Kinetics and mechanism of racemization of Tic-hydantoins, potent sigma-1 agonists

Article abstract of DOI:10.1016/j.tetasy.2011.01.015

The configurational stability of seven Tic-hydantoin sigma-1 agonists 1-7 was studied in organic and aqueous media to mimic conditions encountered during either their synthesis or their evaluation as a drug candidate (physico chemical property determinati

Full text of DOI:10.1016/j.tetasy.2011.01.015

Tetrahedron: Asymmetry 22 (2011) 125–133
Contents lists available at ScienceDirect
Tetrahedron: Asymmetry
journal homepage: www.elsevier.com/locate/tetasy
Kinetics and mechanism of racemization of Tic-hydantoins, potent
sigma-1 agonists
Anne-Claire Cabordery a,b, Marion Toussaint a,c,d, Nathalie Azaroual a,b, Jean-Paul Bonte a,e
,
Patricia Melnyk a,b,c,d, Claud Vaccher a,b, Catherine Foulon a,b,
⇑
a Univ Lille Nord de France, F-59000 Lille, France
b UDSL, EA 4481, UFR Pharmacie, F-59000 Lille, France
c UMR CNRS 8161, F-59000 Lille, France
d Institut Pasteur de Lille, F-59000 Lille, France
e UDSL, EA 4481, ICPAL, F-59000 Lille, France
a r t i c l e i n f o
a b s t r a c t
Article history:
The configurational stability of seven Tic-hydantoin sigma-1 agonists 1–7 was studied in organic and
aqueous media to mimic conditions encountered during either their synthesis or their evaluation as a
drug candidate (physico chemical property determination and pharmacological study). This study was
performed from the enantiomers directly obtained by asymmetric synthesis (hydantoins 1, 3–7) or after
semi-preparative separation of the racemic obtained according to the same asymmetric procedure (thi-
ohydantoin 2), by chiral HPLC. The racemization phenomenon was followed using recently validated chi-
ral HPLC and capillary electrophoresis separation methods using derivatized cellulose and amylose chiral
stationary phases (Chiralcel OD-H and Chiralpak AD) and highly sulphated cyclodextrins [highly S-b-CD
in the background electrolyte (BGE)], respectively. The kinetic parameters (rate constant, half-life and
apparent free energy barriers) of racemization were calculated using a first-order reaction model. The
influence of the acid–base content, pH in aqueous medium, concentration of the buffer, and temperature
were investigated. The fastest racemization rates were observed under basic conditions. All hydantoins
were shown to present the same magnitude of configurational stability, whereas thiohydantoin 2 was
characterized by a high chiral instability, especially in ethanolic and aqueous media: its high instability
in ethanol explains that a racemic mixture was obtained after asymmetric synthesis; its instantaneous
racemization observed from the neutral pH makes the preparation of the enantiomers of 2 not relevant.
Finally, the mechanism of racemization was elucidated using nuclear magnetic resonance (NMR): the
comparison of the kinetics of the deuteration to the kinetics of the racemization suggests the involve-
ment of a common reaction mechanism of SE1 type for hydantoin 1, while an SE2 type reaction seems
to be involved for thiohydantoin 2.
Received 15 September 2010
Revised 8 November 2010
Accepted 4 January 2011
Available online 2 March 2011
Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction
of the enantiomers are either pointless or impossible. If chemical
racemization is not detectable or very slow, both separate enanti-
Chirality is of fundamental importance in drug development.
Since enantiomers often differ in pharmacological activity, toxicity,
and pharmacodynamic characteristics, the use of optically active
(pure) enantiomers is recommended.1,2 In the early stages of the
development of chiral drugs, the possibility of their racemization
must be investigated3 in order to make sure that the enantiomers
remain configurationally stable during the manufacturing process,
their shelf-life and under physiological conditions. If chemical
racemization is very fast, the resolution and separate examination
omers and racemate have to be evaluated for their pharmacoki-
netic and pharmacodynamic properties.
In the field of research of medications to treat cocaine addiction,
modulators of sigma-1 chaperone protein, historically called a ‘r1
receptor’, involved in several acute and chronic effects of cocaine
have already been developed.4,5 These tetrahydrosioquinoline-
hydantoin derivatives (Fig. 1—compounds 1 and 3-7) include an
asymmetric center whose configuration influences their affinity
for
r1 receptors. Among them, (S)-1 and (R)-1 appear to be the lead
compounds with IC50 values in the nanomolar range (8.7 and 13.7
for (S)-1 and (R)-1, respectively) lower than haloperidol, the refer-
⇑
Corresponding author. Address: Laboratoire de Chimie Analytique, Faculté des
Sciences Pharmaceutiques et Biologiques, Université Lille Nord de France, BP 83, 3,
rue du Pr. Laguesse, 59006 Lille Cedex, France.
ence used as a
r1 receptor antagonist. Pharmacomodulation of the
Tic-hydantoin structure has led to compound 2.6 Despite the use of
the same asymmetric synthesis procedure used for the preparation
E-mail address: catherine.foulon@univ-lille2.fr (C. Foulon).
0957-4166/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2011.01.015

126
A.-C. Cabordery et al. / Tetrahedron: Asymmetry 22 (2011) 125–133
O
O
10
11, 12, 13
10a
14
N
n N
N
N
N
N
Y
15
5
Z
O
1 (Z = O, n = 3)
2 (Z = S, n = 3)
3 (Z = O, n = 2)
4 (Y = NPh)
Compounds:
Compounds:
5
(Y = CH2)
O
N
N
R
O
Compounds:
6 (R = Bz)
7 (R = H)
Figure 1. Chemical structures of compounds studied.
of 1 and 3–7, we have previously demonstrated7 that both enanti-
omers of 2 were in fact racemic mixtures. This may result from a
fast racemization phenomenon, as described in the literature.8–17
Nevertheless, compound 2 appears to have a high affinity with
the r1 receptor ligand (in IC50 = 4.5 nM).
hance the throughput of the separation, the loading capacity of
the column using optimal separation conditions previously deter-
mined7 (n-hexane/2-propanol-90/10; 1 mL minꢀ1) was evaluated.
