00 (dd, J ¼ 8.6 Hz and
2.6 Hz, 1H), 7.84 (bd, J ¼ 6.6 Hz, 1H), 7.62 (d, J ¼ 8.4 Hz, 2H), 7.41 (d,
J ¼ 8.4 Hz, 2H), 6.91 (d, J ¼ 8.7 Hz, 1H), 4.42 (bt, J ¼ 4.7 Hz, 2H), 4.11
(dd, J ¼ 8.7 Hz and 5.4 Hz, 1H), 3.68 (bt, J ¼ 4.7 Hz, 2H), 3.24 (s, 3H),
3.17 (t, J ¼ 7.2 Hz, 2H), 2.55 (t, J ¼ 7.4 Hz, 2H), 2.12e1.97 (m, 1H), 0.86
(d, J ¼ 6.5 Hz, 3H), 0.84 (d, J ¼ 6.4 Hz, 3H). 13C NMR (DMSO‑d6,
63 MHz):
d
¼ 174.19, 170.00, 162.54, 144.28, 137.39, 135.32, 134.27,
128.71, 128.64, 126.82, 110.80, 70.19, 64.71, 58.06, 34.85, 30.25,
28.33, 19.39, 18.01; LC/MS (ESI): 430.5 [M ꢃ H]-, 432.6 [MþH]þ;
HRMS (TOF, ESIþ) cald for C22H29N2O5S [MþH]þ 433.1797, found
433.1801; purity ¼ 98%; light orange solid. Yield ¼ 46%.
4.3.2. Design of virtual library
The creation of the virtual focused library relied on the pro-
cedure described in the DOTS methodology [30]. Briefly, a focused
13

M. Garcia, L. Hoffer, R. Leblanc et al.
European Journal of Medicinal Chemistry 223 (2021) 113601
library was generated starting from an activated substructure (here,
fragment 45 that contains an aryl halide reactive function), a
collection of commercially available BBs and a set of in silico
encoded medicinal chemistry relevant chemical reactions. This
chemistry-driven strategy is expected to design accessible com-
pounds with high reaction yields. The single Suzuki rule (#31 ac-
cording to the reference study [36] was enabled for the design
stage. A raw focused library of 1865 compounds was automatically
designed. Then, several fully automated post-processing stages
were applied to this raw focused library to extract a subset of
(Invitrogen) supplemented with 10% fetal calf serum (FCS) (Euro-
bio) at 37 ꢂC under 5% CO2. Syntenin CRISPR/Cas9 knockout (Syn-
tKO) MCF-7 cells were generated following the procedures as
previously described [37]. Briefly, SDCBP-specific gRNA oligos
sgRNASDCBP1 targeting exon1 (F: CACCgCTATCCCTCACGATG-
GAAGT; R: AAACACTTCCATCGTGAGGGATAGc) were cloned into the
pX458 two-in-one CRISPR targeting vector (Addgene, Cambridge,
MA), individually transfected into MCF7 cells and sorted for GFP
expression after 48h. Single cell clones were grown and screened
by western blotting for syntenin expression. One individual clone
has been used in the in vitro experiments. The cDNA encoding
mCherry-syntenin was derived from the eGFP-syntenin cDNA
construct, by restriction-ligation. The expression vector for Ce-
RAB5(Q79L) was received from W. Annaert (K.U. Leuven,
duplicate-free compounds with
a focus on structures with
reasonable physiochemical properties. More precisely, a structural
check was performed with StructureChecker from ChemAxon
(http://www.chemaxon.com) to eliminate compounds with any
structural warning. Duplicates and compounds that contain at least
one undesired element (other than C, H, N, O, P, S, F, Cl, Br or I) were
also discarded. Usual threshold values for common physico-
chemical descriptors (molecular weight ꢄ 550, cLogP ꢄ 5.5,
tPSA ꢄ 150 Å2, H-Bond acceptors ꢄ 10, H-Bond donors ꢄ 5, rota-
tional bond count ꢄ 14, and formal charge count ꢄ 2) were used to
automatically extract a subset of compounds with reasonable
properties. Then, yuck filters, which were used to design the clean
subset of the ZINC database (http://zinc.docking.org), were
employed to reject compounds with undesired functional groups.
