A Substructure Approach toward Polymeric Receptors Targeting Dihydrofolate Reductase Inhibitors. 2. Molecularly Imprinted Polymers against Z-L-Glutamic Acid Showing Affinity for Larger Molecules

Article abstract of DOI:10.1021/jo034588e

The preparation of a molecularly imprinted polymer against N-Z-L-glutamic acid using a novel bis-urea functional monomer is described. The polymer exhibits affinity for the template over N-Z-protected aspartic acid and glycine and, further, is capable of binding larger molecules, e.g., the anti-cancer drug methotrexate, containing the glutamate substructure.

Full text of DOI:10.1021/jo034588e

A Su bstr u ctu r e Ap p r oa ch tow a r d
P olym er ic Recep tor s Ta r getin g
Dih yd r ofola te Red u cta se In h ibitor s. 2.
Molecu la r ly Im p r in ted P olym er s a ga in st
Z-L-Glu ta m ic Acid Sh ow in g Affin ity for
La r ger Molecu les
F IGURE 1. Highly schematic representation of the strategy
of molecular imprinting.
Andrew J . Hall,*,†,‡ Laura Achilli,§
Panagiotis Manesiotis,†,‡ Milena Quaglia,§,|
Ersilia De Lorenzi,§ and Bo¨rje Sellergren†,‡
monomers are employed to ensure that the template
molecule is completely complexed. This, in turn, means
that nonassociated functional monomer is incorporated
into the polymer matrix, leading to high levels of non-
specific binding.
The use of designed functional monomers in molecular
imprinting is not common, although there have been
notable exceptions.3-7 Herein, we wish to report the
synthesis of the novel, functional monomer 1 and its
application in the imprinting of N-Z-L-glutamic acid (2).
Further, as a continuation of an earlier study on sub-
structure approaches to the imprinting of dihydrofolate
reductase inhibitors,8 we wish to report the recognition
properties of the polymers toward larger molecules
containing the glutamate substructure, e.g., the anti-
cancer drug methotrexate (3).
Institut fu¨r Anorganische Chemie und Analytische Chemie,
J ohannes Gutenberg Universita¨t, Duesbergweg 10-14,
D-55099, Mainz, Germany, and Department of
Pharmaceutical Chemistry, University of Pavia, Via
Taramelli 12, 27100 Pavia, Italy
andy@infu.uni-dortmund.de
Received May 6, 2003
Abstr a ct: The preparation of a molecularly imprinted
polymer against N-Z-L-glutamic acid using a novel bis-urea
functional monomer is described. The polymer exhibits
affinity for the template over N-Z-protected aspartic acid and
glycine and, further, is capable of binding larger molecules,
e.g., the anti-cancer drug methotrexate, containing the
glutamate substructure.
The use of molecular imprinting, as a method to
prepare materials showing strong and selective affinity
for the imprinted (templated) structure and closely
related analogues, has received growing interest in recent
years.1 The most flexible approach to imprinting is the
“noncovalent” approach,2 depicted schematically in Fig-
ure 1, where template-functional monomer assemblies,
formed by noncovalent interactions, are “locked” into a
three-dimensional network on co-polymerization with an
excess of cross-linking monomer. Extraction of the tem-
plate from the polymer matrix leaves behind sites which
are complementary in shape and functional to the
template (and analogues).
This approach is very attractive due to its inherent
simplicity, but problems remain, notably with regards to
the imprinting of templates which are not soluble in the
low-polar organic solvents generally used in the imprint-
ing step, e.g., chloroform and toluene. A further problem
is that the use of commercially available functional
monomers generally requires that a large excess of said
Our design for the novel functional monomer reported
herein is based on a report by the group of Hamilton,9
where a suitably spaced bis-urea was found to exhibit
reasonable affinity for glutaric acid (as its bis-tetrabu-
† J ohannes Gutenberg Universita¨t.