First, the effect of overloading was examined with the samples of
concentrations in the range of 0.5–5 mM: enantioseparation re-
mained excellent in this range. Higher concentrations were not
evaluated due to the low solubility of 2 in the mobile phase. Next,
the injected volume (in the maximal concentration) was varied in
Herein, in continuation of our work in the field of racemiza-
tion,18 our first aim was to evaluate the kinetics of racemization
of compounds 1–7 in an organic medium, especially under acidic
and basic conditions at different temperatures, in order to mimic
conditions accounted for, during the synthesis. Secondly, the con-
figurational stability of these compounds was evaluated in an
aqueous medium like that used for their pharmacological study.
The influence of the pH, concentration of the buffer, and the tem-
perature were then investigated. For these purposes, a semi-pre-
parative separation of the enantiomers of 2 was performed. The
mechanism of the racemization was also investigated using nucle-
ar magnetic resonance (NMR). This study may help to explain the
occurrence of a racemic mixture for 2 and to determine if the res-
olution of the racemate by HPLC to prepare enantiopure com-
pounds is useful or not.
the range of 20–200
lL. The ‘touching band phenomenon’ and the
loss of efficient enantioseparation occurring by the injection of
200
100
l
l
L led us to select a final concentration of 5 mM and a
L injection volume, which corresponds to an injection of
0.19 mg (MW = 379.2). The collected fractions of each enantiomer,
trapped in a water/ice bath at 273 K, were evaporated to dryness
under nitrogen steam. The residue was dissolved in 500
lL of eth-
anol; 20 L of this solution were injected in the chromatographic
l
system to estimate the yields that were around 90%. Although a
small amount of each enantiomer was obtained for one injection,
about 2 mg of each enantiomer were finally obtained after 25
injections. This amount was high enough to allow a stability study
on the enantiomers. For the preparation of a greater amount of
each enantiomer, transposition of this method to the preparative
scale should be performed.
2. Results and discussion
The exact enantiomeric purities of the isolated enantiomers of 2
were then determined both in HPLC and in capillary electrophore-
sis using optimal separation methods after the injection of 0.5 mM
of the enantiomers. The enantiomeric purity was calculated from
the respective area and the corrected area in HPLC and capillary
electrophoresis, respectively. The enantiomeric purities obtained
in HPLC and capillary electrophoresis were in accordance. They
were about 91.5% and 89.6% for 2P1 and 2P2, respectively (first
and second eluted enantiomers in HPLC; second and first migrating
enantiomers in capillary electrophoresis; medium values of HPLC
and capillary electrophoresis measurement). The poor enantio-
meric purities of the isolated enantiomers can be explained by
their chiral instabilities.
Previously, the separation of the enantiomers of 1–7 derivatives
was optimized and validated by both HPLC and capillary electro-
phoresis.7 Since the HPLC methods were developed using Chiralcel
OD-H and Chiralpak AD with the mixtures of n-hexane/ethanol or
n-hexane/2-propanol as mobile phases, they were then chosen to
evaluate the kinetics of racemization of the derivatives 1–7 in an
ethanolic medium (separation of the enantiomers in less than
25 min). Moreover, the capillary electrophoresis method devel-
oped in a pH 2.5 phosphate buffer solution using highly S-b-CDs
was used to study the racemization in an aqueous medium (sepa-
ration of the enantiomers in less than 11 min). It is noteworthy
that a preliminary study of the chiral stability of all compounds
in the mobile phases and in the BGE used for their analysis in HPLC
and capillary electrophoresis, respectively, was evaluated at the
temperature used for the analysis. During the time of analysis, no
significant variation of the enantiomeric excess was measured ex-
cept for 2 in HPLC (variation of 2%).
2.2. Kinetics of racemization
Racemization can be described as the formation of a racemate
from a pure enantiomer in an irreversible first-order reaction (1)
2.1. Semipreparative separation of the enantiomers of the
thiohydantoin
krac
ðRÞ !ðRSÞ
or
ð1Þ
Semipreparative enantioseparation of thiohydantoin 2, on the
analytical Chiralpak AD column was investigated. In order to en-
ðSÞ k!racðRSÞ

A.-C. Cabordery et al. / Tetrahedron: Asymmetry 22 (2011) 125–133
127
with:
for (S)-1. Moreover, a general decrease in t1/2
was observed
rac
when the temperature was increased: for 2P1, t1/2
is equal to
rac
ln ee ¼ ꢀkrac
t
ð2Þ
2.5 h and 0.1 h at 298 and 333 K, respectively. This expected phe-
D
G# parameter in
The enantiomeric excesses (ee %) were calculated from the respec-
tive peak areas and the corrected peak areas in HPLC and capillary
electrophoresis, respectively, according to the previous validated
methods.7
nomenon is associated with an increase in the
accordance with the Van’t Hoff equation.
The study of the racemization kinetics was finally extended to
other derivatives (S)-3-(S)-7 using the ethanolic solutions with
1000 equiv of TEA (T = 333 K). All the kinetic parameters shown
in Table 1 appear to be of the same magnitude for these derivatives
½Sꢁ ꢀ ½Rꢁ
ee% ¼
ꢂ 100
ð3Þ
½Sꢁ þ ½Rꢁ
and (S)-1, with t1/2
between 2 h and 2.5 h whereas under the
rac
A linear plot of ln ee versus time indicates a first-order reaction of
racemization and allows us to determine the apparent constant rate
(krac = ꢀslope) and half-life time (t1/2 rac = ln 2/krac), defined as the
time required for an enantiopure solution to reach 50% ee.
same conditions 2P1 completely racemized in less than 5 min.
The racemization kinetics appear to be poorly influenced by the
nature of the side chain. Nevertheless, replacement of the oxygen
atom in (S)-1 or (R)-1 with a sulfur atom to obtain 2P1 and 2P2,
drastically increased the chiral instability of the enantiomers.
