Effect of Physical Properties of Nanogel Particles on the Kinetic

Dec 23, 2013 - Rate constants were quantified by analyzing time course of protein binding ... In living system, biomacromolecules such as enzymes cont...
2 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/Biomac

Effect of Physical Properties of Nanogel Particles on the Kinetic Constants of Multipoint Protein Recognition Process Masahiko Nakamoto, Yu Hoshino,* and Yoshiko Miura* Department of Chemical Engineering, Kyushu University, 744 Motooka, Nishi-ku, Fukuoka 819-0395, Japan S Supporting Information *

ABSTRACT: We report the effect of physical properties, such as flexibility and polymer density, of nanogel particles (NPs) on the association/dissociation rates constant (kon and koff) and equilibrium constants (Kd) of multipoint protein recognition process. NPs having different flexibilities and densities at 25 °C were synthesized by tuning cross-linking degrees and the volume phase transition (VPT) temperature. Rate constants were quantified by analyzing time course of protein binding process on NPs monitored by a quartz crystal microbalance (QCM). Both kon and koff of swollen phase NPs increased with decreasing cross-linking degree, whereas crosslinking degree did not affect kon and koff of the collapsed phase NPs, indicating that polymer density of NPs governs kon and koff. The results also suggest that the mechanical flexibility of NPs, defined as the Young’s modulus, does not always have crucial roles in the multipoint molecular recognition process. On the other hand, Kd was independent of the cross-linking degree and depended only on the phase of NPs, indicating that molecular-scale flexibility, such as side-chain and segmental-mode mobility, as well as the conformation change, of polymer chains assist the formation of stable binding sites in NPs. Our results reveal the rationale for designing NPs having desired affinity and binding kinetics to target molecules.

■

INTRODUCTION Synthetic polymer nanogel particles (NPs) with intrinsic affinity to target biomacromolecules are of significant interest as robust and mass-produced substitutes for antibodies.1 It has been demonstrated that poly(N-isopropylacrylamide) (PNIPAm) NPs synthesized with an optimized combination and feed ratio of functional monomers can capture target proteins2−4 and peptides5−7 thru a combination of hydrophobic, electrostatic, and hydrogen bonding interactions. NPs with greater affinity and specificity to target peptides can further be synthesized by molecular imprint polymerization8−10 and isolated by affinity purification.11 Recently, it has been demonstrated that multifunctional NPs and dendrimers having strong affinity with target molecules (Kd < 10−7 M) are capable of recognizing and neutralizing functions of the target even in the bloodstream of living animals.12−14 However, in in vivo experiments, the activities of NPs are significantly lower than those expected from the results of in vitro experiments.14 It was expected that the differences arose from competition between the designed functions and undesired events such as the creation of protein coronas on NPs and the metabolism of the particles.15−19 To utilize these NPs in a nonequilibrium system, association and dissociation kinetics, as well as the equilibrium constants of the target binding process, must be carefully engineered.14,16,20−22 For instance, to utilize NPs for target neutralization or imaging in living animals, NPs having a fast association rate and a slow dissociation rate with target molecules should be engineered. In applying NPs to © 2013 American Chemical Society

purification, separation, and catalytic conversion processes, not only fast association but also fast dissociation response to environmental changes must be engineered. In living system, biomacromolecules such as enzymes control the binding kinetics as well as equilibrium constant of the target binding process via utilizing chain mobility and conformation changes of proteins.23−25 For instance, large conformation changes of DNA polymerase accelerate the association rate and decelerate the dissociation rate of the DNA capture process, resulting in the formation of a stable DNA−protein complex (“induced-fit” mechanism).26 Recently, it has also been reported that the disordered “flexible” domain in proteins has a crucial role in the determination of the target association rate: Large mobility of the disordered polypeptide chain enhances the association rate constant by diversifying structures that are in contact with target molecules (“fly-casting” mechanism).27−29 Recently, we reported that protein-binding kinetics on synthetic polymer nanoparticles can be controlled by tuning flexibility and inducing conformation changes of polymer chains.30 However, contributions of each factor on the kinetic constants were not clarified, since physical properties, such as polymer density, mechanical flexibility, and molecular scale flexibility, of NPs were not distinguished. In this paper, we aimed to generalize the effect of each physical property of Received: October 16, 2013 Revised: December 23, 2013 Published: December 23, 2013 541

dx.doi.org/10.1021/bm401536v | Biomacromolecules 2014, 15, 541−547

Biomacromolecules

Article

Scheme 1. (a) Preparation of ManNPs; (b) Illustration of Interaction between ManNPs and Con A; (c) Illustration of ManNP1−6

