C. Baggiani et al. / Reactive & Functional Polymers 73 (2013) 833–837
835
(1H, t, CHCOOH); ClSA-Phe, 7: 11.9 (1H, s, COOH), 8.2 (1H, d,
CONH), 7.6 (1H, d, aromatic H), 7.4 (2H, t, aromatic H), 7.3 (2H,
d, aromatic H), 7.0 (1H, d, aromatic H), 5.2 (1H, s, aromatic OH),
4.9 (1H, t, CHCOOH); SA-Phe, 8: 11.9 (1H, s, COOH), 8.2 (1H, d,
CONH), 7.6 (1H, d, aromatic H), 7.6 (1H, t, aromatic H), 7.4 (2H, t,
aromatic H), 7.3 (2H, d, aromatic H), 7.2 (1H, t, aromatic H), 7.0
(1H, d, aromatic H), 5.2 (1H, s, aromatic OH), 4.9 (1H, t, CHCOOH).
structures. Annealing conditions were fixed as 300 K considering
the dynamic equilibrium reached after 2000 fs, with step of
0.1 fs. At the end of the annealing process, the position of metha-
crylic acid around the template was optimized again to reach a
minimum of potential energy.
3. Results and discussion
2.3. Synthesis of molecularly imprinted polymers
The successful mimic template CHNA-Phe was rationally de-
signed to preserve the general structure of the target analyte
OTA. In fact, a comparison between OTA and this mimic showed al-
most complete overlapping of the two molecules, with a high de-
gree of similarity not only structurally, but also as solvent
accessible surfaces, electrostatic potential surfaces and lipophilic/
hydrophilic surfaces [6], while a polymer prepared with the same
mimic, but with completely different functional monomers
showed the same recognition properties towards OTA [7], whereas
a polymer prepared with OTA as a template recognized the mimic
well [9].
Therefore, the rational design of novel mimics for OTA will meet
the same criteria used for Phe-CHNA, including the chirality of the
amino acidic sub-structure and the planarity of naphthalene rings.
Moreover, to assure an efficient imprinting effect, the several dis-
tinct potential points of interaction with functional monomers the
The imprinted polymers were prepared according to a method
previously reported in literature with minor modifications [6]. In
10 ml thick wall borosilicate test tubes, solutions with molar ratio
template:functional monomer:cross-linker 1:6:54 were prepared
by dissolving 0.27 mmol of OTA-mimic in 4.8 ml of anhydrous
chloroform. Then, 0.162 ml (1.62 mmol) of MAA, 2.85 ml of EDMA
(14.6 mmol) and 0.051 g of AIBN were added. The mixtures were
purged with nitrogen and sonicated in an ultrasonic bath for
10 min, sealed, and left to polymerize overnight at 60 °C. The bulk
polymers obtained were broken with a steel spatula, grounded in a
mechanical mortar and mechanically wet-sieved to 15–38 lm.
Then, the template was extracted by packing the polymers in poly-
propylene SPE columns and exhaustively washing with acetic
acid–methanol 1:9 (v/v) till no template was detectable by HPLC
analysis of eluate. The washed polymers were dried under vacuum
at 70 °C for 2 h and stored in a desiccator. Blank polymers (NIPs)
were prepared and treated in the same manner, omitting the
template.
a-carboxyl of L-phenylalanine, the amide bridge and the phenolic
hydroxyl should be retained. These criteria, though valid, seem to
be somewhat restrictive, because it is not possible to change most
of the molecular structure of the reference mimic without the risk
of losing OTA recognition in the resulting imprinted polymer. As a
consequence, we decided to study two different kinds of mimic
2.4. Liquid chromatography
molecules: OTA-analogs with amino acids other than L-phenylala-
Adequate amounts of polymer were suspended in a 1 + 1 v/v
ethanol–water mixture, and the slurry was packed in 3.9 ꢀ 100-
mm stainless-steel HPLC columns. The packing of the stationary
phases was performed by gradually adding the slurry of the poly-
mer to the column and eluting it with 1 + 1 (v/v) ethanol–water at
constant pressure of 20 MPa. The packed columns were washed at
1 ml/min with 9 + 1 (v/v) ethanol–acetic acid until a stable baseline
was reached (254 nm).
nine, obtained by truncation of the amino acid substructure (type-
1 mimics: CHNA-Ala, CHNA-Gly), and OTA-analogs obtained by
the subtraction of an aromatic ring or chlorine to CHNA substruc-
ture (type-2 mimics: HNA-Phe, CHNA-ClSA, CHNA-SA).
3.1. Type-1 mimic templates
When type-1 mimics are used to prepare the imprinted poly-
mers, the experimental results reported in Table 1 show that
OTA is better retained by the imprinted columns than the template
itself. This fact represents further evidence of what we had previ-
ously observed regarding a polymer imprinted with CHNA-Phe:
OTA was recognized better than the template and most of the ana-
logs examined, both in acetonitrile and in chloroform [6]. This fact
may be related to the different strength of nonspecific interactions
between the polymer and ligand. In fact, examining the capacity
factors measured on the non-imprinted column, it is possible to
observe that this polymer retains OTA much more than the mimics.
This confirms the generally accepted view that – to the net of non-
specific interactions – a polymer imprinted with a mimic tends to
bind this molecule better than the target. Interestingly, all the
polymers prepared with type-1 mimics recognize the OTA mole-
cule quite well. It should be considered that a decrease of the
Packed columns were equilibrated at a flow rate of 1 ml/min
with 40 ml of acetonitrile–acetic acid 0.1% (v/v); then, 5
ll of stock
solution of ligand diluted to 50 g/ml with the mobile phase were
l
injected and eluted at 1 ml/min recording absorbance at 344 nm.
Each elution was repeated three times to assure chromatogram
reproducibility. Column void volumes were measured for each mo-
bile-phase formulation by eluting 5 ll of acetone 0.1% (v/v) in ace-
tonitrile, and the absorbance was recorded at 343 nm (332 nm
with OTA, 304 nm with ClSA-Phe and SA-Phe). The capacity factor
(k) was calculated as (t–t0)/t0, where t is the retention time of the
eluted ligand and t0 is the retention time corresponding to the col-
umn void volume. The imprinting factor (IF) is defined as an index
of the imprinting efficacy respect to a NIP prepared in the same
conditions. It was calculated as kMIP/kNIP, where kMIP is the capacity
factor of a ligand eluted on the imprinted column and kNIP is the
capacity factor of a ligand eluted on the NIP.
2.5. Computer simulation
Table 1
capacity factors (k) and imprinting factors (IF) measured in imprinted and
not-imprinted polymers for ochratoxin A and type-1 mimic templates.
The computer used to simulate monomer–template interac-
tions was a Intel Core2 double CPU 3.00 GHz, 5 GB of RAM, running
a Windows 7 operative system with the molecular graphic soft-
ware HyperChem 8.08 (Hypercube Inc., Waterloo, Canada). Molec-
ular models of templates and methacrylic acid were separately
optimized by using a semiempirical quantum method (RM1) and
assembled together. Then, a simulated annealing process was ap-
plied to optimize the arrangement of the resulting supramolecular
NIP
MIP CHNA-Phe
MIP CHNA-Ala
MIP CHNA-Gly
k
k
IF
6.73
18.42
8.27
5.27
k
IF
8.00
17.62
12.93
6.35
k
IF
7.71
15.14
9.63
9.30
OTA
1.08
0.35
0.56
0.74
7.26
6.44
4.66
3.89
8.62
6.16
7.28
4.68
8.31
5.29
5.42
6.86
CHNA-Phe
CHNA-Ala
CHNA-Gly