‡ Current address: Institut fu¨r Umweltforschung, Universita¨t Dort-
mund, D-44221 Dortmund, Germany. Tel: 0049 231 755 4082. Fax:
0049 231 755 4084.
§ University of Pavia.
(3) Tanabe, K.; Takeuchi, T.; Matsui, J .; Ikebukuro, K.; Yano, K.;
Karube, I. J . Chem. Soc., Chem. Commun. 1995, 2003-2304.
(4) Wulff, G.; Gross, T.; Scho¨nfeld, R. Angew. Chem., Int. Ed. Engl.
1997, 36, 1962-1964.
| Current address: Institut fu¨r Anorganische Chemie und Ana-
lytische Chemie, J ohannes Gutenberg Universita¨t, Duesbergweg 10-
14, D-55099, Mainz, Germany.
(1) (a) Sellergren, B., Ed. Molecularly Imprinted Polymers: Man-
Made Mimics of Antibodies and Their Applications in Analytical
Chemistry; Elsevier Science B.V. (NL): Amsterdam, 2001. (b) Wulff,
G. Angew. Chem., Int. Ed. Engl. 1995, 34, 1812-1832.
(2) (a) Arshady, R.; Mosbach, K. Makromol. Chem. 1981, 182, 687-
692. (b) Sellergren, B.; Lepisto¨, M.; Mosbach, K. J . Am. Chem. Soc.
1988, 110, 5853-5860.
(5) Spivak, D. A.; Shea, K. J . J . Org. Chem. 1999, 64, 4627-4634.
(6) Lubke, C.; Lubke, M.; Whitcombe, M. J .; Vulfson, E. N. Macro-
molecules 2000, 33, 5098-5105.
(7) Asanuma, H.; Kajiya, K.; Hishiya, T.; Komiyama, M. Chem. Lett.
1999, 28, 665-666.
(8) Quaglia, M.; Chenon, K.; Hall, A. J .; De Lorenzi, E.; Sellergren,
B. J . Am. Chem. Soc. 2001, 123, 2146-2154.
10.1021/jo034588e CCC: $25.00 © 2003 American Chemical Society
Published on Web 10/22/2003
9132
J . Org. Chem. 2003, 68, 9132-9135

F IGURE 2. 1H NMR titration data for titration of 1 with bis-
TBA-glutarate in DMSO-d6. The CIS of the “outer” urea proton
is shown The inset shows the corresponding J ob plot.
tylammonium (TBA) salt) in a competitive medium (Ka
) 790M-1 via 1H NMR titration at 20 °C in DMSO-d6;
Ka ) 1300 M-1 via isothermal titration calorimetry at
25 °C in DMSO).
F IGURE 3. Retention behavior of the template and related
analytes on the imprinted (MIP) and nonimprinted (NIP)
polymers in TEA-modified ACN mobile phases. The retention
of N-Z-D-Glu was identical to that of N-Z-L-Glu, within
experimental error. Retention is expressed as the retention
factor k ) (t(analyte) - t0)/t0, where t(analyte) and t0 are the
retention times of the analyte and the void volume marker
(acetone), respectively. Data calculated as the average of at
least three injections with an RSD < 4%. Conditions: 20 µL
injections of analytes (10 mM), flow rate ) 1 mL/min, column
dimensions ) 100 mm × 4.6 mm i.d., detection at 260 nm.
The polymerizable bis-urea receptor (1) was readily
prepared in one step, from p-xylylene diamine and
3-isopropenyl-R,R-dimethylbenzyl isocyanate, in 67% yield.
To determine the binding abilities of (1), we performed
1H NMR titrations using bis-TBA-glutarate as our model
dicarboxy anion guest (see Figure 2).10 Addition of
increasing amounts of bis-TBA glutarate (0-10 equiv)
to DMSO-d6 solutions of 1 (concentration ) 5 mM) led
to significant downfield complexation-induced shifts (CIS)
of both the “inner” and “outer” urea proton signals (CIS
≈ 1.8 ppm in each case). A J ob plot titration (inset Figure
1), performed to establish the stoichiometry of the
interaction revealed a small deviation from the expected
1:1 monomer/guest stoichiometry (maximum at mole
fraction 1 ≈ 0.45). Indeed, the data obtained in the first
titration exhibited deviation from the theoretical 1:1
binding isotherm at higher guest concentrations (7.5 and
10 equiv); this is indicative of the formation of higher
order complexes. Exclusion of the deviant points led to a
good fit to a 1:1 binding isotherm, and an apparent
association constant (Kapp) of 1500 M-1 was extracted.