Whereas less than 1 hour was sufficient enough to racemize 2P1
or 2P2, 270 days were required for (S)-1 and (R)-1. The chiral insta-
bility of the enantiomers of 2 may explain the racemic nature of
the compounds obtained by asymmetric synthesis that include
steps at a high temperature. It could also explain the poor enantio-
meric purity of enantiomers isolated by preparative HPLC.
The apparent free energy barrier (
to the Eyring equation (Eq. 4):
D
G#) is calculated according
krac ꢃ h
D
G# ¼ ꢀRT lnð
Þ
ð4Þ
kb ꢃ T
The major part of the kinetic study of the racemization was
achieved for one enantiomer of each derivative. It was nevertheless
verified that similar results were obtained with the other enantio-
mer under the same experimental conditions.
2.2.2. In aqueous media
2.2.2.1. Influence of the pH and buffer concentration. The race-
mization of (S)-1 and 2P2 was first studied in a 33 mM pH 7.4 phos-
phate buffer/ethanol ꢀ75/25 (v/v) mixture and at 298 K. Whereas
the kinetic of racemization of (S)-1 was followed successfully,
our attempt to determine the racemization kinetic parameters of
2P2 was unsuccessful due to its high chiral instability. According
to this result and to the interconversion of the enantiomers ob-
served in the acidic and basic ethanolic media, racemization was
finally studied from pH 2 to 3.5 for 2P2 and from pH 6.4 to 8.4
for (S)-1 (racemization was to slow under acidic conditions), at
298 K, using 33 mM phosphate buffer/ethanol ꢀ75/25 (v/v) solu-
tions. This study was realized both with and without the addition
of convenient amounts of NaCl to maintain a constant ionic
strength of the buffers (I = 0.1 M). It is worth mentioning that no
significant difference in the kinetic parameters was obtained with
or without NaCl (CV = 3%), which indicates that the ionic strength
has no influence on the racemization phenomenon. For example,
Figures 3 and 4 show the electropherograms obtained for 2P2 at
pH 2 and 3.5 and the plot ln ee versus time obtained at pH 2,
respectively. The kinetic parameters obtained through the ln ee
versus time plots are shown in Table 2 and show the large influ-
ence of the pH on the kinetics of the racemization, since for both
2.2.1. In ethanolic media
The effect of acid (acetic acid) or base (triethylamine (TEA))
addition at various concentrations in the ethanolic solution on
the chiral stability was studied at 298 and 333 K (room tempera-
ture and high temperature applied during the synthesis step) for
(S)-1 and 2P1. Table 1 summarizes the results obtained except
for 2P1 with acetic acid or TEA, as the kinetics of racemization
was too high to be followed either by HPLC or EC (complete race-
mization took place after 5 min in all the cases). Regardless of the
composition and the temperature of the ethanolic solution, race-
mization was always observed. For example, Figure 2 shows the
chromatogram versus time obtained for (S)-1 in the presence of
TEA (1000 equiv). The data reveal that the racemization rate con-
stant increases both by the addition of acetic acid and TEA, with
a greater increase in the basic ethanolic solutions. For (S)-1 at
333 K, t1/2 rac decreases from 996 h in ethanol to 88 h and 3 h using
1000 equiv of acetic acid and TEA, respectively. The rate constant
increases with the acetic acid/enantiomer or TEA/enantiomer ra-
tios. As expected, the decrease in the t1/2 rac correlated with a de-
crease in the activation energy barrier: a
D
G# difference of
16.1 kJ molꢀ1 calculated for TEA/enantiomer ratios of 0 and 1000
Table 1
Kinetic parameters (krac and t1/2 rac) for the racemization of (S)-1, 2P1 and (S)-3–(S)-7 in the ethanolic solutions: influence of the acid-base content
Temperature (K)
298
Compound
Solvent
krac (hꢀ1
)
t1/2
(h)
D )
G# (kJ molꢀ1
rac
(S)-1
EtOH
1.25 ꢂ 10ꢀ4
4.73 ꢂ 10ꢀ4
4.73 ꢂ 10ꢀ2
2.78 ꢂ 10ꢀ1
6.96 ꢂ 10ꢀ4
1.71 ꢂ 10ꢀ3
7.86 ꢂ 10ꢀ3
3.17 ꢂ 10ꢀ2
2.80 ꢂ 10ꢀ1
5530.0
1460.0
14.6
115.5
112.3
100.8
EtOH + 1000 equiv CH3COOH
EtOH + 1000 equiv TEA
298
333
2P1
EtOH
2.5
96.5
(S)-1
EtOH
996.0
405.0
88.0
21.8
2.5
124.7
122.2
118.0
114.1
108.1
EtOH + 100 equiv CH3COOH
EtOH + 1000 equiv CH3COOH
EtOH + 100 equiv TEA
EtOH + 1000 equiv TEA
333
333
333
333
333
333
2P1
EtOH
5.32
0.1
2.1
2.0
2.3
2.1
2.4
99.9
107.7
107.5
107.9
107.6
108.0
(S)-3
(S)-4
(S)-5
(S)-6
(S)-7
EtOH + 1000 equiv TEA
EtOH + 1000 equiv TEA
EtOH + 1000 equiv TEA
EtOH + 1000 equiv TEA
EtOH + 1000 equiv TEA
3.24 ꢂ 10ꢀ1
3.41 ꢂ 10ꢀ1
2.99 ꢂ 10ꢀ1
3.31 ꢂ 10ꢀ1
2.90 ꢂ 10ꢀ1

128
A.-C. Cabordery et al. / Tetrahedron: Asymmetry 22 (2011) 125–133
0.20
0.10
0.00
t = 0 h
-0.10
0.20
t = 3 h 00
t = 6 h 00
t = 22 h 30
0.10
0.00
-0.10
0.20
0.10
0.00
-0.10
0.20
0.10
0.00
-0.10
7.00
7.50
8.00
8.50
9.00
9.50
10.00
10.50
11.00
11.50
12.00
t(min)
Figure 2. Chromatograms of (S)-1 after 0, 3, 6 and 22.5 h in ethanol in the presence of 1000 equiv of TEA.
compounds a twofold decrease in the half-lives is observed for a
one unit pH increase. At the same pH (2.5), the kinetics of the race-
mization of (S)-1 were much slower than for 2P2 (t1/2 rac ratio of
400). It is noteworthy, that the high instability of 2P2 at pH 7.4
(t1/2 rac under 5 min), made the preparative separation for the phar-
macological study unnecessary as the enantiomers would racemize
almost instantaneously in the biological medium.
