Water-Soluble molecularly imprinted nanoparticles (MINPs) with tailored, functionalized, modifiable binding pockets

Article abstract of DOI:10.1002/chem.201404919

Construction of receptors with binding sites of specific size, shape, and functional groups is important to both chemistry and biology. Covalent imprinting of a photocleavable template within surface-core doubly cross-linked micelles yielded carboxylic acid-containing hydrophobic pockets within the water-soluble molecularly imprinted nanoparticles. The functionalized binding pockets were characterized by their binding of amine- and acid-functionalized guests under different pH values. The nanoparticles, on average, contained one binding site per particle and displayed highly selective binding among structural analogues. The binding sites could be modified further by covalent chemistry to modulate their binding properties.

Full text of DOI:10.1002/chem.201404919

DOI: 10.1002/chem.201404919
Full Paper
&
Molecular Recognition
Water-Soluble Molecularly Imprinted Nanoparticles (MINPs) with
Tailored, Functionalized, Modifiable Binding Pockets
Joseph K. Awino and Yan Zhao*[a]
Abstract: Construction of receptors with binding sites of
specific size, shape, and functional groups is important to
both chemistry and biology. Covalent imprinting of a photo-
cleavable template within surface–core doubly cross-linked
micelles yielded carboxylic acid-containing hydrophobic
pockets within the water-soluble molecularly imprinted
nanoparticles. The functionalized binding pockets were char-
acterized by their binding of amine- and acid-functionalized
guests under different pH values. The nanoparticles, on aver-
age, contained one binding site per particle and displayed
highly selective binding among structural analogues. The
binding sites could be modified further by covalent chemis-
try to modulate their binding properties.
erogeneous and less structurally defined.[2,6] According to a rep-
resentative review,[7] MIPs ideally are “preparable in one (or
few) high yielding synthetic step(s)”, “able to be post-syntheti-
cally functionalized,” and possess “homogeneous imprinted
sites of high stability” with “high (binding) affinity with possi-
bility to tune.” Although highly desirable and important to the
applications of MIPs, many of these features are yet to be real-
ized with traditional imprinting techniques.
Introduction
The active (binding, transport, or catalytic) sites of proteins are
key to their intended functions. Researchers in the field of
supramolecular chemistry have, over the last few decades, syn-
thesized many synthetic receptors that mimic one or more as-
pects of these active sites, with the majority of them prepared
through molecular synthesis.[1] Although molecular synthesis
ensures synthetically pure and discrete functional molecules,
the significant synthetic efforts required often become an im-
pediment to scale-up and practical applications of the materi-
als.
Herein, we report a method to construct tailor-made, hydro-
phobic binding pockets possessing specific binding functional
groups in water-soluble nanoparticles. Our method has over-
come some key challenges in conventional molecular imprint-
ing including heterogeneous distribution of binding sites and
difficulty in direct characterization of binding properties by
spectroscopic methods. These molecularly imprinted nanopar-
ticles (MINPs) were shown to distinguish guests based on their
size, shape, and functional groups. Most interestingly, the bind-
ing pockets could be modified through standard chemistry to
alter their molecular-recognition properties.[8] Different from
traditional MIPs[2] or other reported imprinted nanoparticles,[9]
our MINPs on average contained one binding pocket per parti-
cle, thus bridging the gap between the discrete receptors
made through molecular synthesis and those less well-defined
receptors made through traditional imprinting.
Molecular imprinting is a conceptually different approach to
synthetic receptors.[2] Instead of building the receptors first
and then trying to fit their guests into the structures, one
simply co-polymerizes appropriate functional monomers (FMs)
and cross-linkers around molecular templates. Removal of the
templates generates guest-complementary binding pockets
within the polymer matrix. Much progress has been made in
molecularly imprinted polymers (MIPs) since Wulff[3] and Mos-
bach[4] respectively pioneered covalent and noncovalent im-
printing (referring to the binding interactions between FMs
and the template). The concept has also been extended
beyond traditional macroporous polymers to imprinted surfa-
ces[2] and even unimolecularly within dendrimers.[5]
Although molecular imprinting can create binding sites
more efficiently than molecular synthesis, it is generally accept- Results and Discussion
ed that the binding sites obtained through imprinting are het-
Materials design and synthesis
The method was a development from our recently reported
[a] J. K. Awino, Prof. Dr. Y. Zhao
Department of Chemistry
Iowa State University
molecular imprinting method using surface-cross-linked mi-
celles (SCMs)[10] prepared from 1 (Scheme 1).[11] This surfactant
(1) has a tripropargylammonium headgroup cross-linkable on
the surface by the click reaction. The methacrylate group at
the hydrophobic tail enables core cross-linking around a hydro-
phobic template solubilized by the micelle in water. In the pre-
Ames, IA 50011-3111 (USA)
Fax: (+1)515-294-0105
E-mail: zhaoy@iastate.edu
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201404919.