Figure 1. Hydrodynamic diameter of ManNPs synthesized with (a) 40% TBAm and 2% BIS (ManNP1; open triangle) or 10% BIS (ManNP2; closed triangle); (b) 20% TBAm and 2% BIS (ManNP3; open square) or 10% BIS (ManNP4; closed square); and (c) 0% TBAm and 2% BIS (ManNP5; open circle) or 10% BIS (ManNP6; closed circle) as a function of temperature. QCM Measurement. A 27 MHz quartz crystal microbalance (QCM) system (Affinix Q4, Initium Inc.) was used to monitor and quantify interactions between the ManNPs and concanavalin A (Con A).30 Immobilization of Con A on the Surface of the QCM Sensor. Con A was immobilized on the QCM electrode via an amine coupling procedure. Gold electrodes were cleaned twice with piranha solution (fresh mixture of H2O2 (aq) and H2SO4; 30% H2O2 (aq)/H2SO4 = 1:3 (v/v)) for 5 min. Next, 3,3′-dithiodipropionic acid (0.2 mL, 1 mM in Milli-Q water) was loaded into the QCM cells and then incubated for more than 30 min. The resulting cells were washed with Milli-Q water and then the carboxylic acid groups on the gold surface were activated for 30 min by loading 0.1 mL of a 1:1 v/v aqueous solution of 1-ethyl3-(3-dimethylaminopropyl)carbodiimide (100 mg·mL−1) and Nhydroxysuccinimide (100 mg·mL−1) to form N-hydroxysuccinimidyl (NHS) esters. After rinsing the activated cells with Milli-Q water several times, 10 mM phosphate buffer (pH 7.4, 137 mM NaCl, 2.68 mM KCl, 1.8 mM CaCl2, and 0.49 mM MgCl2, 500 μL) was added to the cells and the cells were set on the QCM system. After stabilization of the oscillation frequency, Con A (concentration in the cell after injection: 600 nM) was injected on the activated QCM electrodes. Immobilization of ManNPs on the QCM Sensor. After rinsing the Con A-immobilized cells to remove the nonimmobilized Con A

polymer NPs on association and dissociation rate constants and equilibrium constants of the protein recognition. In general, the cross-linking degree and the coil−globule phase transition of PNIPAm affect the flexibility and polymer density of NPs. Hence, in this study, NPs having different cross-linking degrees and positioned in various states in phase transition behavior were prepared and protein binding kinetics to these NPs was analyzed. In addition, we discuss the effect of flexibility, including mechanical flexibility, which is defined as the Young’s modulus, and molecular-scale flexibility, which is defined as the mobility of polymer chains, as well as polymer density of NPs, on the association and dissociation kinetics and equilibrium of target molecules with NPs.

■

EXPERIMENTAL SECTION

Synthesis of p-Acrylamidophenyl-α-D-mannopyranoside (Man) Monomer. Man was synthesized in accordance with our previous study.30,31 Preparation of Man-Incorporated NPs (ManNPs). A pseudoprecipitation polymerization method was employed for the preparation of ManNPs (SI).30,32 542

dx.doi.org/10.1021/bm401536v | Biomacromolecules 2014, 15, 541−547

Biomacromolecules

Article

Table 1. Monomer Feed, Swelling Ratio, Phase Condition, Man Density, and Kinetic/Equilibrium Constants of ManNP NPs

TBAm feed (mol %)

BIS feed (mol %)

Man feed (mol %)

swelling ratio (d25°C/d50°C)

ManNP1

40

2

0.1

1.0

ManNP2

40

10

0.1

1.0

ManNP3 ManNP4 ManNP5

20 20 0

2 10 2

1.7 1.5 9.0

3.1 2.4 4.5

ManNP6

0

10

1.5

2.3

state at 25 °C collapsed phase collapsed phase transition transition swollen phase swollen phase

using the 10 mM phosphate buffer, the cells were filled with the buffer (500 μL) and set on the QCM system. After stabilization of the oscillation frequency, ManNPs (concentration in the cell after injection: 4 μg·mL−1) were repeatedly injected into the cells until saturation of the frequency change was achieved.

density of Man [(10 nm)−3]

kon (103 M−1·s−1)

koff (10−3·s−1)

Kd (nM)