Cxcalc and Jcsearch programs from ChemAxon were respectively
used to compute descriptors and detect undesired substructures.
The last post-processing stage involved various computations
which are required for the subsequent virtual screening: major
microspecies, partial charges, force-field atomic types, and single
3D-conformer were computed for each compound. A maximum
common substructure (MCS) algorithm was also used to flag atoms
that were already present in the fragment 45. These atoms were
constrained in the subsequent virtual screening stage (see below).
The mapping of force-field atomic types was done using programs
from AmberTools (http://ambermd.org), while all the other steps
were performed using in house tools relying on the ChemAxon Java
API. In the end, the virtual focused chemical library contained 402
compounds.
Belgium). The CAY10594 inhibitor (used at 10
from Santa Cruz (sc-223,874) [19].
mM) was purchased
4.4.2. Cancer cell viability
The effects of the compounds C58 and SyntOFF were tested both
on MCF-7 and MCF-7 SyntKO cells. Briefly, the cells were treated
with increasing concentrations of the compounds for 48 h and
early/late apoptosis was measured by using the FITC Annexin V
apoptosis detection kit with 7-AAD (Biolegend, San Diego, CA) ac-
cording to the manufacturer. Annexin V positive but 7-AAD nega-
tive (early apoptotic cells) and Annexin V positive and 7-AAD
positive (late stage apoptosis) was determined by using FACS LSRII
flow cytometer (BD Biosciences) and data were analyzed with
Flowjo software (Tree Star).
4.4.3. Exosomes and total cell lysates
For comparative analyses, exosome-enriched fractions were
collected from equivalent amounts of culture medium, conditioned
by equivalent amounts of cells. MCF-7 cells were treated with
indicated compounds or with DMSO 0.2% as control, in medium
containing exosome-depleted FCS (10%). After 16 h, cell media were
collected and exosomes were isolated by three sequential centri-
fugation steps at 4 ꢂC: 10 min at 500ꢀg, to remove cells; 30 min at
10,000ꢀg, to remove cell debris; and 1h30min at 100,000ꢀg, to
pellet exosomes (exosome-enriched fraction), followed by one
4.3.3. Virtual screening
wash with 1400
serum and secreted proteins. Exosomal pellets were then re-
suspended in 100 L of PBS1X. The lysates from corresponding
cultures were cleared by centrifugation at 1500 rpm for 5 min and
then resuspended in lysis buffer (TrisHCL pH 7.4 30 mM, NaCl
m
L of PBS1X (100,000ꢀg, 1h), to remove soluble
The virtual screening was performed on both prepared binding
sites as described in the DOTS methodology [PMID 29883107].
More precisely, the final subset of compounds was screened using
the S4MPLE conformational tool with constraints on the original
location of the fragment 45. Unflagged atoms, corresponding to
newly added atoms from virtual synthesis, were free to explore
adjacent sub-pockets without any constraints. The sampling stage
mainly consisted in three independent simulations of 400 gener-
ations with a population of 30 individuals. All saved poses were
merged into one file before switching to the post-processing stage
that involved the minimization of all non-redundant poses while
unlocking all ligand atoms. The energy used to rank all screened
compounds was equal to the best energy of the complex minus the
energy of the best conformer of the free ligand. The latter is ob-
tained by performing three independent simulations with the same
parameters on the ligand alone. The goal is to compute a potential
energy difference between bound and free forms to estimate the
interaction energy while considering the strain energy of the
ligand.
m
150 mM, 1% NP40 (IGEPAL), 1 mg/mL aprotinin, 1 mg/mL leupeptin).
