molecules
Article
Hydrolysis and Enantiodiscrimination of (R)- and
(S)-Oxazepam Hemisuccinate by Methylated β-Cyclodextrins:
An NMR Investigation
Andrea Cesari 1 , Federica Balzano 2, * , Gloria Uccello Barretta 2, *






Citation: Cesari, A.; Balzano, F.;
Uccello Barretta, G.; Recchimurzo, A.
Hydrolysis and Enantiodiscrimination
of (R)- and (S)-Oxazepam
Hemisuccinate by Methylated
β-Cyclodextrins: An NMR
Investigation. Molecules 2021, 26, 6347.
https://doi.org/10.3390/molecules
26216347
Academic Editor: Graziella Vecchio
Received: 22 September 2021
Accepted: 15 October 2021
Published: 20 October 2021
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2021 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Molecules 2021, 26, 6347. https://doi.org/10.3390/molecules26216347
and Alessandra Recchimurzo 2
1 Dipartimento di Scienze Chimiche, Università di Padova, Via Marzolo 1, 35131 Padova, Italy;
andrea.cesari@unipd.it
2 Dipartimento di Chimica e Chimica Industriale, Università di Pisa, Via Moruzzi 13, 56124 Pisa, Italy;
alessandra.recchimurzo@phd.unipi.it
* Correspondence: federica.balzano@unipi.it (F.B.); gloria.uccello.barretta@unipi.it (G.U.B.)
Abstract: Partially and exhaustively methylated β-cyclodextrins [(2-methyl)-β-CD (MCD), heptakis-
(2,6-di-O-methyl)-β-CD (DIMEB), and heptakis-(2,3,6-tri-O-methyl)-β-CD (TRIMEB)] have been
compared in the hydrolysis and enantiodiscrimination of benzodiazepine derivative (R)- or (S)-
oxazepam hemisuccinate (OXEMIS), using nuclear magnetic resonance (NMR) spectroscopy as an
investigation tool. After 6 h, MCD induced an 11% hydrolysis of OXEMIS, remarkably lower in
comparison with underivatized β-CD (48%), whereas no hydrolysis was detected in the presence of
DIMEB or TRIMEB after 24 h. DIMEB showed greater ability to differentiate OXEMIS enantiomers
in comparison to TRIMEB, by contrast MCD did not produce any splitting of racemic OXEMIS
resonances. Both enantiomers of OXEMIS underwent deep inclusion of their phenyl pendant into
cyclodextrins cavities from their wider rims, but tighter complexes were formed by DIMEB with
respect to TRIMEB.
Keywords: NMR; chiral discrimination; methylated cyclodextrins; benzodiazepines; hydrolysis;
inclusion complexes
1. Introduction
Benzodiazepines (BDZs) are a class of psychopharmaceuticals employed as sedatives,
anxiolytics, hypnotic, and anticonvulsants with less collateral effects than barbiturate. The
main biological mechanism of BDZs involves a positive allosteric modulation of GABAA
receptors located in the brain, enhancing the effect of γ-aminobutyric acid neurotransmitter,
and hence reducing the excitability of neurons [1].
The skeletal core of BDZs is composed of a benzene ring fused with a 1,4-diazepine
(or a 1,5-diazepine) and decorated with different substituents and/or additional heteroaro-
matic rings (examples of BDZs are given in Figure 1a). Despite BDZs have the same
biological target, the variety of their chemical structures reflects the diverse pharmacokinet-
ics (short-, intermediate-, and long-acting drugs), which can be addressed for the treatment
of different medical conditions [2].
Among commonly used BDZs, Oxazepam (OX, Figure 1b), which has been prescribed
for curing insomnia and anxiety arising from alcohol withdrawal syndrome, is a chiral
BDZ with an asymmetric center at C3 -OH, but it is commercialized as a racemic mix-
ture. However, some studies have reported a different activity of (S)-OX with respect
to (R)-OX, with the former from 100- to 200-fold more affine to the binding site in com-
parison with the latter one [3]. Therefore, administering low amounts of the more active
stereoisomer can be the way to reduce unwanted side effects. Moreover, there are evi-
dence that hepatic glucuronidation (main route of human clearance) produces enriched
(S)/(R)-OX-glucuronide diastereomeric ratios, as found in human plasma and urinary
https://www.mdpi.com/journal/molecules
es 2021, 26, x FOR PEER REVIEW 2 of 11
Molecules 2021, 26, 6347 2 of 11

(S)/(R)-OX-glucuronide diastereomeric ratios, as found in human plasma and urinary ex-

excretions
cretions [4,5]. This [4,5].difference
metabolic This metabolic differencetoisallnot
is not common common
BDZs to all
[6]. The BDZs [6]. The stereoselec-
stereoselectiv-
tivity of glucuronosyltransferase
ity of glucuronosyltransferase isoenzymes can isoenzymes can be by
be also interfered alsoother
interfered by other administered
administered
drugs [7]. drugs [7].

