Rational design of a redox-labeled chiral target for an enantioselective aptamer-based electrochemical binding assay

Article abstract of DOI:10.1002/chem.201302979

A series of redox-labeled L-tyrosinamide (L-Tym) derivatives was prepared and the nature of the functional group and the chain length of the spacer were systematically varied in a step-by-step affinity optimization process of the tracer for the L-Tym aptamer. The choice of the labeling position on L-Tym proved to be crucial for the molecular recognition event, which could be monitored by cyclic voltammetry and is based on the different diffusion rates of free and bound targets in solution. From this screening approach an efficient electroactive tracer emerged. Comparable dissociation constants K_d were obtained for the unlabeled and labeled targets in direct or competitive binding assays. The enantiomeric tracer was prepared and its enantioselective recognition by the corresponding anti-D-Tym aptamer was demonstrated. The access to both enantiomeric tracer molecules opens the door for the development of one-pot determination of the enantiomeric excess when using different labels with well-separated redox potentials for each enantiomer. Trace compounds: Redox tracers have been synthesized for enantioselective electrochemical ligand binding assays by relying on the combined use of an oligonucleotide-aptamer receptor with the detection of the redox label. A rational step-by-step optimization procedure has been developed leading to a redox-labeled L-tyrosinamide derivative (see figure) conserving the high affinity towards the aptamer.

Full text of DOI:10.1002/chem.201302979

DOI: 10.1002/chem.201302979
Full Paper
&
Aptamers
Rational Design of a Redox-Labeled Chiral Target for an
Enantioselective Aptamer-Based Electrochemical Binding Assay
Julie Moreau,[b] Lylian Challier,[b] Noꢀmie Lalaoui,[b] FranÅois Mavrꢀ,[a] Vincent Noꢁl,[b]
Benoꢂt Limoges,[a] Bernd Schçllhorn,*[a] and Claire Fave*[a, b]
Abstract: A series of redox-labeled l-tyrosinamide (l-Tym)
derivatives was prepared and the nature of the functional
group and the chain length of the spacer were systematical-
ly varied in a step-by-step affinity optimization process of
the tracer for the l-Tym aptamer. The choice of the labeling
position on l-Tym proved to be crucial for the molecular rec-
ognition event, which could be monitored by cyclic voltam-
metry and is based on the different diffusion rates of free
and bound targets in solution. From this screening approach
an efficient electroactive tracer emerged. Comparable disso-
ciation constants Kd were obtained for the unlabeled and la-
beled targets in direct or competitive binding assays. The
enantiomeric tracer was prepared and its enantioselective
recognition by the corresponding anti-d-Tym aptamer was
demonstrated. The access to both enantiomeric tracer mole-
cules opens the door for the development of one-pot deter-
mination of the enantiomeric excess when using different
labels with well-separated redox potentials for each
enantiomer.
Introduction
are numerous types of ligand binding assays and, though tech-
niques for direct monitoring of an interaction between a biore-
ceptor and an analyte have been increasingly proposed in
recent years (e.g., surface plasmon resonance,[8] capillary elec-
trophoresis,[9] or other related techniques), indirect methods
associated with the quantitative detection of an extrinsic tracer
(mostly a fluorescent label detected by an appropriate spectro-
scopic method) continue to be the focus of topical interest in
the development of new and improved formats of ligand bind-
ing assays.[10–12] This has been well illustrated in the field of
electrochemical biosensors, notably with the development of
electronic DNA- or aptamer-based sensors that involve
a redox-labeled oligonucleotide capture probe anchored onto
an electrode surface.[13,14] Associated with these developments,
different strategies of nucleic acid labeling by using diverse
redox probes have been proposed.[15,16] In spite of the many
practical advantages of electrochemical detection methods
compared with optical ones, including high sensitivity, low-
cost, robustness, field-portable capability, and ease of imple-
mentation, the principal inconvenience of DNA- or aptamer-
based electrochemical sensors is the need to immobilize the
redox-labeled capture probe on an electrode surface. An alter-
native to overcome this drawback, only scarcely reported so
far, is to electrochemically monitor on a bare electrode the cur-
rent response of a redox label engaged in a homogeneous
ligand binding reaction. The few examples in the literature
have been exclusively demonstrated for competitive homoge-