Having proven that 1 interacts strongly with the model
dicarboxy anion, it was used in the preparation of a MIP
using bis-TBA-N-Z-L-glutamate (prepared by the reac-
tion of N-Z-L-glutamic acid with 2 equiv of TBA hydrox-
ide in methanol) as the template and DMSO as the
polymerization solvent. Thermally initiated polymeriza-
tion at 40 °C was allowed to continue for 24 h. Following
the polymerization, the template was removed by extrac-
tion with methanol in a Soxhlet apparatus. The polymer
was then crushed and sized to yield particles of size 25-
50 µm. A control, nonimprinted polymer (NIP) was
prepared under identical conditions, but with the omis-
sion of the template molecule.
phases based on acetonitrile (ACN) and modified with
triethylamine (TEA). The results are summarized in
Figure 3.
Considering first the behavior of the template, we see
that with pure acetonitrile as the mobile phase the
differences in retention between the MIP and NIP are
relatively small. This is not overly surprising, as the urea
functionality should bind only weakly to nondissociated
acid functions. Modification of the mobile phase with
small amounts of TEA (1-2%) leads to the observance
of a definite imprinting effect; this effect levels off after
addition of 2% TEA to the mobile phase. These observa-
tions act as a further confirmation of the mode of
imprinting and subsequent recognition in these polymers,
i.e., that the urea functionalities interact strongly with
template only after the addition of an agent capable of
causing deprotonation. Unfortunately, we found that the
MIP exhibited no enantioselectivity. Polarimetry mea-
surements indicated that this was not due to racemisa-
tion of the template during any of the preparation
procedures. The binding site was designed to be able to
accommodate 1,5-diacids primarily, while leaving space
for larger substituents at N. Thus, we used a stoichio-
metric amount of 1 to complement the acid groups in the
template. For enhanced enantioselectivity, according to
Dalgleish’s three-point rule,11 three interactions are
required for effective enantiodiscrimination. This may be
achieved in the present case by targeting the carbamate
moiety, but was not the objective in the present work.
Despite the lack of enantioselectivity observed, we
found that the MIP was able to separate 2 from an
equimolar mixture with N-Z-L-Asp and N-Z-Gly, as
shown in Figure 4, while the NIP was not. The retention
order of the analytes on both polymers (single injections)
The recognition properties of the respective polymers
toward the template, its optical antipode, and other
amino acids were examined via HPLC using mobile
(9) (a) Fan, E.; van Arman, S. A.; Kincaid, S.; Hamilton, A. D. J .
Am. Chem. Soc. 1993, 115, 369-370. (b) Linton, B. R.; Goodman, M.
S.; Fan, E.; van Arman, S. A.; Hamilton, A. D. J . Org. Chem. 2001,
66, 7313-7319.
(10) Connors, K. A. Binding Constants: The Measurement of Molec-
ular Complex Stability; J . Wiley & Sons: New York, 1987.
(11) Dalgleish, C. E. J . Chem. Soc. 1952, 3940.
J . Org. Chem, Vol. 68, No. 23, 2003 9133

F IGURE 4. Separation of an equimolar mixture of N-Z-L-Glu
(2), N-Z-L-Asp, and N-Z-Gly (each 3.33 mM) on the MIP (solid
line) and NIP (broken line) columns. Conditions: mobile phase
2% TEA in ACN, flow rate ) 1 mL/min, column dimensions )
100 mm × 4.6 mm i.d., detection at 260 nm.
