New desulfured troglitazone derivatives: Improved synthesis and biological evaluation

Article abstract of DOI:10.1016/j.ejmech.2019.111939

Breast cancer is a major medical threat which cannot be sufficiently addressed by current therapies because of spontaneous or acquired treatment resistance. Besides, triple-negative breast cancer (TNBC) tumors do not respond to targeted therapies, thus new therapeutic strategies are needed. In this context, we designed and prepared new desulfured troglitazone (TGZ)-derived molecules and evaluated them in vitro for their anti-proliferative activity, with a special focus on triple-negative breast cancer cell lines. Optimization of the synthetic strategies and deracemization of the lead compound were performed to give highly active compound 10 with low-micromolar potency. Further studies revealed that this compound triggers apoptosis rather than cell cycle arrest as observed with TGZ.

Full text of DOI:10.1016/j.ejmech.2019.111939

Journal Pre-proof
New desulfured troglitazone derivatives: Improved synthesis and biological evaluation
Dorian Dupommier, Claire Muller, Corinne Comoy, Sabine Mazerbourg, Andrea
Bordessa, Eline Piquard, Manon Pawlak, Flavian Piquard, Hélène Martin, Elia De
Fays, Stéphanie Grandemange, Stéphane Flament, Michel Boisbrun
PII:
S0223-5234(19)31091-8
DOI:
https://doi.org/10.1016/j.ejmech.2019.111939
EJMECH 111939
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 3 October 2019
Revised Date: 29 November 2019
Accepted Date: 1 December 2019
Please cite this article as: D. Dupommier, C. Muller, C. Comoy, S. Mazerbourg, A. Bordessa, E. Piquard,
M. Pawlak, F. Piquard, Héè. Martin, E. De Fays, Sté. Grandemange, Sté. Flament, M. Boisbrun, New
desulfured troglitazone derivatives: Improved synthesis and biological evaluation, European Journal of
Medicinal Chemistry (2020), doi: https://doi.org/10.1016/j.ejmech.2019.111939.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
© 2019 Published by Elsevier Masson SAS.

New desulfured troglitazone derivatives: improved synthesis and
biological evaluation
Dorian Dupommiera,1, Claire Mullerb,1, Corinne Comoya, Sabine Mazerbourgb, Andrea
Bordessaa, Eline Piquarda, Manon Pawlaka, Flavian Piquarda, Hélène Martinc, Elia De Faysb,
b
b
a
*
Stéphanie Grandemange , Stéphane Flament , Michel Boisbrun
*Corresponding author. Tel.: +33 372 74 52 64;
E-mail address: michel.boisbrun@univ-lorraine.fr (M. Boisbrun).
1These authors contributed equally to the work.
1

aUniversité de Lorraine, CNRS, L2CM, F-54000 Nancy, France
bUniversité de Lorraine, CNRS, CRAN, F-54000 Nancy, France
c PEPITE EA4267, Univ. Bourgogne Franche-Comté, F-25000 Besançon, France
ABSTRACT
Breast cancer is a major medical threat which cannot be sufficiently addressed by current
therapies because of spontaneous or acquired treatment resistance. Besides, triple-negative
breast cancer (TNBC) tumors do not respond to targeted therapies, thus new therapeutic
strategies are needed. In this context, we designed and prepared new desulfured troglitazone
(TGZ)-derived molecules and evaluated them in vitro for their anti-proliferative activity, with
a special focus on triple-negative breast cancer cell lines. Optimization of the synthetic
strategies and deracemization of the lead compound were performed to give highly active
compound 10 with low-micromolar potency. Further studies revealed that this compound
triggers apoptosis rather than cell cycle arrest as observed with TGZ.
Keywords: Troglitazone, Chromane, Deracemization, Lipase, Breast cancer, Apoptosis
1. Introduction
Breast cancer is the most commonly diagnosed cancer and the leading cause of cancer death
in women worldwide [1]. Current treatments for breast cancer are multimodal with surgery,
radiotherapy, chemotherapy, and targeted therapy [2]. However, all types of breast cancers
develop resistance to these conventional therapies. A tumor could be resistant to multiple
treatment strategies, such as chemotherapy and targeted therapy based on monoclonal
antibody treatment [3–5]. Besides, the triple-negative breast cancer tumors (TNBCs),
characterized by a lack of expression of estrogen receptor, progesterone receptor, and human
epidermal growth factor receptor HER2, do not respond to targeted therapies. They further
present a higher chance of recurrence after treatment with conventional chemotherapeutic
agents. TNBCs (15% of breast cancer cases worldwide) are a heterogeneous group of higher
grade cancer, with the worst prognosis, compared to other breast cancer subtypes [6]. The
increasing prevalence of drug resistant cancers and the limited treatment efficacy for TNBCs
are strong arguments for the search for new therapeutic agents and optimized therapeutic
strategies. In this context, thiazolidine-2,4-diones (TZDs) are studied in order to develop
alternative or complementary anticancer agents towards breast cancers. Compounds bearing
this moiety were recently proved to be of high interest as they exhibited numerous biological
activities [7–11] and many of them showed promising in vitro and in vivo anticancerous
properties. In particular, our team developed [12–14] Structure-Activity Relationships (SAR)
studies starting from the structure of troglitazone (TGZ) (Figure 1), a molecule formerly used
as an antidiabetic. We showed that some structural features both increased the anti-
proliferative activity toward breast cancer cell lines and decreased the toxicity against primary
cultured non-malignant human hepatocytes. These aspects comprised mainly the introduction
2

of a double bond adjacent to the TZD moiety (yielding ∆2-TGZ, Figure 1), and the presence
of a lipophilic group on the 6-position of the chromane moiety. Meanwhile, Saha and
coworkers [15,16] synthesized some analogs of well-known TZDs such as rosiglitazone
(RGZ), pioglitazone (PGZ), and troglitazone (TGZ) which were all used as antidiabetics, due
to their affinity for the PPAR-γ receptor. These new compounds were devoid of the sulfur
atom in the TZD ring, yielding pyrrolidinedione derivatives. These compounds showed
diminished PPAR-γ binding, but also a lower toxicity toward THLE-2 model of primary
normal liver cells. Furthermore, they showed enhanced anti-cancer properties; in particular, a
desulfured analog of TGZ called trosuccinimide (TSN) was promising.
The aim of the present study was to investigate a possible synergistic effect between the
desulfuration and our previous SAR studies on the anti-cancer properties of the molecules.
Thus, we prepared new desulfured TGZ derivatives bearing or not a double bond adjacent to
the TZD moiety, and a lipophilic group (Boc or alkyl chain) at the 6-position of the chromane
heterocycle. The inhibition of proliferation of breast cancer cell lines was studied, with a
special attention toward cell lines derived from TNBCs, and the impact on cell cycle was also
studied. Furthermore, since the chromane moiety of TGZ bears an asymmetric center, we set
up a deracemization procedure in order to evaluate the influence of stereochemistry on the
activity. Finally, we investigated a vectorization approach in order to still increase the potency
of our molecules.
Fig. 1. Structures of troglitazone, ∆2-troglitazone, and target desulfured molecules
2. Chemistry
In order to prepare the envisioned desulfured troglitazone derivatives, we decided not to
follow the global procedure reported by Saha and coworkers [15], as it was hampered by two
very low-yield steps. Indeed, we protected the phenol function of the chromane heterocycle
by a Boc group which is very easy to introduce and to remove, compared to the allyl group.
Thus, we easily prepared aldehyde 1 using our reported procedure [12] (Scheme 1). It was
then reacted with 3-(triphenylphosphoranylidene)pyrrolidine-2,5-dione 2 [15] to give Boc-
protected desulfured analog of ∆2-TGZ 3 in good yield. Acidic treatment easily afforded the
deprotected counterpart 4 which we could name ∆2-TSN. Both 3 and 4 could be reduced by
hydrogenation to give saturated compounds 5 and 6 (TSN) in good yield.
3

Scheme 1. Reagents and conditions: (a) MeOH, RT; (b) TFA, CH2Cl2; (c) H2, Pd/C, EtOAc; (d) Tf2O, Pyr.,
CH2Cl2, then 4-hydroxybenzaldehyde, Cs2CO3, DMF; (e) HCl gas, CH2Cl2.
We reported [13] that the introduction of an aminoalkyl group linked to the chromane moiety
is even more effective than a Boc group in increasing the anti-proliferative activity. Thus, we
prepared alcohol 7 [13] (Scheme 1) and after activation as a triflate, we reacted it with 4-
hydroxybenzaldehyde, giving aldehyde 8. To improve this reported procedure, the latter step
was done in the presence of cesium carbonate instead of previously used potassium carbonate.
In these conditions, the reaction was complete after 24 h instead of 48 h. Compound 8 was
then reacted with 2 as above to give 9, which was further N-deprotected to afford 10.
Performing nuclear Overhauser enhancement measurements between the methylene protons
of the pyrrolidinedione ring and the phenyl protons of compound 9 showed a 5%
enhancement of the signal, confirming the (E) stereochemistry of the double bond.
As compounds 9 and 10 gave interesting biological results (see later), we intended to go
further with them. Thus, we had to scale up the synthesis and the preparation of 7 [13] had to
be reinvestigated (Scheme 2).
4

Scheme 2. Reagents and conditions: (a) (HCHO)n, methyl methacrylate, AcOH, Bu2NH; (b) Br(CH2)5COCl,
AlCl3, CH2Cl2; (c) LiAlH4, AlCl3, Et2O; (d) NaN3, DMF; (e) H2, Pd/C, EtOAc; (f) Boc2O, CH2Cl2
.
First, we intended to prepare easily chromane 12. Instead of achieving a three-steps procedure
[12] starting from commercial trolox, we applied a reported procedure [17] on 2,3,5-
trimethylphenol by condensing it with methylmethacrylate and paraformaldehyde. This
method afforded 12 in modest (42%) yield, but it was straightforward since after refluxing, a
simple filtration afforded a pure compound. Furthermore, as all the starting materials were
affordable and since this reaction was easily scalable, huge quantities (up to 19 g) of the
desired compound could be obtained per batch. Besides, when we had to prepare compound
13 at the multi-gram scale following our reported procedure [13], it was difficult to isolate it
in sufficient purity. Some by-products were generated and among them, low amounts of the
sulfide 17 (Figure 2) could be identified by mass spectrometry. Its formation might be due to
a reaction of 12 with remaining traces of highly electrophilic thionyl chloride in the freshly
prepared bromohexanoyl chloride. A previous study already described the obtaining of a
sulfide from a reaction between thionyl chloride and an aromatic compound [18]. In addition,
traces of dimer 18 (Figure 2) were also identified by the same technique. As 17 exhibits a
polarity close to the desired compound 13, it was difficult to remove it by chromatography.
Fig. 2. Structures of by-products 17 and 18 obtained during the Friedel-Crafts reaction.
Thus, in order to avoid such side-products formation, we developed a meticulous workup.
Indeed, after careful evaporation of the excess of thionyl chloride, the crude bromohexanoyl
chloride was treated with formic acid in order to remove any trace of remaining reactant (see
experimental section). The isolation of pure 13 was thus achieved at 6 g scale. Reduction to
give 14 and formation of azide 15 proceeded as previously reported [13], but again in order to
scale up the whole procedure, we had to change our protocol for the final steps. Indeed, one-
pot reduction of azide 15 and protection to give 7 gave tedious results at more than 500 mg
scale. Some non-isolated by-products were generated at the hydrogenation step; thus the
5

primary amine 16 could preferentially be chromatographied before protecting it to give multi-
gram amounts of 7 with high purity.
Since the chromane moiety contains an asymmetric carbon, it was challenging to separate
both enantiomers of our lead structures in the search for a difference in anti-proliferative
activity. In our hands, chiral chromatography failed to separate both isomers of either
compound 9 or 10. Thus, we intended to separate them by a kinetic resolution method. Hyatt
and Skelton reported [19] on the separation of both enantiomers of (R,S)-6-hydroxy-2,5,7,8-
tetramethyl-2-chromanmethanol through a succinic anhydride acylation catalyzed by
supported Amano lipase PS. Thereby, we applied this technique to racemic alcohol 7,
considering that the protected aminoalkyl chain should not interfere with the acylation
reaction which would take place at the other side of the heterocycle. Indeed, after 18 hours of
reaction and according to TLC, approximately half of 7 was converted to the expected acid-
ester 19, together with remaining 7 (Scheme 3). Unlike Hyatt’s work, the acid-ester 19 was
too lipophilic to be isolated by aqueous basic extraction, nevertheless both compounds were
easily separated by column chromatography. Subsequently, 19 was saponified to give the
expected alcohol. According to Hyatt and Skelton work, we assume that the stereoselection
should be identical, thus the lipase should preferentially catalyze the acylation of the (S)
isomer of 7, giving (S)-19 (Scheme 3). In addition, the same stereoselection was observed by
others using very similar but different substrates [17,20].
Scheme 3. Reagents and conditions: (a) Succinic anhydride, supported Amano Lipase PS, MTBE; (b) LiOH,
THF/MeOH; (c) (R)-Mosher acyl chloride, CDCl3, C5D5N, C6D6.
The enantiomeric excess of (R)-7 and (S)-7 were evaluated by coupling these compounds
with (R)-Mosher acyl chloride as reported [20] (Scheme 3) to give the corresponding esters
(R,R)-20 and (S,R)-20 (Scheme 3, Figure 3). The relative amounts of the resulting
diastereomers were evaluated by 1H NMR, since each oxymethylene group gave an AB
system between 4.00 and 4.35 ppm, without overlapping with any other signals. After one
lipase reaction, the deduced ee was 83% in one case and 87% for the other case. It reached
96% after two reactions for both isomers, and 98% after three reactions (Figure 3).
Compounds (R)-7 and (S)-7 with 98% ee were subsequently used as substrates for a 3-steps
synthesis leading to (R)-10 and (S)-10 as described before. We payed attention to the
possibility of racemization of the chiral carbon atom during the final acidic deprotection
leading to the desired final compounds. We first examined the stability of the stereochemistry
of compound (S)-7 in acidic conditions. In a first attempt, we noticed that it led to partial
decomposition. We thus easily acetylated (S)-7 and then deprotected it with HCl to give the
6