neous binding immunoassays. Only recently a homogeneous
enantiospecific electrochemical ligand binding assay that relied
on the combined use of an oligonucleotide–aptamer receptor
with the electrochemical detection of a redox label has been
proposed.[17] In this method, the principle of detection recalled
Ligand binding assays take advantage of the selectivity of
a biomolecular interaction (i.e., between a ligand and a recep-
tor) to identify and quantify the presence of a molecule in
a complex matrix.[1] They have found widespread applications
in many fields such as clinical chemistry, drug development,
food analysis, or environmental chemistry.[2–4] In some of these
applications, the detection of chiral molecules might be of cen-
tral importance. This is, for example, the case in the food in-
dustry for assaying food quality through the enantiomeric
analysis of amino acids,[5] in the pharmaceutical industry by de-
termination of enantiomeric excess, in forensic science for de-
termining and quantifying illicit chiral substances,[3,6] or in envi-
ronmental science for monitoring the fate of emerging chiral
pollutants.[7] The development of rapid, simple, sensitive, and
highly selective analytical methods for identifying and quanti-
fying trace amounts of one enantiomer in a complex sample
that contains its mirror image and many other unrelated mole-
cules remains an important goal in analytical sciences. There
[a] Dr. F. Mavrꢀ, Dr. B. Limoges, Prof. B. Schçllhorn, Dr. C. Fave
Laboratoire d’Electrochimie Molꢀculaire, UMR 7591 CNRS
Universitꢀ Paris Diderot, Sorbonne Paris Citꢀ
15 rue Jean-Antoine de Ba¨ıf, 75205 Paris Cedex 13 (France)
Fax: (+33)1-57-27-87-88
E-mail: bernd.schollhorn@univ-paris-diderot.fr
claire.fave@univ-paris-diderot.fr
[b] Dr. J. Moreau, L. Challier, N. Lalaoui, Dr. V. Noꢁl, Dr. C. Fave
ITODYS, UMR 7086 CNRS
Universitꢀ Paris Diderot, Sorbonne Paris Citꢀ
15 rue Jean-Antoine de Ba¨ıf, 75205 Paris Cedex 13 (France)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201302979.
Chem. Eur. J. 2014, 20, 2953 – 2959
2953
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
methylviologen (MV2+) and ferrocene (Fc). Depending on the
electron-donating or -withdrawing substituents, the standard
potentials of methylviologen[18] (i.e., E0 of the first reversible re-
duction of MV2+ into MV +) and ferrocene (i.e., E0 of the one-
C
electron reversible oxidation of FeII to FeIII)[19] can be tuned
from À0.5 to À0.7 V and from À0.2 to +0.5 V (versus Ag/
AgCl), respectively. Another interesting property of methyviolo-
gen is its ability to efficiently electrocatalyze the reduction of
dioxygen,[20] thus opening a way to amplify the sensitivity of
detection of this redox label.
Because the exact structure of the Tym/aptamer complex
has not yet been resolved, the effect of analyte labeling on the
aptamer-binding recognition could not be anticipated. There-
fore a screening approach based on a systematic change of
the linking position as well as the spacer-chain length between
the analyte and the redox label has been undertaken. Among
various possible synthetic strategies, we favored the direct
chemical labeling of the commercially available enantiomeri-
cally pure l-Tym. Such a straightforward approach aims to pro-
vide rapid and modular access to the corresponding redox-la-
beled molecules. Moreover, it could be easily extended to any
other amino acid or analogue target using the same redox
labels.
Figure 1. Principle of the aptamer-based competitive binding assay with
electrochemical detection.
in Figure 1 basically depends on the difference of diffusion
rates between the target molecule and the aptamer/target
complex, and thus on the ability to detect the former more
easily than the latter in homogeneous solution. The proof-of-
concept of this electrochemical method was designed especial-
ly for the enantioselective detection of trace amounts of the
small chiral model analyte l- or d-tyrosinamide (l- or d-Tym)
by a single-stranded 49-mer d- or l-DNA receptor, respectively.
On the basis of a homogeneous competitive binding assay
strategy and of the electrochemical signal that resulted from
the preferential detection of the free redox-labeled l-Tym, as
little as 0.1% of the minor enantiomer in a non-racemic mix-
ture could be achieved (equivalent to the determination of
a 99.8% enantiomeric excess) in less than one minute.[17] Until
now, the method has been validated for the homogeneous
competitive binding assay of a single enantiomeric form of the
analyte, but it is conceivable that the method could be extend-
ed to a simultaneous determination of the two enantiomeric
forms in the same sample by employing two distinct redox
probes, with each labeling one of the corresponding enantio-
mer analytes.