Although only little variations were observed from sample to
sample, exosomal amounts loaded for the western blot were
normalized according to the number of parent cells from where
exosomes were secreted.
4.4.4. Western blots
The proteins were heat-denaturated in Laemmli sample buffer,
fractionated in 12.5% or 15% gels by SDSePAGE and electro-
transferred to nitrocellulose membrane. Membranes were stained
with Ponceau red and immunoblotted with the indicated primary
antibodies for: CD9 (1/5000), CD63 (1/5000), CD81 (1/5000) and
ꢀ
ADAM10 (1/1000) antibodies provided by E. Rubinstein (Universite
Paris-Sud UMRS_935, Villejuif); syntenin (Homemade; 1/3000),
Tubulin (Sigma-Aldrich; 1/10,000), ALIX (Homemade, 1/500),
HSP70 (Santa Cruz Biotechnology; 1/500), Flotillin-1 (BD Bio-
sciences; 1/1000), EGFR (Cell Signaling; 1/1000), EpCAM (Santa
Cruz Biotechnology; 1/200), Fibronectin (BD Biosciences; 1/5000),
Src (Cell signaling; 1/1000), TSP1 (Lab Vision; 1/100), SDC1 intra-
cellular domain (Homemade, 1/1000), SDC4 intracellular domain
(Abnova; 1/2000), and HRP-conjugated secondary antibodies
4.4. Cellular evaluation
4.4.1. Cell culture and reagents
MCF-7 cell lines were purchased from the American Type cul-
ture collection (Manassas, VA). Cells were grown in DMEM-F12
14

M. Garcia, L. Hoffer, R. Leblanc et al.
European Journal of Medicinal Chemistry 223 (2021) 113601
therapies? Immunol. Rev. 263 (2015) 279e301.
(Mouse or Rabbit, Thermofisher scientific; 1/10,000). Signals were
visualized using Amersham ECL Prime Western Blotting Detection
Reagent (GE Healthcare) [19].
[13] N. Bosc, C. Muller, L. Hoffer, D. Lagorce, S. Bourg, C. Derviaux, M.E. Gourdel,
J.C. Rain, T.W. Miller, B.O. Villoutreix, M.A. Miteva, P. Bonnet, X. Morelli,
O. Sperandio, P. Roche, Fr-PPIChem, An academic compound library dedicated
to protein-protein interactions, ACS Chem. Biol. 15 (2020) 1566e1574.
[14] P. Filippakopoulos, J. Qi, S. Picaud, Y. Shen, W.B. Smith, O. Fedorov, E.M. Morse,
T. Keates, T.T. Hickman, I. Felletar, M. Philpott, S. Munro, M.R. McKeown,
Y. Wang, A.L. Christie, N. West, M.J. Cameron, B. Schwartz, T.D. Heightman,
N. La Thangue, C.A. French, O. Wiest, A.L. Kung, S. Knapp, J.E. Bradner, Selective
inhibition of BET bromodomains, Nature 468 (2010) 1067e1073.
[15] O. Ulucan, S. Eyrisch, V. Helms, Druggability of dynamic protein-protein in-
terfaces, Curr. Pharmaceut. Des. 18 (2012) 4599e4606.
Declaration of competing interest
The authors declare that they have no known competing
financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
[16] M. Jung, K.A. Gelato, A. Fernandez-Montalvan, S. Siegel, B. Haendler, Targeting
BET bromodomains for cancer treatment, Epigenomics 7 (2015) 487e501.
[17] S. Milhas, B. Raux, S. Betzi, C. Derviaux, P. Roche, A. Restouin, M.J. Basse,
E. Rebuffet, A. Lugari, M. Badol, R. Kashyap, J.C. Lissitzky, C. Eydoux, V. Hamon,
M.E. Gourdel, S. Combes, P. Zimmermann, M. Aurrand-Lions, T. Roux,
C. Rogers, S. Muller, S. Knapp, E. Trinquet, Y. Collette, J.C. Guillemot, X. Morelli,
Protein-protein interaction inhibition (2P2I)-Oriented chemical library ac-
celerates hit discovery, ACS Chem. Biol. 11 (2016) 2140e2148.