Figure 1. Figure 1. Chemical

structuresstructures of (a) Diazepam
(Valium):(Valium):
R1 =CH3R, =CH
1 2 3, R2=O,
R =O, R3 =H,R R
3=H, R4=phenyl,
4 =phenyl, R5=Cl;
Chemical of (a) Diazepam R5 =Cl; Alprazolam
Alprazolam (Xanax): R 1=CH(CH 3)=N-N=R 2, R3=H, R4=phenyl, R5=Cl;1 Flurazepam (Dolmane):
(Xanax): R =CH(CH3 )=N-N=R , R =H, R =phenyl, R =Cl; Flurazepam (Dolmane): R =(CH2 )2 N(Et)2 , R =O, R3 =H, R4 =2-
1 2 3 4 5 2
R1=(CH2)2N(Et)2, R2=O, R3=H, R41=2-fluorophenyl, R5=Cl; Lorazepam (Ativan): R1=H, R2=O, R3=OH,
fluorophenyl,4 R5 =Cl; Lorazepam (Ativan): R =H, R2 =O, R3 =OH, R4 =2-chlorophenyl, R5 =Cl; Clonazepam (Klonopin): R1 =H,
R =2-chlorophenyl, R 5=Cl; Clonazepam (Klonopin): R1=H, R2=O, R3=H, R4=2-chlorophenyl, R5=NO2.
R2 =O, R3 =H, R4 =2-chlorophenyl, R5 =NO2 . (b) Oxazepam (Serax), and (c) Oxazepam succinate half-ester (hemisuccinate).
(b) Oxazepam (Serax), and (c) Oxazepam succinate half-ester (hemisuccinate).
Attempts of separation of BDZs enantiomers are made difficult by the occurrence of
Attempts racemization
of separationequilibria
of BDZs enantiomers are madeconditions
in enantioseparation difficult by the occurrence
(aqueous medium). of The inversion
racemization equilibria in enantioseparation conditions (aqueous medium). The inversion
of the chiral center occurs very rapidly in vitro but the situation might be different in vivo,
of the chiral center
sinceoccurs very rapidly
the strong in vitro binding
stereospecific but the situation
to human mightserumbe different
albuminincan vivo,be involved [8].
since the strong stereospecific binding to human serum albumin
Recently, four different mechanisms of racemization of Oxazepam have beencan be involved [8]. Re- proposed and
cently, four different
compared mechanisms
with a DFT of approach,
racemization of Oxazepam
denoting have been
the ring-chain proposed and
tautomerization as the best fitting
compared withmechanism
a DFT approach, denoting the ring-chain
of the experimental data available [9]. tautomerization as the best fit-
ting mechanism of the experimental
To avoid data available
rapid racemization of [9].
BDZs during separation processes, the synthesis of
To avoid their
rapidstereochemically
racemization of stable BDZs and duringmore separation
soluble esterprocesses,
derivativesthe synthesis of
has been proposed [10,11].
their stereochemically stable
Interestingly, and more
conjugation of soluble
oxazepam ester derivatives(OXEMIS,
hemisuccinate has beenFigureproposed1c) with dopamine
[10,11]. Interestingly,
improved conjugation of oxazepam
the neurotransmitter hemisuccinate
delivery (OXEMIS,
into the brain Figure 1c)GABAergic
and enhanced with transmis-
dopamine improved the neurotransmitter delivery into the brain and enhanced
sion [12]. Enantioseparation of BDZs hemisuccinates was achieved by using cyclodextrins GABAer-
gic transmission [12]. Enantioseparation
(CDs)-based chromatographic of BDZs
column hemisuccinates
for HPLC [13].was As achieved
a matter of byfact,
usinginteraction at the
cyclodextrins (CDs)-based
apolar cavitychromatographic column fora HPLC
of cyclodextrins constitutes [13].interaction
privileged As a matter of fact, with several
mechanism
interaction at the apolar
kinds cavity of leading
of substrates cyclodextrins constitutes
to the inclusion a privileged
of their interaction
apolar moieties mech-
[14–18]. Moreover, being
anism with several kinds ofchiral,
cyclodextrins substrates leading to
the formation of the inclusion ofinclusion
diastereomeric their apolar moieties
complexes produces poten-
[14–18]. Moreover,
tiallybeing cyclodextrins
differentiable signalschiral,
for thethe
twoformation
includedofenantiomers
diastereomeric inclusion
by chromatographic [19–25]
complexes produces
and/orpotentially
spectroscopic differentiable signals for
techniques [24–28]. the two included
Commonly, inclusion enantiomers
does not involve polar pen-
dants of
by chromatographic enantiomeric
[19–25] substrates which
and/or spectroscopic may protrude
techniques [24–28].from the large and
Commonly, small rims of the
inclu-
cyclodextrins
sion does not involve and interact
polar pendants with cyclodextrin
of enantiomeric hydroxyl
substrates whichfunctions
may protrudeincurring
frominto hydrolytic
processes,
the large and small rims ofwhich have been reported
the cyclodextrins for ester
and interact withcompounds
cyclodextrin[29–31].
hydroxyl func-
tions incurring intoIn consideration
hydrolytic of above
processes, premises,
which have been we focused
reported ourforattention on partially and exhaus-
ester compounds
[29–31]. tively methylated β-cyclodextrins [(2-methyl)-β-CD (MCD), heptakis-(2,6-di-O-methyl)-
β-CD (DIMEB),
In consideration of above and heptakis-(2,3,6-tri-O-methyl)-β-CD
premises, we focused our attention on partially (TRIMEB)] andtoex- be compared to
native β-CD
haustively methylated with regard to
β-cyclodextrins the ability to differentiate
[(2-methyl)-β-CD OXEMIS enantiomers and propensity
(MCD), heptakis-(2,6-di-O-me-
to hydrolyze
thyl)-β-CD (DIMEB), their ester function. Both aspects(TRIMEB)]
and heptakis-(2,3,6-tri-O-methyl)-β-CD were investigated by nuclear magnetic
to be compared
to native β-CDresonance
with regard (NMR) to thespectroscopy. As a matterOXEMIS
ability to differentiate of fact, CDs are very popular
enantiomers and pro-chiral solvating
agents their
pensity to hydrolyze (CSAs) for function.
ester NMR, which Bothhave beenwere
aspects employed to differentiate
investigated by nuclearNMR signals of several
mag-
classes of enantiomeric substrates [26–28]. Furthermore,
netic resonance (NMR) spectroscopy. As a matter of fact, CDs are very popular chiral quantitative response of NMR
makes this technique very appealing both for
solvating agents (CSAs) for NMR, which have been employed to differentiate NMR sig-the determination of enantiomeric excesses
nals of severaland for the
classes detection andsubstrates
of enantiomeric quantification
[26–28].of degradation
Furthermore, products. Finally,
quantitative re- enormous po-
tentialities of NMR in the field of investigations
sponse of NMR makes this technique very appealing both for the determination of enan- of complexation phenomena have been
exploited
tiomeric excesses and for to the
investigate
detection theandinteraction mechanisms
quantification involving the
of degradation cyclodextrin-based
products. Fi- chiral
auxiliaries and enantiomeric OXEMIS at a molecular
nally, enormous potentialities of NMR in the field of investigations of complexation phe- level and, hence, ultimately, the chiral
nomena have beendiscrimination
exploited to processes.
investigate the interaction mechanisms involving the cy-
clodextrin-based chiral auxiliaries and enantiomeric OXEMIS at a molecular level and,
hence, ultimately, the chiral discrimination processes.
Molecules 2021, 26, x FOR PEER REVIEW 3 of 11
Molecules
Molecules 2021,
2021, 26,26, x FOR PEER REVIEW
6347 3 3ofof11
11

2.2.Results
Results and and Discussion
Discussion
2.
2.1.Results
OXEMIS and Discussion
Characterization
2.1.OXEMIS
2.1. OXEMISCharacterization
Characterization
1H NMR spectrum of OXEMIS (1 mM, Figure 2a) showed three sets of signals. The
1H NMR spectrum of OXEMIS (1 mM, Figure 2a) showed three sets of signals. The
1 H NMR spectrum of OXEMIS (1 mM, Figure 2a) showed three sets of signals. The
aromatic
aromatic protons resonances
protons resonances between
resonances between
between 7.17.1 and
7.1 and 7.6
7.6 ppm
ppm were easily
easily distinguishedon on the
aromatic
basis of protons
multiplicity, while H and 7.6 ppm were
were easily distinguished
distinguished onthe
3 was the only isolated singlet clearly observable at 5.89 ppm.
the
basisof
basis ofmultiplicity,
multiplicity,while
whileH H3 was
was the
the only
only isolated
isolated singlet
singlet clearly
clearly observable
observable atat5.89
5.89ppm.
ppm.
Methylene group adjacent to3 the chiral center at C3 (CH2b, 2.73 ppm) was differentiated
Methylenegroup
Methylene group adjacent
adjacent to
to the
the chiral
chiral center
center atat CC3 (CH
(CH2b,, 2.73 ppm)
ppm) was
was differentiated
from CH 2b 2.73
effect3produced
differentiated
from CH 2a (2.43
2a (2.43 ppm)
ppm) on
on the
the basis
basis of the
of the ROE
ROE effect produced atthe
at thefrequency
frequencyofofHH3 3(Fig-
(Fig-
from CH2a (2.43 ppm) on the basis of the ROE effect1 produced at the frequency of H3
ure S1a, Supplementary
ure S1a,S1a,
Supplementary Material).
Material). The dependence
The dependence of H NMR
of H NMR chemical shifts of OXEMIS
of 1 Hchemical shifts of OXEMIS
1
(Figure Supplementary Material). The dependence NMR chemical shifts of
on concentration witnessed the presence of self-association
self-associationprocesses
processes(Figure
(Figure2a,b),
2a,b), which
OXEMIS on concentration witnessed the presence of self-association processes (Figurewhich
on concentration witnessed the presence of 2a,b),
could
couldin inprinciple
principle play
play a play
relevant role in the hetero-complexation
hetero-complexationbehavior.behavior.
which could in principle a relevant role in the hetero-complexation behavior.