1.1

9

1.2

133

0.9

8

1.2

150

0.6 0.9 0.9

39 21 26

1.4 0.8 1.9

36 38 73

1.1

18

1.4

78

NPs, respectively. Volume densities of Man in each ManNPs at 25 °C were calculated to be 0.9 ± 0.3 molecules per 10 nm3 NPs from the polymer density of NPs and incorporation of Man quantified by 1H NMR data (Table 1).30 Thus, ManNPs having identical volume density, different phase conditions (collapsed, transition, and swollen), and different cross-linking degrees (2 or 10 mol % cross-linked) were prepared (Scheme 1c, ManNPs1−6). Setup of Experimental Conditions to Monitor the Time Course of Con A Binding on ManNPs. ManNPs were immobilized on the Au surface of 27 MHz quartz crystal microbalance (QCM) sensors. Time courses of Con A binding process in phosphate buffered saline (pH 7.4, 137 mM NaCl, 2.6 mM KCl, 1.8 mM CaCl2, and 0.49 mM MgCl2) were monitored at 25 °C. To analyze the effect of flexibility and polymer density of ManNPs on the association and dissociation rate constants of the Con A−ManNP interaction, it is important that nonspecific interaction such as Con A−Au sensor interaction and Con A−NP (which does not have Man) interaction are negligible. To prevent nonspecific interaction between Con A and the Au sensor, we used Con A as blocking agent as well as ligand to immobilize ManNPs on the sensor: Con A was immobilized on the surface of the sensor until saturation via amine-coupling reaction prior to immobilizing ManNPs. Subsequently, ManNPs were immobilized on the Con A-saturated surface via mannose−Con A interaction. Approximately 2000 Hz frequency decrease was observed as a response to injection of 600 nM Con A in a QCM cell, in which the sensor surface was activated by succinimide ester. In contrast, no frequency decreasing was observed when 600 nM Con A was injected in a QCM cell in which the sensor was saturated by Con A (Figure 2a). These results indicate that the nonspecific interaction between Con A and the Au sensor was negligible up to 600 nM. The binding specificity between ManNPs and Con A was investigated by two independent experiments. First, when ManNPs (ManNP4) were injected into the QCM cells on which Con A had been immobilized, an approximately 2000 Hz frequency decrease was observed. In contrast, no interaction was observed when the NPs without Man (same phase and cross-linking degree to ManNP4 except for without Man) were injected (Figure 2b). A frequency decrease was also observed when ManNPs (ManNP2) were injected on the Con A surface and no interaction was observed when the NPs were without Man (same phase and cross-linking degree to ManNP2). These results indicate that the interaction between Con A and ManNPs arises only from the specific Con A−Man interaction and that other regions in ManNPs are inert to Con A in the experimental conditions.

■

RESULTS AND DISCUSSION Design and Preparation of Nanogel Particles. As a model of target molecules, Con A, a 104 kDa α-D-mannose binding protein was selected. It is a homotetrameric protein having four mannose-binding sites on the corners of its tetrahedral structure, and each side is 10 nm long.33 Man monomer was synthesized as described30,31 and copolymerized into PNIPAm NPs (Scheme 1a) to prepare ManNPs that recognize Con A through multipoint interactions (Scheme 1b). To reveal the effects of phase and cross-linking degree of NPs on the target binding process, volume density of the mannose groups in each NP must be equal.22 It has been reported that incorporation of Man in NPs is proportional to the feed ratio of Man in the polymerization process.30 It has also been reported that the polymer density of collapsed NPs are independent of cross-linking degree and VPTT.34 Thus, NPs having different cross-linking degrees and phases at 25 °C were first prepared without Man to determine the appropriate feed ratio of Man to yield NPs having equal Man density (Supporting Information).30 On the basis of those preliminary results, we carefully tuned the feed ratio of Man in each NP to yield NPs having equal volume density of Man of around 1.0 molecules per 10 nm3 NPs at 25 °C. The volume phase transition temperature (VPTT) of NPs were tuned by incorporating a hydrophobic monomer N-tertbutylacrylamide (TBAm)) into PNIPAm NPs.32 Effects of cross-linking degree (2 mol % and 10 mol %) on the temperature-dependent VPT behavior of ManNPs that were copolymerized with 40, 20, and 0 mol % TBAm, are shown in Figure 1. It should be noted that incorporation of Man in NPs did not affect swelling behavior of NPs significantly. As expected, NPs with 40, 20, and 0 mol % TBAm showed VPTT below 15 °C, around 25 °C, and above 35 °C, respectively.30,32 Thus, we defined NPs that are copolymerized with 40, 20, and 0 mol % TBAm as collapsed NPs, transition NPs, and swollen NPs, respectively. Swelling ratios of the NPs were defined as diameter at 25 °C/diameter at 50 °C. The swelling ratio of collapsed NPs was independent of cross-linking degree (Figure 1a). However, the swelling ratio of transition NPs and swollen NPs decreased from 3.1 to 2.3 and from 4.5 to 2.4, respectively, with increasing cross-linking degree from 2 to 10 mol % (Figure 1b,c), as reported previously.34 We defined NPs that are copolymerized with 2 and 10% BIS as low cross-linked and high cross-linked 543