Fig. 3. Parts of NMR spectra (400MHz, 4.00-4.35 ppm range) showing the AB system of the oxomethylene
group of ester 20 obtained from 7 as a racemic or after one, two, and three kinetic resolution reactions.
primary ammonium chloride (S)-21 (Scheme 4). It was then re-protected and deacetylated to
give back the starting compound. The latter was coupled with Mosher acyl chloride as
previously and the deduced ee was the same than the starting material one. Furthermore, it
was also comforting that (R)-10 and (S)-10 obtained from (R)-7 and (S)-7 exhibited opposite
optical rotation (see experimental part).
Scheme 4. Reagents and conditions: (a) Ac2O, Pyr., DMAP, CH2Cl2; (b) HCl gas, CH2Cl2; (c) Boc2O, TEA,
CH2Cl2; (d) LiOH, THF/MeOH.
With the aim of confirming the attributed stereochemistry, we did many attempts in order to
get suitable crystals for X-ray diffraction. Since ferrocenyl esters were reported to facilitate
the formation of crystals from carboxylic acids [21], ester 22 was prepared from racemic
7

compound 7, but unfortunately it showed no tendency to give crystals, and acidic removal of
the protecting group resulted in decomposition of the structure (Scheme 5). As we had success
in getting crystals from dinitrophenylhydrazones [22], we condensed aldehyde (R)-8 with 2,4-
dinitrophenylhydrazine to give hydrazone (R)-23 in good yield, which could be deprotected to
give (R)-24. Besides, compound (R)-7 was converted to benzoate (R)-25 which was further
N-deprotected to give (R)-26 (Scheme 5). Among these compounds, (R)-23 and (R)-26 gave
crystals which were submitted to X-ray analysis. For both crystals, diffraction was observed,
but with a too-low resolution to be resolved.
Scheme 5. Reagents and conditions: (a) Ferrocenoyl chloride, iPr2NEt, DMAP, CH2Cl2; (b) 2,4-
dinitrophenylhydrazine, MS 4Å, toluene; (c) HCl gas, CH2Cl2; (d) Benzoyl chloride, pyridine, CH2Cl2.
Besides, we intended to vectorize our lead compound toward cancerous cells by grafting a
glucose unit. For this purpose, we envisioned a copper-mediated Huisgen click reaction
between an alkyne-derived glucose and an analog of 10 bearing an azide function. Thus, azide
15 was coupled to 4-hydroxybenzaldehyde to give aldehyde 27, which was in turn reacted
with phosphoranylidene 2 to give compound 28 (Scheme 6).
D-glucose was peracetylated as
previously reported [23] to give β- -glucopyranose pentaacetate which was reacted with
D
propargyl alcohol to give 29 as described by Murray [24]. Cycloaddition reaction between 28
and 29 was conducted in the presence of in situ generated Cu(I) to give adduct 30, which
could be deacetylated to give target compound 31 in good yield (Scheme 6).
8

Scheme 6. Reagents and conditions: (a) Tf2O, Pyr., CH2Cl2, then 4-hydroxybenzaldehyde, Cs2CO3, DMF; (b) 2,
MeOH, RT; (c) Ac2O, AcONa, then propargyl alcool, BF3.OEt2, CH2Cl2; (d) 28, CuSO4, sodium ascorbate, t-
BuOH/MeOH; (e) MeONa, MeOH.
3. Biological studies and discussion
The synthesized molecules were tested for their anti-proliferative activity against 4 breast
cancer cell lines: one hormone-dependent (MCF-7) and 3 TNBC derived (hormone-
independent) (MDA-MB-231, MDA-MB-435, and Hs578T) breast cancer cell lines. The
results are expressed as IC50 in the Table 1. TGZ and ∆2-TGZ, prepared as previously
reported [12], were also tested for comparison. The representative dose-dependent responses
of MDA-MB-231 cancer cells to TGZ and compound 10 are presented in supplementary data.
Table 1. Anti-proliferative activity of TGZ and some derivatives on different breast cancer cell lines.
a
IC50
Compound
MCF-7b
MDA-MB-231c MDA-MB-453c
Hs578Tc
> 100
74.3 ± 3.9
35.9 ± 3.1
15.8 ± 3.9
57.0 ± 4.4
13.2 ± 1.3
6.7 ± 0.6
4.0 ± 0.2
5.8 ± 0.4
45.5 ± 3.7
61.2 ± 2.9
28.0 ± 2.0
56.5 ± 3.3
32.4 ± 3.3
20.6 ± 1.0
5.4 ± 0.7
67.1 ± 6.1
35.8 ± 4.1
13.6 ± 2.0
58.3 ± 4.1
13.2 ± 1.0
3.9 ± 0.4
TGZ
6 (TSN)
72.6 ± 6.8
38.2 ± 3.4
> 100
5
∆2-TGZ
38.9 ± 2.2
30.8 ± 1.3
6.4 ± 0.8
8.8 ± 0.6
4
3
1.04 ± 0.2
3.5 ± 0.1
9
5.9 ± 0.1
10
a Concentration (µM) required to decrease tumor cell number by 50% ± SEM (Standard Error of the Mean).
b Hormone-dependent breast cancer cell line.
c Hormone-independent breast cancer cell line.
First of all, we could notice that the highest IC50 values (corresponding to the lowest
activities) were obtained with the Hs578T cell line, while the most sensitive cell line (except
for TGZ) was MDA-MB-453. Hs578T cells were not sensitive to TGZ and ∆2-TGZ, but the
IC50 of all the other molecules were lower than 100 µM, whatever the cell line. Removal of
the sulfur atom of TGZ yielded 6 and our tests confirmed that this molecule (named TSN in
other studies) was more active than TGZ, except against the MDA-MB-231 cell line. The
lipophilic Boc-derived compound 5 was approximately two-fold more active than 6, in
accordance with our previous results in the sulfured series [12]. ∆2-TGZ showed a slightly
better potency than TGZ as expected, and the removal of the sulfur afforded 4 (which we
could name ∆2-TSN), which was more potent than both ∆2-TGZ and the related saturated
9

compound 6. Thus, the presence of the double bond shows a synergistic effect with the
removal of the sulfur atom. It is interesting to notice that the introduction of the double bond
on 6 to give compound 4, has the same enhancing effect on the activity than the adjunction of
the Boc group on 6, as 4 and 5 have c.a. the same anti-proliferative activity toward each cell
line. Furthermore, the adjunction of the Boc group on 4 afforded very active compound 3,
with low micromolar activity on two cell lines. Finally, the addition of a N-Boc protected
aminoalkyl chain gave compound 9, which showed a further fourfold increase in potency
compared to 3 for the three hormone-independent breast cancer cell lines, and a lower effect
against MCF-7. Removal of the Boc protecting group yielded 10 which showed a slightly
lower activity than 9. Altogether, one can notice that the removal of the sulfur atom has a
synergistic effect with the chemical modifications which were previously showed to increase
the potency, i.e. the presence of a double bond and of a lipophilic group at the 6-position of
the chromane heterocycle. This approach led to very active compounds 9 and 10 with low-
micromolar activities. Because of its lower lipophilicity, we chose compound 10 for further
consideration. We then studied the effect of compound 10 on two other cancer cell lines:
HCT116 (colon cancer) and U87 (glioblastoma). This molecule displayed an anticancer effect
with IC50 values of 6.4 ± 0.4 µM and 8.7 ± 1.1 µM on these two cell lines respectively. These
values were in a similar range than those measured in breast cancer cells. The sensibility of
U87 cell to compound 10 is to be emphasized when considering the actual difficulty to treat
the very aggressive glioblastoma tumors[25]. In contrast, exposure of non tumoral primary
hepatocytes to compound 10 afforded a much higher IC50 of 42.1 µM, showing an interesting
selectivity.
Besides, we intended to separate both enantiomers of compound 10 by using a kinetic
resolution method. It allowed us to evaluate the anti-proliferative activity of each isomer. We
tested them against two representative breast cancer cell lines: a hormone-dependent one
(MCF-7) and a triple negative (hormone-independent) one (MDA-MB-231). The measured
IC50 were unexpected (Table 2). Concerning MCF-7, both (R) (10.9 µM) and (S) (9.9 µM)
isomers were significantly less active than the racemic compound (5.8 µM). On the other
hand, the opposite was seen concerning the MDA-MB-231 cell line: both isomers were more
active than the racemic compound (5.9 µM), (S) derivative being the most potent one with
IC50 = 2.7 µM. It is quite difficult to rationalize, since we expected to have the anti-
proliferative activity of the racemic compound as being roughly the mean of the IC50 of the
two enantiomers in both cell lines. Without knowledge about the exact target of this family of
compounds, one may however estimate that concerning the hormone-dependent cell line, both
enantiomers have a synergistic action whereas in the hormone-independent one, when they
are mixed in the racemic mixture, they somehow impede each other.
Furthermore, we envisioned to vectorize our lead compound in order to preferentially target
cancer cells. Due to the famous Warburg effect [26,27], the rate of the aerobic glycolysis is
higher in cancerous cells than in normal cells, leading to an overexpression of the glucose
receptors. Thus, we chose a strategy consisting in the addition of a glucose unit to our
compound. Numerous examples of this strategy have been reported [28]. As depicted above,
azide analog 28 was easily prepared, which was linked to an alkyne-derived peracetylated
glucose via a copper-mediated cycloaddition, giving 30. Final deacetylation gave target
compound 31. As depicted in Table 2, the azide intermediate 28 kept a fair anti-proliferative
activity, but lower than compound 10. Grafting the peracetylated glucose unit (giving 30)
totally abrogated the anti-proliferative activity toward MDA-MB-231 hormone-independent
10