Figure2.
Figure 2.2.111H
H NMR
H NMR (600
NMR (600 MHz,
(600 MHz, K
MHz, HPO44/D
K222HPO 22O 50 mM,
mM, 25
25 °C)
◦ spectra
°C) spectraof
ofOXEMIS:
OXEMIS:(a)
(a)11mM
mMand
and(b)
(b)1212
Figure 4 /D2 O 50 mM, 25 C) spectra of OXEMIS: (a) 1 mM and (b)
mM.
mM.
12 mM.

Lowsolubility
Low
Low solubility of
solubility of OXEMIS
OXEMIS allowed
allowed to to investigate
investigate only only the
the13.7
13.7mMmMto to0.3
0.3mM mMcon- con-
centration gradient, inside which the greatest chemical shift variations (∆δ = δ[13.7 δmM]
centration
centration gradient,
gradient, inside
inside which
which the
the greatest
greatest chemical
chemical shift
shift variations
variations (Δδ
(Δδ = =δδ
[13.7 mM]
[13.7 mM]− − δ[0.3[0.3
−mM]
mM]δ ) were
)[0.3
were detected
mM]detected
for H
for H3/Hfor
) were detected 3 /H (≈−27
9 (≈−27
9 H3 /H Hz)
Hz) and
≈−27
9 (and
H
H6/H
6 /H (≈−12
8 (≈−12
Hz)
8 and H Hz)
Hz) (Figure
6 /H 8 (≈−12
(Figure S2, Supplementary
S2,Hz)
Supplementary
(Figure S2,
Material). Protons
Supplementary
Material). Protons of the
of the lateral
Material). lateral
Protonschain
chainof (CH 2a/CH2b)chain
2a/CH2b) were
the lateral
(CH affected
were(CH 2a /CH
affected to aaless
to2b )less extent.
were Interest-
affected
extent. to a
Interest-
ingly,
less
ingly, every
extent. proton underwent
everyInterestingly,
proton underwent low-frequency
every low-frequency
proton underwent shifts with concentration
shiftslow-frequency
with concentration increasing,
shifts increasing,
with concentra- hint-
hint-
ingaaincreasing,
tion
ing π-π stacking-like
π-π stacking-like self-association
hintingself-association
a π-π stacking-like of
of OXEMIS, which
whichwas
self-association
OXEMIS, was ofwitnessed
OXEMIS,by
witnessed bythe
thenature
which was wit-
nature ofof
dipolarby
nessed
dipolar correlations detected
the naturedetected
correlations of dipolar in the
in the 2D
2D ROESY
correlations map
map (Figure
ROESYdetected in the
(Figure S1b,
2DSupplementary
S1b, ROESY map (Figure
Supplementary Material).S1b,
Material).
As a matter
Supplementary of fact, H
Material).3 produced
As a unexpected
matter
As a matter of fact, H3 produced unexpected ROEs of fact,ROEs
H 3
at the
produced frequencies
unexpected of H 10/H11at
ROEs
at the frequencies of H 10/H11 (C-ring) (C-ring)
the fre-
and H
quencies 6/H 9 (B-ring).
of H
and H6/H9 (B-ring).
10 /H 11
A representation
(C-ring) and
A representation of H of
/Hself-associated
(B-ring).
6 self-associated
9 A OXEMIS is
representationshown ofin Figure 3.
self-associated
OXEMIS is shown in Figure 3.
OXEMIS is shown in Figure 3.

Figure 3. 3D representation of OXEMIS dimer as obtained from NMR data.

Figure 3.
Figure 3. 3D
3D representation
representation of
of OXEMIS
OXEMIS dimer
dimer as
as obtained
obtained from
from NMR
NMR data.
data.
An attempt to calculate self-association constant (Ka), based on the non-linear fitting
An
An attempt
attempt to calculate self-association constant (K
(Kaa), based on
on the
the non-linear fitting
of dilution data, to calculate
failed due toself-association constant
the fact that limited solubility based
of OXEMIS non-linear
did not allowfitting
to
of
of dilution data, failed due to the fact that limited solubility of OXEMIS did not allow to
explore a sufficiently wide range of concentrations. Indeed, an almost linear dependenceto
dilution data, failed due to the fact that limited solubility of OXEMIS did not allow
explore
explore aa sufficiently
sufficiently wide
wide range
range of
of concentrations.
concentrations. Indeed,
Indeed, an
an almost
almost linear
linear dependence
dependence
of δobs on concentration was found (Figure S3a, Supplementary Material), leading to great
of
of δδobs on concentration
obs on concentration waswas found (Figure S3a, Supplementary Material), leading to great
errors in Ka determination. On the other hand, a quite small K aMaterial), leading
was calculated in to great
CDCl 3
errors in K determination. On the other hand, a quite small K was calculated
errors in Ka determination. On the other hand, a quite small K a was calculated in CDCl33
a a in CDCl
(Ka = 5.1 ± 0.2 M−1 ) where a wider range of concentrations could be explored (0.5 to
Molecules 2021, 26, x FOR PEER REVIEW 4 of 11
Molecules 2021, 26, 6347 4 of 11

(Ka = 5.1 ± 0.2 M−1) where a wider range of concentrations could be explored (0.5 to 132
mM),
132 as shown
mM), in Figure
as shown S3b,
in Figure Supplementary
S3b, Supplementary Material. It isIt noteworthy
Material. that,
is noteworthy based
that, basedon
such
on K value,
such the the
K value, molar fraction
molar of dimer
fraction was less
of dimer was than 3%, and
less than 3%,hence negligible,
and hence in the
negligible,
concentration
in conditions
the concentration (6 mM)(6we
conditions employed
mM) for the investigation
we employed of OXEMIS
for the investigation complex-
of OXEMIS
ation by cyclodextrins.
complexation by cyclodextrins.

2.2.
2.2. CDs
CDs Characterization
Characterization
The
The three
three cyclodextrins
cyclodextrins candidates
candidates for
for OXEMIS
OXEMIS enantiodiscrimination
enantiodiscrimination were
were MCD,
MCD,
DIMEB, and TRIMEB (11H NMR spectra shown in Figure 4a–c).
DIMEB, and TRIMEB ( H NMR spectra shown in Figure 4a–c).

Figure4.
Figure 4. 11H
H NMR
NMR (600
(600 MHz,
MHz,KK22HPO
HPO44/D
/D22OO50
50mM,
mM, 25
25 °C,
◦ C, 12
12 mM)
mM) spectra of: (a)
spectra of: MCD, (b)
(a) MCD, (b) DIMEB,
DIMEB,
and (c) TRIMEB.
and (c) TRIMEB.