F IGURE 5. Retention of methotrexate (3) on the imprinted
and nonimprinted polymers in different mobile phases. The
retentions are given as retention factors, k, as defined in
Figure 2. Conditions: 20 µL injections of 3 (1 mM), flowrate )
1 mL/min, column dimensions ) 100 mm × 4.6 mm i.d.,
detection at 260 nm.
follows a similar order, perhaps reflecting the predipo-
sition of the functional monomer toward 1,5-dicarboxy
anions. However, retention of 2 is much stronger on the
MIP, suggesting that the imprinting process enhances
this predisposition by “locking in” the functionality of the
monomer and reducing the inherent flexibility of the
monomeric “binding pocket”.
As a continuation of our earlier work on substructure
imprinting approaches for the recognition of dihydrofolate
reductase inhibitors,8 we were further interested in
assessing the ability of the MIP to recognize larger
molecules containing the glutamate substructure, e.g.,
methotrexate (3), folic acid (4), and leucovorin (5).
tion of 3 is reported to have pKa values of 3.4 and 4.7,12
while for the acid form of 5 (folinic acid) the values are
3.1 and 4.8.12 In our study, 5 is present as the calcium
salt. Thus, for 5 to bind effectively to the polymers, the
binding energy needs to overcome that of the calcium
complex; we feel that this is unlikely to occur, hence the
elution of 5 without retention on the MIP and NIP
columns. No pKa values are available for the glutamate
portion of 4, as it is soluble only between pH 5-10.5.13
While it is reasonable to assume that the glutamate
portion of 4 has pKa values similar to those of 3, we
attribute its lack of retention on the polymers to the
nature of the pteridone substructure of the molecule,
although the reason is not fully clear.
Con clu sion s
A novel bis-urea functional monomer has been pre-
pared, and its ability to form strong intermolecular
hydrogen bonds to 1,5-dicaboxy anions in a competitive
medium (DMSO-d6) has been quantified. In the subse-
quently prepared MIP, imprinting effects are observed
only when the analyte is in the deprotonated state, thus
confirming the mode of binding to the pendant polymer
functionality. The MIP is selective for N-Z-L-Glu over
N-Z-L-Asp and N-Z-Gly. Further, the MIP is able to bind
the drug methotrexate, thus providing further evidence
that substructure approaches to molecular imprinting
offer a viable alternative to the imprinting of the entire
target structure. This may be important in cases where
the target exhibits limited solubility or is prohibitively
expensive. Due to the predisposition of the bis-urea
monomer to bind R,ω-C5 diacids, its use in combination
with competing monomers to imprint larger targets
containing the glutamic acid substructure may also be
feasible. This is the subject of our current investigations.
The three molecules exhibit vastly different behavior
when applied to the MIP and NIP columns. In all three
mobile phases, 3 is retained much more strongly on the
MIP than on the NIP, the retention increasing with the
amount of TEA in the mobile phase (see Figure 5).
Conversely, both 4 and 5 eluted either at the same time
or before the void marker (acetone).
In the case of 3, the basicity of the pteridine part of
the molecule is likely to be sufficient to cause at least
some deprotonation of the glutamate moiety, thus leading
to retention of 3 even in the unmodified ACN mobile
phase. The dissociated state is ensured in the TEA-
modified mobile phases and, accordingly, leads to in-
creased retention of 3. The enhanced recognition of 3 by
the MIP when compared to the NIP provides further
evidence for the success of the imprinting process. The
behaviors of 4 and 5 on both MIP and NIP columns are
somewhat more difficult to explain. The glutamate por-
Exp er im en ta l Section
Ma ter ia ls a n d Meth od s. All materials were of reagent grade
unless otherwise stated. 3-Isopropenyl-R,R-dimethylbenzyl iso-
cyanate and N-Z-L-glutamic acid (2, 99%) were purchased from
(12) Merck Index, 11th ed.; Merck & Co., Inc.: Rahway, NJ , 1989.