Chem. Eur. J. 2014, 20, 1 – 8
1
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
celle of 1 through electrostatic interactions, it orients the tem-
plate to make its methacrylate group point inward. Because
the methacrylate group is nearly at the opposite end of mole-
cule from the sulfate that is anchored at the micelle surface,
the methacrylate (after polymerization and photo-deprotec-
tion) is expected to position the carboxylic acid deep inside
the hydrophobic pocket of the final MINP–COOH (Scheme 1).
MINP–COOH synthesis was adapted from our earlier proce-
dures.[11a] Click-cross-linking of the template-containing mi-
celles by the water-soluble diazide 3 in the presence of a CuI
catalyst created alkynyl–SCMs, which were surface-functional-
ized through another round of click reactions with sugar-de-
rived 4. The reactions were performed at 10 mm of 1 in water,
above its critical micelle concentration (CMC) of 0.55 mm. Each
SCM, according to our dynamic light scattering (DLS) study,
contained approximately 50 surfactant molecules. Thus, a ratio
of [1]/[2]=1:0.02 in theory placed one template within each
SCM, a feature verified in our previous bile salt-binding
MINPs.[11a]
In the previous procedure, we employed photo-polymeri-
zation to cross-link the methacrylate of 1 with divinylbenzene
(DVB) solubilized in the core. The method was clearly unsuita-
ble with 2 having the photocleavable o-nitrobenzyl ether. We
thus decided to solubilize a small amount of AIBN (azobisiso-
butyronitrile, a thermal initiator) at the beginning of the proce-
dure and carried out thermal polymerization of the methacry-
late and DVB at 758C for 16 h after the “surface-clicking”. For-
tunately, as hydrophobic interactions generally remain effective
at high temperatures (and generally change from entropically
driven to enthalpically driven with increasing temperatures),[13]
template 2 was successfully cross-linked with the rest of the
structure. Alkenic protons disappeared completely at the end
of the thermal polymerization process as shown by 1H NMR
spectroscopy (Figure S1 in the Supporting Information).
Our last step in the materials synthesis was the photolytic
cleavage of the o-nitrobenzyl linkage to remove the nitroso de-
rivative 5 and vacate the binding site. The reaction progress
could be monitored easily by fluorescence spectroscopy (Fig-
ure S2 in the Supporting Information) because of the fluores-
cent naphthalene group of 5.[14]
According to DLS, the SCM, surface-functionalized SCM, and
the final MINP-COOH averaged 3.5, 6.3, and 4.7 nm in diameter,
respectively (Figure S3 in the Supporting Information). The size
of MINP-COOH translated to approximately 50,000 Daltons in
molecular weight (Figure S4 in the Supporting Information),
comparable to many proteins in this regard. Overall, the MINP-
COOH bears a resemblance to a water-soluble protein receptor
with a hydrophilic exterior, a hydrophobic core, a specifically
shaped hydrophobic binding site, and an internal functional
group. It is worth mentioning that the sugar-derived surface
ligand 4 was installed not just to make the MINP mimic
a water-soluble protein in its surface hydrophilicity; its high
crystallinity allowed the MINPs to be easily precipitated from
solvents such as acetone while maintaining complete solubility
in water and polar solvents such as DMF. As will be shown
later, solubility in selected organic solvents was critical to the
covalent modification of the MINPs.
Scheme 1. Preparation of MINP–COOH.
vious work, we demonstrated that selective binding pockets
could be created within the SCM for bile salts. The limitation
of the previous method lies in the fact that only hydrophobic
pockets with prescribed shapes could be created within the
MINPs. Without specific binding groups, the pockets recognize
guests primarily based on their size, shape, and hydrophobici-
ty.
To install a functional group within a molecularly imprinted
binding pocket, one normally has to employ FMs that bind the
template through specific noncovalent interactions. Although
such a method works well for conventional MIPs in organic
media, it is completely unsuitable in our case because the
entire imprinting takes place in aqueous solution. Not only do
FMs with polar binding groups (e.g., methacrylic acid) tend to
stay in water instead of within the hydrophobic core of the mi-
celle, the intended noncovalent template–FM complex is also
unlikely to be stable when a large amount of water is pres-
ent.[12] Even if the template–FM complex is somehow made
stable inside the micelle, the polar FM most likely would stay
at the surfactant/water interface instead of in the hydrophobic
core of the micelle. As a result, even if such imprinting is made
to work, it will be difficult to have the polar functional group
deep within the hydrophobic core of the resulting MINP. Need-
less to say, for molecular recognition in water, polar binding in-
teractions typically are stronger in a more deeply imbedded
hydrophobic microenvironment.