The influence of the buffer concentration was investigated at pH
7.4 for (S)-1 and at pH 2.5 for 2P2 at 298 K. As summarized in Ta-
ble 3, an increase in the buffer concentration from 33 mM to
166 mM leads to a large decrease in the half-lives, that is, from
28.8 h to 12.1 h for (S)-1 and from 4.4 h to 1.1 h for 2P2. This de-
crease is associated, as expected, with a decrease in the free energy
barriers. It appears that the first-order rate of racemization con-
stants (in hours) is linearly dependant on the phosphate buffer
concentration (in mM) as shown by the linear equation obtained
for both compounds (r2 >0.988) (Eqs. 5 and 6) (Fig. 5).
0
-0.2
-0.4
-0.6
-0.8
-1
0
1
2
3
4
5
6
7
-1.2
-1.4
t(h)
Figure 4. Variations of ln ee versus time obtained at 298 K for 2P2 in a 33 mM pH 2
phosphate buffer/ethanol—75/25 (v/v) solution.
2.2.2.2. Influence of the temperature. The influence of the tem-
perature was investigated in the range of 288–310.5 K, under the
experimental conditions that allowed the enantiomeric excess to
be followed easily, that is, in 33 mM pH 7.4 phosphate buffer/eth-
anol—75/25 (v/v) for (S)-1 and in 33 mM of pH 2.5 phosphate buf-
fer/ethanol—75/25 (v/v) for 2P2. The results shown in Table 4 are in
krac1ðSÞ ¼ 2:55 ꢂ 10ꢀ4½bufferꢁ þ 1:56 ꢂ 10ꢀ2
krac2P2 ¼ 3:46 ꢂ 10ꢀ3½bufferꢁ þ 3:26 ꢂ 10ꢀ2
ð5Þ
ð6Þ
accordance with the Van’t Hoff equation (increase in
temperature). It is also possible to calculate the enthalpic and
D
G# with
It appears that the influence of the phosphate buffer concentration
is more drastic for derivative 2P2.
0.020
0.015
0.010
0.005
0.000
0.020
0.015
0.010
0.005
0.000
0.020
0.015
0.010
0.005
0.000
0.020
0.015
0.010
0.005
0.000
(a)
(b)
t = 6 h
t = 6 h
t = 2 h
t = 0 h
t = 2 h
t = 0 h
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50
t (min)
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50
t (min)
Figure 3. Electropherograms of 2P2 after 0, 2 and 6 h at 298 K in: (a) 33 mM pH 2 phosphate buffer/ethanol—75/25 (v/v) solution; (b) 33 mM pH 3.5 phosphate buffer/
ethanol—75/25 (v/v) solution.

A.-C. Cabordery et al. / Tetrahedron: Asymmetry 22 (2011) 125–133
129
Table 2
Kinetic parameters (krac and t1/2 rac) and free energy barriers (
buffer (I = 0.1 M)/ethanol—75/25 (v/v) at 298 K: influence of the pH
D
G#) for the racemization of (S)-1, 2P2 and (S)-3-(S)-7 in 33 mM phosphate
Compound
pHa
krac (hꢀ1
)
t1/2
(h)
D )
G# (kJ molꢀ1
rac
(S)-1
2.5
6.4
6.9
7.4
7.9
8.4
4.10 ꢂ 10ꢀ4
1.16 ꢂ 10ꢀ2
1.80 ꢂ 10ꢀ2
2.41 ꢂ 10ꢀ2
4.31 ꢂ 10ꢀ2
6.96 ꢂ 10ꢀ2
1689.0
59.5
38.5
28.8
16.1
10.0
112.6
104.3
103.2
102.5
101.1
99.9
2P2
2
1.37 ꢂ 10ꢀ1
1.58 ꢂ 10ꢀ1
2.40 ꢂ 10ꢀ1
3.71 ꢂ 10ꢀ1
5.0
4.4
2.9
1.9
98.2
97.8
96.8
95.7
2.5
3.1
3.5
(S)-3
(S)-4
(S)-5
(S)-6
(S)-7
7.4
7.4
7.4
7.4
7.4
2.33 ꢂ 10ꢀ2
1.55 ꢂ 10ꢀ2
1.80 ꢂ 10ꢀ2
1.51 ꢂ 10ꢀ2
1.12 ꢂ 10ꢀ2
29.7
44.7
38.5
45.9
61.9
102.6
103.6
103.2
103.7
104.4
a
pH of the phosphate buffer.
Table 3
Kinetic parameters (krac and t1/2 rac) and free energy barriers (
(v/v) at 298 K: influence of the buffer concentration
D
G#) for the racemization of (S)-1and 2P2 in phosphate buffer/ethanol—75/25
krac (hꢀ1 G# (kJ molꢀ1
Compound
pHa
7.4
[Buffer] (mM)
)
t1/2
(h)
D
)
rac
(S)-1
33
66
100
133
166
2.41 ꢂ 10ꢀ2
3.14 ꢂ 10ꢀ2
4.23 ꢂ 10ꢀ2
4.95 ꢂ 10ꢀ2
5.75 ꢂ 10ꢀ2
28.8
22.1
16.4
14.0
12.1
102.5
101.9
101.1
100.7
100.4
2P2
2.5
33
66
100
133
166
1.58 ꢂ 10ꢀ1
2.35 ꢂ 10ꢀ1
4.03 ꢂ 10ꢀ1
4.80 ꢂ 10ꢀ1
6.12 ꢂ 10ꢀ1
4.4
2.9
1.7
1.4
1.1
97.8
96.9
95.5
95.1
94.5
a
pH of the phosphate buffer.