dx.doi.org/10.1021/bm401536v | Biomacromolecules 2014, 15, 541−547

Biomacromolecules

Article

Since the amount of Con A in the solution is in large excess compared with that of the binding sites on the surface of the QCM, eq 2 is rewritten as eq 3 by defining τ in eq 4. The amount of Con A/binding-site complex at time t after injection is given as eq 5 by integrating eq 3 ([C]0 and [B]0 are initial concentrations of Con A and binding sites in NPs, respectively). In the QCM measurement, the time courses of [C·B] can be monitored as the frequency change (ΔF). Thus, τ can be obtained by fitting the time course of ΔF via Con A binding with the single exponential function (eq 5) as binding relaxation time. Association and dissociation rate constants are given as the slope and y-intercept of the linear correlations between Con A concentration and the reciprocal of τ (eq 4), respectively. The dissociation equilibrium constant is given as koff/kon.

Figure 2. Time courses of QCM frequency change. (a) Con A (600 nM) was injected on the surface of succinimide ester (red) and Con A saturated surfaces (gray); (b) 4, 8, and 24 μg/mL of high cross-linked (10% BIS) transition state NPs with (green, ManNP4) and without Man (gray) were injected on the Con A immobilized surface at each arrow. (c) Con A (240 nM) was injected on the ManNP4 immobilized surface in the presence of 0 (red), 1 (light green), 10 (blue), and 100 mM (gray) Man. (d) Time course of QCM frequency changes, normalized by frequency changes at 30 min, corresponding to the addition of 5 (gray), 10 (green), 30 (black), 60 (blue), 90 (orange), 120 (light blue), and 240 nM (red) Con A on the surface of ManNP4. All of the QCM experiments in this study were performed at 25 °C, 10 mM phosphate buffer (pH 7.4, 137 mM NaCl, 2.68 mM KCl, 1.8 mM CaCl2, and 0.49 mM MgCl2).

kon

koff

d [C·B] = kon[C ][B] − koff [C·B] dt

(3)

1 = kon[C ]0 + koff τ

(4)

⎛ ⎛ −t ⎞⎞ k [C ] [B] [C·B] = ⎜1 − exp⎜ ⎟⎟ on 0 0 ⎝ τ ⎠⎠ kon[C ]0 + koff ⎝

(5)

The observed time course of Con A binding could be fitted very well by the single exponential function (Figure 2d). Reciprocals of the binding relaxation time (τ), acquired by the exponential function of each of the binding processes, were proportional to the concentration of Con A (Figure 3). These results indicate that the binding process can be approximated as the Langmuir-type binding process, although results of affinity purification experiments have indicated that randomly functionalized NPs have heterogeneous binding sites and affinities to target molecules.11 Thus, we analyzed the kinetics of binding of Con A on ManNPs by the Langmuir model in this study. kon, koff, and Kd obtained from the analysis are listed in Table 1. Mode of Con A Recognition Observed on ManNPs. Equilibrium constants of Con A binding to ManNPs1−6 were in the submicromolar range (Table 1). This is more than 2−3 orders of magnitude smaller than the reported value for monovalent mannose-Con A interaction,35 indicating that all Con A binding processes observed in this study were multipoint interactions between more than two mannose molecules in ManNPs and one Con A molecule. Since the volume density of the recognition point in all ManNPs are equal, the difference in kinetic and equilibrium constants can be interpreted as a result of accessibility (steric hindrance) around binding sites and the probability of polymer chains to form a stable binding structure before and after contact with Con A. Effect of Cross-Linking Degree of NPs to the Multipoint Con A Recognition Process. The association and the dissociation rate constants of low cross-linked and collapsed ManNP1 (kon = 9 M−1 s−1 and koff = 1.2 s−1) and high cross-linked and collapsed ManNP2 (kon = 8 M−1 s−1 and koff = 1.2 s−1)) were almost equal (Figure 3a,b), resulting in equal equilibrium constants (Kd = 133 and 150 nM, respectively; Table 1). In contrast, both association and dissociation rate constants of low cross-linked transition ManNP3 (kon = 39 M−1 s−1 and koff = 1.4 s−1)) were 1.7fold larger than those of high cross-linked and transition ManNP4 (kon = 21 M−1 s−1 and koff = 0.8 s−1); Figure 3c,d).