MCD is
MCD is aa cyclodextrin
cyclodextrinderivatized
derivatizedatatOH OH 2, with
2 , witha substitution
a substitution degree of 3.5,
degree of as care-
3.5, as
fully determined
carefully determined in thein1HtheNMR1 H NMR
spectrum by comparing
spectrum the integrated
by comparing areas ofareas
the integrated the ano-
of
meric protons resonances between 5.3 to 4.9 ppm and the
the anomeric protons resonances between 5.3 to 4.9 ppm and the superimposed proton superimposed proton signals of
ring and methoxy groups in the spectral region 4.0 to 3.2
signals of ring and methoxy groups in the spectral region 4.0 to 3.2 ppm. Two kinds ofppm. Two kinds of anomeric
protons were
anomeric protonsdistinguished, which originated
were distinguished, by non-methylated
which originated (4.94 ppm)
by non-methylated (4.94and
ppm)methyl-
and
ated (5.13 ppm)
methylated (5.13glucopyranose
ppm) glucopyranose units. As a matter
units. As aofmatter
fact, selective
of fact, perturbation at the fre-
selective perturbation
quency
at of 5.13 ppm
the frequency resulted
of 5.13 in the expected
ppm resulted ROE effect
in the expected ROEat the frequency
effect of the methoxy
at the frequency of the
group (see
methoxy [32] for
group (seea[32]
detailed
for a characterization of MCD).of MCD).
detailed characterization
In the 11H NMR spectrum
In the spectrum of of DIMEB
DIMEB (Figure
(Figure 4b),4b), two
two singlets
singlets attributable
attributable toto methoxy
methoxy
groups were clearly identified at 3.44 ppm and 3.27 ppm.
groups were clearly identified at 3.44 ppm and 3.27 ppm. Comparison of chemical Comparison of chemical shifts of
shifts
DIMEB
of DIMEB andand native β-CD
native β-CD (Table
(Table1) showed
1) showed thatthat
majormajordifferences
differences between
betweentheir chemical
their chemi-
shifts werewere
cal shifts found for Hfor
found 1, HH21,,and H6 protons,
H2, and as expected
H6 protons, as expectedfor a for
2,6-disubstitution. Perturba-
a 2,6-disubstitution. Per-
tion at the at
turbation frequency of anomeric
the frequency protonproton
of anomeric H1 at 5.11H1 at ppm5.11produced
ppm producedROE atROE3.44 ppm,
at 3.44which
ppm,
was
whichattributed to OMeto
was attributed 2 (Figure S4, Supplementary
OMe2 (Figure S4, Supplementary Material), thus allowing
Material), to attribute
thus allowing to at-
OMe
tribute at 3.27 ppm. The ratio between integrated
6 OMe6 at 3.27 ppm. The ratio between integrated areas areas of H and ring protons (including
1 of H1 and ring protons (in-
methoxy signals) confirmed
cluding methoxy the presence
signals) confirmed of two methoxy
the presence groups per
of two methoxy glucopyranose
groups unit
per glucopyra-
of (DS ≈
nose unit of DIMEB (DS ≈ 14). Remaining ring protons were assigned (Table 1) by 2D
DIMEB 14). Remaining ring protons were assigned (Table 1) by 1D ROESY, 1D
COSY,
ROESY, and2D2D HSQC
COSY, andexperiments (Figures S5–S7,
2D HSQC experiments Supplementary
(Figures Material). Material).
S5–S7, Supplementary
TRIMEB produced a single set of well-resolved signals (Figure 4c), resonances of
which were attributed (Table 1) by comparing scalar and dipolar interactions detected by
1D ROESY, 2D COSY, and 2D HSQC experiments (Figures S8–S10, Supplementary Mate-
rial).
Molecules 2021, 26, 6347 5 of 11

Molecules 2021, 26, x FOR PEER REVIEW 5 of 11

Table 1. 1 H NMR (600 MHz, 12 mM, K2 HPO4 /D2 O 50 mM, 25 ◦ C) chemical shifts (ppm) of β-CD,
MCD, DIMEB, and TRIMEB protons.
Table 1. 1H NMR (600 MHz, 12 mM, K2HPO4/D2O 50 mM, 25 °C) chemical shifts (ppm) of β-CD,
MCD, DIMEB, and TRIMEB protons.
β-CD MCD 1 DIMEB TRIMEB
H1 β-CD4.94 MCD 4.94/5.13
1 DIMEB 5.11 5.16
TRIMEB
HH1 2 4.94 3.52 3.53/3.26
4.94/5.13 5.113.26 3.22
5.16
H3 3.84 3.81/3.88 3.85 3.56
HH2 3.52 3.46 3.53/3.26
3.45/3.48 3.263.47 3.22
3.62
4
HH3 5 3.84 3.74 3.81/3.88
3.73/3.70 3.853.77 3.56
3.74
H
H6/6
4 0 3.46 3.75 3.74/3.74
3.45/3.48 3.473.63 3.53/3.72
3.62
OMe
H5 2 3.74 - 3.73/3.70 3.44 3.773.44 3.39
3.74
OMe3 - - 3.27 3.48
H6/6′
OMe6
3.75 - 3.74/3.74- 3.63 - 3.53/3.72
3.25
OMe2 - 3.44 3.44 3.39
1 Chemical shifts relative to non-methylated and 2-methylated MCD glucopyranose units, both present in MCD.
OMe3 - - 3.27 3.48
TRIMEB
OMe6 produced- a single set of well-resolved - signals (Figure
- 4c), resonances3.25of which
were attributed
1 Chemical (Table to
shifts relative 1) non-methylated
by comparing and scalar and dipolar
2-methylated MCD interactions
glucopyranosedetected by 1D
units, both
ROESY, 2D
present in COSY, and 2D HSQC experiments (Figures S8–S10, Supplementary Material).
MCD.

2.3.
2.3. OXEMIS
OXEMIS Hydrolysis
Hydrolysis by by Cyclodextrins
Cyclodextrins
The
The chemical
chemicalstability
stabilityofofOXEMIS
OXEMISover overtime
timewas
wascompared
compared as as
pure compound
pure compound andandin
the presence
in the of 2ofequivalents
presence 2 equivalents of each
of eachcyclodextrin,
cyclodextrin,in in
buffered
buffered solutions
solutions (K(K2 HPO
2HPO 4 /D
4/D22O,
O,
50 mM). OXEMIS hydrolyzation was followed directly in the NMR tube (T = 25 ◦ C) and in
50 mM). OXEMIS hydrolyzation was followed directly in the NMR tube (T = 25 °C) and
static conditions.
in static TheThe
conditions. resonance of CH
resonance group
of2aCH located on hemisuccinic chain was selected
2a group located on hemisuccinic chain was se-
for monitoring the process. The parent cyclodextrin
lected for monitoring the process. The parent cyclodextrin β-CD wasβ-CD includedwas for comparison
included for com- as
well. Less than 1% of hydrolysis product (OX) was detected after 6 h in
parison as well. Less than 1% of hydrolysis product (OX) was detected after 6 h in the casethe case of pure
OXEMIS, whereas whereas
of pure OXEMIS, a rapid hydrolyzation was detected
a rapid hydrolyzation in the presence
was detected of β-CD, of
in the presence with an
β-CD,
exponential decay during
with an exponential decay theduring
very first
the points of the
very first kinetic
points (firstkinetic
of the 2 h, Figure
(first 5).
2 h, Hydrolysis
Figure 5).
was observed even in the presence of MCD, but with a remarkably
Hydrolysis was observed even in the presence of MCD, but with a remarkably slower slower rate (Figurerate5).
As an example, after 6 h, the concentration of hydrolysis product in presence
(Figure 5). As an example, after 6 h, the concentration of hydrolysis product in presence of β-CD was
five-fold (48%) than that detected in the presence of MCD (11%).
of β-CD was five-fold (48%) than that detected in the presence of MCD (11%).

Figure5.
Figure 5. OXEMIS
OXEMIS concentration
concentrationin
inthe
thepresence
presenceof
of2-equivalents
2-equivalentsofofβ-CD
β-CDor
orMCD
MCD(12
(12mM)
mM)ininDDO2O
2
(K2HPO4 50 mM) over time, as detected by 11H NMR (600 MHz, 25 °C).
◦
(K2 HPO4 50 mM) over time, as detected by H NMR (600 MHz, 25 C).