(13) Pohland, A.; Flynn, E. H.; J ones, R. G.; Shove, W. J . Am. Chem.
Soc. 1951, 73, 3247-3252.
9134 J . Org. Chem., Vol. 68, No. 23, 2003

Aldrich (Deisenhofen, Germany), while N-Z-D-glutamic acid
(98+%) was from Lancaster (Morecombe, UK). 1,4-Xylylenedi-
amine, glutaric acid, 1 M methanolic tetrabutylammonium
hydroxide, and tetrabutylammonium benzoate were purchased
from Acros Organics (Geel, Belgium). Methotrexate (3, (+)-
amethopterin, >98%), folic acid (4, 98%), leucovorin (5, dl-L-form
of folinic acid, as its calcium(II) salt, 95%), N-Z-L-aspartic acid
(99%), and N-Z-glycine (99%) were purchased from Sigma (St.
Louis, MO). Anhydrous solvents, tetrahydrofuran and dimethyl
sulfoxide, were obtained from Fluka (Deisenhofen, Germany)
and stored over appropriate molecular sieves. All solvents used
in the chromatographic experiments were of HPLC grade and
were purchased from Carlo Erba (Milan, Italy).
Syn th esis of Bis-tetr a bu tyla m m on iu m Glu ta r a te. This
compound was prepared from glutaric acid by the procedure
described above and used without further purification.
Associa tion Con sta n t Deter m in a tion . The model com-
pound used as titration partner against 1 was bis-tetrabuty-
lammonium glutarate, prepared as described above. The stoi-
chiometry of the interaction between 1 and bis-TBA-glutarate
was determined by the method of continuing variance (J ob plot).
Thus, equimolar solutions (5 mM) of each monomer and the
glutarate in DMSO-d6 were added together in the ratios 0:1, 0.1:
0.9, 0.2:0.8, ..., 0.0.8:0.2, 0.9:0.1, and 1:0, respectively, and plots
of ∆δ‚a versus a were constructed, where ∆δ refers to the
complexation induced shift of the urea protons in the monomer
and a is the mole fraction of monomer. Association constants
for the interactions between 1 and bis-TBA-glutarate were
determined by titrating an increasing amount of 2 into a
constant amount of host. The concentration of 1 was 5 mM, and
the amounts of glutarate added were 0, 0.5, 1, 2, 3, 4, 5, 7.5,
and 10 equiv (i.e., 0-50 mM), respectively. The CIS of both sets
of urea protons were monitored, and titration curves were
constructed. The raw titration data were fitted to a 1:1 binding
isotherm and the association constants were obtained by least-
squares fitting of the isotherms using Microcal Origin.
All NMR spectra were obtained using a Bruker Advance DRX
400 spectrometer. 1H spectra were obtained at 400 MHz, and
13C spectra were obtained at 100 MHz. Elemental microanalysis
was performed using a “CHN-rapid” HERAEUS analyzer (Ha-
nau, Germany).
All HPLC evaluations were performed using a HP1090
instrument (Hewlett-Packard, Palo Alto, CA) equipped with a
Rheodyne sample valve (20 mL loop) and a diode array detector,
connected to a HP ChemStation (Version A.06.03). Mobile phases
were degassed by sparging with He for 10 min prior to column
conditioning. All analyses were conducted at room temperature
at a flow rate of 1 mL/min, with detection at 260 nm. Retention
factors are expressed as k ) (t - t0)/t0, where t is the retention
time of the analyte and t0 is the retention time void marker
(acetone). Analyte sample preparation was as follows: Z-pro-
tected amino acids were all as solutions in ACN; for 3, a 5 mM
solution in N-methylpyrrolidinone (NMP) was diluted to 1 mM
with ACN; a 1 mM solution of 4 was prepared in NMP; a 1 mM
solution of 5 was prepared in water.