To overcome these challenging problems, we designed
a photocleavable template (2) containing an o-nitrobenzyl link-
age (Scheme 1). The overall hydrophobicity of 2 allowed it to
be easily incorporated into the micelle of 1. The sulfonate
group of the template had a strategic purpose in our design:
in addition to strengthening the binding with the cationic mi-
&
&
Chem. Eur. J. 2014, 20, 1 – 8
www.chemeurj.org
2
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Characterization of the carboxylic acid-containing binding
pockets
To further understand the role of the MINP carboxyl in the
binding, we performed similar fluorescence titrations of amine
6 at different pH values (2.2–9.5) in citrate phosphate and Tris
buffers. The concentration (10–50 mm) of the buffers showed
negligible effect on the obtained binding constants. Such re-
sults generally suggest that ionic strength does not play any
significant role in the binding.[16] In the bile salt-binding MINPs,
the negligible effect of ionic strength seemed to come from
two opposing effects of the salts on the hydrophobic interac-
tions and electrostatic forces involved in the binding, respecti-
vely.[11a]
To characterize the carboxyl-functionalized MINP receptors, we
first studied the binding of a template analogue, 6, which con-
tained an amino group in the position of the methacrylate
group in template 2. As the carboxyl group in the MINP bind-
ing pocket was generated from the polymerized methacrylate
of 2, compound 6 upon binding should have its amino group
in close proximity to the MINP carboxyl group. The host–guest
binding thus should be driven by a combination of hydropho-
bic interactions (between the hydrophobic portion of 6 and
the MINP) and an ammonium–carboxylate salt bridge. Being
located in a hydrophobic microenvironment, the salt bridge
should be particularly strong.[15] As in our bile salt-binding
MINPs, electrostatic interactions between the positively
charged cross-linked micelle and the sulfonate group should
contribute as well.
Because both the host and the guests contain removable
protons, we have to consider the acid/base properties of all
the reactants in the binding.[17] Scheme 2 shows the acid–base
Scheme 2. The acid–base equilibria and the binding of amine 6 by MINP–
COOH.
The water solubility of our MINPs allowed us to study their
binding properties by using standard titration methods, a fea-
ture difficult to achieve with conventional imprinting that
tends to yield insoluble macroporous polymers.[1c] As shown
by Figure 1a, upon titration of 6 by MINP–COOH in Tris buffer
equilibria involved and the binding of amine 6 by MINP–
COOH. The acidity constant of (protonated) amine 6 (pKNH ) in
3
solution is probably similar to that for 1-phentylethylamine
(pKa =9.4).[18] Assuming MINP–COOH deprotonates more easily
than protonated 6, we anticipate that the strongest binding
between MINP–COOH and 6 would occur in between pKMINP
and pKNH . Below pKMINP, the dominant forms of the reactants
3
+
are MINP–COOH and RNH3 (i.e., protonated 6). These two
species cannot form the ammonium–carboxylate salt bridge di-
rectly. Instead, MINP–COOH needs to undergo an unfavorable
deprotonation (in the acidic medium) in order for the binding
to occur. The binding, thus, would be compromised by the de-
protonation of the MINP carboxyl. Above pKNH , MINP–COOÀ
3
and RNH2 (i.e., 6 itself) will dominate on the other hand. To
form the ammonium–carboxylate salt bridge, 6 has to undergo
an unfavorable protonation under the now basic conditions
and, thus, would also weaken the binding.
Figure 1. (a) Emission spectra of 6 (lex =300 nm) upon addition of different
concentrations of MINP–COOH in 50 mm Tris buffer (pH 7.4). [6]=0.5 mm.
(b) Nonlinear least squares fitting of the emission intensity of 6 at 415 nm to
a 1:1 binding isotherm.
Our titration confirmed the predictions. As shown by
Figure 2, the binding was undetectable by fluorescence spec-
troscopy at ꢀpH 5, became stronger with increasing pH, and
at pH 7.4, the emission at 410 nm decreased and a weaker
peak at 470 nm appeared gradually. Although we could not be
certain why the MINP-bound 6 emitted at a longer wave-
length, it is possible that the binding slowed down the rota-
tion around the s-bond between the triazole and the naphthyl
rings and enhanced the conjugation between the two aromat-
ic groups. For the same reason, although environmentally sen-
sitive dansyl fluorophores are often used to probe the local po-
larity of the binding pocket, we could not do so with 6, as its
emission depends on multiple factors. Nonetheless, the fluo-
rescence data fit nicely to a 1:1 binding isotherm to afford
a binding constant of Ka =(1.5Æ0.3)ꢁ106 mÀ1 (Figure 1b).
Figure 2. The apparent binding constants of MINP–COOH for 6 in citrate
phosphate buffer (pH 2.2–6.2) and Tris buffer (pH 7.4–9.5) obtained by fluo-
rescence titration.