Therefore,the
Therefore, thehydrolysis
hydrolysisofofOXEMIS
OXEMIS is is correlated
correlated with
with thethe availability
availability of free
of free reac-
reactive
tive -OH groups in the cyclodextrin, able to establish hydrogen bonds, probably
-OH groups in the cyclodextrin, able to establish hydrogen bonds, probably with the with the
hemisuccinic pendant. Accordingly, DIMEB and TRIMEB did not
hemisuccinic pendant. Accordingly, DIMEB and TRIMEB did not produce hydrolysis produce hydrolysis
withinthe
within the24
24h,h,pointing
pointingoutout
thethe relevance
relevance of of
thethe interaction
interaction of hemisuccinic
of hemisuccinic chainchain
withwith
the
the hydroxyls
hydroxyls atCthe
at the 2 C
and2 and
C6 C 6 sites.
sites.
Molecules 2021, 26, 6347 6 of 11
Molecules 2021, 26, x FOR PEER REVIEW 6 of 11

2.4.
2.4.OXEMIS
OXEMISEnantiodiscrimination
Enantiodiscrimination
When
When MCDwas
MCD was added
added to pure rac-OXEMIS,
to pure rac-OXEMIS, only variations in
only variations in chemical
chemical shifts
shiftswere
were
detected without signal splitting (Figure 6a,b, Table 2). Hence, a complexation occurred
detected without signal splitting (Figure 6a,b, Table 2). Hence, a complexation occurred
which
whichdid
didnot
notproduce
producechiral
chiraldiscrimination.
discrimination.

Figure6.6.11H
H NMR
NMR (600
(600 MHz,
MHz, K
K22HPO
HPO4/D 2O 50 mM, 25 °C)◦ spectra of: (a) pure OXEMIS (6 mM), (b)
Figure 4 /D2 O 50 mM, 25 C) spectra of: (a) pure OXEMIS (6 mM),
OXEMIS/MCD (1:2), (c) (R)-OXEMIS/DIMEB (1:2), (d) (S)-OXEMIS/DIMEB (1:2), (e) (R)-
(b) OXEMIS/MCD (1:2), (c) (R)-OXEMIS/DIMEB (1:2), (d) (S)-OXEMIS/DIMEB (1:2), (e) (R)-
OXEMIS/TRIMEB (1:2), and (f) (S)-OXEMIS/TRIMEB (1:2).
OXEMIS/TRIMEB (1:2), and (f) (S)-OXEMIS/TRIMEB (1:2).
Table2.2.1 H
Table
1H NMR (600 MHz, K2HPO4/D2O 50 mM, 25 °C) chemical shifts (ppm) of pure OXEMIS (6
NMR (600 MHz, K2 HPO4 /D2 O 50 mM, 25 ◦ C) chemical shifts (ppm) of pure OXEMIS
mM) and complexation shifts (Δδ = δmixture − δpure, ppm) of OXEMIS in mixture with 2-equivalents of
(6 mM) and complexation shifts (∆δ = δmixture − δpure , ppm) of OXEMIS in mixture with 2-equivalents
MCD, DIMEB, and TRIMEB.
of MCD, DIMEB, and TRIMEB.
OXEMIS/MCD OXEMIS/DIMEB OXEMIS/TRIMEB
OXEMIS/MCD OXEMIS/DIMEB OXEMIS/TRIMEB
Ring δ Δδ Δδ(S) Δδ(R) Δδ(S) Δδ(R)
Ring
A H3 5.89 δ −0.06∆δ ∆δ(S) −0.03
−0.03 ∆δ(R) −0.05 ∆δ(S) ∆δ(R)
−0.04
BA H6 H3 7.25 5.89 −0.06
−0.16 −0.03 −0.19
−0.18 −0.03 −0.04−0.05 −0.04
−0.06
BB H8 H6 7.54 7.25 −
0.02 0.16 −
0.06 0.18 −
0.03 0.19 − 0.04
0.00 −0.06
0.00
B H 7.54 0.02 0.06 0.03 0.00 0.00
B H9 8 7.20 0.05 0.10 0.09 0.01 0.02
B H9 7.20 0.05 0.10 0.09 0.01 0.02
CC H10H 7.40 7.40 −0.03
−0.03 0.060.06 −0.04−0.04 0.05 0.05 0.01
0.01
10
CC H11H11 7.37 7.37 −0.05
−0.05 −0.13
−0.13 −0.08−0.08 0.02 0.02 0.01
0.01
CC H12H12 7.49 7.49 0.010.01 0.010.01 −0.01−0.01 −0.01−0.01 −0.01
−0.01
-- CHCH
2a 2a 2.43 2.43 −0.01
−0.01 −0.03 −0.01
−0.03 −0.01 −0.01−0.01 0.00
0.00
- CH2b 2.73 0.05 0.11 0.06 −0.01 −0.01
- CH2b 2.73 0.05 0.11 0.06 −0.01 −0.01

By
Bycontrast,
contrast, DIMEB
DIMEB produced
produced complexation
complexation shifts shifts(Δδ (∆δ==δδmixture − δ−
mixture δ)pure
pure ) along
along withwith
en-
enantiodifferentiation
antiodifferentiation ofofseveral severalOXEMIS
OXEMISprotonsprotons(Figure
(Figure6a,c,d).
6a,c,d).InInparticular,
particular,significant
significant
complexation
complexation shifts shifts were
were measured
measured for for HH66, H99,, H
H1010, ,and
andHH1111(0.04
(0.04toto0.19
0.19ppm,
ppm,Table
Table2).
2).
Major
Majornon-equivalences
non-equivalences (∆∆δ (ΔΔδ == |Δδ Δδ∆δ
|∆δ(S)(S)− − (R)|)
(R) |)
werewere detected
detected (Table
(Table 2) 2)
for for
the the protons
protons be-
belonging
longing to to ring
ring C and
C and for methylene
for the the methylene group group CH2btonear
CH2b near to the
the ester ester moiety.
moiety. Inter-
Interestingly,
estingly, (S)-enantiomer protons were more affected by the presence
(S)-enantiomer protons were more affected by the presence of DIMEB, being its complex- of DIMEB, being its
complexation shifts greater
ation shifts greater thanofthose
than those of (R)-enantiomer,
(R)-enantiomer, with the with
solethe sole exception
exception of protonof proton
H6. It
His6 .noteworthy
It is noteworthy
that, inthat,
eachincase,
eachinternal
case, internal
protonsprotons
of DIMEB of DIMEB
were more wereaffected
more affected
by the
by the presence
presence of OXEMISof OXEMIS than external
than external ones wereones(Table
were 3), (Table
as a 3),
clearas aindication
clear indication of the
of the occur-
occurrence of inclusion
rence of inclusion phenomena.
phenomena.
Molecules 2021, 26, 6347 7 of 11

Table 3. 1 H NMR (600 MHz, K2 HPO4 /D2 O 50 mM, 25 ◦ C) complexation shifts (∆δ = δmixture − δpure ,
ppm) of cyclodextrin protons (12 mM) in the presence of OXEMIS (6 mM).