8 x 10-1
6 x 10-1
The study of the racemization kinetics was finally extended to
other derivatives (S)-3-(S)-7 in 33 mM pH 7.4 phosphate buffer/
ethanol—75/25 (v/v). The kinetic parameters shown in Table 2 ap-
pear to be all of the same magnitude for these derivatives and (S)-
1, with t1/2 rac between 28 h and 62 h. Kinetics of racemization ap-
pear to be poorly influenced by the nature of the side chain. The
smallest rate of racemization obtained for compound 7 is in accor-
dance with the observations of Bovarnick and Clarke.15 They have
in fact observed a sixfold higher rate of racemization for 3-phenyl-
5-p-hydroxybenzylhydantoin than for 5-p-hydroxybenzylhydan-
toin. Moreover, under the same conditions 2(P1) is completely
racemized during the analysis time, that is, in less than 5 min:
replacement of the oxygen atom in (S)-1 or (R)-1 with a sulfur
atom to obtain (S)-2 and (R)-2, respectively, drastically increased
the chiral instability of the enantiomers. Its instant racemization
makes the preparation of the enantiomers of 2 not relevant.
4 x 10-1
k
2 x 10-1
0
0
50
100
150
200
[buffer] (mM)
Figure 5. Influence of the phosphate buffer concentration on the rate constant of
racemization (krac) of (S)-1 at pH 7.4 (N) and 2P2 at pH 2.5 (j).
2.2.2.3. Mechanism of the racemization. The kinetics of the
racemization of 5-substitued hydantoins have previously been
studied.8–17 Several authors have suggested that the racemization
mechanism occurs via a keto-enol tautomerism.10–14 Bovarnik sug-
gested that racemization is similar to the enolization in its depen-
dence on the ionization of hydrogen on the asymmetric carbon
atom; that is, the ionization is dependent on the alkalinity of the
medium and the activating effects of the groups attached to the
asymmetric carbon atom.15 As Lazarus,16 Reist et al.17 proposed a
general-base catalysis. A careful study of the racemization and
entropic terms according to the Eyring equation (Eq. 7) by plotting
ln (krac/t) versus 1/T (Fig. 6).
ꢀ
ꢁ
ꢀ
ꢁ
krac
T
D
RT
H
kb
h
DS
R
ln
¼ ꢀ
þ ln
þ
ð7Þ
The
D
H and
D
S obtained are 68.4 kJ molꢀ1 and ꢀ113 molꢀ1 Kꢀ1 for
(S)-1 and 78.1 kJ molꢀ1 and ꢀ66.8 molꢀ1 Kꢀ1 for 2P2, respectively.
These parameters seem to be significantly different. Nevertheless,
it is not possible to conclude a distinct racemization mechanism
for these two derivatives as pH are different.

130
A.-C. Cabordery et al. / Tetrahedron: Asymmetry 22 (2011) 125–133
Table 4
Kinetic parameters (krac and t1/2 rac) and free energy barriers (D
G#) for the racemization of (S)-1 and 2P2 in 33 mM phosphate buffer/
ethanol—75/25 (v/v): influence of the temperature
Compound
pHa
7.4
T (K)
krac (hꢀ1
)
t1/2 (h)
D )
G# (kJ molꢀ1
(S)-1
288
293
298
303
310.5
1.04 ꢂ 10ꢀ2
1.63 ꢂ 10ꢀ2
2.41 ꢂ 10ꢀ2
4.81 ꢂ 10ꢀ2
8.46 ꢂ 10ꢀ3
66.4
42.5
28.8
14.4
81.9
101.0
101.7
102.5
102.5
109.6
2P2
2.5
288
293
298
303
310.5
4.6 ꢂ 10ꢀ2
9.12 ꢂ 10ꢀ2
1.58 ꢂ 10ꢀ1
2.39 ꢂ 10ꢀ1
5.56 ꢂ 10ꢀ1
15.2
7.6
4.4
2.9
1.2
97.5
97.5
97.8
98.5
98.8
a
pH of the phosphate buffer.
formed at 288 K in a 33 mM pD 2.6 deuterated phosphate buffer/
ethanol—75/25 (v/v) solution using the racemic mixture (total con-
centration 3.5 mM). At the same time, the kinetics of racemization
of 2P2 at 3.5 mM in the deuterated buffer solution used in NMR
were investigated in the capillary electrophoresis at 288 K.
-12
0,0032 0,00325
-13
0,0033 0,00335
0,0034 0,00345
0,0035
pH 2.5
-14
-15
-16
-17
-18
-19
y = -9392,6x + 15,731
R2 = 0,9955
In the literature, the 1H and 13C spectra of 1 and 2 in chloroform
CDCl3 are described.5,6 Herein, the attribution of the 1H signals was
performed in deuterated phosphate buffer/ethanol—75/25 (v/v);
the results are displayed in Table 5. Moreover, as illustrated in Fig-
ure 7 for compound (S)-1, the deuteration of proton 10a was iden-
tified by the integration decrease of the signal at 4 ppm
(corresponding to protons 10a, 14 and 140) and the coupling disap-
pearing. After deuteration, the two signals corresponding to pro-
tons 10 and 100 (doublet of doublet) at 2.28 ppm and 3.01 ppm,
respectively, were simplified to a doublet with a coupling constant
k
pH 7.4
y = -8236x + 10,142
2 = 0,99
R
1/T (K-1)
Figure 6. Influence of the temperature on the rate constant of racemization (krac) of
(S)-1 in 33 mM pH 7.4 phosphate buffer/ethanol—75/25 (v/v) (N) and 2P2 in 33 mM
pH 2.5 phosphate buffer/ethanol—75/25 (v/v) (j).
0
loss of 13 Hz (J10a–10 ) and 4 Hz (J10a–10), respectively.
For a first-order H/D exchange reaction, ln It/I0 must vary, in a
linear fashion versus the time plot according to Eq. 8:
deuteration of 5-monosubstituted hydantoin led them to propose
an SE2 push–pull mechanism for the racemization.