[18] E.J. Faivre, K.F. McDaniel, D.H. Albert, S.R. Mantena, J.P. Plotnik, D. Wilcox,
L. Zhang, M.H. Bui, G.S. Sheppard, L. Wang, V. Sehgal, X. Lin, X. Huang, X. Lu,
T. Uziel, P. Hessler, L.T. Lam, R.J. Bellin, G. Mehta, S. Fidanze, J.K. Pratt, D. Liu,
L.A. Hasvold, C. Sun, S.C. Panchal, J.J. Nicolette, S.L. Fossey, C.H. Park,
K. Longenecker, L. Bigelow, M. Torrent, S.H. Rosenberg, W.M. Kati, Y. Shen,
Selective inhibition of the BD2 bromodomain of BET proteins in prostate
cancer, Nature 578 (2020) 306e310.
[19] R. Leblanc, R. Kashyap, K. Barral, A.L. Egea-Jimenez, D. Kovalskyy, M. Feracci,
M. Garcia, C. Derviaux, S. Betzi, R. Ghossoub, M. Platonov, P. Roche, X. Morelli,
L. Hoffer, P. Zimmermann, Pharmacological inhibition of syntenin PDZ2
domain impairs breast cancer cell activities and exosome loadifing with
syndecan and EpCAM cargo, J. Extracell. Vesicles 10 (2020), e12039.
[20] X.S. Chen, J.C. Longgood, C. Michnoff, S.G. Wei, D.E. Frantz, I. Bezprozvanny,
High-throughput screen for small molecule inhibitors of Mint1-PDZ domains,
Assay Drug Dev. Technol. 5 (2007) 769e783.
[21] J.P. Borg, PDZ Mediated Interactions, Springer Protocols, 2021, p. 2256.
[22] E.Y.S. Lin, L.F. Silvian, D.J. Marcotte, C.C. Banos, F. Jow, T.R. Chan, R.M. Arduini,
F. Qian, D.P. Baker, C. Bergeron, C.A. Hession, R.L. Huganir, C.F. Borenstein,
I. Enyedy, J. Zou, E. Rohde, M. Wittmann, G. Kumaravel, K.J. Rhodes,
R.H. Scannevin, A. Dunah, K.M. Guckian, Potent PDZ-domain PICK1 inhibitors
that modulate amyloid beta-mediated synaptic dysfunction, Sci Rep-Uk 8
(2018).
Acknowledgements
This work was supported by grants from the National Research
Agency (ANR, Investissements d’Avenir, A*MIDEX project ANR-11-
IDEX-0001-02), the Institut National du Cancer (INCa, subvention
2013e105), the Belgian Foundation against cancer (STK,FA/2014/
294), the French National Research Agency (ANR-18-CE13-0017,
Project SynTEV), the European Union's Horizon 2020 research and
innovation program under the Marie Sklodowska-Curie grant
agreement No 747025, the French Infrastructure for Integrated
Structural Biology (FRISBI) ANR-10-INSB-05-01 with the use of the
AFMB laboratory Xray facility. The authors acknowledge the Euro-
pean Synchrotron Radiation Facility (ESRF) as well as the SOLEIL
synchrotron (Proxima 1 and Proxima 2 beamlines) for provision of
synchrotron radiation facilities. M. Garcia was the recipient of La
Ligue, French Foundation for Cancer Research. R. Leblanc was the
recipient of the Fondation ARC, French Foundation for Cancer
Research.
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.ejmech.2021.113601.
[23] T.P. Kegelman, B. Wu, S.K. Das, S. Talukdar, J.M. Beckta, B. Hu, L. Emdad,
K. Valerie, D. Sarkar, F.B. Furnari, W.K. Cavenee, J. Wei, A. Purves, S.K. De,
M. Pellecchia, P.B. Fisher, Inhibition of radiation-induced glioblastoma inva-
sion by genetic and pharmacological targeting of MDA-9/Syntenin, Proc. Natl.