Although three likely labeling sites (Figure 2a, b, and c) can
be easily identified in l-Tym, only positions a and b were inves-
tigated in the present work. In fact, modification of position c
Figure 2. Redox labeling sites on l-Tym.
To move in this direction, a significant effort of synthesis and
molecular engineering was required. Herein we describe the
rational design and the systematic optimization of the tracer
molecule properties.
was previously shown by us[17] to completely impede the spe-
cific aptamer-binding recognition, thereby suggesting the
amide function to be an indispensable motif for molecular rec-
ognition. Labeling through the amino group of l-Tym was the
most natural since it was previously used in the systematic
evolution of ligands by exponential enrichment (SELEX) to
afford the final l-Tym aptamer (l-Tym was coupled to the
SELEX solid phase through this function).[21] The primary amine
was transformed into an amide linkage by means of a peptidic
coupling or, alternatively, to a secondary amine link by means
of reductive amination. Labeling of position b by means of
a Mitsunobu coupling was used to afford ether derivatives,
thereby suppressing the inherent electroactivity of Tym related
to the oxidizable phenolic group.
For successful achievement of a homogeneous competitive
binding assay, it is important that the labeling does not signifi-
cantly affect the aptamer binding. Ideally, the redox-labeled
analyte should have the same aptamer affinity binding as the
unlabelled analyte and its size and solubility should not be sig-
nificantly altered. The design of the labeled Tym derivative was
thus guided by the following considerations. First, redox labels
that showed fully reversible electron exchange with common
electrodes in buffered aqueous solution were preferentially
chosen because they are more amenable to a sensitive and re-
producible detection. Clearly, the choice of chemically stable
redox labels with an accessible electroactivity within the po-
tential window of aqueous solutions is another important crite-
rion. Because of their distinct standard potential (E0) as well as
of their high stability and fast reversible one-electron transfer
process at an electrode, we selected as redox labels dicationic
Each redox-labeled l-Tym compound was then systematical-
ly tested for its enantiospecific recognition and affinity binding
to the aptamer through a simple electrochemical titration
assay, thereby providing immediate feedback on each pre-
pared tracer as well as an accurate picture of the binding.
Chem. Eur. J. 2014, 20, 2953 – 2959
www.chemeurj.org
2954
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Once the best tracers were identified, their enantiomers were
synthesized, thus opening a route to the development of an
electrochemical competitive binding assay for the determina-
tion of the enantiomeric excess (ee) of l/d-Tym mixtures in
one pot.
Scheme 2. Synthesis of viologen l-Tym derivatives l-1 to l-3: i) Br-
(CH2)nCO2H, TBTU, DIPEA, EtOAc, 50–85%; ii) N-methyl-4-pyridylpyridinium
iodide salt, CH3CN, 808C, 48 h, 36–95%.
Results and Discussion
The rational step-by-step procedure that led to the design and
synthesis of tyrosinamide derivatives is detailed below. The O-
ferrocenyl l-Tym tracer l-16 (Scheme 5 below) was previously
shown to be well recognized by the aptamer.[17] Its dissociation
constant (Kd =(1.2Æ0.4) mm) was even found to be slightly
better than that of the unlabeled l-Tym (Kd =(3.8Æ0.7) mm),
thus demonstrating that the labeling of l-Tym through the
active hydroxyl group does not impede the aptamer-binding
recognition.[17] This result also indicates that the hydrogen
atom on the phenolic function does not significantly partici-
pate in the aptamer molecular recognition.
Scheme 3. Synthesis of ferrocenyl-labeled l-Tym derivatives l-4 to l-6:
i) FcCHO or C3H7CHO, NEt3, MeOH; ii) NaBH4, 48C, 40–85%; iii) FcCOOH, TBTU,
DIPEA, EtOAc, 60%.
Amine labeling
Synthesis
termediate imine with sodium borohydride to afford the sec-
ondary amine l-4 in 40% yield (Scheme 3). Compound l-6 was
prepared in a similar way to assess the influence of the steric
hindrance on the amino position on the binding affinity to the
aptamer. Derivative l-5 was obtained in 60% yield by means
of peptide coupling that involved the ferrocene carboxylic acid
and both diisopropylethylamine (DIPEA) and TBTU.