Chem. Eur. J. 2014, 20, 1 – 8
www.chemeurj.org
3
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
weakened again above pH 7.4. According to Scheme 2, the
occurred at low pH values (2.2–2.6) and showed a sharp de-
crease as the solution became less acidic. The binding weak-
ened significantly at pH 3 and became completely undetecta-
ble by fluorescence at ꢁpH 6.2. If we take the midpoint be-
tween pH 2.6 (where the binding was the maximum) and
pH 6.2 (where the binding first became zero) as the acidity
constant of 7, a value of 4.4 is obtained, exactly as predicted
from the pKa of 3,4-dimethoxybenzoic acid.
maximum Ka (1.5ꢁ106 mÀ1) should reflect the binding when
+
MINP–COOÀ and RNH3 predominate in the solution. If we
take the midpoint between pH 5 (where the binding was still
zero but began to rise) and pH 7.4 (where the binding was the
strongest) as pKMINP, the acidity constant of the MINP carboxyl
is estimated to be 6.2. This value is significantly higher than
acetic acid (pKa =4.76) or benzoic acid (pKa =4.20) in water.
The larger pKa for MINP–COOH is very reasonable and strongly
supports the location of the acidic group in a hydrophobic mi-
croenvironment. It is well known from protein chemistry that
a carboxyl group located in a hydrophobic pocket is more diffi-
cult to deprotonate than in aqueous solution, as the resulting
carboxylate cannot be solvated properly in a hydrophobic mi-
croenvironment.[19]
Because the maximum binding constants for 6 and 7 were
quite similar (1.5 and 1.3ꢁ106 mÀ1, respectively), it seems the
ammonium–carboxylate salt bridge and the carboxylic acid
dimer make similar contributions to the overall binding. This is
a useful piece of information for molecular recognition in
water. Recent work of ours shows that, although a (guanidini-
um–carboxylate) salt bridge could be strong, for molecular rec-
ognition in self-assembled hydrophobic entities such as mi-
celles or lipid bilayers, a carboxylic acid dimer could be more
useful. This is because the uncharged nature of a carboxylic
acid dimer makes migration into a hydrophobic microenviron-
ment easier.[21] Charged functional groups (e.g., ammonium,
guanidinium, or carboxylate) often have a strong tendency to
stay within or at least close to water to satisfy their solvation
needs.[22]
The presence of the carboxylic acid in the hydrophobic
binding pocket of MINP–COOH was verified additionally by its
binding of the carboxylic acid guest 7. Scheme 3 shows the
Binding selectivity of MINP–COOH
Scheme 3. The acid–base equilibria and the binding of acid 7 by MINP–
An important property of molecularly imprinted materials is
their binding selectivity. The above study demonstrated bind-
ing selectivity for acid- and base-functionalized template ana-
logues in a pH-dependent manner. For example, at pH 7.4, the
MINP receptor bound 6 with micromolar affinity but did not
bind 7 at all. Under acidic conditions (pH 2.2–2.6), the exact
opposite selectivity was achieved.
COOH.
various acid–base equilibria involved in the binding. Because
the carboxyl group of 7 is exposed to solvent, its pKa is expect-
ed to be similar to that of 3,4-dimethoxybenzoic acid (pKa =
4.43).[20] As the binding between two carboxylic acids occurs
through the hydrogen-bonded carboxylic acid dimer, both
MINP–COOH and 7 need to be in the protonated form to ach-
ieve the strongest binding.
To understand the binding selectivity of the MINP receptor
for other non-acidic/basic analogues (8–11), we switched to
isothermal titration calorimetry (ITC) to determine the binding
constants. One reason for the change was that fluorescence ti-
tration is unsuitable for bindings with lower binding affinities.
Additionally, ITC could afford other useful information includ-
ing binding enthalpy, entropy, and
the number of binding sites per
particle (N). Because these guests
Indeed, as shown by Figure 3, the pH profile for the binding
do not contain acid/base groups,
of 7 was nearly opposite to that of 6: the strongest binding
we performed ITC titrations under
neutral conditions in 50 mm Tris
buffer.
ITC confirmed both the 1:1 bind-
ing stoichiometry and the binding
affinity for 6 and yielded a Ka value
very similar to that obtained by
fluorescence titration (i.e., 1.5ꢁ
106 mÀ1, see Table 1, entry 1). The
titration curve and the fitting of
the experimental data are shown
Figure 4. ITC titration curve
in Figure 4. ITC was able to detect
Figure 3. The apparent binding constants of MINP–COOH for 7 in citrate
phosphate buffer (pH 2.2–6.2) and Tris buffer (pH 7.4–9.5) obtained by fluo-
rescence titration.
obtained at 298 K for the
the binding of 7 (which could not
binding of 6 in 50 mm Tris
be measured by fluorescence titra- buffer (pH 7.4).