Acad. Sci. U. S. A. 114 (2017) 370e375.
References
[24] I. Elidrisi, P.V. Bhatt, T. Govender, H.G. Kruger, G.E.M. Maguire, Synthesis and
NMR elucidation of novel amino acid cavitand derivatives, Tetrahedron 70
(2014) 7057e7066.
[25] I. Elidrisi, S. Negin, P.V. Bhatt, T. Govender, H.G. Kruger, G.W. Gokel,
G.E.M. Maguire, Pore formation in phospholipid bilayers by amphiphilic
cavitands, Org. Biomol. Chem. 9 (2011) 4498e4506.
[26] C. Lamberth, H.J. Kempf, M. Kriz, Synthesis and fungicidal activity of N-2-(3-
methoxy-4-propargyloxy) phenethyl amides. Part 3: stretched and hetero-
cyclic mandelamide oomyceticides, Pest Manag. Sci. 63 (2007) 57e62.
[27] Y.L. Song, Y.F. Dong, T. Yang, C.C. Zhang, L.M. Su, X. Huang, D.N. Zhang,
G.L. Yang, Y.X. Liu, Synthesis and pharmacological evaluation of novel bisin-
dolylalkanes analogues, Bioorg. Med. Chem. 21 (2013) 7624e7627.
[28] K. Matsumoto, S. Suga, J. Yoshida, Organic reactions mediated by electro-
chemically generated ArSþ, Org. Biomol. Chem. 9 (2011) 2586e2596.
[29] Y.X. Liu, Z.P. Cui, Y.H. Li, Y.C. Gu, Q.M. Wang, 3-Hydrazido and 3-hydrazono
derivatives of tenuazonic acid and their herbicide evaluation, J. Heterocycl.
Chem. 51 (2014) E197eE201.
[1] D.W. Greening, S.K. Gopal, R. Xu, R.J. Simpson, W. Chen, Exosomes and their
roles in immune regulation and cancer, Semin. Cell Dev. Biol. 40 (2015)
72e81.
[2] I. Wortzel, S. Dror, C.M. Kenific, D. Lyden, Exosome-mediated metastasis:
communication from a distance, Dev. Cell 49 (2019) 347e360.
[3] C. Rajagopal, K.B. Harikumar, The origin and functions of exosomes in cancer,
Front Oncol 8 (2018).
[4] J.J. Grootjans, P. Zimmermann, G. Reekmans, A. Smets, G. Degeest, J. Durr,
G. David, Syntenin, a PDZ protein that binds syndecan cytoplasmic domains,
Proc. Natl. Acad. Sci. U. S. A. 94 (1997) 13683e13688.
[5] K. Lambaerts, S. Van Dyck, E. Mortier, Y. Ivarsson, G. Degeest, A. Luyten,
E. Vermeiren, B. Peers, G. David, P. Zimmermann, Syntenin,
a syndecan
adaptor and an Arf6 phosphatidylinositol 4,5-bisphosphate effector, is
essential for epiboly and gastrulation cell movements in zebrafish, J. Cell Sci.
125 (2012) 1129e1140.
[6] J.R. Couchman, Transmembrane signaling proteoglycans, Annu. Rev. Cell Dev.
Biol. 26 (2010) 89e114.
[30] L. Hoffer, Y.V. Voitovich, B. Raux, K. Carrasco, C. Muller, A.Y. Fedorov,
C. Derviaux, A. Amouric, S. Betzi, D. Horvath, A. Varnek, Y. Collette, S. Combes,
P. Roche, X. Morelli, Integrated strategy for lead optimization based on frag-
ment growing: the diversity-oriented-target-focused-synthesis approach,
J. Med. Chem. 61 (2018) 5719e5732.
[31] L. Hoffer, J.P. Renaud, D. Horvath, In silico fragment-based drug discovery:
setup and validation of a fragment-to-lead computational protocol using
S4MPLE, J. Chem. Inf. Model. 53 (2013) 836e851.