It
lnð Þ ¼ ꢀkdeut
t
ð8Þ
I0
By considering the chemical structure of the hydantoin deriva-
tives studied, including a hydrogen atom, at the
a-position of the
with It, the integral of exchanging proton at time t, Io the integral of
exchanging proton at time 0 and kdeut the apparent rate constant of
deuteration. The half-life time of deuteration is defined as the time
for which It = I0/2 (t1/2 deut = ln 2/kdeut).
Variations of ln It/I0 versus time were then studied for (S)-1 and
2. Since the plots are linear (r2 >0.988), the H/D exchange is first-
order. The apparent deuteration and racemization rate constants,
and the corresponding half-lives obtained using NMR and CE,
respectively, are shown in Table 6. As reported by Reist et al.17 dur-
ing their study of the mechanism of racemization of 5-monosubsti-
tuted, hydantoins a small isotopic effect (1.5-fold factor between
rate constants), was observed (see Table 6) for both compounds.
ketone function (H10a) and in accordance with the mechanisms
proposed in the literature for the racemization of hydantoin deriv-
atives substituted at the 5-position, NMR experiments in D2O were
then performed to study the possible lability of H10a. Experimental
conditions were chosen in order to obtain a medium half-life and
to minimize the deuteration of the analyte during the acquisition
of the spectra (around 10 min). For hydantoin (S)-1, the kinetics
of deuteration (H/D exchange) and racemization were studied in
a 33 mM pD 7.3 deuterated phosphate buffer/ethanol—75/25 (v/
v) solution at 298 K, using NMR and capillary electrophoresis,
respectively. The study of deuteration of thiohydantoin 2 was per-
Table 5
Chemical shifts (d), multiplicity and coupling constant (J) of 1H for (S)-1 at 298 K and for 2 at 288 K, in 33 mM pD 7.3 phosphate buffer/
EtOD—75/25 (v/v) 298 K and 33 mM pD 2.6 phosphate buffer/EtOD—75/25, respectively
Proton
1
2
d (ppm)
Multiplicity
J (Hz)
d (ppm)
Multiplicity
J (Hz)
12
1.86
2.28
2.62
m
dd
s
1.88
2.29
2.6
m
dd
s
100
13 and 15
13 and 15
15
13
10
11
140
2.69
3.01
3.44
3.98
m
dd
m
d
2.76
2.96
3.65
3.98
m
dd
m
d
4 and 15
4 and 15
13
13
13
13
14
4.04
d
4.04
d
10a
4.06
4.29
4.72
m
d
d
4.13
4.36
5.05
m
d
d
50
16
16
17
17
5
Aromatics
6.95 and 7.25
m
6.93 and 7.20
m

A.-C. Cabordery et al. / Tetrahedron: Asymmetry 22 (2011) 125–133
131
Figure 7. 1H NMR spectrum of (S)-1 after 0 and 96 h in 33 mM pD7.3 phosphate buffer/ethanol—75/25 (v/v) at 298 K.
using enantiomers obtained directly by an asymmetric synthesis
(hydantoins 1, 3–7) or after a semi-preparative separation of the
racemic obtained according to the same asymmetric procedure
(thiohydantoin 2), by chiral HPLC using Chiralpak AD stationary
phase. In the ethanolic solutions, the racemisation rate was shown
to increase with the acetic acid/enantiomer or TEA/enantiomer ra-
tios and with temperature. Whereas configurational stability ap-
pears to be poorly influenced by the nature of the side chain
(compounds 1, 3–7) replacement of the oxygen atom in 1 by a sul-
fur atom to obtain 2 drastically increased the chiral instability of
the enantiomers [complete racemization of 2P1 or 2P2 occurred
in 1 h versus 270 days for (S)-1 and (R)-1]. Chiral instability of
the enantiomers of 2 in the ethanolic media may explain the race-
mic nature of the compounds obtained by asymmetric synthesis
that include different steps at a high temperature. In aqueous med-
ia, the racemization rate appears to increase with pH, ionic
strength, and temperature. As observed in the ethanolic solutions,
the nature of the side chain of the hydantoins poorly influences
their chiral stability, whereas the replacement of the oxygen atom
in 1 with a sulfur atom to obtain 2, drastically increases the chiral
instability of the enantiomers. The instant racemization of the
enantiomers of thiohydantoin 2 makes their preparative separa-
tion by HPLC irrelevant. Comparison of the kinetics of deuteration
of the H10a proton located at the stereogenic carbon to the kinetics
of racemization suggests the involvement of a common reaction
mechanism of the SE1 type for hydantoin 1, whereas an SE2 type
reaction seems involved for thiohydantoin 2. This difference in
mechanism may explain the great difference in configurational sta-
bility of hydantoin 1, 3–7 versus thiohydantoin 2.
Table 6
Kinetic parameters for the deuteration (kdeut, t1/2
and DG#deut) and racemization
deut
(krac D, t1/2 rac D and DG#racD) at 298 K, in 33 mM pD 7.3 deuterated phosphate buffers/
EtOD—75/25 (v/v) for (S)-1 and at 288 K, in 33 mM pD 2.6 deuterated phosphate
buffers/EtOD—75/25 (v/v) for 2 ; kinetic parameters for the racemization (krac H, t1/2
rac H and DG#racH ) at 298 K, in 33 mM pH 7.4 non-deuterated phosphate buffers/EtOD—
75/25 (v/v) for (S)-1 and at 288 K, in 33 mM pH 2.5 non-deuterated phosphate
buffers/EtOD—75/25 (v/v) for 2
Compound
kdeut (hꢀ1
krac (hꢀ1
(S)-1
2
)
)
)
1.60 ꢂ 10ꢀ2
3.11 ꢂ 10ꢀ2
5.91 ꢂ 10ꢀ2a
4.56 ꢂ 10ꢀ2a
1.61 ꢂ 10ꢀ2
D
krac (hꢀ1
2.41 ꢂ 10ꢀ2
H
t1/2
t1/2
t1/2
(h)
(h)
(h)
43.3
43.1
28.8
22.9
deut
11.7a
rac
rac
D
a
15.2
H
G#deut (kJ molꢀ1
G#rac D (kJ molꢀ1
G#rac H (kJ molꢀ1
)
103.6
103.5
102.5
96.8
98.4
D
D
D
a
)
97.5a
)
a
From 2P2.