DIMEB/OXEMIS TRIMEB/OXEMIS
Proton ∆δ(S) ∆δ(R) ∆δ(S) ∆δ(R)
H1 −0.02 −0.02 −0.01 −0.01
H2 −0.03 0.00 −0.01 −0.01
H3 −0.09 −0.07 −0.02 −0.01
H4 −0.03 −0.01 −0.01 −0.01
H5 −0.16 −0.09 −0.03 −0.01
H6/60 −0.02 −0.06 0.00 0.00
OMe2 0.00 0.00 −0.01 −0.01
OMe3 −0.01 −0.01 −0.03 −0.03
OMe6 - - 0.00 0.00

In presence of TRIMEB, smaller complexation shifts and non-equivalences were de-

tected (Table 2). In particular, methylene protons of the hemisuccinate fragment were
almost unaffected by TRIMEB (∆δ = 0.00 to 0.01 ppm, Table 2). Correspondingly, small
complexation shifts were also measured for TRIMEB protons (Table 3, Figure S11 in
Supplementary Material), with the internal ones more affected, in particular in the mixture
containing the (S)-enantiomer.

2.5. OXEMIS/CD Interaction Mechanism

To give a better understanding of the origin of chiral discrimination process of OX-
EMIS due to its interaction with DIMEB and TRIMEB, stoichiometries and association
constants of the diastereomeric complexes were determined in mixtures containing each cy-
clodextrin and OXEMIS enantiomer. By using the Job method [33], both OXEMIS/DIMEB
and OXEMIS/TRIMEB complexes showed 1-to-1 interaction stoichiometries for both the
guest enantiomers (Figure S12, Supplementary Material). Association constants (Ka ) of
the diastereomeric complexes were calculated by using the Foster-Fyfe method [34] in
solutions containing a fixed amount of OXEMIS and increasing amounts of the cyclodex-
trin (Figure S13, Supplementary Material). (S)-OXEMIS formed a tighter complex with
DIMEB with respect to (R)-OXEMIS (Ka = 930 ± 38 M−1 vs. 137 ± 11 M−1 , respectively),
whereas association constants for TRIMEB were significantly lower for both enantiomers
(Ka = 40.5 ± 3.0 M−1 for (S)-OXEMIS vs. 2.5 ± 0.9 M−1 for (R)-OXEMIS).
The interaction stereochemistry was investigated by 1D and 2D ROESY analyses.
As shown in Figure 7a, intermolecular ROEs were detected between internal protons of
DIMEB and phenyl C-ring protons along with H6 and H3 of (S)-OXEMIS. In particular, H3
protons of DIMEB (located on the internal wider rim) produced the most intense cross-peak
on H10 and significant dipolar interactions with H6 and H3 protons of (S)-OXEMIS. By
contrast, H5 protons of DIMEB (on the internal narrower rim) gave the main ROEs with H11
and H10 of phenyl C-ring. Such proximity constraints support a deep inclusion of phenyl
C-ring of (S)-OXEMIS from the wider cavity of DIMEB, according to the most relevant
complexation shifts measured for (S)-OXEMIS protons (phenyl C-ring, H6 and CH2b ) and
for internal protons (∆δH5 > ∆δH3 ) of DIMEB (Tables 2 and 3).
In the mixture (R)-OXEMIS/DIMEB analogous ROEs between internal protons of
DIMEB and H10 /H11 (phenyl C-ring), H6 and H3 of (R)-OXEMIS were detected (Figure 7b),
but with lower intensities.
The analysis of (S)-OXEMIS/TRIMEB and (R)-OXEMIS/TRIMEB mixtures showed
ROE patterns for phenyl C-ring similar to those observed in the presence of DIMEB (Figure 7),
confirming the inclusion of ring C from the wider rim of TRIMEB. However, H6 and H3
protons of OXEMIS did not give ROE effects with internal protons of TRIMEB and H12
proton of phenyl C-ring was in spatial proximity of H5 protons of TRIMEB, in according to a
less deep inclusion of the guest with respect to DIMEB complexes.
Molecules 2021, 26, 6347 8 of 11
Molecules 2021, 26, x FOR PEER REVIEW 8 of 11

Figure 7. 1H NMR (600 MHz, K2HPO4/D2O 50 mM, 25 °C) and 1D ROESY spectra (H3 and H5 pro-
Figure 7. 1 H NMR (600 MHz, K2 HPO4 /D2 O 50 mM, 25 ◦ C) and 1D ROESY spectra (H3 and H5
tons) of cyclodextrins for: (a) (S)-OXEMIS/DIMEB (1:2), (b) (R)-OXEMIS/DIMEB (1:2), (c) (S)-
protons) of cyclodextrins for: (a) (S)-OXEMIS/DIMEB (1:2), (b) (R)-OXEMIS/DIMEB (1:2), (c) (S)-
OXEMIS/TRIMEB (1:2), and (d) (R)-OXEMIS/TRIMEB (1:2).
OXEMIS/TRIMEB (1:2), and (d) (R)-OXEMIS/TRIMEB (1:2).
The relative magnitude of ROE effects for the two enantiomers in presence of DIMEB
The relative magnitude of ROE effects for the two enantiomers in presence of DIMEB
orTRIMEB
or TRIMEBwellwellreflects
reflectsthe
thestrength
strengthof
ofOXEMIS/cyclodextrin
OXEMIS/cyclodextrin interaction
interaction as
as quantified
quantifiedby
by
association constants determination.
association constants determination.

3. Materials
3. Materials and
and Methods
Methods
3.1.
3.1. Materials
Materials
Heptakis-(2,6-di-O-methyl)-β-CD
Heptakis-(2,6-di-O-methyl)-β-CD(DIMEB, (DIMEB, 98%) andand
98%) K2 HPO 4 , β-Cyclodextrin
K2HPO (β-CD),
4, β-Cyclodextrin (β-
heptakis-(2,3,6-tri-O-methyl)-β-CD
CD), heptakis-(2,3,6-tri-O-methyl)-β-CD (TRIMEB, 98%) 98%)
(TRIMEB, were were
purchased
purchasedfrom from
Sigma-Aldrich
Sigma-Al-
(St. Louis,
drich MO, USA).
(St. Louis, Deuterated
MO, USA). water (D
Deuterated 2 O) and
water deuterated
(D2O) chloroform
and deuterated (CDCl3(CDCl
chloroform ) were3)
purchased from Deutero GmbH (Kastellaun, Germany). Oxazepam (7-chloro-3-hydroxy-5-
were purchased from Deutero GmbH (Kastellaun, Germany). Oxazepam (7-chloro-3-hy-
phenyl-1,3-dihydro-2H-1,4-benzodiazepin-2-one,
droxy-5-phenyl-1,3-dihydro-2H-1,4-benzodiazepin-2-one, OX) was kindly
OX) wasdonated by Profarmaco
kindly donated by
Spa (Milan, Italy). Racemic hemisuccinate half-ester of Oxazepam was
Profarmaco Spa (Milan, Italy). Racemic hemisuccinate half-ester of Oxazepam was prepared by acy-
pre-
lation
paredwith succinicwith
by acylation anhydride
succinicinanhydride
presence of inpyridine
presenceaccording
of pyridine toaccording
[13]. The pure enan-
to [13]. The
tiomers of OX (enantiomeric
pure enantiomers excess (e.e.)excess
of OX (enantiomeric >98%,(e.e.)
determined by HPLC on
>98%, determined byβ-cyclodextrin-
HPLC on β-cy-
based CSP—Cyclobond
clodextrin-based I) were obtained
CSP—Cyclobond by fractional
I) were obtainedcrystallization
by fractional(acetone/water
crystallization100:1)
(ace-
of diastereomeric salts with ( − )-(1R,2S)-ephedrine.
tone/water 100:1) of diastereomeric salts with (−)-(1R,2S)-ephedrine.
3.2. NMR Experiments
3.2. NMR Experiments
NMR measurements were performed on Varian INOVA600 spectrometer (Varian Inc.,
NMR measurements were performed on Varian INOVA600 spectrometer (Varian
Palo Alto, CA, USA) operating at 600 MHz for 1 H, equipped with Varian temperature
Inc., Palo Alto, CA, USA) operating at 600 MHz for 1H, equipped with Varian temperature
control unit (25 ± 0.1 ◦ C). 2D NMR spectra were obtained by using standard sequences.
control unit (25 ± 0.1 °C). 2D NMR spectra were obtained by using standard sequences.
Spectral width used was the minimum required in both dimensions. 2D gCOSY (gradient
Spectral width used was the minimum required in both dimensions. 2D gCOSY (gradient
correlated spectroscopy) spectra were recorded in the absolute mode acquiring 8 scans
correlated spectroscopy) spectra were recorded in the absolute mode acquiring 8 scans
with a 1 s relaxation delay between acquisitions and 4K data points for each of 200 FIDs.
with a 1 s relaxation delay between acquisitions and 4K data points for each of 200 FIDs.
2D TOCSY (total correlation spectroscopy) spectra were recorded acquiring 8 scans with
2D TOCSY (total correlation spectroscopy) spectra were recorded acquiring 8 scans with
a 1 s relaxation delay, 200 increments, 4K data points, and a mixing time of 120 ms. 1D
a 1 s relaxation delay, 200 increments, 4K data points, and a mixing time of 120 ms. 1D
ROESY (rotating-frame Overhauser enhancement spectroscopy) spectra were recorded
ROESY (rotating-frame Overhauser enhancement spectroscopy) spectra were recorded
Molecules 2021, 26, 6347 9 of 11