Second, when Con A (240 nM) was injected into the cells in which ManNP4 had been immobilized, an approximately 1500 Hz frequency decrease was observed. However, the interaction was almost completely inhibited by Man monomer (Figure 2c). These results indicate that the interaction between Con A and ManNPs arises from the specific Con A−Man interaction and that other regions in ManNPs are inert to Con A in the experimental conditions. Analysis of Con A Binding to NPs. Con A (5−240 nM) was injected into the cells in which ManNPs were immobilized and Con A−ManNPs interactions were monitored for 30 min. Figure 2d shows the typical frequency changes as a function of time. Time course of frequency changes were normalized by frequency changes at 30 min. It is obvious that the association rate of Con A depended on the concentration of Con A. If the binding event can be approximated as the Langmuirtype absorption defined as eq 1, the increasing amount of Con A/binding site complex per unit time is given by eq 2 ([C], [B], and [C·B] are the concentrations of Con A, binding-site in NPs, and Con A/binding site complex, respectively). Con A + binding site XooY Con A·binding site

k [C ] [B] ⎞ d −1 ⎛ [C·B] = ⎜[C·B] − on 0 0 ⎟ dt kon[C ]0 + koff ⎠ τ ⎝

(1) (2) 544

dx.doi.org/10.1021/bm401536v | Biomacromolecules 2014, 15, 541−547

Biomacromolecules

Article

higher association rate of swollen NPs (26 and 18 M−1 s−1, respectively) compared with that of collapsed NPs (9 and 8 M−1 s−1, respectively). Equilibrium constants of transition NPs (ManNPs3, 36 nM, and ManNPs4, 38 nM; Table 1) were 50% smaller than that of swollen NPs, as listed above. This was a result of faster association of transition NPs (39 and 21 M−1 s−1, respectively) compared to that of swollen NPs (26 and 18 M−1 s−1, respectively) and slower dissociation of transition NPs (1.4 and 0.8 s−1, respectively) compared to that of swollen NPs (1.9 and 1.4 s−1, respectively). In summary, equilibrium constants of Con A to ManNPs are independent of cross-linking degree and are controlled solely by the phase of ManNPs. Swollen NPs bind to Con A more strongly than do collapsed NPs, since swollen NPs capture Con A faster than do collapsed ones. Transition NPs bind to Con A far more strongly than do swollen NPs because transition NPs capture faster and release more slowly than do swollen NPs. Effect of Mechanical Flexibility of NPs to the Multipoint Con A Recognition Process. It has been reported that the Young’s modulus of both collapsed and swollen PNIPAm particles increases (get less flexible) with increasing cross-linking degree.36 However, kinetic and equilibrium constants are not always dependent on the crosslinking degree: kon, koff, and Kd of collapsed ManNPs are independent of cross-linking degree, indicating that the mechanical flexibility of nanogel particles, defined as the Young’s modulus, does not always have crucial roles in the multipoint molecular recognition process. Effect of Polymer Density of NPs to the Multipoint Con A Recognition Process. It has been reported that the polymer density of collapsed phase NPs are independent of the cross-linking degree, but that of swollen and transition phase NPs increases with increasing cross-linking degree.34 Our results revealed that kinetic constants could be tuned by the cross-linking degree only when the polymer density of ManNPs was altered by the cross-linking degree: The association and dissociation rate constants of transition and swollen ManNPs can be increased by lowering the polymer density of NPs (by lowering the cross-linking degree), whereas that of collapsed ManNPs cannot be tuned by cross-linking degree. It has been reported that diffusion rates of molecules in hydrogels increase with decreasing the polymer density as a result of increasing free volume in the gel network.37−39 It has also been reported that diffusion rates are decreased by weak interactions between diffusing molecules and polymers forming the hydrogel network.40,41 However, little has been reported about relationship between polymer density and association/ dissociation rate constants and equilibrium constants. Our results clearly suggest that the polymer density of NPs affects only on the exchange rate of Con A binding to the sites on NPs without affecting the equilibrium constants. Effect of Molecular-Scale Flexibility of NPs to the Multipoint Con A Recognition Process. Our results revealed that swollen NPs interacted with Con A stronger than collapsed NPs regardless of cross-linking degree. It has been reported that PNIPAm chains in the swollen state are in a random coil conformation with large mobility of segments and side chains, whereas PNIPAm chains in the collapsed state are in the globule conformation with lower mobility of segments and side chains.42,43 The smaller Kd (strong affinity) of swollen NPs than that of collapsed NPs has been explained by the difference in the molecular-scale flexibility of PNIPAm chains in