&
&
Chem. Eur. J. 2014, 20, 1 – 8
www.chemeurj.org
4
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Table 1 shows that all the bindings studied by ITC were
largely enthalpically driven, with generally unfavorable entrop-
ic terms. We do not believe the results imply that the contribu-
tion of hydrophobic interactions were insignificant. Although
the classical hydrophobic effect is considered to be entropical-
ly driven,[24] the effect is multifaceted and the energetic charac-
teristics may be different depending on the (aliphatic/aromatic)
nature of the guests and the size/shape of the hydrophobic
surfaces.[25]
Table 1. Binding data for MINP–COOH obtained by ITC.[a]
Entry Guest Ka
[104 MÀ1
DG
[kcalmolÀ1
DH
[kcalmolÀ1
ÀTDS
N
]
]
]
[kcalmolÀ1
]
1
2
3
4
5
6
7
8
9
6
7
8
9
10
11
12
6[b]
152Æ6
8.8Æ0.5
76Æ3
À8.4
À6.7
À8.0
À8.0
À5.2
À6.4
À70.2
À15.3
À29.8
À32.1
À7.2
61.8
8.6
21.8
24.1
2.0
1.0Æ0.1
0.6Æ0.1
0.7Æ0.1
0.6Æ0.1
0.4Æ0.1
0.4Æ0.1
1.2Æ0.1
0.6Æ0.1
0.7Æ0.1
73Æ6
0.6Æ0.5
5.2Æ0.7
À8.2
2.8
0.27Æ0.06 À4.7
À3.5
À1.1
À1.1
2.2
Notably, MINP–COOH contained one binding pocket per
nanoparticle, evident from the binding studies for those
guests with strong bindings (in which the ITC curve fitting
would be more reliable). This feature came directly from the
surfactant aggregation number and the ratio of surfactant to
template during the imprinting. As demonstrated in the bile
salt-binding MINPs, if needed, the binding stoichiometry could
be tuned by the surfactant/template ratio quite easily.[11a]
3.1Æ0.5
À6.1
À7.4
À5.0
12[b] 25Æ2
À9.6
[a] The titrations were generally performed in duplicate in 50 mm Tris
buffer (pH 7.4) and the errors between the runs were <20% except in
very weak bindings. [b] The host was MINP–CONHNaph.
tion) at pH 7.4 and, as expected, gave a much weaker Ka, ap-
proximately 1/17 of that for 6. Ketone 8 and ester 9 were
bound similarly as expected (entries 3 and 4) and were bound
more strongly than acid 7.[23] We were delighted to see the
moderate selectivity for 6 over 8 or 9. After all, these com-
pounds were very similar guests in many regards.
Covalent modification of the binding pockets
Similar to an active site of a protein, the binding pocket of
MINP–COOH could be modified through covalent chemistry.
This could be a great way to tune the binding properties of
the MINP receptor. To demonstrate this feature, we dissolved
the nanoparticles in DMF and activated the MINP carboxyl
with 1-ethyl-3-(3-dimethylamino-propyl)carbodiimide hydro-
chloride (EDCI), a standard amide-coupling reagent. After treat-
ment with 2-aminonaphthalene, the resulting MINP–CON-
HNaph displayed characteristic naphthalene emission at
410 nm (Figure S5 in the Supporting Information).
Most interestingly, MINP–COOH and MINP–CONHNaph dis-
played (anticipated) distinctly different molecular-recognition
properties. As described earlier, amine 6 was the best guest for
MINP–COOH at pH 7.4, as everything including size, shape, and
functional groups matched perfectly between the host and the
guest. Not surprisingly, dansyl sulfate 12, which at most repre-
sented a “half-matched” guest, was too small to bind strongly
to MINP–COOH (Table 1, entry 7, Ka =0.27ꢁ104 mÀ1 or 560
times weaker than that of 6). Once naphthylated, however, the
binding pocket was expected to bind dansyl sulfonate much
better, as the 2-aminonaphthalene group was chosen to make
up for what was missing in dansyl sulfate 12 from 6 (i.e., the
phenyl–triazole spacer between the amino and the naphthyl
group; compare the two structures in Figure 5). As shown by
the data in Table 1 (entries 8 and 9), MINP–CONHNaph showed
It is interesting to consider the ionic state of the carboxyl
group on MINP–COOH during binding. The binding study for
amine 6 yields a pKa of 6.2 for the MINP carboxyl. Upon “plug-
ging” the binding pocket with a hydrophobic guest such as 8
or 9, the immediate environment around the MINP carboxyl
becomes more hydrophobic upon the expulsion of water from
the binding pocket. We would not be surprised that the bind-
ing should further increase its pKa so that the carboxyl stays
“comfortably” protonated even when the bulk aqueous phase
has a pH of 7.4. In other words, for an acidic (or basic) group
relatively deep inside in a hydrophobic pocket in water, its
acid or base property is not a fixed constant as in solution but
is an intimate function of the guest present in the binding
pocket. These properties apparently are critical to the binding
and catalytic properties of proteins.[19]
Compound 10 is overall quite similar to 8 and 9 but lacks
a methoxy group and a methyl ester or acetyl group. Its bind-
ing was two orders of magnitude weaker, testifying to the ex-
cellent shape/size selectivity of the MINP (Table 1, entry 5).