Syn th esis of 1-[1-(3-Isop r op en ylp h en yl)-1-m eth yleth yl]-
3-[4-[3-[1-(3-isop r op en ylp h en yl)-1-m et h ylet h yl]u r eid om -
eth yl]ben zyl]u r ea (1). 1,4-Xylylene diamine (2.72 g, 20 mmol)
was dissolved in anhydrous tetrahydrofuran (125 mL), and the
solution was stirred at ambient temperature under a stream of
nitrogen. 3-Isopropenyl-R,R-dimethylbenzyl isocyanate (8.85 mL,
44 mmol) was then added dropwise, such that the heat of
reaction did not cause the solvent to boil. The solution was
stirred overnight at ambient temperature under a stream of
nitrogen, whereafter the solvent was removed in vacuo. The
residue was recrystallized from ethanol to yield the product as
a white, crystalline solid (6.87 g, 67%): mp 210-211 °C; 1H NMR
(DMSO-d6) δ 1.51 (s, 12H), 2.06 (s, 6H), 4.08 (d, 4H), 5.04 (s,
2H), 5.32 (s, 2H), 6.20 (t, 6.20, 2H), 6.35 (s, 2H), 7.09 (s, 2H),
7.19-7.26 (m, 6H), 7.44 (s, 2H); 13C NMR (DMSO-d6) δ 21.85,
30.43, 42.53, 54.43, 112.52, 122.02, 123.06, 124.5, 127.01, 128.09,
139.50, 140.28, 143.33, 149.35, 157.29; HRMS calcd 538.3308,
found 538.3284. Anal. Calcd for C34H42N4O2‚0.5H2O: C, 74.56;
H, 7.91; N, 10.23. Found: C, 74.6; H, 7.7; N, 10.1.
P olym er P r ep a r a tion . The molecularly imprinted polymer
(MIP) was prepared as follows. Bis-tetrabutylammonium-N-Z-
L-Glu (0.281 g, 1 mmol), functional monomer 1 (0.538 g, 1 mmol),
ethyleneglycol dimethacrylate (3.96 g, 20 mmol), and the initia-
tor ABDV (45 mg, 1% w/w total monomers) were dissolved in
anhydrous DMSO (5.6 mL). The solution was transferred to a
glass tube and then degassed by purging with nitrogen for 15
min. Polymerization was initiated by placing the stoppered tube
in a water bath thermostated at 40 °C and allowed to continue
for 48 h. After this time, the tubes were broken and the polymers
crushed lightly. The crushed polymers were extracted with
methanol in
a Soxhlet apparatus for 21 h. The extracted
polymers were then further crushed and sieved, and particles
of size 25-50 µm were collected. This fraction was then repeat-
edly sedimented to remove fine particles using methanol/water
(80/20 v/v), and the sedimented particles were used for the
chromatographic evaluation. A control, nonimprinted polymer
(NIP) was prepared in exactly the same manner, but with the
omission of the template molecule from the pre-polymerization
mixture. For the HPLC evaluation, the MIP and NIP were each
slurry-packed into stainless steel columns (100 mm × 4.6 mm
i.d., Alltech, Milan, Italy) using MeOH/water (80/20 v/v) as
pushing solvent using Haskel air drive fluid pump (Haskel Inc.,
Burbank, CA).
Ack n ow led gm en t. We gratefully acknowledge fi-
nancial support from the European Union under the
program “Training and Mobility of Researchers” (EU-
TMR project MICA, Contract No. FMRX-CT-98-0173),
the DAAD-Vigoni project, and the PRIN protject
(2002034857_003).
Syn th esis of Bis-tetr a bu tyla m m on iu m -N-Z-L-glu ta m a te.
N-Z-L-Glutamic acid (0.563 g, 2 mmol) was dissolved in methanol
(50 mL) and 1 M methanolic tetrabutylammonium hydroxide
(4 mL, 4 mmol) was added in one portion. The solution was
stirred at ambient temperature for 1 h, and then the solvent
was removed in vacuo. The oily residue was dried overnight at
80 °C in vacuo and was used without further purification.
J O034588E
J . Org. Chem, Vol. 68, No. 23, 2003 9135