Labeling of the l-Tym on the primary amine group afforded
the water-soluble compounds l-1 to l-5 represented in
Scheme 1. For the three dicationic derivatives l-1 to l-3, the
Electrochemical properties
Electrochemical experiments were carried out by using cyclic
voltammetry (CV) in a deaerated tris(hydroxymethyl)aminome-
thane (Tris) buffer (pH 7.4) and using carbon-based screen-
printed electrodes. For compounds l-1 to l-3, two monoelec-
tronic reversible waves were systematically observed in the
cyclic voltammograms (Figure 3A, Table 1), which were charac-
Scheme 1. N-Redox-labeled l-Tym targets.
redox-active methylviologen was separated from l-Tym by dif-
ferent chain lengths (one to ten methylene groups) in such
a way as to anticipate the possible influence of positive charg-
es and/or steric hindrance brought about by the pyridinium
groups on molecular recognition. In the case of the ferrocenyl
label, only the nature of the linking group was varied through
the formation of a secondary amine (l-4) or a peptide link (l-5)
(Scheme 1).
teristic of the two successive one-electron reductions of the di-
+
cation MV2+ to a monocation MV and then to a neutral com-
C
The synthesis of compounds l-1 to l-3 (Scheme 2) involved
the formation of a bromide intermediate (l-7 to l-9). The re-
spective bromides were obtained in fairly good yields (50–
85%) by the action of the corresponding bromocarboxylic acid
derivative on l-Tym in the presence of the coupling reagent
N,N,N’,N’-tetramethyl-o-(benzotriazol-1-yl)uranium tetrafluoro-
borate (TBTU). Substitution of the bromides with 1-methyl-4-
pyridylpyridinium iodide in acetonitrile led to the formation of
the corresponding MV2+-labeled Tym l-1 to l-3. The pure
product could be easily isolated by precipitation in a mixture
of acetonitrile and diethyl ether.
Figure 3. A) Cyclic voltammograms of l-1 (solid line), l-2 (dotted), and l-3
(dashed). B) Cyclic voltammograms of l-4 (dotted) and l-5 (solid). The ex-
periments were achieved in l-Tym (50 mm) derivative in Tris buffer solution
(pH 7, v=0.1 VsÀ1).
Compound l-4 was obtained by the reaction of l-Tym with
ferrocenyl carboxaldehyde and subsequent reduction of the in-
Chem. Eur. J. 2014, 20, 2953 – 2959
www.chemeurj.org
2955
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 1. Redox potentials (versus Ag/AgCl) and dissociation constants of
the redox-labeled Tym derivatives.[a]
Compound l-1
l-2
l-3
l-4
l-5
l-15 l-16
l-17
E01 [V]
À0.55 À0.62 À0.62 +0.15 +0.37 À0.57 +0.17 +0.02
E02 [V]
À0.90 À0.97 À0.88
–
80
–
–
58
–
À0.89
60
–
62
–
–
64
–
DEp1 [mV]
DEp2 [mV]
Kd [mm]
52
60
56
65
43
82
68
n.r.
n.r.
n.r.
n.r.
n.r.
0.6
1.2
n.r.
[a] DEp =peak potential difference. n.r.=no recognition.
pound at approximately À0.6 and À1.0 V (versus Ag/AgCl), re-
spectively.
Figure 4. Normalized catalytic peak current as a function of aptamer concen-
&
*
tration for the following ligand/receptor couples: ( ) l-Tym/d-Apt49, ( ) l-6/
~
d-Apt49, and ( ) l-Tym/49-mer scrambled oligonucleotide. Each data point
The increasing alkyl chain length between the viologen
entity and l-Tym induces only weak anodic shifts of the stan-
dard potentials for both reversible waves (Figure 3A, Table 1).
Cyclic voltammograms of the ferrocene-labeled compounds l-
4 and l-5 (Figure 3B, Table 1) show both a single monoelec-
tronic reversible oxidation wave located at +0.15 and +0.37 V
(versus Ag/AgCl), respectively, followed by an ill-defined irre-
versible anodic wave at approximately +0.7 V. The standard
potential difference of the ferrocenyl groups is in agreement
with the electron-withdrawing effect of the differently substi-
tuted cyclopentadienyl groups. The carbonyl substituent in l-5
is a stronger electron-withdrawing group than the methylene
substituent in l-4, thereby shifting the E0 of ferrocene to
a higher value. The ill-defined irreversible anodic peak ob-
served for each molecule at approximately +0.7 V is typical
of the irreversible oxidation of the phenol group (see the cyclic
voltammogram of l-Tym in Figure S1 of the Supporting
Information).[22]
represents the average of three measurements. The indicated error bars are
standard deviations. (Data are extracted from CV of Tris buffer (5 mm, pH 7)
solutions containing [Os(bpy)3]2+ (1 mm), l-Tym, or l-6 (5 mm), NaCl (50 mm),
MgCl2 (10 mm), and increasing concentrations of Apt49.)