Moreover, for each compound, the activation energies of deutera-
tion and racemization are identical within an experimental error
(CV of 2%), confirming a common reaction mechanism. For hydan-
toin 1, the deuteration and racemization were investigated in the
same buffer and showed similar kinetic parameters (kdeut/krac ꢄ 1).
This result seems to show that deuteration takes place with com-
plete racemization. Both deuteration and racemization share a
common mechanism involving the formation of a carbanion cap-
tured from either side with equal probability, that is, an SE1 mech-
anism.17 For 2, the deuteration constant was about half that of the
racemization constant suggesting that deuteration occurs with an
inversion of configuration.17
4. Experimental
4.1. Reagents and solutions
3. Conclusion
Each of the enantiomers of Tic-hydantoin derivatives 1–7
(Fig. 1) was synthesized in our laboratory according to previously
The configurational stability of seven Tic-hydantoin sigma-1
agonists 1–7 was studied in organic and aqueous media, either

132
A.-C. Cabordery et al. / Tetrahedron: Asymmetry 22 (2011) 125–133
published methods,10,13,15 starting from L-Tic-OH for compounds
(S)-1, (S)-3-(S)-7 and D-Tic-OH for compounds (R)-1, (R)-3-(R)-7.
Racemic compound 2 was obtained either from L-Tic-OH or from
D-Tic-OH.
hydrodynamic injection was made with a 5 s injection time at
0.5 psi (cathodic injection). The applied field was of 0.40 kV cmꢀ1
normal polarity. All compounds were detected at 200 nm. New
capillaries were flushed for 20 min with 0.1 M sodium hydroxide
(NaOH) (P = 20 psi) and 5 min with water (P = 20 psi). When using
the neutral CDs, the capillary was successively flushed each day
with NaOH (5 min, 20 psi), water (1 min, 20 psi) and BGE (3 min,
20 psi). The capillary was successively flushed each day with NaOH
(5 min, 20 psi), water (1 min, 20 psi), polyethylene oxide (PEO)
(1 min, 20 psi), water (1 min, 20 psi) and then with BGE (3 min,
20 psi). Between each run, it was treated with water (1 min,
20 psi) and BGE (3 min, 20 psi).
Ethanol, 2-propanol,n-hexane(allofHPLCgrade), phosphoricacid
(85% w/w), triethylamine, triethanolamine (TEA), sodium hydroxyde
(NaOH), sodium chloride (NaCl), sodium hydrogenophosphate and
sodium dihydrogenophosphate were obtained from Merck (Nogent
sur Marne, France) or Baker (Noisy le Sec, France). NaOH 0.1 M, PEO
(0.4%, Mw = 300,000), highly sulfated b cyclodextrins (highly S-b-
CD, Mw = 2381; 20% w/v in a 50 mM phosphate buffer at pH 2.5,
which corresponds to a 84.0 mM solution) were purchased from
Beckman (Beckman Coulter France, Villepinte, France). Deuterium
oxide (100%; D2O) and ethanol D6 (EtOD) were purchased from Eur-
iso-top (Gif sur Yvette, france). The phosphate buffers in the 2.1–3.6
and 6.4–8.5 pH ranges were prepared, respectively, from phosphoric
acid solution and sodium dihydrogenophosphate solution adjusted
to the desired pH by the addition of sodium hydroxide (1 M). An
appropriate volume of sodiumchloride (1 M) was occasionally added
to set the ionic strength to the desired value (0.1 M).
4.4. Kinetics of racemization
The ethanolic solutions (1 M) of TEA and acetic acid were pre-
pared; an appropriate volume of these solutions were mixed to
0.15 mM ethanolic solutions of the enantiomers (S)-1, (R)-2, (S)-
3-(S)-7 to study their configurational stability at a concentration
of 0–1000 equiv of TEA or acetic acid. The kinetic experiments in
the ethanolic medium were conducted in a thermostatic bath to
maintain the temperature of the samples during the racemization
process either at 298 K or at 333 K. The samples were taken at reg-
ular intervals and were analyzed by HPLC. In aqueous medium, ki-
netic experiments were conducted in the temperature controlled
samples storage unit of the capillary electrophoresis apparatus that
permits both to maintain a constant temperature along the racemi-
zation process and also an easy automation. The chiral stability of
the enantiomers (S)-1, (R)-2, (S)-3-(S)-7 was studied in different
phosphate buffer/ethanol—75/25 (v/v) solutions, prepared by the
dilution of 0.4 mM ethanolic stock solutions in the corresponding
phosphate buffers. The analyte concentration was 0.1 mM in all
cases unless otherwise specified. All displayed results are mean
values of triplicate kinetic results.
4.2. HPLC
Chromatographic analyses were carried out using a gradient
Waters 600E metering pump model equipped with a Waters 996
photodiode array spectrophotometer, a 7125 Rheodyne injector
with a 20 lL loop, and a Waters in-line degasser apparatus. Chro-
matographic data were collected and processed on a computer run-
ning with Millennium 2010. Chiral analyses were performed on a
polysaccharide Chiralcel OD-H cellulose column (tris-3,5-dimethyl-
phenylcarbamate; 250 ꢂ 4.6 mm I.D.; 5
amylose column (tris-3,5-dimethylphenylcarbamate; 250 ꢂ
4.6 mm I.D.; 10 m) (Daicel Chemical Industries, Baker France)
lm) and a Chiralpak AD
l
using previously optimized and validated methods7 that is, Chir-
alpak AD amylose column with a n-hexane/2-propanol-40/60 mo-
bile phase (293 K 0.1; 0.8 mL minꢀ1) for the enantiomers of
compounds 1, 4, 6, 7; Chiralpak AD amylose column with a n-hex-
ane/2-propanol-90/10 (1 mL minꢀ1) for the enantiomers of com-
pounds 2; Chiralcel OD-H cellulose column with a n-hexane/
ethanol-80/20 mobile phase (293 K 0.1; 1 mL minꢀ1) for the
enantiomers of 3 and 5. Mobile phases used according to an iso-
4.5. Kinetics of the H/D exchange
The NMR spectroscopy experiments were performed on a Bru-
ker Avance 500 with a TXI probe operating at 500.13 MHz and
the data were analyzed using the Topspin software. The attribu-
tions of the 1H signals of hydantoin 1 and thiohydantoin 2 were
performed by various NMR experiments: COSY (correlation spec-
troscopy) (correlation homonuclear 1H–1H), HSQC (heteronuclear
single quantum coherence) (correlation heteronuclear 1H–13C)
and HMBC (heteronuclear multiple bond correlation) (correlation
cratic mode, were previously filtered through
a membrane
(0.45 m) and degassed with a Waters in-line degasser apparatus.
l
The column eluant was monitored at 206 nm for all compounds.