Figure 3. Plots of binding relaxation time (τ) on collapsed ((a) ManNP1, low cross-linked, and (b) ManNP2, high cross-linked), transition ((c) ManNP3, low cross-linked, and (d) ManNP4, high cross-linked), and swollen ((e) ManNP5, low cross-linked, and (f) ManNP6, high cross-linked) NPs as a function of Con A concentration.

Interestingly, equilibrium constants of the transition NPs (Kd = 36 and 38 nM, respectively) were equal as a result of the same extent of increase in both association and dissociation rates (Table 1). Increase in the association and the dissociation rates were also shown in swollen NPs: both association and dissociation rate constants of low cross-linked and swollen ManNP5 (kon = 26 M−1 s−1 and koff = 1.9 s−1) were 1.4-fold larger than these of high cross-linked and swollen ManNP6 (kon = 18 M−1 s−1 and koff = 1.4 s−1; Figure 3e,f). Equilibrium constants of swollen NPs (Kd = 73 and 78 nM, respectively) were equal as a result (Table 1). From these results, we concluded that equilibrium constant is independent of cross-linking degree, and depends only on the phase of ManNPs. The exchange rate of transition and swollen ManNPs can be increased by lowering the cross-linking degree, whereas that of collapsed ManNPs cannot be tuned by crosslinking degree. Effect of Phase of NPs to the Multipoint Con A Recognition Process. Equilibrium constants of swollen NPs (ManNPs5, 73 nM, and ManNPs6, 78 nM; Table 1) are about 50% smaller than that of collapsed NPs (ManNPs1, 133 nM, and ManNPs2, 150 nM; Table 1). This was a result of the 545

dx.doi.org/10.1021/bm401536v | Biomacromolecules 2014, 15, 541−547

Biomacromolecules

■

NPs:30 the molecular-scale mobility of the PNIPAM chains expands the diversity of polymer structure, allowing Man on the polymer chains to map onto mannose-binding sites on Con A more rapidly than the do the collapsed NPs, resulting in the formation of a stable Con A−ManNP complex. Effect of phase of NPs on the equilibrium constants might be comparable to the decreasing of equilibrium binding constant and/or initial binding rate of Con A to glycol-polymers with lowering of mobility of binding sites attached to polymers.22,44,45 Increasing kon of NPs in random coil conformation might also be comparable to the acceleration of association rate in intrinsically unstructured proteins through the “fly-casting” mechanism.28,46 Transition NPs bind to Con A far more strongly than do swollen NPs regardless of cross-linking degree. This result has been explained by the “induced-fit” type mechanism.30 It has been observed that the volume of ManNPs gradually decreased to ∼40% by binding to Con A. Shrinkage of the particles results in an increase in polymer density (steric hindrance) around the binding sites on the polymer chains in the transition-state NPs after Con A binding, resulting in a decrease in the dissociation rate. We did not observe allosteric binding behavior from the QCM charts of transition ManNPs in this study. This suggests that shrinkage of ManNPs induced by Con A binding was not cooperative shrinkage of NPs but local structural change of PNIPAm chains around the Con A binding site as a result of polymer flexibility.