with a mixing time of 0.3 s, a delay of 1 s, and 1024 scans. 2D ROESY map was recorded
with mixing time of 0.3 s and 32 scans. gHSQC (gradient Heteronuclear Single Quantum
Coherence) spectra were obtained with 1.2 s relaxation delay and 16 scans for each of the
128 increments.

3.3. NMR Samples Preparation for Stoichiometries and Association Constants

The NMR samples for stoichiometries determination were prepared by mixing differ-
ent volumes of stock solutions of each component (8 mM) to a prefixed volume directly
into the NMR tube.
The NMR samples for the determination of association constants of the diastereomeric
complexes formed by (S)- and (R)-OXEMIS with DIMEB or TRIMEB were prepared at
constant concentration of OXEMIS (0.2 mM) and CD/OXEMIS ratio ranging from 1 to 200.

4. Conclusions
Among methylated cyclodextrins, DIMEB was the more suited both to interact with the
two enantiomers of OXEMIS and differentiate them, without hydrolysis of the hemisuccinic
arm. Phenyl ring C of OXEMIS well fits in the cavity sizes both of DIMEB and TRIMEB,
but enhanced stabilization of the two enantiomers inside the cavity of DIMEB is correlated
to the presence of residual underivatized OH functions, probably due to their hydrogen
bond interactions with the hemisuccinate pendant protruding from the cyclodextrin cavity.
Degradation processes are mainly due to the presence of hydroxyl groups at C2 and C6 , as
found in the case of MCD and native β-CD, with a particular relevance of C2 -hydroxyls.

Supplementary Materials: The following are available online: Figure S1: 2D ROESY map expansion
of OXEMIS; Figure S2: Dependence of OXEMIS 1 H NMR chemical shifts variations on concentration;
Figure S3: Fitting of dilution data of OXEMIS in D2 O and CDCl3 ; Figure S4: 1D ROESY and 1 H
NMR spectra of DIMEB; Figure S5: 1D ROESY spectrum of DIMEB for H1; Figure S6: 2D COSY
map of DIMEB; Figure S7: 2D HSQC map of DIMEB; Figure S8: 1D ROESY spectrum of TRIMEB
for H1; Figure S9: 2D COSY map of TRIMEB; Figure S10: 2D HSQC map of TRIMEB; Figure S11:
1 H NMR spectra of pure DIMEB, (R)-OXEMIS/DIMEB, (S)-OXEMIS/DIMEB, pure TRIMEB, (R)-

OXEMIS/TRIMEB, and (S)-OXEMIS/TRIMEB; Figure S12: Job plot of (S)-OXEMIS in the presence of
TRIMEB; Figure S13: Foster-Fyfe determination of association constant for (R)- and (S)-OXEMIS with
DIMEB and TRIMEB.
Author Contributions: Conceptualization, G.U.B. and F.B.; methodology, G.U.B. and F.B.; software,
A.C.; validation, A.C, A.R. and F.B.; formal analysis, A.C. and A.R.; investigation, A.C. and A.R.;
data curation, A.C. and A.R.; writing—original draft preparation, A.C.; writing—review and editing,
A.C., A.R., F.B. and G.U.B.; visualization, A.C.; supervision, G.U.B. and F.B.; project administration,
G.U.B.; funding acquisition, G.U.B. All authors have read and agreed to the published version of the
manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Data are included in this published article.
Conflicts of Interest: The authors declare no conflict of interest.
Sample Availability: Not available.

References
1. Griffin, C.E., III; Kaye, A.M.; Kaye, A.D. Benzodiazepine Pharmacology and Central Nervous System–Mediated Effects. Ochsner J.
2013, 13, 214–223.
2. Schmitz, A. Benzodiazepine Use, Misuse, and Abuse: A Review. Ment. Health Clin. 2016, 6, 120–126. [CrossRef]
3. Pham-Huy, C.; Villain-Pautet, G.; Hua, H.; Chikhi-Chorfi, N.; Galons, H.; Thevenin, M.; Claude, J.-R.; Warnet, J.-M. Separation of
oxazepam, lorazepam, and temazepam enantiomers by HPLC on a derivatized cyclodextrin-bonded phase: Application to the
determination of oxazepam in plasma. J. Biochem. Biophys. Methods 2002, 54, 287–299. [CrossRef]
Molecules 2021, 26, 6347 10 of 11