breast cancer cell line, but some activity was kept toward MCF-7 cell line, at the same level as
TSN (6). Finally, the deprotected target compound 31 showed no activity against both cancer
cell lines (Table 2). It seems clear that the nature of the substituents located at the end of the
alkyl chain have a strong influence on the anti-proliferative activity. Replacing the primary
amine of 10 by an azide group has a negative effect on the activity which could be attributed
to the loss of H-bond donor. Unfortunately, the grafting of a glucose moiety is even more
detrimental to the activity, in opposition to what was initially planned. In particular, the total
absence of activity of compound 31 could find two explanations. First, one could consider
that the high polarity of 31 prevents its passive diffusion into the cancerous cells, contrary to
the other compounds. From another point of view, the glucose moiety of 31 could lead to a
strong anchorage of the molecule on the cell surface via linkage to the glucose receptor, once
again preventing any passive diffusion. This time, the loss of activity could be attributed to
an unexpected inability of the whole molecular structure to enter the cell through the glucose
receptor after anchoring to it.
Table 2. Anti-proliferative activity of TGZ derivatives on different breast cancer cell lines.
a
IC50
Compound
MCF-7b
5.8 ± 0.4
10.9 ± 1.7
9.9 ± 0.9
18.1 ± 1.7
33.4 ± 7.9
> 100
MDA-MB-231c
5.9 ± 0.1
4.3 ± 0.3
2.7 ± 0.4
21.7 ± 3.4
> 100
rac-10
(R)-10
(S)-10
28
30
> 100
31
a Concentration (µM) required to decrease tumor cell number by 50% ± SEM (Standard Error of the Mean).
b Hormone-dependent breast cancer cell line.
c Hormone-independent breast cancer cell line.
The activity of compound 10 was further characterized in the triple-negative breast cancer cell
line MDA-MB-231. Cell cycle analysis performed by flow cytometry revealed that 23.4 ±
5.8% and 43.9 ± 7% of compound 10-treated cells were detected in the sub-G1 fraction,
corresponding to the apoptotic cells, after 48 and 72 h, respectively (figure 4A). By contrast,
in cells exposed to TGZ, there was no significant increase in the number of cells in the sub-
G1 fraction compared to control cells (DMSO) (figure 4B). These data are in accordance with
our previous results [29] showing that the related compound ∆2-TGZ induced only a slight
increase (< 5%) in the proportion of MDA-MB-231 cells number in the sub-G1 fraction. After
48 h of TGZ treatment, in comparison to control cells, we observed an increase in the
proportion of cells in the G0/G1 phase (85.5 ± 0.3% versus 62.6 ± 0.5%) and a decrease in the
proportion of cells in the S phase (2.3 ± 0.2% versus 15.7 ± 0.4%) and G2/M phase (10.4 ±
0.6% versus 20.5 ± 0.6%). The situation was similar after 72 h: 80 ± 3.8% versus 62 ± 2.7%
11

in G0/G1 phase; 1.9 ± 0.1% versus 13.1 ± 1.2% in S phase; 12.2 ± 0.8% versus 19.5 ± 0.5% in
G2/M phase. This indicates a cell cycle arrest in the G0/G1 phase induced by TGZ that we also
previously reported in MDA-MB-231 cells after exposure to the related compound ∆2-TGZ
[29]. In compound 10-treated cells, we observed a global decrease in the cell number in the
three phases of the cell cycle after 48 and 72 h of treatment. After a 48 h treatment, this
decrease was significant only for cells in the S phase (8 ± 0.8% versus 14.8 ± 0.2%). After 72
h of treatment, the decrease was significant for cells in the G0/G1 phase (35 ± 3.9% versus
62.4 ± 1%) and for cells in the S phase (6.1 ± 0.9% versus 11.8 ± 0.9%). Thus, in contrast to
TGZ and ∆2-TGZ which are cytostatic agents inducing a cell cycle arrest in G0/G1 phase,
compound 10 appears as a potent pro-apoptotic molecule.
Fig. 4. Compound 10 increases the sub-G1 population of MDA-MB-231 breast cancer cells whereas TGZ
induces cell cycle arrest. MDA-MB-231 cells were treated for indicated times with 6 µM of 10 (A), 45 µM of
TGZ (B) or DMSO for control cells. Cell cycle was analyzed by flow cytometry. Bar diagrams represent the
percentage of cells in the different phases of the cell cycle. Data are represented as mean ± standard error of the
mean of at least three independent experiments. Statistical significances of treated cells compared to control cells
are indicated (*, p<0.05; **, p<0.01; ***, p<0.001).
To confirm the pro-apoptotic activity of compound 10, caspase 3/7 activity was measured
after 48 and 72 hours of treatment (figure 5A). Compound 10 induced a 2.7-fold and 1.4-fold
increase of MDA-MB-231 caspase 3/7 activity after 48h and 72h treatment respectively,
compared to control cells. The lower, but significant, increase in caspase 3/7 activity after 72
hours of treatment could be explained by the 1.7- fold increase in the basal level of caspase
3/7 activity in control cells between 48 h and 72 h conditions. To comfort these data, we
further determined the percentage of apoptotic cells by propidium iodide (nuclear labelling of
dead cells) and FITC-Annexin V double-staining followed by flow cytometry analysis (figure
5B and 5C). Annexin V is an apoptosis marker able to bind to phosphatidyl serine, a
membrane component translocated from the cytoplasmic face of the plasma membrane to the
12

cell surface during apoptosis. Compound 10 induced a 2.3-fold and 3.7-fold (23.8 ± 4.3%
versus 10.4 ± 1.3% and 37.4 ± 4.3% versus 10.2 ± 0.9%) significant increase of the
proportion of MDA-MB-231 labelled with FITC-Annexin V after 48h and 72h treatment
respectively, compared to control cells (figure 5B and 5C). Overall, these convergent results
show that compound 10 is not a cytostatic but a pro-apoptotic agent in the TNBC cell line
MDA-MB-231.
Fig. 5. Compound 10 induces apoptosis of MDA-MB-231 breast cancer cells. MDA-MB-231 cells were treated
for indicated times with 6 µM of 10 or DMSO for control cells. (A) Caspase 3/7 activity was measured by a
luminescent method. (B) Double-staining intensity with Propidium Iodide and Annexin V was quantified by
flow cytometry and bar graphs show the percentage of apoptotic cells. (C) Dot-plot graphs representative of the
flow cytometry analysis: the annexin V labeled apoptotic cells are localized in the lower and upper right
quadrants. Data are represented as mean ± standard error of the mean of at least three independent experiments.
Statistical significances of treated cells compared to control cells are indicated (*, p<0.05; **, p<0.01; ***,
p<0.001).
4. Conclusion
13

In this study, numerous desulfured derivatives of TGZ were prepared and tested for their anti-
proliferative activity against hormone-dependent and hormone-independent breast cancer cell
lines. A clear synergistic effect on the activity was observed between the desulfuration and
our previously reported key structural modifications of the TGZ template. This led to very
active compounds from which the best ones exhibited low-micromolar activities. Synthetic
efforts were done in order to scale up initial procedures at the multi-gram scale and a kinetic
resolution method was used to prepare and evaluate both isomers of the most promising
compound (10). Flow cytometry, caspase 3/7 activity measurement and FITC-Annexin V
double staining showed that this compound is clearly a pro-apoptotic agent. Further studies
are ongoing in our team to better understand the underlying mode of action of this compound.
5. Experimental protocols
5.1. Chemistry
5.1.1. General methods
Diethylether and toluene were dried using a MBRAUN MB-SPS-800 solvent purification
system. Other solvents and liquid reagents were purified and dried according to recommended
procedures. Chemical reagents were purchased from Merck, Fisher Scientific or TCI Europe
and were used as received. Amano lipase PS from Burkholderia cepacia and Hyflo Supercel®
were purchased from Sigma-Aldrich. Analytical thin-layer chromatography was performed
with silica gel 60 F254, 0.25 mm pre-coated TLC plates (Merck). Compounds were visualized
using UV light (254 nm) and a solution of cerium sulfate tetrahydrate and phosphomolybdic
acid in 10% aqueous sulfuric acid as developing agent. Alternatively, a 1% ethanolic solution
of ninhydrin or a potassium permanganate aqueous solution were used. Column
chromatographies were performed using a Grace Reveleris® apparatus using 40 µM silica gel.
NMR spectra were recorded at 303K on a Bruker Avance III 200 (200 MHz for 1H and 50.3
MHz for 13C), DPX250 (250 MHz for 1H and 62.9 MHz for 13C) or on a Bruker DRX400
(400 MHz for 1H and 100.6 MHz for 13C) spectrometer. The chemical shifts are reported in
ppm (δ) relative to residual solvent peak [30]. Mass spectra (MS) were recorded on a
micrOTOFq (Bruker) ESI/QqTOF spectrometer. Melting points (M.p.) were determined with
a Kofler bench and are uncorrected. FTIR spectra were recorded on a Perkin-Elmer spectrum
1000 spectrophotometer using a thin film deposition on NaCl window or KBr pellets, or on a
Shimadzu IRAffinity-1 apparatus equipped with an ATR PIKE diamond gladiATR. Optical
rotations were measured on an Anton Paar MCP 300 polarimeter (sodium D-line: 589 nm)
using a 10 cm cell at 20 °C in CHCl3 or 10% v/v MeOH in CHCl3, and the concentration is
expressed in g/100mL. Elemental analyses were performed using a Thermofinnigan Flash EA
1112 apparatus.
5.1.2. (E)-tert-butyl (2-((4-((2,5-dioxopyrrolidin-3-ylidene)methyl)phenoxy)methyl)-2,5,7,8-
tetramethylchroman-6-yl) carbonate (3)
14

To a solution of compound 1 (2.30 g, 5.22 mmol) in dry MeOH (160 mL) was added the
phosphoranylidene 2 (1.88 g, 5.22 mmol). The suspension was stirred at room temperature
under argon for 24 h, filtered, and washed with cooled MeOH. The crude compound was
purified by column chromatography (eluent: cyclohexane/EtOAc, 80:20) to give 2.01 g (3.85
mmol, 74% yield) of white crystals. M.p. 206-208 °C. IR (film) υ (cm-1): 3163, 2918, 1745,
1703, 1601, 1514, 1342, 1234, 1170, 1155, 1093, 839. 1H NMR (400 MHz, CDCl3): δ 1.42 (s,
3H, CH3), 1.55 (s, 9H, t-Bu), 1.89 (m, 1 H, chromane 3-HaHb), 2.04 (s, 3H, CH3), 2.06 (s,
3H, CH3), 2.08 (s, 3H, CH3), 2.13 (m, 1 H, chromane 3-HaHb), 2.64 (m, 2 H, chromane 4-
CH2), 3.57 (d, J = 1.9 Hz, 2H, CH2CO), 3.94, 4.04 (AB system, J = 9.5 Hz, 2H, OCH2), 6.99
(d, J = 8.7 Hz, 2H, Harom), 7.42 (d, J = 8.7 Hz, 2H, Harom), 7.55 (br t, 1 H, CH=C), 8.40 (br s,
1H, NH). 13C NMR (100 MHz, CDCl3): δ 11.9 (CH3), 12.0 (CH3), 12.8 (CH3), 20.2 (CH2),
22.9 (CH3), 27.8 (t-Bu), 28.4 (CH2), 35.2 (CH2), 72.8, 74.5, 83.0, 115.6 (CHarom), 117.3,
121.6, 123.3, 125.6, 127.0, 127.6, 132.2 (CHarom), 135.0 (CH=C), 141.5, 148.9, 152.4, 160.9,
171.3 (C=O), 174.2 (C=O). ESI-MS (pos. mode): m/z = 544.23 ([M+Na]+). Anal. Calcd for
C30H35NO7, 1/3 H2O (527.62): C, 68.29; H, 6.81; N, 2.65. Found: C, 68.35; H, 6.66; N, 2.64.
5.1.3. (E)-3-(4-((6-hydroxy-2,5,7,8-tetramethylchroman-2-yl)methoxy)benzylidene)
pyrrolidine-2,5-dione (4)
To a solution of compound 3 (959 mg, 1.84 mmol) in CH2Cl2 (25 mL) was added
trifluoroacetic acid (8 mL). The mixture was stirred at room temperature for 1.5 h and the
solvent was evaporated. The residue was dissolved in EtOAc (70 mL), the solution was
washed with 5% aqueous NaHCO3 solution (2 x 50 mL), water (2 x 50 mL), dried (MgSO4)
and concentrated to dryness. The residue was crystallized in EtOH to give 565 mg (1.28
mmol, 70% yield) of white crystals. M.p. 174-176 °C. IR (film) υ (cm-1): 3415, 2975, 1749,
1705, 1651, 1251. 1H NMR (400 MHz, CDCl3): δ 1.42 (s, 3H, CH3), 1.92 (m, 1 H, chromane
3-HaHb), 2.09 (s, 3H, CH3), 2.10 (m, 1 H, chromane 3-HaHb), 2.12 (s, 3H, CH3), 2.16 (s, 3H,
CH3), 2.66 (m, 2 H, chromane 4-CH2), 3.58 (d, J = 2.2 Hz, 2H, CH2CO), 3.95, 4.04 (AB
system, J = 9.2 Hz, 2H, OCH2), 4.27 (s, 1H, OH), 7.00 (d, J = 8.8 Hz, 2H, Harom), 7.42 (d, J =
8.8 Hz, 2H, Harom), 7.55 (br t, 1 H, CH=C), 8.11 (br s, 1H, NH). 13C NMR (100 MHz, CDCl3):
δ 11.4 (CH3), 12.0 (CH3), 12.4 (CH3), 20.4 (CH2), 22.7 (CH3), 28.9 (CH2), 35.3 (CH2), 73.0,
74.1, 115.6 (CHarom), 117.3, 118.7, 121.4, 121.5, 122.9, 126.9, 132.2 (CHarom), 135.1 (CH=C),
145.1, 145.2, 161.0, 171.1 (C=O), 174.0 (C=O). ESI-MS (pos. mode): m/z = 460.15
([M+K]+). Anal. Calcd for C25H27NO5, H2O (439.51): C, 68.32; H, 6.65; N, 3.19. Found: C,
68.61; H, 6.28; N, 3.30.
5.1.4 tert-butyl (2-((4-((2,5-dioxopyrrolidin-3-yl)methyl)phenoxy)methyl)-2,5,7,8-
tetramethylchroman-6-yl) carbonate (5)
In a round bottom flask was first introduced 10% palladium on carbon (10 mg), in order to
avoid any ignition of the solvent. A solution of 3 (100 mg, 0.19 mmol) in EtOAc (10 mL) was
then added. The mixture was stirred under hydrogen atmosphere for 18 h at room
temperature, then filtered on celite® and the solvent was removed under reduced pressure. The
residue was purified by column chromatography (eluent: cyclohexane/EtOAc, 100:0 to 70:30)
to afford 5 as a colorless foam (45 mg, 0.09 mmol, 47 % yield). M.p. 69-71 °C. IR (ATR): ν
(cm-1) 3169, 2928, 1749, 1714, 1609, 1510, 1456, 1367. 1H NMR (400 MHz, CDCl3): δ 1.41
(s, 3H, CH3), 1.55 (s, 9H, t-Bu), 1.88 (m, 1H, chromane 3-HaHb), 2.03 (s, 3H, CH3), 2.08 (s,
6H, 2×CH3), 2.13 (m, 1H, chromane 3-HaHb), 2.49 (A part of an ABX system, J = 18.4, 5.0
Hz, 1H, PhCHaHb), 2.63 (m, 2H, chromane 4-CH2), 2.72 (B part of an ABX system, J = 18.4,
15