■

Article

ASSOCIATED CONTENT

S Supporting Information *

Detailed preparation and characterization procedure of ManNPs are described. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]; [email protected]. Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS Financial support from MEXT (23111716 and 25107726) and NEDO (11B10002d) is greatly appreciated. REFERENCES

(1) Haag, R.; Kratz, F. Angew. Chem., Int. Ed. 2006, 45, 1198−1215. (2) Lee, S. H.; Hoshino, Y.; Randall, A.; Zeng, Z.; Baldi, P.; Doong, R.-A.; Shea, K. J. J. Am. Chem. Soc. 2012, 134, 15765−15772. (3) Yonamine, Y.; Yoshimatsu, K.; Lee, S. H.; Hoshino, Y.; Okahata, Y.; Shea, K. J. ACS Appl. Mater. Interfaces 2013, 5, 374−379. (4) Yoshimatsu, K.; Lesel, B. K.; Yonamine, Y.; Beierle, J. M.; Hoshino, Y.; Shea, K. J. Angew. Chem., Int. Ed. 2012, 51, 2405−2408. (5) Hoshino, Y.; Urakami, T.; Kodama, T.; Koide, H.; Oku, N.; Okahata, Y.; Shea, K. J. Small 2009, 5, 1562−1568. (6) Hoshino, Y.; Koide, H.; Furuya, K.; Haberaecker, W. W.; Lee, H.; Oku, N.; Shea, K. J. Proc. Natl. Acad. Sci. U.S.A. 2012, 109, 33−38. (7) Cabaleiro-Lago, C.; Quinlan-Pluck, F.; Lynch, I.; Lindman, S.; Minogue, A. M.; Thulin, E.; Walsh, D. M.; Dawson, K. A.; Linse, S. J. Am. Chem. Soc. 2008, 130, 15437−15444. (8) Wulff, G. Angew. Chem., Int. Ed. 1995, 34, 1812. (9) Hoshino, Y.; Kodama, T.; Okahata, Y.; Shea, K. J. J. Am. Chem. Soc. 2008, 130, 15242−15243. (10) Zeng, Z.; Hoshino, Y.; Rodriguez, A.; Yoo, H.; Shea, K. J. ACS Nano 2010, 4, 199−204. (11) Hoshino, Y.; Haberaecker, W. W., III.; Kodama, T.; Zeng, Z.; Okahata, Y.; Shea, K. J. J. Am. Chem. Soc. 2010, 132, 13648−13650. (12) Dernedde, J.; Rausch, A.; Weinhart, M.; Enders, S.; Tauber, R.; Licha, K.; Schirner, M.; Zügel, U.; Bonim, A. V.; Haag, R. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 19679−19684. (13) Hoshino, Y.; Koide, H.; Urakami, T.; Kanazawa, H.; Kodama, T.; Oku, N.; Shea, K. J. J. Am. Chem. Soc. 2010, 132, 6644−6645. (14) Hoshino, Y.; Koide, H.; Furuya, K.; Haberaecker, W. W.; Lee, H.; Oku, N.; Shea, K. J. Proc. Natl. Acad. Sci. U.S.A. 2012, 109, 33−38. (15) Cedervall, T.; Lynch, I.; Foy, M.; Berggard, T.; Donnelly, S. C.; Cagney, G.; Linse, S.; Dawson, K. A. Angew. Chem. 2007, 119, 5856− 5858. (16) Cedervall, T.; Lynch, I.; Lindman, S.; Berggard, T.; Thulin, E.; Nilsson, H.; Dawson, K. A.; Linse, S. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 2050−2055. (17) Sun, X.; Rossin, R.; Turner, J. L.; Becker, M. L.; Joralemon, M. J.; Welch, M. J.; Wooley, K. L. Biomacromolecules 2005, 6, 2541−2554. (18) Owens, D. E., III.; Peppas, N. A. Int. J. Pharm. 2006, 307, 93− 102. (19) Monopoli, M. P.; Åberg, C.; Salvati, A.; Dawson, K. A. Nat. Nanotechnol. 2012, 7, 779−786. (20) Hoshino, Y.; Imamura, K.; Yue, M.; Inoue, G.; Miura, Y. J. Am. Chem. Soc. 2012, 134, 18177−18180. (21) Yue, M.; Hoshino, Y.; Imamura, K.; Miura, Y. Angew. Chem., Int. Ed. 2013, DOI: 10.1002/anie.201309758. (22) Kiessling, L. L.; Grim, J. C. Chem. Soc. Rev. 2013, 42, 4476− 4491. (23) Mittal, S.; Cai, Y.; Nalam, M. N. L.; Bolon, D. N. A.; Schiffer, C. A. J. Am. Chem. Soc. 2012, 134, 4163−4168.