Ester 11 has a hexyl group instead of the methyl group in 8.
Although its binding was stronger than that of 10, the one
order of magnitude reduction in Ka from methyl ester 8 indi-
cated that the binding pocket was quite discriminating.
Figure 5. Binding of amine 6 by MINP–COOH versus binding of dansyl sulfo-
nate 12 by MINP–CONHNaph.
Chem. Eur. J. 2014, 20, 1 – 8
www.chemeurj.org
5
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
(8.0 mL). The precipitate formed was washed five times with 1:4
water/acetone mixture and dried overnight in the dark to give an
off-white powder. The power was dissolved in Millipore water
(1 mL) and irradiated in a Rayonet reactor for 12 h. Water was re-
moved under reduced pressure and the residual sample was
washed five times with 1:4 water/acetone mixture in a centrifuge
tube and dried to give the product as a white powder (15 mg,
75%).
a remarkable reversal of binding selectivity for 6 and 12:
whereas the bigger guest was preferred by MINP–COOH by
560 times, the smaller guest was bound more strongly by the
naphthylated receptor than the larger one by eight times.
Conclusion
Molecular imprinting in surfactant micelles is a powerful
method to create nanoparticle receptors that resemble water-
soluble proteins. Their hydrophilic exterior,[10a,c,d,g] hydrophobic
core,[26] and internal tailor-made binding sites[11] all could be
tuned easily with the surface-cross-linked micelle platform. Pre-
vious noncovalent imprinting in the micelles only yielded hy-
drophobic pockets with predefined shape and size.[11] By com-
bining covalent imprinting with a photoprotection strategy, we
now can install specific functional groups within the binding
pockets. Despite the many protein-like features, the MINP re-
ceptors are highly cross-linked materials with robust properties
and long-term stability. Importantly, the entire preparation and
purification of MINPs could be done in 2–3 days without spe-
cial techniques. With their excellent molecular-recognition
properties and facile preparation, we anticipate MINPs to
become very useful in many applications where custom-made,
specific binding sites are needed.
Synthesis of MINP–CONHNaph
1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDCI, 10 mL of
61.0 mgmLÀ1 in dry DMF, 0.004 mmol) was added to a stirred solu-
tion of MINP–COOH (20.0 mg, 0.0004 mmol) in dry DMF (1 mL) at
08C under nitrogen. After 2 h, 2-aminonaphthalene (10 mL of
56.2 mgmLÀ1 in DMF, 0.004 mmol) was added and the mixture was
stirred for 24 h at room temperature. The mixture was concentrat-
ed in vacuo and poured into 2 mL of acetone. The precipitate was
collected by centrifugation and rinsed several times with 2 mL of
acetone to afford the product as an off-white powder (16 mg,
80%).
Acknowledgements
We thank NSF (CHE-1303764) for supporting the research.
Keywords: amphiphilic
·
host–guest
·
hydrophobic
·
molecularly imprinting · receptors
Experimental Section
General
[1] a) J. L. Atwood, J. M. Lehn, Comprehensive Supramolecular Chemistry,
Pergamon, New York, 1996; b) J. W. Steed, P. A. Gale, Supramolecular
Chemistry: From Molecules to Nanomaterials, Wiley, Weinheim, 2012;
c) H.-J. Schneider, A. K. Yatsimirsky, Principles and methods in supra-
molecular chemistry, Wiley, New York, 2000.
Methanol, methylene chloride, and ethyl acetate were of HPLC
grade and were purchased from Fisher Scientific. All other reagents
and solvents were of ACS-certified grade or higher, and were used
as received from commercial suppliers. Routine 1H and 13C NMR
spectra were recorded on a Bruker DRX-400 or on a Varian VXR-
400 spectrometer. ESI-MS was recorded on a Shimadzu LCMS-2010
mass spectrometer. UV/Vis spectra were recorded at ambient tem-
perature on a Cary 100 Bio UV/Visible spectrophotometer. Fluores-
cence spectra were recorded at ambient temperature on a Varian
Cary Eclipse Fluorescence spectrophotometer. ITC was performed
by using a MicroCal VP-ITC Microcalorimeter with Origin 7 software
and VPViewer2000 (GE Healthcare, Northampton, MA). Syntheses
of the compounds are reported in the Supporting Information.
[2] a) G. Wulff, Angew. Chem. Int. Ed. Engl. 1995, 34, 1812–1832; Angew.