4. Court, M.H.; Duan, S.X.; Guillemette, C.; Journault, K.; Krishnaswamy, S.; Von Moltke, L.L.; Greenblatt, D.J. Stereoselective
Conjugation of Oxazepam by Human UDP-Glucuronosyltransferases (UGTs): S-Oxazepam Is Glucuronidated by UGT2B15,
while R-Oxazepam Is Glucuronidated by UGT2B7 and UGT1A9. Drug Metab. Dispos. 2002, 30, 1257–1265. [CrossRef] [PubMed]
5. Patel, M.; Tang, B.K.; Grant, D.M.; Kalow, W. Interindividual variability in the glucuronidation of (S) oxazepam contrasted with
that of (R) oxazepam. Pharmacogenetics 1995, 5, 287–297. [CrossRef] [PubMed]
6. Uchaipichat, V.; Suthisisang, C.; Miners, J.O. The Glucuronidation of R- and S-Lorazepam: Human Liver Microsomal Kinetics,
UDP-Glucuronosyltransferase Enzyme Selectivity, and Inhibition by Drugs. Drug Metab. Dispos. 2013, 41, 1273–1284. [CrossRef]
7. Patel, M.; Tang, B.K.; Kalow, W. (S)Oxazepam glucuronidation is inhibited by Ketoprofen and other substrates of UGT2B7.
Pharmacogenetics 1995, 5, 43–49. [CrossRef]
8. Nguyen, L.A.; He, H.; Pham-Huy, C. Chiral drugs: An overview. Int. J. Biomed. Sci. 2006, 2, 85–100.
9. Hok, L.; Božičević, L.; Sremec, H.; Šakić, D.; Vrček, V. Racemization of oxazepam and chiral 1,4-benzodiazepines. DFT study of
the reaction mechanism in aqueous solution. Org. Biomol. Chem. 2019, 17, 1471–1479. [CrossRef]
10. Yang, S.K.; Lu, X.-L. Racemization Kinetics of Enantiomeric Oxazepams and Stereoselective Hydrolysis of Enantiomeric Oxazepam
3-Acetates in Rat Liver Microsomes and Brain Homogenate. J. Pharm. Sci. 1989, 78, 789–795. [CrossRef]
11. Maksay, G.; Tegyey, Z.; Kemeny, V.; Lukovits, I.; Otvos, L.; Palosi, E. Oxazepam esters. 1. Correlation between hydrolysis rates
and brain appearance of oxazepam. J. Med. Chem. 1979, 22, 1436–1443. [CrossRef]
12. Cassano, T.; Lopalco, A.; de Candia, M.; Laquintana, V.; Lopedota, A.; Cutrignelli, A.; Perrone, M.; Iacobazzi, R.M.; Bedse, G.;
Franco, M.; et al. Oxazepam–Dopamine Conjugates Increase Dopamine Delivery into Striatum of Intact Rats. Mol. Pharm. 2017,
14, 3178–3187. [CrossRef]
13. Bertucci, C.; Domenici, E.; Uccello-Barretta, G.; Salvadori, P. High-performance liquid chromatographic resolution of racemic
1,4-benzodiazepin-2-ones by means of a β-cyclodextrin silica bonded chiral stationary phase. J. Chromatogr. A 1990, 506, 617–625.
[CrossRef]
14. Szente, L.; Szemán, J. Cyclodextrins in Analytical Chemistry: Host–Guest Type Molecular Recognition. Anal. Chem. 2013, 85,
8024–8030. [CrossRef]
15. Kfoury, M.; Landy, D.; Fourmentin, S. Characterization of Cyclodextrin/Volatile Inclusion Complexes: A Review. Molecules 2018,
23, 1204. [CrossRef]
16. Morin-Crini, N.; Fourmentin, S.; Fenyvesi, E.; Lichtfouse, E.; Torri, G.; Fourmentin, M.; Crini, G. 130 years of cyclodextrin
discovery for health, food, agriculture, and the industry: A review. Environ. Chem. Lett. 2021, 19, 2581–2617. [CrossRef]
17. Lee, J.-U.; Lee, S.-S.; Lee, S.; Bin Oh, H. Noncovalent Complexes of Cyclodextrin with Small Organic Molecules: Applications and
Insights into Host–Guest Interactions in the Gas Phase and Condensed Phase. Molecules 2020, 25, 4048. [CrossRef]
18. Periasamy, R. Cyclodextrin-based molecules as hosts in the formation of supramolecular complexes and their practical ap-
plications—A review. J. Carbohydr. Chem. 2021, 135–155. [CrossRef]
19. Chankvetadze, B. Application of enantioselective separation techniques to bioanalysis of chiral drugs and their metabolites. TrAC
Trends Anal. Chem. 2021, 143, 116332. [CrossRef]
20. Alvarez-Rivera, G.; Bueno, M.; Vivas, D.B.; Cifuentes, A. Chiral analysis in food science. Trends Anal. Chem. 2019, 123, 115761.
[CrossRef]
21. Juvancz, Z.; Bodáné-Kendrovics, R.; Szente, L.; Maklári, D. Cyclodextrins as Dominant Chiral Selective Agents in the Capillary
Separation Techniques. Period. Polytech. Chem. Eng. 2021, 65, 580–594. [CrossRef]
22. Scriba, G.K. Chiral recognition in separation sciences: Part I: Polysaccharide and cyclodextrin selectors. Trends Anal. Chem. 2019,
120, 115639. [CrossRef]
23. Hancu, G.; Papp, L.; Tóth, G.; Kelemen, H. The Use of Dual Cyclodextrin Chiral Selector Systems in the Enantioseparation of
Pharmaceuticals by Capillary Electrophoresis: An Overview. Molecules 2021, 26, 2261. [CrossRef]
24. Guo, C.; Xiao, Y. Negatively charged cyclodextrins: Synthesis and applications in chiral analysis-A review. Carbohydr. Polym.
2020, 256, 117517. [CrossRef]
25. Gunjal, P.; Singh, S.K.; Kumar, R.; Kumar, R.; Gulati, M. Role of Chromatograph-based Analytical Techniques in Quantification of
Chiral Compounds: An Update. Curr. Anal. Chem. 2021, 17, 355–373. [CrossRef]
26. Wang, S.-Y.; Li, L.; Xiao, Y.; Wang, Y. Recent advances in cyclodextrins-based chiral-recognizing platforms. Trends Anal. Chem.
2019, 121, 115691. [CrossRef]
27. Wenzel, T.J. Differentiation of Chiral Compounds Using NMR Spectroscopy, 2nd ed.; John Wiley & Sons: Hoboken, NJ, USA, 2018;
ISBN 9781119324768.
28. Balzano, F.; Uccello-Barretta, G.; Aiello, F. Chiral Analysis by NMR Spectroscopy: Chiral Solvating Agents. In Chiral Analysis:
Advances in Spectroscopy, Chromatography and Emerging Methods; Polavarapu, P.L., Ed.; Elsevier: Amsterdam, The Netherlands,
2018; pp. 367–427, ISBN 9780444640277.
29. Chin, T.-F.; Chung, P.-H.; Lach, J.L. Influence of Cyclodextrins on Ester Hydrolysis. J. Pharm. Sci. 1968, 57, 44–48. [CrossRef]
30. Popielec, A.; Loftsson, T. Effects of cyclodextrins on the chemical stability of drugs. Int. J. Pharm. 2017, 531, 532–542. [CrossRef]
31. Trotta, F.; Moraglio, G.; Rapposelli, A. The effect of cyclodextrins on the hydrolysis of carboxylic acid esters. J. Incl. Phenom.
Macrocycl. Chem. 1994, 20, 353–361. [CrossRef]
Molecules 2021, 26, 6347 11 of 11

32. Cesari, A.; Piras, A.M.; Zambito, Y.; Barretta, G.U.; Balzano, F. 2-Methyl-β-cyclodextrin grafted ammonium chitosan: Synergistic
effects of cyclodextrin host and polymer backbone in the interaction with amphiphilic prednisolone phosphate salt as revealed by
NMR spectroscopy. Int. J. Pharm. 2020, 587, 119698. [CrossRef]
33. Homer, J.; Perry, M.C. Molecular complexes: Part 18—A nuclear magnetic resonance adaptation of the continuous variation (job)
method of stoicheiometry determination. J. Chem. Soc. Faraday Trans. 1 Phys. Chem. Condens. Phases 1986, 82, 533–543. [CrossRef]
34. Foster, R.; Fyfe, C.A. Interaction of electron acceptors with bases: Part 15—Determination of association constants of organic
charge-transfer complexes by n.m.r. spectroscopy. Trans. Faraday Soc. 1965, 61, 1626–1631. [CrossRef]