served in contrast to the well-defined asymptotic signal de-
crease obtained with l-Tym, thus demonstrating the absence
of l-6 recognition by the aptamer and confirming the critical
role of the free amino group in the recognition of labeled de-
rivatives by the aptamer (the small signal diminution observed
with l-6 in Figure 4 is attributed to a nonspecific response de-
crease since it is similar to the one obtained with l-Tym in the
presence of an increasing amount of a 49-mer random DNA
sequence).[17]
Phenol labeling
Synthesis of the l-Tym tracers
Aptamer-binding recognition tests
To corroborate our previous results obtained with l-16,[17] la-
beling of the l-Tym phenolic position (Figure 2) was explored
with different redox probes, leaving the primary amino group
unchanged. The viologen-labeled derivative l-15 was obtained
according to a five-step synthesis. The strongly nucleophilic
amino group first had to be protected. The phenolic group of
the N-tert-butoxycarbonyl (Boc)-protected derivative l-11 was
then treated with 1-bromo-3-chloropropane in the presence of
potassium carbonate to lead to compound l-12 in 65% yield.
Substitution of chloride l-12 by 4,4’-bipyridine followed by N-
methylation afforded the viologen derivative l-14 as an iodide
salt. Removal of the Boc protecting group was obtained in
a mixture of trifluoroacetic acid and dichloromethane, thus
leading to l-15 as a trifluoroacetate salt (Scheme 4).
The analogue ferrocenyl derivative l-16 (Scheme 5) was ob-
tained by means of a Mitsunobu reaction of l-Tym and ferro-
cene methanol using the conditions of a previously published
procedure.[23] Because of its low oxidation potential (close to
0 V versus Ag/AgCl), pentamethylferrocene is an interesting al-
ternative redox label as well. The synthesis of pentamethylfer-
rocene methanol was performed according to an adaptation
of a previously described protocol.[24] Its coupling to l-Tym was
performed under the same conditions as described above.
To evaluate whether the redox-labeled l-Tym derivatives were
recognized by the 49-mer anti-l-Tym, binding recognition tests
were carried out. For such purposes, the magnitude of the
anodic or cathodic CV peak current of a redox-labeled l-Tym
compound (5 mm in solution) was monitored as a function of
increasing aptamer concentration. The CV peak current of the
redox label was expected to decrease when the redox-labeled
target was recognized by the aptamer. Unfortunately, for all
tracer compounds l-1 to l-5, a stable signal was observed,
even for large excess amounts of aptamer, thus indicating
a complete loss of the specific binding recognition. This result
suggests the crucial role of the primary amino group in the
recognition process. To confirm this hypothesis and also to
verify if it might not be the result of a steric effect on the mo-
lecular recognition owing to the presence of the bulky ferro-
cenyl group, compound l-6 was tested. To suitably electro-
chemically detect l-6 in the presence of d-Apt49, we used the
redox-mediated catalytic oxidation of the phenolic group of l-
Tym (this was achieved by adding a minute amount of the
redox mediator [OsII(bpy)3]2+ (bpy=2,2’-bipyridyl) to the bind-
ing solution as previously described).[17] However, as shown in
Figure 4, no significant current response decrease could be ob-
Chem. Eur. J. 2014, 20, 2953 – 2959
www.chemeurj.org
2956
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
d-16 were the same as their
enantiomeric forms l-15 and l-
16 (data not shown).
Aptamer-binding recognition
tests
In contrast to compound l-16,
for which a good affinity binding
to d-Apt49 was obtained (Kd =
1.2 mm),[17] the aptamer binding
test with l-17 was unsuccessful,
Scheme 4. Synthesis of the viologen l-Tym derivative l-15: i) (Boc)2O, dioxane, NaOH, 58%; ii) Br(CH2)3Cl, K2CO3,
DMF, 65%; iii) 4,4’-bipyridine, CH3CN, KI, 58%; iv) MeI, CH3CN, 25%; v) TFA/CH2Cl2 (50:50), 95%. TFA=trifluoroace-
tic acid.
thus suggesting complete loss of recognition (Table 1). Pres-
ently there is no explanation for this result, but we might pos-
tulate that the high hydrophobicity of the pentamethylferro-
cenyl label as well as its bulky size should interfere in the rec-
ognition process. The viologen derivative l-15 was also tested
for its binding to the d-Apt49, but instead of CV we preferred
to use square-wave voltammetry (SWV). This technique has the
advantage of being more sensitive than CV and providing
better-defined and more reproducible current responses when
working with micromolar concentrations of MV2+. The magni-
tude of the SWV peak currents of 2 mm l-15 (the second reduc-
Scheme 5. Ferrocenyl-labeled derivatives l-16 and l-17.