Chem. 1995, 107, 1958–1979; b) G. Wulff, Chem. Rev. 2001, 101, 1–28;
c) K. Haupt, K. Mosbach, Chem. Rev. 2000, 100, 2495–2504; d) K. J. Shea,
Trends Polym. Sci. 1994, 2, 166–173; e) B. Sellergren, Molecularly im-
printed polymers: man-made mimics of antibodies and their applications
in analytical chemistry, Elsevier, Amsterdam, 2001; f) B. Sellergren,
Angew. Chem. Int. Ed. 2000, 39, 1031–1037; Angew. Chem. 2000, 112,
1071–1078; g) M. Komiyama, Molecular imprinting: from fundamentals
to applications, Wiley-VCH, Weinheim, 2003; h) M. Yan, O. Ramstrçm,
Molecularly imprinted materials: science and technology, Marcel Dekker,
New York, 2005; i) C. Alexander, H. S. Andersson, L. I. Andersson, R. J.
Ansell, N. Kirsch, I. A. Nicholls, J. O’Mahony, M. J. Whitcombe, J. Mol. Rec-
ognit. 2006, 19, 106–180.
[3] G. Wulff, A. Sarhan, K. Zabrocki, Tetrahedron Lett. 1973, 14, 4329–4332.
[4] B. Sellergren, M. Lepistoe, K. Mosbach, J. Am. Chem. Soc. 1988, 110,
5853–5860.
[5] a) S. C. Zimmerman, M. S. Wendland, N. A. Rakow, I. Zharov, K. S. Suslick,
Nature 2002, 418, 399–403; b) S. C. Zimmerman, I. Zharov, M. S. Wend-
land, N. A. Rakow, K. S. Suslick, J. Am. Chem. Soc. 2003, 125, 13504–
13518.
Synthesis of MINP–COOH
DVB (2.8 mL, 0.02 mmol), AIBN in DMSO (10 mL of 8.2 mgmLÀ1
0.0005 mmol), 2 in D2O (10 mL, 0.0004 mmol) were added to
a 2.0 mL micellar solution of surfactant 1 (9.3 mg, 0.02 mmol) in
D2O. (D2O was used instead of H2O to facilitate monitoring of the
,
1
reaction progress by H NMR spectroscopy.) The mixture was ultra-
sonicated for 10 min. Compound 3 (4.13 mg, 0.024 mmol), CuCl2 in
D2O (10 mL of 6.7 mgmLÀ1, 0.0005 mmol), and sodium ascorbate in
D2O (10 mL of 99 mgmLÀ1, 0.005 mmol) were then added and the
reaction mixture was stirred slowly at room temperature (25 8C).
After 12 h, compound 4 (10.6 mg, 0.04 mmol), CuCl2 in D2O (10 mL
of 6.7 mgmLÀ1, 0.0005 mmol), and sodium ascorbate in D2O (10 mL
of 99 mgmLÀ1, 0.005 mmol) were added and the mixture was
stirred for another 6 h. The reaction vial was sealed with a rubber
stopper and the reaction mixture was purged with nitrogen for
15 min before it was stirred at 758C for 16 h. The resultant solution
(2.0 mL) was cooled to room temperature and poured into acetone
[6] X. Wu, W. R. Carroll, K. D. Shimizu, Chem. Mater. 2008, 20, 4335–4346.
[7] S. C. Zimmerman, N. G. Lemcoff, Chem. Commun. 2004, 5–14.
[8] a) S. McNiven, Y. Yokobayashi, S. H. Cheong, I. Karube, Chem. Lett. 1997,
26, 1297–1298; b) N. Kirsch, C. Alexander, M. Lꢂbke, M. J. Whitcombe,
E. N. Vulfson, Polymer 2000, 41, 5583–5590; c) R. J. Umpleby, G. T. Rush-
ton, R. N. Shah, A. M. Rampey, J. C. Bradshaw, J. K. Berch, K. D. Shimizu,
Macromolecules 2001, 34, 8446–8452; d) B. Sellergren, Macromolecules
2006, 39, 6306–6309; e) T. Takeuchi, N. Murase, H. Maki, T. Mukawa, H.
Shinmori, Org. Biomol. Chem. 2006, 4, 565–568.
[9] a) Y. Hoshino, T. Kodama, Y. Okahata, K. J. Shea, J. Am. Chem. Soc. 2008,
130, 15242–15243; b) A. Cutivet, C. Schembri, J. Kovensky, K. Haupt, J.
Am. Chem. Soc. 2009, 131, 14699–14702; c) K. G. Yang, M. M. Berg, C. S.
&
&
Chem. Eur. J. 2014, 20, 1 – 8
www.chemeurj.org
6
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Zhao, L. Ye, Macromolecules 2009, 42, 8739–8746; d) Z. Y. Zeng, J. Patel,
S. H. Lee, M. McCallum, A. Tyagi, M. D. Yan, K. J. Shea, J. Am. Chem. Soc.
2012, 134, 2681–2690; e) Y. Ma, G. Q. Pan, Y. Zhang, X. Z. Guo, H. Q.