9.0 Hz, 1H, PhCHaHb), 2.87 (m, 1H, TZD CHCHaHb), 3.09-3.19 (m, 2H, TZD CHCHaHb +
CHCHaHb), 3.84, 3.96 (AB system, J = 9.1 Hz, OCH2), 6.85 (d, J = 8.7 Hz, 2H, Harom), 7.07
(d, J = 8.7 Hz, 2H, Harom), 7.95 (br s, 1H, NH). 13C NMR (100 MHz, CDCl3): δ 11.9 (CH3),
12.0 (CH3), 12.8 (CH3), 20.3 (CH2), 23.1 (CH3), 27.8 (t-Bu), 28.4 (CH2), 34.4 (CH2), 35.5
(CH2), 43.0 (CH), 72.7, 74.7, 82.9, 115.2 (CHarom), 117.4, 123.2, 125.1, 127.4, 129.2, 130.2
(CHarom), 141.4, 149.0, 152.4, 158.3, 176.2 (C=O), 179.3 (C=O). ESI-MS (pos. mode): m/z =
446.25 [M+Na]+. Anal. Calcd for C25H29NO5, H2O (432.52): C, 69.43; H, 6.99; N, 3.24.
½
Found: C, 69.20; H, 7.06; N, 3.05.
5.1.5. 3-(4-((6-hydroxy-2,5,7,8-tetramethylchroman-2-yl)methoxy)benzyl)pyrrolidine-2,5-
dione (6)
In a round bottom flask was first introduced 10% palladium on carbon (10 mg), in order to
avoid any ignition of the solvent. A suspension of 4 (100 mg, 0.24 mmol) in EtOAc (10 mL)
was then added. The mixture was stirred under hydrogen atmosphere for 18 h at room
temperature, then filtered on celite® and the solvent was removed under reduced pressure to
give a pale yellow foam (91 mg, 0.21 mmol, 89% yield). M.p. 75-77 °C. IR (ATR): ν (cm-1)
3486, 3200, 2922, 1771, 1699, 1608, 1510, 1456, 1418, 1352. 1H NMR (400 MHz, CDCl3) : δ
1.41 (s, 3H, CH3), 1.89 (m, 1H, chromane-3-HaHb), 2.10 (s, 3H, CH3), 2.11 (s, 3H, CH3), 2.12
(m, 1H, chromane-3-HaHb), 2.16 (s, 3H, CH3), 2.48 (A part of an ABX system, J = 18.5, 4.9
Hz, 1H, PhCHaHb), 2.64 (m, 2H, chromane 4-CH2), 2.71 (B part of an ABX system, J = 18.5,
9.0 Hz, 1H, PhCHaHb), 2.88 (m, 1H, TZD CHCHaHb), 3.07-3.19 (m, 2H, TZD CHCHaHb +
CHCHaHb), 3.86-3.96 (AB system, J = 9.1 Hz, CH2O), 4.33 (br s, 1H, OH), 6.86 (d, J = 8.6
Hz, 2H, Harom), 7.07 (d, J = 8.6 Hz, 2H, Harom), 8.20 (br s, 1H, NH). 13C NMR (100 MHz,
CDCl3): δ 11.4 (CH3), 12.0 (CH3), 12.3 (CH3), 20.5 (CH2), 22.9 (CH3), 28.8 (CH2), 34.4
(CH2), 35.4 (CH2), 43.0 (CH), 72.8, 74.2, 115.2 (CHarom), 117.5, 118.7, 121.4, 122.8, 129.1,
130.1 (CHarom), 145.1, 145.2, 158.4, 176.5 (C=O), 179.5 (C=O). ESI-MS (pos. mode): m/z =
446.19 [M+Na]+. Anal Calcd for C25H29NO5, 1/3 H2O (429.52): C, 69.91; H, 6.96; N, 3.26.
Found: C, 70.18; H, 7.02; N, 3.19.
5.1.6 tert-butyl (6-(2-((4-formylphenoxy)methyl)-2,5,7,8-tetramethylchroman-6-yl)hexyl)
carbamate (8)
Preparation of the triflate intermediate: to an ice cooled solution of anhydrous pyridine (445
µL, 5.5 mmol) in dry CH2Cl2 (5 mL) was added dropwise trifluoromethanesulfonic anhydride
(211 µL, 1.3 mmol) under argon. A solution of 7 (385 mg, 0.9 mmol) in CH2Cl2 (5 mL) was
added 5 minutes later to the reaction mixture. The solution was stirred for 20 minutes at 0 °C
and the solvent was evaporated. The residue was dissolved in EtOAc (15 mL), the solution
was washed with water (2 x 10 mL), 5% aqueous NaHCO3 solution (2 x 10 mL), and brine (2
x 10 mL). The organic phase was then dried (MgSO4) and the solvent was removed under
vacuum to afford a brownish liquid.
Condensation with 4-hydroxybenzaldehyde: to a solution of the above triflate in dry DMF (4
mL) were added 4-hydroxybenzaldehyde (112 mg, 0.9 mmol) and Cs2CO3 (449 mg, 1.4
mmol). The reaction mixture was stirred under argon at room temperature for 24 h, and then
EtOAc (20 mL) was added. The organic layer was washed with water (2 x 20 mL), 5%
aqueous NaHCO3 solution (2 x 20 mL), water (2 x 20 mL), dried (MgSO4) and the solvent
16

was evaporated under vacuum to give a yellow liquid. The residue was purified by column
chromatography (eluent: cyclohexane/EtOAc, 100:0 to 85:15) to afford 8 (328 mg, 0.63
mmol, 68 % yield) as a colorless liquid. 1H NMR (400 MHz, CDCl3): δ 1.43 (s, 3H, CH3),
1.35-1.53 (m, 8H, linker CH2), 1.44 (s, 9H, t-Bu), 1.90 (m, 1H, chromane 3-HaHb), 2.10 (s,
3H, CH3), 2.13 (m, 1H, chromane 3-HaHb), 2.15 (s, 3H, CH3), 2.20 (s, 3H, CH3), 2.57-2.71
(m, 4H, chromane 4-CH2 + linker CH2), 3.12 (m, 2H, CH2NHBoc), 3.98, 4.08 (AB system, J
= 9.3 Hz, 2H, CH2O), 4.51 (br s, 1H, NHBoc), 7.03 (d, J = 8.7 Hz, 2H, Harom), 7.82 (d, J = 8.7
Hz, 2H, Harom), 9.88 (s, 1H, CHO). Other analytical data were in accordance with the previous
ones [13].
5.1.7. tert-butyl (E)-(6-(2-((4-((2,5-dioxopyrrolidin-3-ylidene)methyl)phenoxy)methyl)-
2,5,7,8-tetramethylchroman-6-yl)hexyl)carbamate (9)
To a solution of compound 8 (208 mg, 0.40 mmol) in anhydrous MeOH (10 mL) was added
compound 2 (143 mg, 0.40 mmol). The suspension was stirred at room temperature under
argon for 12 h. The mixture was filtered and washed with cooled MeOH to give 153 mg (0.25
mmol, 63% yield) of white powder. M.p. 205-207 °C. IR (ATR) υ (cm-1): 3381, 2924, 1769,
1705, 1688, 1603, 1514, 1308, 1252, 1163. 1H NMR (400 MHz, CDCl3) : δ 1.43 (s, 3H, CH3),
1.44 (m, 8H, linker CH2), 1.45 (s, 9H, t-Bu), 1.90 (m, 1H, chromane-3-HaHb), 2.10 (s, 3H,
CH3), 2.12 (m, 1H, chromane-3-HaHb), 2.15 (s, 3H, CH3), 2.20 (s, 3H, CH3), 2.59 (m, 2H,
linker CH2), 2.66 (m, 2H, chromane 4-CH2), 3.12 (m, 2H, CH2NHBoc), 3.58 (d, J = 2.2 Hz,
2H, CH2CO), 3.95, 4.05 (AB system, J = 9.2 Hz, 2H, OCH2), 4.50 (br s, 1H, NHBoc), 7.00
(d, J = 8.9 Hz, 2H, Harom), 7.42 (d, J = 8.9 Hz, 2H, Harom), 7.55 (br t, J = 2.2 Hz, 1H, CH=C),
8.09 (s, 1H, NH). 13C NMR (100 MHz, CDCl3): δ 12.2 (CH3), 15.1 (CH3), 16.0 (CH3), 20.8
(CH2), 22.9 (CH3), 26.8 (CH2), 28.6 (t-Bu), 28.8 (CH2), 30.0 (CH2), 30.1 (CH2), 30.2 (CH2),
30.3 (CH2), 35.3 (CH2), 40.7 (CH2), 73.0, 74.2, 115.6 (CHarom), 116.9, 121.4, 122.3, 126.8,
131.1, 131.7, 132.2 (CHarom), 133.6, 135.1 (CH=C), 149.1, 156.1, 161.0, 171.0 (C=O), 174.0
(C=O). ESI-MS (pos. mode): m/z = 627.34 [M+Na]+. Anal. Calcd for C36H48N2O6 (604.79):
C, 71.50; H, 8.00; N, 4.63. Found: C, 71.75; H, 8.08; N, 4.54.
5.1.8. (E)-3-(4-((6-(6-aminohexyl)-2,5,7,8-tetramethylchroman-2-yl)methoxy)
benzylidene)pyrrolidine-2,5-dione hydrochloride (10)
To a solution of compound 9 (172 mg, 0.28 mmol) in CH2Cl2 (5 mL) was added a saturated
solution of HCl in EtOAc (5 mL). The mixture was stirred at room temperature for 1 h and the
resulting suspension was filtered and washed with Et2O, giving after drying 89 mg (0.16
mmol, 57% yield) of white powder. M.p. 250-252 °C. IR (KBr) υ (cm-1): 3422, 3218, 2931,
1774, 1698, 1652, 1602, 1510, 1343, 1249, 1186, 1165, 1117, 1071, 835. 1H NMR (400 MHz,
DMSO-d6): δ 1.31 (s, 3H, CH3), 1.36 (m, 6H, linker CH2), 1.56 (m, 2H, linker CH2), 1.86 (m,
1H, chromane-3-HaHb), 1.98 (s, 3H, CH3), 2.03 (m, 1H, chromane-3-HaHb), 2.10 (s, 6H, CH3
x 2), 2.59 (m, 2H, CH2), 2.62 (m, 2H, CH2), 2.76 (m, 2H, CH2), 3.59 (d, J = 2.0 Hz, 2H,
CH2CO), 4.06 (m, 2H, OCH2), 7.06 (d, J = 8.1 Hz, 2H, Harom), 7.33 (br s, 1H, CH=C), 7.55 (d,
+
J = 8.1 Hz, 2H, Harom), 7.85 (br s, 3H, NH3 ), 11.34 (br s, 1H, NH). 13C NMR (100 MHz,
DMSO-d6): δ 11.9 (CH3), 14.6 (CH3), 15.6 (CH3), 20.0 (CH2), 21.6 (CH3), 25.7 (CH2), 27.0
(CH2), 28.1 (CH2), 29.0 (CH2), 29.3 (CH2), 29.6 (CH2), 34.8 (CH2), 38.8 (CH2), 72.6, 74.0,
115.3 (CHarom), 116.6, 121.0, 124.1, 127.0, 130.0, 131.0, 131.4 (CH=C), 132.0 (CHarom),
17