CONCLUSIONS

In the present study, we clarified the effect of flexibility and polymer density of NPs on the multipoint target binding process. We achieved this by breaking up flexibility into mechanical flexibility and molecular-scale flexibility. Reducing or increasing the polymer density by tuning the cross-linking degree resulted in increasing or decreasing the association− dissociation exchange rate, respectively, without changing the equilibrium constant Kd. Mechanical flexibility did not always affect on molecular recognition kinetics and equilibrium. The change in molecular-scale flexibility induced by coil−globule phase transition dominated the equilibrium constant. The increase in kon may arise from the increasing molecular-scale mobility of polymer chains in swollen NPs. Transition NPs showed far stronger affinity to targets as a result of greater structural change of polymers induced by target binding. Our results provide strategies for designing NPs with desired binding kinetics and equilibrium for various applications. For instance, in the case of target neutralization and imaging in living animals, both fast association and slow dissociation are required; thus, transition NPs are suitable. In contrast, if fast turnover (both high association rate and high dissociation rate) is required in drug delivery, separation, or catalysis processes, the exchange rate has to be accelerated by decreasing the polymer density. Results of this study provide not only the rationale for the preparation of strong plastic antibodies in the test tube, but also the comprehensive rationale for the preparation of highly active plastic antibodies in dynamic systems. 546

dx.doi.org/10.1021/bm401536v | Biomacromolecules 2014, 15, 541−547

Biomacromolecules

Article

(24) Kosugi, T.; Hayashi, S. J. Am. Chem. Soc. 2012, 134, 7045−7055. (25) Schreiber, G.; Haran, G.; Zhou, H.-X. Chem. Rev. 2009, 109, 839−859. (26) Li, Y.; Korolev, S.; Waksman, G. EMBO J. 1998, 17, 7514−7524. (27) Wright, P. E.; Dyson, H. J. J. Mol. Biol. 1999, 293, 321−331. (28) Prakash, M. K. J. Am. Chem. Soc. 2011, 133, 9976−9979. (29) Shoemaker, B. A.; Portman, J. J.; Wolynes, P. G. Proc. Natl. Acad. Sci. U.S.A. 2000, 97 (16), 8868−8873. (30) Hoshino, Y.; Nakamoto, M.; Miura, Y. J. Am. Chem. Soc. 2012, 134, 15209−15212. (31) Seto, H.; Ogata, Y.; Murakami, T.; Hoshino, Y.; Miura, Y. ACS Appl. Mater. Interfaces 2012, 4, 411−417. (32) Debord, J. D.; Lyon, L. A. Langmuir 2003, 19, 7662−7664. (33) Becker, J. W.; Reeke, G. N.; Cunningham, B. A.; Edelman, G. M. Nature 1976, 259, 406−409. (34) Varga, I.; Gilányi, T.; Mészáros, R.; Filipcsei, G.; Zrínyi, M. J. Phys. Chem. B 2001, 105, 9071−9076. (35) Mann, D. A.; Kanai, M.; Maly, D. J.; Kiessling, L. L. J. Am. Chem. Soc. 1998, 120 (41), 10575−10582. (36) Burmistrova, A.; Richter, M.; Uzum, C.; Klitzing, R. V. Colloid Polym. Sci. 2011, 289, 613−624. (37) Bae, Y. H.; Okano, T.; Hsu, R.; Kim, S. W. Makromol. Chem. Rapid Commun. 1987, 8, 481−485. (38) Okano, T.; Bae, Y. H.; Jacobs, H.; Kim, S. W. J. Controlled Release 1990, 11, 255−265. (39) Amsden, B. Macromolecules 1998, 31, 8382−8395. (40) Lynch, I.; Dawson, K. A. J. Phys. Chem. B 2004, 108, 10893− 10898. (41) Kim, S. W.; Bae, Y. H.; Okano, T. Pharm. Res. 1992, 9, 283− 290. (42) Tokuhiro, T.; Amiya, T.; Mamada, A.; Tanaka, T. Macromolecules 1991, 24, 2936−2943. (43) Chee, C. K.; Rimmer, S.; Soutar, I.; Swanson, L. Polymer 2001, 42, 5079−5087. (44) Ooya, T.; Utsunomiya, H.; Eguchi, M.; Yui, N. Bioconjugate Chem. 2005, 16, 62−69. (45) Hasegawa, T.; Kondoh, S.; Matsuura, K.; Kobayashi, K. Macromolecules 1999, 32, 6595−6603. (46) Rogers, J. M.; Steward, A.; Clarke, J. J. Am. Chem. Soc. 2013, 135, 1415−1422.

547

dx.doi.org/10.1021/bm401536v | Biomacromolecules 2014, 15, 541−547