[32] L. Hoffer, C. Chira, G. Marcou, A. Varnek, D. Horvath, S4MPLE-Sampler for
multiple protein-ligand entities: methodology and rigid-site docking bench-
marking, Molecules 20 (2015) 8997e9028.
[33] A.L. Egea-Jimenez, R. Gallardo, A. Garcia-Pino, Y. Ivarsson, A.M. Wawrzyniak,
R. Kashyap, R. Loris, J. Schymkowitz, F. Rousseau, P. Zimmermann, Frizzled 7
and PIP2 binding by syntenin PDZ2 domain supports Frizzled 7 trafficking and
signalling, Nat. Commun. 7 (2016) 12101.
[7] H.C. Christianson, M. Belting, Heparan sulfate proteoglycan as a cell-surface
endocytosis receptor, Matrix biology, journal of the International Society for
Matrix Biology 35 (2014) 51e55.
[8] V. Friand, G. David, P. Zimmermann, Syntenin and syndecan in the biogenesis
of exosomes, Biol. Cell. 107 (2015) 331e341.
[9] T.P. Kegelman, S.K. Das, L. Emdad, B. Hu, M.E. Menezes, P. Bhoopathi,
X.Y. Wang, M. Pellecchia, D. Sarkar, P.B. Fisher, Targeting tumor invasion: the
roles of MDA-9/Syntenin, Expert Opin. Ther. Targets 19 (2015) 97e112.
[10] S.K. Das, S. Maji, S.L. Wechman, P. Bhoopathi, A.K. Pradhan, S. Talukdar,
D. Sarkar, J. Landry, C.Q. Guo, X.Y. Wang, W.K. Cavenee, L. Emdad, P.B. Fisher,
MDA-9/Syntenin (SDCBP): novel gene and therapeutic target for cancer
metastasis, Pharmacol. Res. 155 (2020).
[11] L.N. Makley, J.E. Gestwicki, Expanding the number of 'druggable' targets: non-
enzymes and protein-protein interactions, Chem. Biol. Drug Des. 81 (2013)
22e32.
[34] D. Kozakov, L.E. Grove, D.R. Hall, T. Bohnuud, S.E. Mottarella, L. Luo, B. Xia,
D. Beglov, S. Vajda, The FTMap family of web servers for determining and
[12] T. Cierpicki, J. Grembecka, Targeting protein-protein interactions in hemato-
logic malignancies: still
a challenge or a great opportunity for future
15

M. Garcia, L. Hoffer, R. Leblanc et al.
European Journal of Medicinal Chemistry 223 (2021) 113601
characterizing ligand-binding hot spots of proteins, Nat. Protoc. 10 (2015)
733e755.
reactions for in silico molecule design, J. Chem. Inf. Model. 51 (2011)
3093e3098.
[35] M. Matziari, F. Beau, P. Cuniasse, V. Dive, A. Yiotakis, Evaluation of P-1'-
diversified phosphinic peptides leads to the development of highly selective
inhibitors of MMP-11, J. Med. Chem. 47 (2004) 325e336.
[36] M. Hartenfeller, M. Eberle, P. Meier, C. Nieto-Oberhuber, K.H. Altmann,
G. Schneider, E. Jacoby, S. Renner, A collection of robust organic synthesis
[37] N.S. Imjeti, K. Menck, A.L. Egea-Jimenez, C. Lecointre, F. Lembo, H. Bouguenina,
A. Badache, R. Ghossoub, G. David, S. Roche, P. Zimmermann, Syntenin me-
diates SRC function in exosomal cell-to-cell communication, Proc. Natl. Acad.
Sci. U. S. A. 114 (2017) 12495e12500.
16