132.5, 148.5, 160.0, 172.1 (C=O), 175.8 (C=O). ESI-MS (pos. mode): m/z = 505.30 [M+H]+.
Anal. Calcd for C31H41ClN2O4, ½ H2O (550.14): C, 67.68; H, 7.70; N, 5.09. Found: C, 67.34;
H, 7.47; N, 5.14.
(S)-10 and (R)-10 were prepared in three steps from respectively (S)-7 and (R)-7. These
compounds showed the same spectral data than the racemic compound 10. Specific data:
(S)-10: [α]20D +11.6 (c 0.1 in CHCl3:MeOH, 9:1). Anal. Calcd for C31H41ClN2O4, 2/3 H2O: C,
67.31; H, 7.71; N, 5.06. Found: C, 67.33; H, 7.54; N, 5.04.
(R)-10: [α]20D -11.6 (c 0.1 in CHCl3:MeOH, 9:1). Anal. Calcd for C31H41ClN2O4, H2O: C,
66.59; H, 7.75; N, 5.01. Found: C, 66.53; H, 7.49; N, 4.88.
5.1.9 methyl 2,5,7,8-tetramethylchromane-2-carboxylate (12)
A solution of 2,3,5-trimethylphenol (25.00 g, 183.57 mmol) in methyl methacrylate (100 mL,
937.87 mmol) was cooled in an ice bath and paraformaldehyde (6.00 g, 199.80 mmol) was
added. The mixture was stirred under argon for 15 minutes and acetic acid (5.50 mL, 95.99
mmol) and dibutylamine (3.80 mL, 22.55 mmol) were added and the reaction medium was
stirred at 0 °C for further 15 minutes. Ice bath was removed and the mixture was heated at
reflux for 48 h. The yellow solution was cooled down to room temperature, and then let at 4
°C for 12 h. The resulting suspension was filtered and the crystals were washed with cooled
MeOH and cooled Et2O to give after drying 19.15 g (77.12 mmol, 42% yield) of white
crystals. M.p. 105-107 °C. IR (ATR): ν (cm-1): 2965, 2920, 1726, 1568, 1447, 1310, 1281,
1206, 1194, 1128, 1109. 1H NMR (400 MHz, CDCl3): δ 1.63 (s, 3H, CH3), 1.88 (m, 1H,
chromane 3-HaHb), 2.13 (s, 3H, CH3), 2.16 (s, 3H, CH3), 2.22 (s, 3H, CH3), 2.40-2.68 (m, 3H,
chromane 3-HaHb + chromane 4-CH2), 3.69 (s, 3H, OCH3), 6.58 (s, 1H, Harom). 13C NMR (100
MHz, CDCl3): δ 11.5 (CH3), 18.8 (CH3), 19.8 (CH3), 20.5 (CH2), 25.6 (CH3), 30.4 (CH2),
52.5 (OCH3), 77.5, 116.5, 122.0, 123.4 (CHarom), 133.3, 135.0, 151.5, 174.5 (C=O). ESI-MS
(pos. mode): m/z = 271.14 [M+Na]+. Anal. Calcd for C15H20O3 (248.14): C, 72.55; H, 8.12.
Found: C, 72.69; H, 8.16.
5.1.10 methyl 6-(6-bromohexanoyl)-2,5,7,8-tetramethylchromane-2-carboxylate (13)
Preparation of the intermediate acyl chloride: a solution of 6-bromohexanoic acid (19.60 g,
100 mmol) and thionyl chloride (21.6 mL, 291 mmol) in dry CH2Cl2 (50 mL) was refluxed
under argon for 3 hours. The solvent was removed under vacuum for 3 h to remove the excess
of thionyl chloride. The resulting colorless liquid was then dissolved in CH2Cl2 (15 mL) and
formic acid (0.75 mL, 20.00 mmol) was added to remove the last traces of thionyl chloride.
The mixture was stirred 2 h at room temperature until no bubbling occurred and used without
further purification.
Friedel-Crafts acylation: To the above solution were added AlCl3 (10.50 g, 78.00 mmol) and
anhydrous dichloromethane (10 mL). A solution of 12 (5.00 g, 20.13 mmol) in CH2Cl2 (35
mL) was then added dropwise in 20 minutes at 0 °C. 15 minutes after the end of the addition
were added EtOAc (250 mL) and water (80 mL). The organic layer was then washed twice
with water (80 mL), and with a 5% aqueous solution of potassium bicarbonate (10 x 30 mL)
to remove the excess of hydrolyzed acyl chloride, and then with water (50 mL). The organic
layer was dried (MgSO4), filtered and concentrated under vacuum to give a brownish liquid
which was further purified by column chromatography (eluent: cyclohexane/EtOAc, 100/0 to
18

88/12) to afford 6.00 g (14.11 mmol, 70% yield) of a colorless liquid which crystallized upon
storage at 4 °C. M.p. 68-70 °C. IR (ATR): ν (cm-1): 2930, 2864, 1746, 1693, 1460, 1435,
1240, 1138, 1111. 1H NMR (400 MHz, CDCl3): δ 1.51 (m, 2H, linker-CH2), 1.61 (s, 3H,
CH3), 1.73 (m, 2H, linker-CH2), 1.79-1.95 (m, 3H, linker-CH2 + chromane 3-HaHb), 1.98 (s,
3H, CH3), 2.06 (s, 3H, CH3), 2.14 (s, 3H, CH3), 2.39-2.50 (m, 2H, chromane 4-HaHb +
chromane 3-HaHb), 2.60 (m, 1H, chromane 4-HaHb), 2.67 (t, J = 7.3 Hz, 2H, CH2CO), 3.42 (t,
J = 6.6 Hz, 2H, CH2Br), 3.67 (s, 3H, OCH3). 13C NMR (100 MHz, CDCl3): δ 11.5 (CH3),
15.7 (CH3), 16.7 (CH3), 20.6 (CH2), 22.7 (CH2), 25.5 (CH3), 27.9 (CH2), 30.4 (CH2), 32.8
(CH2), 33.7 (CH2), 45.6 (CH2), 52.6 (OCH3), 77.5 (Cquat), 117.1, 122.9, 128.0, 129.7, 135.6,
151.6, 174.1 (ester C=O), 211.5 (ketone C=O). ESI-MS (pos. mode): m/z = 447.12 [M+Na,
79Br]+, 449.12 [M+Na, 81Br]+. Anal. Calcd for C21H29BrO4 (425.36): C, 59.30; H, 6.87. Found:
C, 59.36; H, 6.87.
5.1.11 (6-(6-bromohexyl)-2,5,7,8-tetramethylchroman-2-yl)methanol (14)
To a suspension of LiAlH4 (1.82 g, 48.00 mmol) in dry Et2O (30 mL) under argon was added
AlCl3 (4.27 g, 32.02 mmol). Five minutes later, a solution of AlCl3 (2.14 g, 16.05 mmol) and
13 (6.82 g, 16.03 mmol) in dry Et2O (90 mL) was slowly added. The mixture was stirred
under argon at room temperature for 2 h, and the reaction was then quenched with cold water
(140 mL) and 6 M aqueous H2SO4 solution (80 mL). The aqueous layer was extracted twice
with EtOAc (100 mL), and the combined organic layers were washed with 2 M Na2CO3
aqueous solution (70 mL), dried (MgSO4) and the solvent was removed under vacuum to give
a yellow liquid which was purified by column chromatography (eluent: cyclohexane/EtOAc,
100/0 to 88/12) to afford 4.42 g (11.5 mmol, 72% yield) of colorless liquid. 1H NMR (200
MHz, CDCl3): δ 1.23 (s, 3H, CH3), 1.46 (m, 6H, linker CH2), 1.66-2.08 (m, 5H, linker CH2 +
chromane 3-CH2 + CH2OH), 2.12 (s, 3H, CH3), 2.17 (s, 3H, CH3), 2.21 (s, 3H, CH3), 2.55-
2.79 (m, 4H, linker CH2 + chromane 4-CH2), 3.43 (t, J = 6.8 Hz, 2H, CH2Br), 3.58, 3.68 (AB
system, J = 11.4 Hz, 2H, CH2OH). Other analyses were in accordance with published results
[13].
5.1.12. (6-(6-azidohexyl)-2,5,7,8-tetramethylchroman-2-yl)methanol (15)
A solution of 14 (4.81 g, 12.55 mmol) and NaN3 (1.63 g, 25.1 mmol) in dry DMF (50 mL)
was heated at 80 °C for 3 h. The mixture was cooled to room temperature then diluted with
EtOAc (60 mL), washed with water (5 x 30 mL) and with brine (2 x 30 mL). The organic
layer was then dried (MgSO4), and the solvent was removed under vacuum to afford 3.76 g
(10.88 mmol, 87% yield) of 15 as a colorless liquid which was used without further
purification. 1H NMR (200 MHz, CDCl3): δ 1.23 (s, 3H, chromane CH3), 1.45 (m, 6H, linker
CH2), 1.54-1.82 (m, 4H, linker CH2 + chromane 3-HaHb + CH2OH), 1.98 (m, 1H, chromane
3-HaHb), 2.12 (s, 3H, CH3), 2.17 (s, 3H, CH3), 2.21 (s, 3H, CH3), 2.66 (m, 4H, linker CH2 +
chromane 4-CH2), 3.28 (t, J = 6.8 Hz, 2H, CH2N3), 3.62 (m, 2H, CH2O). Other analyses were
in accordance with published results [13].
5.1.13. (6-(6-aminohexyl)-2,5,7,8-tetramethylchroman-2-yl)methanol (16)
19