Zhang, Angew. Chem. Int. Ed. 2013, 52, 1511–1514; Angew. Chem. 2013,
125, 1551–1554.
[15] I. Gitlin, J. D. Carbeck, G. M. Whitesides, Angew. Chem. Int. Ed. 2006, 45,
3022–3060; Angew. Chem. 2006, 118, 3090–3131.
[16] S.-J. Moon, J. W. Jeon, H. Kim, M. P. Suh, J. Suh, J. Am. Chem. Soc. 2000,
122, 7742–7749.
[17] B. Sellergren, K. J. Shea, J. Chromatogr. A 1993, 654, 17–28.
[18] D. D. Perrin, Dissociation constants of organic bases in aqueous solution,
Butterworths, London, 1965.
[19] F. H. Westheimer, Tetrahedron 1995, 51, 3–20.
[20] M. M. Byrne, N. H. P. Smith, J. Chem. Soc. B 1968, 809–812.
[21] L. Widanapathirana, X. Li, Y. Zhao, Org. Biomol. Chem. 2012, 10, 5077–
5083.
[22] Y. Zhao, H. Cho, L. Widanapathirana, S. Zhang, Acc. Chem. Res. 2013, 46,
2763–2772.
[23] As discussed in the previous section, the weak binding for 7 under
pH 7.4 was caused by the deprotonation of both MINP–COOH and 7 at
this pH. In order to bind, both have to undergo an unfavorable proto-
nation under this condition.
[24] M. H. Abraham, J. Am. Chem. Soc. 1982, 104, 2085–2094.
[25] D. Chandler, Nature 2005, 437, 640–647.
[10] a) S. Zhang, Y. Zhao, Macromolecules 2010, 43, 4020–4022; b) S. Zhang,
Y. Zhao, J. Am. Chem. Soc. 2010, 132, 10642–10644; c) X. Li, Y. Zhao, Bio-
conjugate Chem. 2012, 23, 1721–1725; d) H.-Q. Peng, Y.-Z. Chen, Y.
Zhao, Q.-Z. Yang, L.-Z. Wu, C.-H. Tung, L.-P. Zhang, Q.-X. Tong, Angew.
Chem. Int. Ed. 2012, 51, 2088–2092; Angew. Chem. 2012, 124, 2130–
2134; e) S. Zhang, Y. Zhao, Chem. Commun. 2012, 48, 9998–10000; f) Y.-
Z. Chen, P.-Z. Chen, H.-Q. Peng, Y. Zhao, H.-Y. Ding, L.-Z. Wu, C.-H. Tung,
Q.-Z. Yang, Chem. Commun. 2013, 49, 5877–5879; g) G. Chadha, Y.
Zhao, Org. Biomol. Chem. 2013, 11, 6849–6855.
[11] a) J. K. Awino, Y. Zhao, J. Am. Chem. Soc. 2013, 135, 12552–12555;
b) J. K. Awino, Y. Zhao, Chem. Commun. 2014, 50, 5752–5755.
[12] G. V. Oshovsky, D. N. Reinhoudt, W. Verboom, Angew. Chem. Int. Ed.
2007, 46, 2366–2393; Angew. Chem. 2007, 119, 2418–2445.
[13] a) A. Ben-Naim, Hydrophobic interactions, Plenum Press, New York,
1980; b) C. Tanford, The Hydrophobic Effect: Formation of Micelles and
Biological Membranes, Krieger, Malabar, Fla., 1991; c) W. Blokzijl,
J. B. F. N. Engberts, Angew. Chem. Int. Ed. Engl. 1993, 32, 1545–1579;
Angew. Chem. 1993, 105, 1610–1650.
[26] a) X. Li, Y. Zhao, Langmuir 2012, 28, 4152–4159; b) G. Chadha, Y. Zhao,
Chem. Commun. 2014, 50, 2718–2720.
[14] The cross-linking chemistry and covalent structure of the SCMs have
been previously characterized by mass spectrometry and TEM, see
Ref. 10a for details.
Received: August 19, 2014
Published online on && &&, 0000
Chem. Eur. J. 2014, 20, 1 – 8
www.chemeurj.org
7
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
FULL PAPER
&
Molecular Recognition
J. K. Awino, Y. Zhao*
&& – &&
Water-Soluble Molecularly Imprinted
Nanoparticles (MINPs) with Tailored,
Functionalized, Modifiable Binding
Pockets
Protein receptor mimics: Covalent im-
printing within surface-cross-linked mi-
celles in combination with a photode-
protection strategy yielded water-solu-
ble molecularly imprinted nanoparticles
with a single carboxylic acid group in
the binding pocket. The acid-functional-
ized imprinted nanoparticles had excel-
lent molecular recognition properties
and can be modified covalently in the
binding pocket.
&
&
Chem. Eur. J. 2014, 20, 1 – 8
www.chemeurj.org
8
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!