The semipreparative separation of enantiomers of 2 was per-
formed at 293 K 0.1 on the analytical Chiralpak AD amylose col-
long range 1H–13C). To begin 500
lL of a 6 mM (S)-1 solution in a
umn, with n-hexane/2-propanol-90/10 (1 mL minꢀ1
)
using
a
33 mM pD 7.3 deuterated phosphate buffer/EtOD—75/25 (v/v)
mixture and 3.5 mM 2 solution in a 33 mM pD 2.6 deuterated
phosphate buffer/EtOD—75/25 (v/v)) were, respectively, intro-
duced into standard 5 mm NMR tubes. For the study of the H/D ex-
change, spectra with water presaturation during acquisition were
recorded regularly for several days at 298 K for (S)-1and 288 K
for 2. The temperature control of the probe was approximately
0.5 °C. Eighty scans for (S)-1 and one hundred and sixty scans
for 2 with 32 k data points were acquired with a spectral width
of 5000 Hz. The FIDs were transformed with a line broadening of
0.3 Hz. The kinetics of the H/D exchange were investigated by
the integration of the signal at 4 ppm corresponding with the
exchanging proton H10a and protons H5’ at the initial time and at
t time (the calibration of integrals was performed from the signal
of an unexchangeable reference aromatic protons).
100
lL loop.
4.3. Capillary electrophoresis
Capillary electrophoresis experiments were performed on a
Beckman P/ACE MDQ Capillary Electrophoresis system (Beckman
Coulter France, Villepinte, France), including an on-column
diode-array UV-detector. The whole system was driven by a PC
with the 32 Karat software (Beckman Coulter France) package for
system control, data collection, and analysis.
The capillary electrophoresis used a method for the enantiosep-
aration of 1–7 enantiomers as had previously been developed and
validated.7 This method is based on the use of a 50.2 cm (effective
length: 10 cm) ꢂ 50
lm ID untreated fused-silica capillary (Com-
posite Metal Services, Worcestershere, UK), mounted in a cartridge
and thermostated at 298 K 0.1 K. The background electrolyte was
composed of a 50 mM phosphate buffer pH 2.5 (H3PO4 + trietha-
nolamine) containing 10 mM of highly S-b-CD 10 mM for the sep-
aration of 1–3 and 5–7 and 2 mM for the separation of 5. A
Acknowledgments
The 500 MHz NMR facilities were funded by the Région Nord-
Pas de Calais (France), the Ministère de la Jeunesse, de l’Education

A.-C. Cabordery et al. / Tetrahedron: Asymmetry 22 (2011) 125–133
133
7. Cabordery, A. C.; Toussaint, M.; Bonte, J. P.; Melnyk, P.; Vaccher, C.; Foulon, C. J.
Chromatogr., A 2010, 24, 3871–3875.
8. Dinelli, D.; Marconi, W.; Cecere, F.; Galli, G.; Morsi, F. Enzyme Eng. 1978, 3, 477–
481.
Nationale et de la Recherche (MJENR) and the Fonds Européens de
Développement Régional (FEDER).
9. Cabrera, K.; Jung, M.; Fluck, M.; Schurig, V. J. Chromatogr., A 1996, 731, 315–321.
10. Dakin, H. D. Am. Chem. J. 1910, 44, 48.
References
11. Dudley, K. H.; Bius, D. L. Drug Metab. Dispos. 1976, 4, 340–348.
12. Pietzsch, M.; Syldatk, C.; Wagner, F. Ann. N.Y. Acad. Sci. 1992, 672, 478–483.
13. Syldatk, C.; Pietzsch, M. In Enzyme Catalysis in Organic Synthesis; Drauz, K.,
Waldmann, H., Eds.; VCH-Verlag: Weinheim, 1995; pp 409–431.
14. Lee, C. K.; Fan, C. H. Enzyme Microb. Technol. 1999, 24, 659–666.
15. Bovarnick, M.; Clarke, H. T. J. Am. Chem. Soc. 1938, 60, 2426–2430.
16. Lazarus, R. A. J. Org. Chem. 1990, 55, 4755–4757.
17. Reist, M.; Carrupt, P. A.; Testa, B. Helv. Chim. Acta 1996, 79, 767–778.
18. Danel, C.; Foulon, C.; Goossens, J. F.; Bonte, J. P.; Vaccher, C. Tetrahedron:
Asymmetry 2006, 17, 2317–2321.
1. For a policy statement, see: Chirality 1992, 4, 338–340.
2. DeCamp, W. H. Chirality 1989, 1, 2–6.
3. Testa, B.; Carrupt, P. A.; Gal, J. Chirality 1993, 5, 105–111.
4. Cazenave Gassiot, A.; Charton, J.; Girault-Mizzi, S.; Gilleron, P.; Debreu-
Fontaine, M.-A.; Sergheraert, C.; Melnyk, P. Bioorg. Med. Chem. Lett. 2005, 15,
4828–4832.
5. Toussaint, M.; Delair, B.; Foulon, C.; Lempereur, N.; Vaccher, C.; Maurice, T.;
Melnyk, P. Eur. Neuropsychopharmacol. 2009, 19, 504–515.
6. Toussaint, M.; Mousset, D.; Foulon, C.; Jacquemard, V.; Vaccher, C.; Melnyk, P.
Eur. J. Med. Chem. 2010, 45, 256–263.