In a round bottom flask was first introduced 10% palladium on carbon (352 mg), in order to
avoid any ignition of the solvent. A solution of 15 (3.52 g, 10.19 mmol) in EtOAc (30 mL)
was then added. The mixture was stirred under a hydrogen atmosphere for 18 h at room
temperature, then filtered on celite®, and the resulting solution was concentrated. Column
chromatography (eluent: CH2Cl2 (+1 % Et3N)/MeOH, 100/0 to 90/10 in 10 minutes, then 30
minutes in isocratic mode) afforded 2.61 g (8.17 mmol, 80% yield) of slightly yellow liquid.
IR (ATR): 2920, 2851, 1570, 1449, 1312, 1105, 1063. 1H NMR (400 MHz, CDCl3): δ 1.23 (s,
3H, CH3), 1.43 (m, 6H, linker CH2), 1.48 (m, 2 H, linker CH2), 1.74 (m, 1H, chromane 3-
HaHb), 2.00 (m, 3H, chromane 3-HaHb + NH2), 2.12 (s, 3H, CH3), 2.16 (s, 3H, CH3), 2.20 (s,
3H, CH3), 2.58-2.75 (m, 6H, chromane 4-CH2 + linker CH2 + CH2NH2), 3.58, 3.65 (AB
system, 2H, J = 11.3 Hz, CH2OH). 13C NMR (100 MHz, CDCl3): δ 12.2 (CH3), 15.1 (CH3),
16.0 (CH3), 20.7 (2 pics: CH3 + CH2), 26.9 (CH2), 28.0 (CH2), 30.1 (CH2), 30.3 (CH2), 33.2
(CH2), 40.7 (CH2), 42.0 (CH2), 69.6, 75.3, 117.0, 122.2, 131.1, 131.8, 133.5, 149.0. ESI-MS
(pos. mode): m/z = 320.26 [M+H]+. Anal. Calcd for C20H33NO2, ¼ H2O (323.99): C, 74.14;
H, 10.42; N, 4.32. Found: C, 74.42; H, 10.32; N, 4.41.
5.1.14 tert-butyl (6-(2-(hydroxymethyl)-2,5,7,8-tetramethylchroman-6-yl)hexyl)carbamate (7)
To a solution of compound 16 (1.59 g, 4.98 mmol) in dry CH2Cl2 (15 mL) was added under
stirring at room temperature Boc2O (773 µL, 5.0 mmol). After 45 minutes, the solvent was
removed under vacuum, and the crude product was diluted in EtOAc (20 mL), the solution
was washed with an aqueous 5% citric acid solution (2 x 20 mL), water (10 mL), and brine
(10 mL). The organic layer was dried (MgSO4) and the solvent was removed under vacuum to
afford a yellow liquid which was purified by column chromatography (eluent:
cyclohexane/EtOAc, 100/0 to 80/20 in 20 minutes then 20 minutes in isocratic mode) to
afford 7 (2.08 g, 4.96 mmol, 100% yield) as a colorless liquid. 1H NMR (400 MHz, CDCl3):
1.23 (s, 3H, CH3), 1.45 (s, 9H, t-Bu), 1.36-1.53 (m, 6 H, linker CH2), 1.64 (br s, 3H, linker
CH2 + CH2OH), 1.72 (m, 1H, chromane 3-HaHb), 2.01 (m, 1H, chromane 3-HaHb), 2.12 (s,
3H, CH3), 2.16 (s, 3H, CH3), 2.20 (s, 3H, CH3), 2.57-2.70 (m, 4H, chromane 4-CH2 + linker
CH2), 3.12 (t, J = 6.2 Hz, CH2NHBoc), 3.59, 3.65 (AB system, J = 11.2 Hz, 2H, CH2OH),
4.48 (br s, 1H, NHBoc). Other analyses were in accordance with published results [13].
5.1.15. Deracemization of compound 7
5.1.15.1. Preparation of the supported Lipase
According to Couladouros et al. [17], lipase PS-30 (3 g) and Hyflo Supercel® (10 g) were
added to a buffered solution (10 mL) of Na2HPO4 (117.12 mg) and KH2PO4 (82.66 mg) in
distilled water (30 mL, pH 7). The suspension was shaken vigorously then dried under high
vacuum for 24 h.
20

5.1.15.2. (S)-4-((6-(6-((tert-butoxycarbonyl)amino)hexyl)-2,5,7,8-tetramethylchroman-2-
yl)methoxy)-4-oxobutanoic acid: ((S)-enriched 19), first lipase reaction.
To a stirred solution of 7 (1.00 g, 2.38 mmol) in MTBE (6 mL) were added succinic
anhydride (346 mg, 3.46 mmol) and supported Amano Lipase PS (720 mg). After 18 h of
stirring at room temperature, the precipitate was filtered, washed with MTBE (5 mL) and the
filtrate was evaporated under reduced pressure. The residue was purified by flash column
chromatography (eluent: cyclohexane/EtOAc 100/0 to 80/20 in 20 minutes, then 20 minutes
in isocratic mode) to give 415 mg of (R)-enriched 7 (0.98 mmol, 44% yield) as a white
powder and 600 mg of (S)-enriched 19 (1.15 mmol, 52% yield) as a slightly trouble liquid. IR
(ATR) ν (cm-1): 3342, 2974, 2924, 2851, 1736, 1711, 1649, 1572, 1510, 1448, 1366, 1246. 1H
NMR (400 MHz, CDCl3): δ 1.29 (s, 3H, CH3), 1.43 (s, 9H, t-Bu), 1.33-1.52 (m, 8H, linker
CH2), 1.79 (m, 1H, chromane 3-HaHb), 1.93 (m, 1H, chromane 3-HaHb), 2.09 (s, 3H, CH3),
2.14 (s, 3H, CH3), 2.19 (s, 3H, CH3), 2.56-2.70 (m, 8H, succinate CH2 x 2 + chromane 4-CH2
+ linker CH2), 3.11 (m, 2H, CH2NHBoc), 4.09, 4.21 (AB system, J = 11.2 Hz, 2H, CH2O),
4.52 (br s, 1H, NHBoc). 13C NMR (100 MHz, CDCl3): δ 12.1 (CH3), 15.0 (CH3), 16.0 (CH3),
20.7 (CH2), 22.2 (CH3), 26.8 (CH2), 27.1 (CH2), 28.6 (t-Bu), 28.9 (CH2), 29.1 (CH2), 29.9
(CH2), 30.1 (CH2), 30.2 (CH2, 2 peaks), 40.8 (CH2), 69.3, 73.4, 116.7, 122.4, 131.0, 131.5,
133.4, 149.1, 172.0, 176.6. ESI-MS (pos. mode): m/z = 542.31 [M+Na]+. Anal. Calcd for
C29H45NO7 (519.68): C, 67.03; H, 8.73; N, 2.70. Found: C, 66.95; H, 8.88; N, 2.56.
5.1.15.3. General procedure for the saponification of 19.
To a stirred solution of (S)-enriched 19 in a 1:1 mixture of THF and MeOH (10 mL per
mmol) was added a 3 M aqueous solution of LiOH (3.0 eq). After 30 minutes of stirring at
room temperature, the reaction mixture was quenched with a saturated aqueous solution of
NH4Cl (10 mL), and diluted in EtOAc (25 mL per mmol). The solution was washed with
distilled water (3 x 15 mL per mmol), the organic layer was dried (MgSO4), and the solvent
was removed under vacuum to afford (S)-enriched 7 which was used without further
purification. The NMR spectrum was in accordance with the expected data.
5.1.15.4. General procedure for the resolution of the racemic compound 7 (second and third
lipase reactions).
Both (S)-enriched compound 7 and (R)-enriched compound 7 were subjected twice to the
same procedure as before. The final yields after the three enzymatic resolutions are given
below, taking into account that the maximum theoretical yield for each is 50%. The resolved
compounds exhibited identical spectroscopic data to the racemic alcohol 7.
(S)-7 : Yield : 22 % [α]20D +1.1° (c 0.1 in CHCl3), ee = 98%
(R)-7 : Yield : 23 %, [α]20D -1.1° (c 0.1 in CHCl3), ee = 98%
5.1.15.5. General procedure for the measurement of enantiomeric excess of 7.
According to reported procedure [20], in a NMR tube was introduced a solution of 7 (3-7 mg,
7.15 µmol - 16.68 µmol) in C5D5N (0.1 mL) and CDCl3 (0.1 mL), and (R)-Mosher acyl
chloride (5 µL, 26.72 µmol) was added. After 3 h reaction, C6D6 (0.5 mL) and CDCl3 (0.1
mL) were added. The NMR signals of interest are displayed in the 4.00-4.35 area as two AB
systems for both diastereomers.
21

5.1.16. Verification of the conservation of the enantiomeric excess in acidic conditions
5.1.16.1. (S)-(6-(6-aminohexyl)-2,5,7,8-tetramethylchroman-2-yl)methyl acetate
hydrochloride ((S)-21).
To a stirred solution of (S)-7 (100 mg, 0.24 mmol) in dry CH2Cl2 (6 mL) were added Ac2O
(24 µL, 0.26 mmol), pyridine (21 µL, 0.26 mmol) and DMAP (3 mg, 0.024 mmol). The
mixture was stirred under argon for 16 h and the solvents were removed under reduced
pressure. The crude product was dissolved in dry CH2Cl2 (4 mL) and a 1.5 M solution of HCl
in EtOAc was added (2 mL, 3.0 mmol). After 3 h stirring at room temperature, the solvents
were removed under reduced pressure, and the HCl was co-evaporated twice with CH2Cl2 to
give a pale-yellow powder, which was suspended in CH2Cl2, filtered, washed with Et2O (2×5
mL) and with CH2Cl2 (2×2 mL) to give 87 mg (0.22 mmol, 92% yield) of white powder.
M.p.: 130-132 °C. IR (ATR): ν (cm-1) 2920, 1732, 1658, 1520, 1383, 1240. 1H NMR (400
MHz, CDCl3): δ 1.28 (s, 3H, CH3), 1.45 (br s, 6H, linker CH2), 1.80 (m, 3H, chromane 3-
HaHb + linker CH2), 1.93 (m, 1H, chromane 3-HaHb), 2.08 (s, 3H, CH3), 2.09 (s, 3H, CH3),
2.14 (s, 3H, CH3), 2.18 (s, 3H, CH3), 2.55-2.67 (m, 4H, chromane 4-CH2 + linker CH2), 3.00
+
+
(m, 2 H, CH2NH3 ), 4.07, 4.15 (AB system, J = 11.2 Hz, 2H, CH2O), 8.28 (br s, 3H, NH3 ).
13C NMR (100 MHz, CDCl3): δ 12.1 (CH3), 15.0 (CH3), 16.0 (CH3), 20.6 (CH2), 21.0 (CH3),
21.9 (CH3), 26.6 (CH2), 27.9 (CH2), 28.7 (CH2), 29.6 (CH2), 30.0 (CH2), 30.1 (CH2), 40.1
(CH2), 69.0, 73.5, 116.6, 122.4, 130.6, 131.5, 133.4, 149.1, 171.0 (C=O). ESI-MS (pos.
mode): m/z = 362.27 [M]+. Anal. Calcd for C22H36ClNO3 (397.98): C, 66.40; H, 9.12; N, 3.52.
Found: C, 66.28; H, 9.15; N, 3.48.
5.1.16.2. Recovery of compound (S)-7 after acidic treatment: conservation the absolute
configuration.
To a solution of (S)-21 (55 mg, 0.14 mmol) in dry CH2Cl2 (2 mL) were added Et3N (195 µL,
1.4 mmol) and Boc2O (61 mg, 0.28 mmol) at room temperature. The mixture was stirred for 2
h, and the solvent was removed under reduced pressure. The crude product was diluted in a
1:1 mixture of THF and MeOH (4 mL) and an aqueous solution of 3 M LiOH (0.23 mL, 0.70
mmol) was added. After 1 h stirring at room temperature, the reaction was quenched by
addition of a saturated solution of NH4Cl (10 mL), diluted with EtOAc (50 mL) and washed
with water (3 x 10 mL). The organic layer was dried (MgSO4) and the solvent was removed
under reduce pressure. The residue was purified by column chromatography (eluent:
cyclohexane/EtOAc 100/0 to 30/70 in 20 minutes) to give 53 mg of colorless liquid (0.13
mmol, 93% yield). This compound exhibited the same spectral data than the previously
described molecule 7. Furthermore, its optical rotation and the deduced ee after reaction with
(R)-Mosher acyl chloride were identical to starting compound (S)-7.
5.1.17 4-((6-(6-azidohexyl)-2,5,7,8-tetramethylchroman-2-yl)methoxy)benzaldehyde (27)
Preparation of the triflate intermediate: to an ice cooled solution of anhydrous pyridine (1.15
mL, 14.29 mmol) in dry CH2Cl2 (10 mL) was added dropwise trifluoromethanesulfonic
anhydride (560 µL, 3.3 mmol) under argon. A solution of 15 (822 mg, 2.38 mmol) in CH2Cl2
(10 mL) was added 5 minutes later to the reaction mixture. The solution was stirred for 30
22