Compound l-17 was obtained in a modest yield (13%), proba-
bly due to its weak electrophilicity and fair solubility.
tion peak current that corresponded to the MV +/MV0 couple
C
was used as the analytical response) was monitored as a func-
tion of the aptamer concentration. The decrease in the two
peaks upon addition of d-Apt49 clearly shows that the labeled
target l-15 is well recognized by the aptamer. From the data
fitting, a value of Kd =(0.6Æ0.2) mm was inferred. This value is
twice as low as the analogous ferrocenyl-labeled compound l-
16 and approximately six times better than the unlabeled l-
Tym (3.8 mm).
These results show that labeling of l-Tym through the phe-
nolic position, even if not without its limitations, is the most
suitable way for preserving the specific binding to the
aptamer.
Electrochemical properties
CV of a deaerated solution of l-15 shows (Table 1), as expect-
ed, the presence of two reversible waves at À0.57 and À0.89 V
(versus Ag/AgCl), which are characteristic of the presence of vi-
ologen on the molecule (data not shown). These E0 values
were not significantly different from those previously obtained
for l-1 to l-3. Oxidation of the ferrocenyl- and pentamethylfer-
rocenyl-labeled compounds l-16 and l-17 occurred at +0.17
and +0.02 V, respectively (Figure 5, Table 1). These results are
With all these results in hand, we next performed an electro-
chemical competitive binding assay to confirm the Kd value ob-
tained for l-15 with the previous approach. In these experi-
ments, the competitive assay was achieved from a starting so-
lution that contained 5 mm l-16 and 10 mm d-Apt49, and an in-
creasing amount of l-15 (Figure 6A) or l-2 (Figure 6B) analytes.
The decrease in the anodic and cathodic peak currents upon
the addition of d-Apt49 was anticipated due to the previously
well-established enantiospecific recognition of l-16 by the ap-
tamers (black to gray in Figure 6A and B). Adding increasing
amounts of l-15 to this solution gave a steady increase in the
reversible voltammetric peak currents of ferrocene (Figure 6A),
whereas an increasing addition of l-2 did not change anything
(Figure 6B). The resulting competitive calibration plots are
shown in Figure 7C. These results definitely demonstrate the
absence of recognition between l-2 and d-Apt49. From the fit
of the experimental data obtained with l-15 to the theoretical
equation used for a competitive binding,[17] a value of Kd =
(0.4Æ0.1) mm was inferred, which is in very good agreement
with the value determined directly (Kd =(0.6Æ0.2) mm) from
the titration plot in Figure 6B.
Figure 5. Cyclic voltammograms of l-16 (dotted line) and l-17 (solid) at
5 mm in Tris buffer (pH 7, v=0.05 VsÀ1).
consistent with the oxidation potentials E0 = +0.25 and
À0.05 V obtained for the corresponding precursor compounds,
that is, ferrocene methanol and pentamethylferrocene metha-
nol. The lower E0 value of the latter results from the strong
electron-donating effect of the pentamethylcyclopentadienyl
ligand. The electrochemical properties of compounds d-15 and
Chem. Eur. J. 2014, 20, 2953 – 2959
www.chemeurj.org
2957
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
(1 gmLÀ1, MeOH). Compound d-15 was obtained from d-11
with a global yield of 54% by following the same procedure as
described for the synthesis of l-15.
The synthesis of the ferrocenyl tracer d-16 was achieved by
nucleophilic substitution of a ferrocenyl trimethylammonium
salt with the commercial d-Tym hydrochloride in DMSO in the
presence of a 10% aqueous NaOH solution (2 equiv).[25] DMSO
was used to enhance the base-catalyzed reactions. A similar
procedure had been previously reported by Solar and Schu-
maker for the selective O-alkylation of tyrosine derivatives.[26]
After 2 h of heating at 808C, only 20% conversion was re-
1
vealed by H NMR spectroscopy, and longer reaction times re-
sulted in complete decomposition of the product. Liquid chro-
matography of the crude reaction mixture afforded d-16 in
20% yield. The specific rotation values reported in Table 2 con-
firm the presence of the enantiomeric tracer molecules l-/d-15
and l-/d-16.