minutes at 0 °C and the solvent was evaporated. The residue was dissolved in EtOAc (30
mL), the solution was washed with water (3 x 20 mL), 5% aqueous NaHCO3 solution (3 x 20
mL), and brine (3 x 20 mL). The organic phase was then dried (MgSO4) and the solvent was
removed under vacuum to afford an orange liquid.
Condensation with 4-hydroxybenzaldehyde: to a solution of the above triflate in dry DMF (5
mL) were added 4-hydroxybenzaldehyde (291 mg, 2.38 mmol) and K2CO3 (658 mg, 4.76
mmol). The reaction mixture was stirred under argon at room temperature for 5 days, and then
EtOAc (50 mL) was added. The organic layer was washed with water (5 x 20 mL), 5%
aqueous NaHCO3 solution (2 x 20 mL), water (2 x 20 mL), dried (MgSO4) and the solvent
was evaporated under vacuum to give a yellow liquid. The residue was purified by column
chromatography (eluent: cyclohexane/EtOAc, 100:0 to 90:10) to afford 26 (623 mg, 1.39
mmol, 58 % yield) as a colorless liquid. IR (film) υ (cm-1): 3379, 2930, 2094, 1691, 1601,
1574, 1508, 1310, 1248, 1157, 1109, 1038. 1H NMR (250 MHz, CDCl3): δ 1.44 (br s, 9H,
CH3 + linker CH2), 1.63 (m, 2H, linker CH2), 1.92 (m, 1H, chromane-3-HaHb), 2.10 (s, 3H,
CH3), 2.12 (m, 1H, chromane-3-HaHb), 2.16 (s, 3H, CH3), 2.20 (s, 3H, CH3), 2.63 (m, 4H,
chromane 4-CH2 + linker CH2), 3.28 (t, J = 6.8 Hz, 2H, CH2N3), 3.98, 4.08 (AB system, J =
9.3 Hz, 2H, OCH2), 7.03 (d, J = 8.6 Hz, 2H, Harom), 7.82 (d, J = 8.6 Hz, 2H, Harom), 9.89 (s,
1H, CHO). 13C NMR (62.9 MHz, DMSO-d6): δ 12.2 (CH3), 15.1 (CH3), 16.0 (CH3), 20.7
(CH2), 22.9 (CH3), 26.8 (CH2), 28.8 (CH2), 29.0 (CH2), 29.8 (CH2), 30.1 (CH2), 30.2 (CH2),
51.6 (CH2-N3), 73.1, 74.1, 115.2 (CHarom), 116.9, 122.4, 130.3, 131.0, 131.7, 132.1 (CHarom),
133.6, 149.1, 164.2, 190.9 (CHO). ESI-MS (pos. mode): m/z = 472.25 [M+Na]+. Anal. Calcd
for C27H35N3O3 (449.60): C, 72.13; H, 7.85; N, 9.35. Found: C, 72.50; H, 7.92; N, 9.09.
5.1.18. (E)-3-(4-((6-(6-azidohexyl)-2,5,7,8-tetramethylchroman-2-yl)methoxy)benzylidene)
pyrrolidine-2,5-dione (28)
To a solution of 26 (0.52 mmol) in anhydrous MeOH (10 mL) was added compound 2 (187
mg, 0.52 mmol) and the mixture was stirred under argon at room temperature for 12 h. The
suspension was filtered and washed with cooled MeOH to give 189 mg (0.36 mmol, 69%
yield) of white powder. M.p. 130-132 °C. IR (KBr) υ (cm-1): 2927,2853, 2085, 1746, 1699,
1601, 1514, 1456, 1352, 1306, 1244, 1175, 1111. 1H NMR (400 MHz, CDCl3): δ 1.43 (s, 3H,
CH3), 1.45 (br s, 6H, linker CH2), 1.63 (m, 2H, linker CH2), 1.91 (m, 1H, chromane-3-HaHb),
2.10 (s, 3H, CH3), 2.13 (m, 1H, chromane 3-HaHb), 2.16 (s, 3H, CH3), 2.21 (s, 3H, CH3), 2.61
(m, 2H, linker CH2), 2.67 (m, 2H, chromane 4-CH2), 3.28 (t, J = 6.9 Hz, 2H, CH2N3), 3.58 (d,
J = 2.2 Hz, 2H, CH2CO), 3.95, 4.05 (AB system, J = 9.3 Hz, 2H, OCH2), 7.00 (d, J = 8.9 Hz,
2H, Harom), 7.42 (d, J = 8.9 Hz, 2H, Harom), 7.56 (m, 1H, CH=C), 8.38 (br s, 1H, NH). 13C
NMR (100 MHz, CDCl3): δ 12.2 (CH3), 15.1 (CH3), 16.0 (CH3), 20.7 (CH2), 22.9 (CH3), 26.8
(CH2), 28.8 (CH2), 29.0 (CH2), 29.8 (CH2), 30.1 (CH2), 30.2 (CH2), 35.3 (CH2), 51.6
(CH2N3), 73.0, 74.2, 115.6 (CHarom), 116.9, 121.4, 122.4, 126.8, 130.9, 131.7, 132.2 (CHarom),
133.5, 135.0 (CH=C), 149.1, 161.0, 171.3 (C=O), 174.2 (C=O). ESI-MS (pos. mode): m/z =
531.29 [M+H]+, 553.28 [M+Na]+. Anal. Calcd for C31H38N4O4 (530.67): C, 70.16; H, 7.22; N,
10.56. Found: C, 70.50; H, 7.22; N, 10.20.
23

5.1.19. (2R,3R,4S,5R,6R)-2-(acetoxymethyl)-6-(prop-2-yn-1-yloxy)tetrahydro-2H-pyran-
3,4,5-triyl triacetate (29)
Acetylation of α-
as follows: to a suspension of α-
D
-glucose: procedure reported by Wolfrom and Thompson [23] was adapted
-glucose (3.00 g, 16.65 mmol) in acetic anhydride (20.46
D
mL, 216.47 mmol) was added sodium acetate (1.50 g, 18.32 mmol) under argon. The mixture
was refluxed during 2 h and cooled to room temperature before being cooled in an ice bath.
Water (50 mL) was added and the suspension was filtered, washed with water (3 x 20 mL),
and carefully dried to give 3.618 g (9.27 mmol, 56% yield) of white powder which was used
in the next step without further purification.
Propargylation at the 1-position: this step was realized according to the work of Murray et
al. [24]. Spectral data of the compound agreed with the reported data.
5.1.20. (2R,3R,4S,5R,6R)-2-(acetoxymethyl)-6-((1-(6-(2-((4-((E)-(2,5-dioxopyrrolidin-3-
ylidene)methyl)phenoxy)methyl)-2,5,7,8-tetramethylchroman-6-yl)hexyl)-1H-1,2,3-triazol-4-
yl)methoxy)tetrahydro-2H-pyran-3,4,5-triyl triacetate (30)
To a mixture of 28 (138 mg, 0.26 mmol) and 29 (100 mg, 0.26 mmol) were added t-BuOH
(20 mL) and MeOH (2 mL) and sonication was applied. To the resulting solution was added a
mixture of CuSO4 (58 mg, 0.36 mmol) and sodium ascorbate (71 mg, 0.36 mmol) in water
(0.5 mL). The reaction mixture was stirred under argon at room temperature for 12 h, and the
solvents were evaporated. The residue was dissolved in EtOAc (30 mL) and the solution was
washed with water (3 x 20 mL), dried (MgSO4) and concentrated under vacuum to give a
yellow residue which was purified by column chromatography (eluent: cyclohexane/EtOAc,
100:0 to 20:80) to afford 171 mg (0.19 mmol, 74 % yield) of white powder. M.p. 88-90 °C.
IR (ATR): ν (cm-1) 2928, 2853, 1748, 1713, 1601, 1510, 1217, 1034. 1H NMR (400 MHz,
CDCl3): δ 1.42 (m, 9H, chromane CH3 + linker CH2), 1.91 (m, 3H, chromane 3-HaHb +
triazole-CH2-CH2), 1.98 (s, 3H, COCH3), 2.00 (s, 3H, COCH3), 2.03 (s, 3H, COCH3), 2.09 (s,
6H, COCH3 + chromane CH3), 2.12 (m, 1H, chromane 3-HaHb), 2.14 (s, 3H, chromane CH3),
2.19 (s, 3H, chromane CH3), 2.59 (m, 2H, linker CH2), 2.66 (m, 2H, chromane 4-H), 3.59 (d,
J = 2.3 Hz, 2H, CH2CO), 3.74 (X part of an ABX system, 1H, glucose 5-H), 3.95, 4.04 (AB
system, J = 9.3 Hz, 2H, OCH2), 4.15, 4.27 (AB part of an ABX system, J = 12.3, 4.7, 2.1 Hz,
2H, glucose 6-H + 6’-H), 4.35 (br t, J = 7.1 Hz, 2H, linker triazole-CH2), 4.69 (d, J = 8.0 Hz,
1H, glucose 1-H), 4.84, 4.95 (AB system, J = 12.6 Hz, 2H, OCH2), 5.01 (dd, J = 9.5, 8.0 Hz,
1H, glucose 2-H), 5.10 (t, J = 9.6 Hz, 1H, glucose 4-H), 5.20 (t, J = 9.5 Hz, 1H, glucose 3-H),
6.99 (d, J = 8.9 Hz, 2H, Harom), 7.42 (d, J = 8.9 Hz, 2H, Harom), 7.50 (br s, 1H, triazole H),
7.55 (t, J = 2.2 Hz, 1H, C=CH), 7.88 (s, 1H, NH). 13C NMR (100 MHz, CDCl3): δ 12.2
(CH3), 15.1 (CH3), 16.0 (CH3), 20.7 (2 x COCH3), 20.8 (COCH3), 20.9 (COCH3), 22.9 (CH3),
26.6 (CH2), 28.8 (CH2), 29.6 (CH2), 30.0 (CH2), 30.1 (CH2), 30.5 (CH2), 35.3 (CH2), 50.7
(CH2N), 62.0, 63.2, 68.5 (glucose CH), 71.4 (glucose CH), 72.0 (glucose CH), 72.9 (glucose
CH), 73.0, 74.2, 100.2 (glucose 1-CH), 115.6 (phenyl CH) 116.9, 121.4, 122.4, 122.8 (triazole
CH), 126.9, 130.8, 131.7, 132.2 (phenyl CH), 133.5, 135.0 (CH=C), 144.2, 149.1, 161.0,
169.5 (C=O), 169.6 (C=O), 170.4 (C=O), 170.8 (C=O), 171.0 (C=O), 174.0 (C=O). ESI-MS
(pos. mode): m/z = 917.41 [M+H]+, 939.40 [M+Na]+, 955.37 [M+K]+. Anal. Calcd for
C48H60N4O14 (917.01): C, 62.87; H, 6.59; N, 6.11. Found: C, 62.39; H, 6.68; N, 6.05.
24

5.1.21. 3-((E)-4-((2,5,7,8-tetramethyl-6-(6-(4-((((2R,3R,4S,5S,6R)-3,4,5-trihydroxy-6-
(hydroxymethyl)tetrahydro-2H-pyran-2-yl)oxy)methyl)-1H-1,2,3-triazol-1-yl)hexyl)chroman-
2-yl)methoxy)benzylidene)pyrrolidine-2,5-dione (31)
A suspension of compound 30 (87 mg, 0.09 mmol) in anhydrous MeOH (5 mL) was
sonicated until dissolution. A solution of MeONa in MeOH (2.09 mL, 1.15 mg/mL) was
added at room temperature and stirred under argon for 1 h. Amberlite® IR 120H was added
until neutral pH and the suspension was filtered to give after evaporation of the filtrate an
orange liquid which was purified by reverse-phase column chromatography (eluent: H2O
(0.1% TFA)/MeOH, 90:10 to 10:90) to afford after lyophilisation 71 mg (0.09 mmol, 100 %
yield) of white powder. M.p. 126-128 °C. IR (ATR): ν (cm-1) 3387, 2914, 2855, 1717, 1684,
1600, 1340, 1306, 1250, 1165, 1076, 1043, 525. 1H NMR (400 MHz, DMSO-d6): δ 1.31 (m,
7H, chromane CH3 + linker CH2), 1.39 (m, 2H, linker CH2), 1.82 (m, 3H, chromane 3-HaHb +
triazole-CH2CH2), 1.97 (s, 3H, chromane CH3), 2.00 (m, 1H, chromane 3-HaHb), 2.08 (s, 6H,
chromane CH3), 2.50 (2H, linker CH2), 2.61 (m, 2H, chromane 4-H), 2.97 (t, J = 8.7 Hz, 1H,
glucose 2-H), 3.01-3.17 (m, 3H, glucose 3-H + 4-H + 5-H), 3.45, 3.70 (AB part of an ABX
system, J = 11.8, 6.2, 1.6 Hz, 2H, glucose 6-H + 6’-H), 3.59 (d, J = 2.1 Hz, 2H, CH2CO),
4.03, 4.08 (AB system, J = 10.1 Hz , 2H, OCH2), 4.14 (br s, 4H, 4 x OH), 4.25 (d, J = 7.8 Hz,
1H, glucose 1-H), 4.34 (t, J = 7.0 Hz, 2H, linker triazole-CH2), 4.62, 4.83 (AB system, J =
12.2 Hz, 2H, OCH2), 7.06 (d, J = 8.8 Hz, 2H, Harom), 7.33 (s, 1H, C=CH), 7.55 (d, J = 8.8 Hz,
2H, Harom), 8.11 (s, 1H, triazole H), 11.34 (s, 1H, NH). 13C NMR (100 MHz, DMSO-d6): δ
11.9 (CH3), 14.6 (CH3), 15.5 (CH3), 20.0 (CH2), 21.6 (CH3), 25.8 (CH2), 28.1 (CH2), 28.9
(CH2), 29.3 (CH2), 29.5 (CH2), 29.8 (CH2), 34.8 (CH2CO) 49.3 (CH2N triazole), 61.1
(CH2OH), 61.6 (OCH2 triazole), 70.1 (glucose CH), 72.6 (OCH2), 73.4 (glucose CH), 73.9
(OCquat), 76.7 (glucose CH), 77.0 (glucose CH), 102.1 (glucose 1-CH), 115.3 (phenyl CH),
116.6, 120.9, 124.1 (triazole CH), 124.1, 127.0, 130.0, 131.0, 131.4 (CH=C), 132.0 (phenyl
CH), 132.5, 143.7, 148.5, 160.0, 172.1 (C=O), 175.8 (C=O). ESI-MS (pos. mode): m/z =
771.36 [M+Na]+. Anal. Calcd for C40H52N4O10, 2H2O (784.39): C, 61.21; H, 7.19; N, 7.14.
Found: C, 61.01; H, 7.32; N, 6.95.
5.2. Biological evaluation on cancerous cell lines
5.2.1. Cell Lines and Reagents
MCF-7, MDA-MB-231, MDA-MB-453 and Hs578T human breast cancer cell lines, HCT116
colon cancer cell line, and U87 glioblastoma cell line were obtained from American Type
Culture Collection (Manassas, VA, USA). These cell lines were grown at 37 °C under 5%
CO2 in Dulbecco’s Modified Eagle Medium (DMEM, Gibco®, Thermo Fisher Scientific,
Villebon-sur-Yvette, France) for MCF-7, Hs578T, U87, and HCT116, and Roswell Park
Memorial Institute (RPMI) 1640 medium (Gibco®) for MDA-MB-231 and MDA-MB-453.
These phenol red containing media were supplemented with 2 mM L-glutamine (Sigma-
Aldrich, Saint-Quentin Fallavier, France) and 10% fetal calf serum (FCS, Sigma-Aldrich). In
addition, the medium was supplemented with 10 µg/mL insulin for Hs578T cell line. For U87
cells, the medium was supplemented with 1.25 mM sodium pyruvate, 50 µM non-essential
amino acid, 10.8 µg/mL L-serine, 20µg/mL L-asparagine, 0.4% vitamins (100X stock
25

solution) and 1% essential amino acids (50X stock solution) (Thermo Fisher Scientific).
Compounds to be tested were dissolved in Dimethyl sulfoxide (DMSO, Sigma-Aldrich).
5.2.2. Cell Treatments
The different cell lines were seeded at the appropriate density in their respective medium
supplemented with 10% FCS and incubated for 24 h to allow for cell attachment. For crystal
violet staining, cells were seeded in 96-well plates at a density of 1.5 x 104 cells/well (MDA-
MB-231, MCF-7, Hs578T, and U87 cells) or 1.2 x 104 cells/well (HCT116) or 2 x 104
cells/well (MDA-MB-453 cells). For flow cytometry analyses, MDA-MB-231 cells were
seeded in 60 mm-diameter culture dishes at 9 x 105 cells/dish. The cells were then treated in
their respective medium for 48 or 72 hours with the considered compounds at the indicated
concentrations. Control cells were treated with DMSO 0.090 % or 0.012 % to compare with
cells exposed with 45 µM TGZ or 6 µM compound 10, respectively.
5.2.3. Crystal Violet Staining Assay
At the end of the treatment, cells were washed with phosphate-buffered saline (PBS). Cells
were then fixed and stained for 20 min with 0.2% crystal violet (Sigma-Aldrich) in 2%
ethanol. This staining solution was filtered prior to use with a syringe-driven filter unit (0.2
µm pore size, VWR international, Radnor, PA, USA) to eliminate dye precipitates. Cells were
then washed with distilled water to remove excess dye. The plates were dried and the dye was
dissolved in 10% acetic acid. Cell number was determined by absorbance at 595 nm with a
VICTOR™ X3 multilabel plate reader (PerkinElmer, Courtaboeuf, France). In each
condition, the concentration leading to a decrease of 50% of the cell number (IC50) was
determined.
5.2.4. Cell Cycle Analysis
At the end of the treatment, culture medium containing the non-adherent cells was collected.
Adherent cells were washed with PBS, trypsined (0.05% trypsin-EDTA in PBS buffer,
Gibco®) and resuspended in a medium containing 10% FCS. Non-adherent cells and trypsined
cells were pooled and counted. 1 x 106 cells were centrifuged at 500 g for 5 min at room
temperature and the pellet was rinsed with PBS. Cells were then resuspended in 600 µL of
PBS and 1,400 µL of cold 100% ethanol were added drop by drop while vortexing the
suspension to permeabilize and fix the cells efficiently. 5 x 105 cells were then centrifugated
at 800 g for 10 min at room temperature. The pellet was resuspended in 500 µL of DNA
staining solution containing 200 µg of RNAse A (Invitrogen, Thermo Fisher Scientific) and 2
µg of propidium iodide (PI) (Thermo Fisher Scientific). Samples were incubated for 15 min in
the dark at room temperature. Finally, cellular DNA content was determined with a BD
FACSCaliburTM flow cytometer (BD Biosciences) using CellQuest™ Pro software (BD
Biosciences). 20,000 events were analyzed and the percentage of cells in each phase of the
cell cycle was determined.
26

5.2.5. Annexin V/propidium iodide (PI) apoptosis assay
At the end of the treatment, culture medium containing the non-adherent cells was collected.
Adherent cells were washed with PBS, trypsined (0.05% trypsin-EDTA in PBS buffer,
Gibco®) and resuspended in a medium containing 10% FCS. Non-adherent cells and trypsined
cells were pooled and centrifuged at 200 g for 10 min at room temperature and the pellet was
rinsed with PBS. 1 x 106 cells were centrifuged at 200 g for 10 min at room temperature and
then added to a binding buffer solution (0.1 M HEPES pH 7.4, 1.4 M NaCl, 25 mM CaCl2, pH
7.4) to facilitate Annexin V binding to phosphatidylserine. 2 µg of PI (Thermo Fisher
Scientific) and 2.5 µL of FITC Annexin V (BD Biosciences) were added to the cell
suspension and incubated for 15 min in the dark at room temperature. Action of Annexin V
was inhibited by addition of binding buffer solution. 10,000 events were analyzed and the
percentage of Annexin V and/or PI positive cells was determined with a BD FACSCaliburTM
flow cytometer (BD Biosciences) using CellQuest™ Pro software (BD Biosciences).
5.2.6. Statistical Analysis
The results of cell cycle analysis and Annexin V/PI apoptosis assay were analyzed using a
two-tailed unpaired t test with Welch’s correction. Differences in which P-value was less than
0.05 were statistically significant.
5.3. Biological evaluation on non tumoral hepatocytes
5.3.1. Cell culture
The human liver resection was collected in accordance with Article L.1211-2 of the French
Public Health Code, amended by Law No. 2004-800 of August 6, 2004. Human hepatocytes
were isolated from macroscopically healthy surgical liver waste after tumor resection by a
two-step collagenase perfusion. Single donor (58 years old Caucasian female)-derived
hepatocytes were seeded onto collagen-coated 96-well plates (7×104 cells/well) in DMEM
medium supplemented with FCS (10%), insulin (4 µg/ml), dexamethasone (1 µM) and
gentamycin (50 µg/ml). Cells were incubated overnight at 37°C under a humidified 5%
CO2/95% air atmosphere.
5.3.2. Cell treatment
The medium was replaced by a serum-free medium, consisted of DMEM supplemented with
insulin (4 µg/ml), dexamethasone (1 µM) and gentamycin (50 µg/ml). The cultures were
exposed to increasing doses of compound 10 at 9 concentrations ranging from 0 to 200 µM.
After 24 h of treatment, hepatocytes were incubated with MTT (0.5 mg/ml) for 2 h at 37°C.
The formazan crystals were then dissolved in DMSO and the absorbance was measured at 570
nm.
27

Acknowledgments
The authors thank Fabien Lachaud and François Dupire for spectroscopic measurements,
Sandrine Adach for elemental analysis, MENRT for a fellowship to Dorian Dupommier and
Claire Muller. The authors are grateful to Alicia Chateau and Muriel Barberi-Heyob for
having provided the U87 cell line. Technical help from Jessica Camera, Marianna Saribekyan,
and Marie Zaffino is also gratefully acknowledged.
Appendix A. Supplementary data
Supplementary data associated with this article can be found in the online version, at
http://dx.doi.org/xxxxx. These data include 1H and 13C NMR spectra of tested compounds,
and synthetic protocols in the frame of the attempts to prepare derivatives suitable for X-ray
analysis (preparation of compounds 22-26). Representative dose-dependent responses of
MDA-MB-231 cells to TGZ and compound 10 were also included. Representative dose-
dependent responses of U87 and HCT116 cells to compound 10 were also inserted.
References
[1] F. Bray, J. Ferlay, I. Soerjomataram, R.L. Siegel, L.A. Torre, A. Jemal, Global cancer statistics
2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185
countries, CA: A Cancer Journal for Clinicians. 68 (2018) 394–424.
https://doi.org/10.3322/caac.21492.
[2] K.L. Maughan, M.A. Lutterbie, P. Ham, Treatment of Breast Cancer, AFP. 81 (2010) 1339–1346.
[3] E.A. Musgrove, R.L. Sutherland, Biological determinants of endocrine resistance in breast cancer,
Nature Reviews Cancer. 9 (2009) 631–643. https://doi.org/10.1038/nrc2713.
[4] F.X. Claret, T.T. Vu, Trastuzumab: Updated Mechanisms of Action and Resistance in Breast
Cancer, Front. Oncol. 2 (2012). https://doi.org/10.3389/fonc.2012.00062.
[5] K.G. Chen, B.I. Sikic, Molecular Pathways: Regulation and Therapeutic Implications of
Multidrug Resistance, Clin Cancer Res. 18 (2012) 1863–1869. https://doi.org/10.1158/1078-
0432.CCR-11-1590.
[6] C. Yam, S.A. Mani, S.L. Moulder, Targeting the Molecular Subtypes of Triple Negative Breast
Cancer: Understanding the Diversity to Progress the Field, The Oncologist. 22 (2017) 1086–1093.
https://doi.org/10.1634/theoncologist.2017-0095.
[7] Mohd.J. Naim, Md.J. Alam, S. Ahmad, F. Nawaz, N. Shrivastava, M. Sahu, O. Alam, Therapeutic
journey of 2,4-thiazolidinediones as a versatile scaffold: An insight into structure activity
relationship, European Journal of Medicinal Chemistry. 129 (2017) 218–250.
https://doi.org/10.1016/j.ejmech.2017.02.031.
[8] Sucheta, S. Tahlan, P.K. Verma, Biological potential of thiazolidinedione derivatives of synthetic
origin, Chemistry Central Journal. 11 (2017) 130. https://doi.org/10.1186/s13065-017-0357-2.
[9] N. Chadha, M.S. Bahia, M. Kaur, O. Silakari, Thiazolidine-2,4-dione derivatives: Programmed
chemical weapons for key protein targets of various pathological conditions, Bioorganic &
Medicinal Chemistry. 23 (2015) 2953–2974. https://doi.org/10.1016/j.bmc.2015.03.071.
[10]S. Prost, F. Relouzat, M. Spentchian, Y. Ouzegdouh, J. Saliba, G. Massonnet, J.-P. Beressi, E.
Verhoeyen, V. Raggueneau, B. Maneglier, S. Castaigne, C. Chomienne, S. Chrétien, P. Rousselot,
P. Leboulch, Erosion of the chronic myeloid leukaemia stem cell pool by PPARγ agonists, Nature.
525 (2015) 380–383. https://doi.org/10.1038/nature15248.
28