Table 2. Specific rotation of the redox-labeled Tym compounds in metha-
nol.
Compound
[a]2D5 [8mLcmÀ3 gÀ1)]
C [gmLÀ1
]
Figure 6. A) Cyclic voltammograms (v=0.05 VsÀ1) in Tris buffer (5 mm,
pH 7.4, 50 mm NaCl, 10 mmMg2+) of l-16 (5 mm; black curve), l-16 (5 mm),
and d-Apt49 (10 mm) (a) and an increasing amount of l-15: 5, 10, 15,
40 mm (gray). B) Same as in (A) but for an increasing amount of l-2. C) Com-
l-15
d-15
l-16
d-16
+1.84
À3.56
+2.92
À3.10
0.40
0.34
0.33
0.31
~
*
petitive binding curves for l-15 ( ) and l-2 ( ). Dashed line: Fit to the theo-
retical equation of a competitive binding.
The aptamer-binding recognition tests were next performed
for each ferrocene-labeled enantiomer of l-Tym. As expected,
d-16 was not recognized by d-Apt49 but rather only by the
mirror image aptamer l-Apt49 (Figure 8). A Kd value of 1.4 mm
was obtained for the latter, which is nearly the same value as
for the l-enantiomeric ligand/receptor couple.
Figure 7. A) Square-wave voltammograms (f=25 Hz, DE=20 mV,
E
step =5 mV) of l-15 (2 mm) in Tris buffer (5 mm, pH 7.4) solution containing
NaCl (50 mm), Mg2+ (10 mm), and a) 0, b) 1, c) 2, d) 3, and e) 7.5 mm d-Apt49.
B) Normalized peak (at À1.0 V) current response as a function of aptamer
concentration. Each data point is the average of three measurements. Error
bars are standard deviations.
Synthesis of the d-Tym tracers
Figure 8. Normalized catalytic peak current as a function of aptamer concen-
~
*
( ), d-16/l-Apt49 ( ), l-
tration for the ligand/receptor couples: d-16/d-Apt49
Given the proper binding recognition of l-15 and l-16 by d-
Apt49, the mirror image tracers d-15 and d-16 were prepared.
The synthesis of the viologen-labeled d-Tym d-15 was ob-
tained from the commercially available starting compound
d-tyrosine. Measurement of the optical rotation of the Boc-
&
16/d-Apt49 ( ).
Conclusion
protected derivative d-11 gave
a
value of [a]D25 =
A series of redox-labeled l-tyrosinamide derivatives was pre-
pared. The nature of the functional group and the chain
length of the spacer were systematically varied in a step-by-
À20.18mLcmÀ3 gÀ1 (1 gmLÀ1, MeOH), which is in good agree-
ment with its enantiomer l-11 [a]D25 = +22.5 8mLcmÀ3 g
Chem. Eur. J. 2014, 20, 2953 – 2959
www.chemeurj.org
2958
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[1] M. N. Khan, J. W. Findlay, Ligand-Binding Assays: Development, Validation
and Implementation in the Drug Development Arena, Wiley, Hoboken,
2009.
[2] V. Garcꢄa-CaÇas, C. Simo, M. Herreo, E. Ibanez, A. Cifuentes, Anal. Chem.
2012, 84, 10150–10159.
[3] J. T. Liu, R. H. Liu, J. Biochem. Biophys. Methods 2002, 54, 115–146.
[4] E. Eljarrat, P. Guerra, D. Barcelo, TrAC Trends Anal. Chem. 2008, 27, 847–
861.
[5] M. Herrero, C. Simo, V. Garcia-Canas, S. Fanali, A. Cifuentes, Electrophore-
sis 2010, 31, 2106–2114.
[6] F. P. Smith in Handbook of Forensic Drug Analysis (Ed.: J. A. Siegel), Elsev-
ier, Amsterdam (The Netherlands), 2005.
[7] J. R. Kucklick, P. A. Helm, Anal. Bioanal. Chem. 2006, 386, 819–836.
[8] J. Homola, Anal. Bioanal. Chem. 2003, 377, 528–539.
[9] V. Poinsot, M. A. Carpꢀnꢀ, J. Bouajila, P. Gavard, B. Feurer, F. Couderc,
Electrophoresis 2012, 33, 14–35.
[10] D. S. Smith, S. A. Eremin, Anal. Bioanal. Chem. 2008, 391, 1499–1507.
[11] J. Liu, Z. Cao, Y. Lu, Chem. Rev. 2009, 109, 1948–1998.
[12] H. Zhang, F. Li, B. Dever, X.-F. Li„ X. C. Le, Chem. Rev. 2013, 113, 2812–
2841.
[13] A. A. Lubin, K. W. Plaxco, Acc. Chem. Res. 2010, 43, 496–505.
[14] J.-O. Lee, H.-M. So, E.-K. Jeon, H. Chang, K. Won, Y. H. Kim, Anal. Bioanal.
Chem. 2008, 390, 1023–1032.
[15] M. Hocek, M. Fojta, Chem. Soc. Rev. 2011, 40, 5802–5814.
[16] J. Moreau, N. Dendane, B. Schçllhorn, N. Spinelli, C. Fave, E. Defrancq,
Bioorg. Med. Chem. Lett. 2013, 23, 955–958.
[17] L. Challier, F. Mavrꢀ, J. Moreau, C. Fave, B. Schçllhorn, D. Marchal, E.
Peyrin, V. Noe¨l, B. Limoges, Anal. Chem. 2012, 84, 5415–5420.
[18] C. P. Andrieux, P. Hapiot, J. M. Savꢀant, J. Electroanal. Chem. 1985, 189,
121–133.
[19] S. Lu, V. V. Strelets, M. F. Ryan, W. J. Pietro, A. B. P. Lever, Inorg. Chem.
1996, 35, 1013–1023.
[20] J. M. Savꢀant, Chem. Rev. 2008, 108, 2348–2378.
[21] E. Vianini, M. Palumbo, B. Gatto, Bioorg. Med. Chem. 2001, 9, 2543–
2548.
step optimization process of the tracer for its anti-l-Tym apta-
mer. The homogeneous binding recognition events were effi-
ciently monitored by cyclic voltammetry by taking advantage
of the difference in diffusion rates between the free and
bound targets in solution. The results demonstrate the crucial
role of the primary amino group of tyrosinamide in the apta-
mer binding recognition process. In contrast, alkylation of the
phenolic group did not significantly influence the affinity prop-
erties. Comparable dissociation constants Kd were finally ob-
tained for the unlabeled and labeled targets in a direct or com-
petitive binding assay. Furthermore, the redox-labeled enantio-
meric form of l-Tym was prepared and its enantioselective rec-
ognition by the corresponding mirror image aptamer was
demonstrated. The possibility afforded by the labeling of the
two enantiomeric forms of the tyrosinamide target by two dis-
tinct redox labels (i.e., with well-separated redox potentials)
opens the opportunity to develop an enantiomeric competi-
tive binding assay for the determination of the enantiomeric
excess in a single analysis. Beyond this particular example of
an enantiospecific competitive binding assay, the present elec-
trochemical detection method is of more general interest since
it can be relatively easily extended to a wide range of other
small target molecules able to specifically bind to an aptamer.
Moreover, combined with multiple redox labeling, it would be
promising for the development of a multiple analysis in the
same sample.
Acknowledgements
[22] C. Costentin, C. Louault, M. Robert, J. M. Savꢀant, Proc. Natl. Acad. Sci.
USA 2009, 106, 18143–18148.
This work was supported by Agence Nationale pour la Recher-
che (ANR project ECSTASE). Noꢀmie Lalaoui thanks the ITODYS
Laboratory for a grant. We thank David Clainquart (Chemistry
Department, Universitꢀ Paris Diderot) for his help with GC-MS
measurements.
[23] C. Baldoli, L. Falciola, E. Licandro, S. Maiorana, P. Mussini, P. Ramani, C.
Rigamonti, G. Zinzalla, J. Organomet. Chem. 2004, 689, 4791–4802.
[24] F. M. Geisler, G. Helmchen, Synthesis 2006, 13, 2201–2205.
[25] S. L. Solar, R. R. Schumaker, J. Org. Chem. 1966, 31, 1996–1997.
[26] G. Tang, X. Tang, M. Wang, L. Luo, M. Gan, Z. Huang, J. Labelled Compd.
Radiopharm. 2003, 46, 661–668.
Keywords: aptamers
·
chirality
·
electrochemistry
·
Received: July 29, 2013
enantioselectivity · redox labeling
Published online on February 12, 2014
Chem. Eur. J. 2014, 20, 2953 – 2959
www.chemeurj.org
2959
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim