Molecular Recognition and Immobilization of Ligand-Conjugated

Research Article www.acsami.org

Molecular Recognition and Immobilization of Ligand-Conjugated Redox-Responsive Polymer Nanocontainers Wilke C. de Vries, Matthias Tesch, Armido Studer, and Bart Jan Ravoo* Organic Chemistry Institute and Center for Soft Nanoscience, Westfälische Wilhelms-Universität Münster, Corrensstr. 40, D-48149 Münster, Germany S Supporting Information *

ABSTRACT: We present the preparation of ligand-conjugated redox-responsive polymer nanocontainers by the supramolecular decoration of cyclodextrin vesicles with a thin redox-cleavable polymer shell that displays molecular recognition units on its surface. Two widely different recognition motifs (mannose−Concanavalin A and biotin− streptavidin) are compared and the impact of ligand density on the nanocontainer surface as well as an additional functionalization with nonadhesive poly(ethylene glycol) is studied. Aggregation assays, dynamic light scattering, and a fluorometric quantification reveal that the molecular recognition of ligand-conjugated polymer nanocontainers by receptor proteins is strongly affected by the multivalency of interactions and the association strength of the recognition motif. Finally, microcontact printing is used to prepare streptavidin-patterned surfaces, and the specific immobilization of biotin-conjugated nanocontainers is demonstrated. As a prototype of a nanosensor, these tethered nanocontainers can sense a reductive environment and react by releasing a payload. KEYWORDS: nanocontainer, self-assembly, molecular recognition, surface design, stimulus-responsive, immobilization

■

INTRODUCTION The rational design of the surface decoration of nanoparticles allows a versatile modulation of interactions among nanoparticles and enables a tunable interaction with their environment. Such sophisticated nanostructures have found widespread applications including targeted delivery,1,2 sensing,3 and imaging.4 In general, two approaches can be used to engineer the surface decoration of nanoparticles. (1) Functional units can be included during the synthesis of the nanomaterials, as for example in the self-assembly of (endgroup-)functionalized copolymers.5,6 (2) The surface can be designed by postsynthetic approaches such as physical adsorption7−9 or by chemical conjugation.10,11 Among the various motifs for surface decoration, the functionalization of nanoparticles with ligands facilitates the recognition by proteins via ligand−receptor binding and results in a highly specific interaction of the nanoparticles with their environment. In nature, for example, the recognition and adhesion of cells as well as the infection of cells with viruses are mainly driven by carbohydrate−protein interactions. Proteins that bind to a densely arranged coating of polysaccharides on the cell surface (“glycocalyx”) are termed lectins and typically bind in a multivalent fashion, whereby the association constant Ka for a single carbohydrate−lectin complex usually is 103−104 M−1.12,13 To mimic and study these multivalent interactions, the self-assembly of carbohydrate-conjugated polymers, so-called glycopolymers, into nanostructures exhibiting multiple carbohydrate ligands on their surface and their multivalent interplay with lectins have gained increasing interest in the past.14−16 Among others, these © 2017 American Chemical Society

platforms have been utilized in drug delivery as a promising approach to target specific cells.9,17−19 Considering surface modifications where a highly robust and selective conjugation is desired, the recognition of the small vitamin biotin by the tetrameric protein streptavidin (STA) constitutes a widely used ligand−receptor pair.5,20 The very high association constant of Ka ∼ 1015 M−1 represents the strongest noncovalent biological interaction known and has resulted in extensive use in the field of nanodevices and surface functionalized nanostructures.21−23 Recently, we have introduced novel redox-responsive nanocontainers applying a supramolecular stabilization of vesicles by the self-assembly of a redox-responsive polymer shell around the vesicle templates. These disulfide cross-linked polymer shelled vesicles (PSVSS) can recognize a reducing microenvironment and react by releasing hydrophilic payloads, for example inside the reductive intracellular environment.24 Given the potential and significance of surface-functionalized responsive materials, we herein introduce a method for the surface functionalization of PSV SS . Importantly, PSV SS represent an ideal scaffold for the conjugation of bioactive ligands because their thin and flexible polymer shell is easily accessible. To this end, we carefully design the surface properties of PSVSS by decorating them with carbohydrate or biotin ligands as well as with short poly(ethylene glycol) (PEG) chains to further stabilize them. These nanocontainers should Received: October 12, 2017 Accepted: November 15, 2017 Published: November 15, 2017 41760

DOI: 10.1021/acsami.7b15516 ACS Appl. Mater. Interfaces 2017, 9, 41760−41766

Research Article

ACS Applied Materials & Interfaces

containers to the STA solution. By comparison of this value with the number of biotin-binding sites in the calibration measurement, the concentration of biotin at the surface of PSVSSBiotin was calculated. Preparation of Biotin-Patterned Surfaces by Microcontact Printing (μCP). Biotin-patterned surfaces were prepared following a recently published procedure.22 Briefly, silicon or glass substrates were cleaned by sonication in pentane, acetone, and ultrapure water. The cleaned substrates were immersed in a freshly prepared piranha solution (concentrated H2SO4/H2O2 (30%) = 3/1 (v/v)) for 30 min and extensively washed with ultrapure water and EtOH and dried in a stream of argon. These oxidized substrates were used to prepare an alkene self-assembled monolayer (SAM) by incubation in a solution of 7-octenyltrichlorosilane (0.1 vol %) in toluene for 40 min. The substrates were washed with dichloromethane and EtOH, dried, and used for μCP of biotin-SH. Therefore, poly(dimethylsiloxane) (PDMS) stamps were prepared by casting a 10:1 (v/v) mixture of poly(dimethylsiloxane) (Dow Corning) and curing agent (Sylgards 184, Dow Corning) on a patterned silicon master. After curing at 80 °C for 24 h, stamps were cut out and put into an UV ozonizer for 55 min directly prior to use. The stamps were incubated with biotin-SH ink (10 mM biotin-SH, 20 mM 2,2-dimethoxy-2-phenylacetophenone (DMPA) in MeOH) for 30 s before excess ink was removed in a stream of argon, and the stamps were carefully placed on the alkenemodified substrates. After the thiol-ene reaction was initiated photochemically using a high-power UV-light-emitting diode (LED), which was placed approximately 2 cm above the stamp for 5 min, the stamp was removed, and the substrate was rinsed with EtOH and dried. Finally, unreacted alkenes were saturated with mercaptoethanol by applying a solution of 3 wt % DMPA in mercaptoethanol to the substrate, which was then covered with a microscopy cover slide and irradiated by using the high-power UV-LED for 10 min. Upon washing with EtOH and drying, biotin-patterned surfaces were obtained. Immobilization of PSVSSBiotin. Biotin-patterned substrates were blocked with 3 wt % bovine serum albumin (BSA) in HBS for 30 min and washed with HBS afterward (2 × 5 min). Incubation with STA (50 μg mL−1) for 15 min and careful rinsing with HBS yielded STApatterned surfaces. The STA-patterned surfaces were incubated with PSVSSBiotin for 15 min and subsequently carefully and repeatedly washed with HBS. PSVSSBiotin with a maximum functionalization of 1% of the carboxylic acid groups by biotin-NH2 and a concentration of 20 μM β-cyclodextrin amphiphiles in HBS were used in the immobilization procedure. For rhodamine B-labeled PSVSSBiotin, 5 mol % of amphiphilic β-cyclodextrin was replaced by rhodamine Blabeled amphiphilic β-cyclodextrin. For carboxyfluorescein (CF) loaded PSVSSBiotin, the preparation of the nanocontainers was performed in HBS containing 100 μM CF. Payload Encapsulation and Release. CDV (100 μM amphiphilic β-cyclodextrin) were prepared in a buffered solution of the payload (self-quenching solution of 5 mM CF in HBS), and PSVSSBiotin were prepared by the above-described procedure. After cross-linking of the polymer shell, nonencapsulated dye and byproducts were removed by Sephadex G50 size-exclusion chromatography with HBS as the eluent. A fraction of ∼5.0 mL was collected and used for the immobilization on a small glass slide (8 mm × 15 mm surface area) by the above-described procedure (instead of the μCP step, the surface was completely functionalized by applying 20 μL of biotin-SH ink capped by a microscopy cover slide). The glass slide with immobilized PSVSSBiotin was placed in a cuvette with 10 mm path length, and 1.5 mL HBS was added to cover the immobilized PSVSSBiotin. The HBS was replaced four times to remove any residual nonimmobilized material. The time-dependent measurement of CF fluorescence (λex = 480 nm, λem = 520 nm) in the supernatant solution was started and 400 μM tris(2-carboxyethyl)phosphine (TCEP) was added when indicated. Release at time t was calculated by dividing the fluorescence intensity at time t by the intensity after a complete release.

be specifically recognized by corresponding proteins and the effect of the different recognition motifs and ligand densities is investigated. Finally, the surface-functionalized nanocontainers are assembled into patterned redox-responsive surface coatings, which can release a payload upon applying a reductive trigger.

■

EXPERIMENTAL SECTION

Materials Synthesis. The synthesis and characterization of adamantane-terminated poly(acrylic acid) (Ad-PAA, degree of polymerization ≈ 119, Mw/Mn = 1.2), amphiphilic β-cyclodextrin, and amine-functionalized ligands (biotin-NH2, mannose-NH2) is described in detail the Supporting Information (SI). Preparation of Cyclodextrin Vesicles (CDV). Unilamellar bilayer CDV were prepared by hydration of a thin film and extrusion. Briefly, a 2 mM stock solution of amphiphilic β-cyclodextrin in chloroform was added in a round-bottom flask and the solvent was evaporated in a stream of argon to obtain a thin film. The residual solvent was removed under high vacuum. The film was hydrated by the addition of 4-(2-hydroxyethyl)piperazine-1-ethanesulfonic acid (HEPES) buffered saline (HBS, 20 mM HEPES, 150 mM NaCl, pH 7.4) and vigorous stirring for at least 2 h. This solution was sonicated for 15 s and repetitively passed through a polycarbonate membrane with 100 nm pore size in a Liposofast manual extruder (AVESTIN) to yield CDV. Preparation of Surface-Functionalized PSVSSLigand. To a buffered solution of CDV (100 μM amphiphilic β-cyclodextrin in HBS), 25.0 μM Ad-PAA (50% coverage of total β-cyclodextrin cavities at the outward surface of vesicles) was added and this mixture was gently stirred for 30 min to obtain polymer-decorated vesicles (PDV). For cross-linking and functionalization of the polymer shell, 9.00 mM 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC· HCl, corresponding to 3.0 equiv of total carboxylic acid groups at the PDV surface) was added. After 20 min, 600 μM cystamine (corresponding to 0.20 equiv of total carboxylic acid groups at the PDV surface) and, for ligand-conjugated PSVSSLigand, 30−120 μM ligand-NH2 (biotin-NH2 or mannose-NH2, corresponding to 0.01− 0.04 equiv of total carboxylic acid groups at the PDV surface) were added. For a PEG functionalization, 1.20 mM PEG-NH2 (0.40 equiv of total carboxylic acid groups at the PDV surface) was added 2 h after the addition of cystamine and ligands. The colloid was stirred slowly overnight, and the byproducts were removed by dialysis (Spectra/Por regenerated cellulose dialysis membranes, molecular weight cut-off 6− 8 kDa) against HBS (3× buffer exchange within 48 h) to yield PSVSSLigand (approximately 0.75 mg mL−1). Aggregation Assay. Aggregation assays were performed by the measurement of optical density at λ = 600 nm (OD600) collecting one data point every 5 s and dynamic light-scattering (DLS) measurements before and after each experiment. The measurement procedure was as follows. The OD600 of 700 μL PSVSSLigand (synthesized from 100 μM β-cyclodextrin amphiphiles in HBS) was measured for 5 min before 35 μL of streptavidin (2 mg mL−1, Sigma-Aldrich) or Concanavalin A (ConA, 2 mg mL−1, Sigma-Aldrich) were added to make a resulting protein concentration of 0.1 mg mL−1 and the measurement was continued for at least 30 min. For experiments with ConA, the HBS buffer medium was supplemented with 1 mM CaCl2 and 1 mM MnCl2. Fluorescence Titration of Biotin. The quantification of biotin conjugated to the surface of PSVSSBiotin was performed following a method introduced by Parce and co-workers.25 Briefly, a solution of 8 mg L−1 streptavidin (STA) in HBS was prepared. For calibration to 1.5 mL of this STA solution, biotin (2.5 μM) was added in 5 μL increments and fluorescence of STA (λex = 300 nm, λem = 340 nm) was measured. The break point of this titration gives the number of biotin-binding sites in 1.5 mL STA solution. For the quantification of biotin at the surface of PSVSSBiotin, 15 μL of a 10-fold diluted nanocontainer solution was added to 1.5 mL of the STA solution. Titration curves of this mixtures with biotin (2.5 μM) are depicted in the SI in Figure S5. Here, the break point gives the number of freebiotin-biding sites after the addition of the functionalized nano41761


Research Article


Scheme 1. Schematic Representation of the Preparation of Ligand-Functionalized PSVSSLigand and Molecular Structures of Amine-Functionalized Ligands

Figure 1. (a) Size distribution according to DLS and (b) ζ-potentials of CDV, PDV, and PSVSSLigand functionalized with biotin-NH2, mannose-NH2, or PEG-NH2. CDV, PDV, and PSVSSLigand concentrations correspond to 100 μM β-cyclodextrin amphiphiles in HBS (20 mM HEPES, 150 mM NaCl, pH 7.4) and measurements were performed at 25 °C. ζ-Potential represents mean ± SD (n = 5).

■

RESULTS AND DISCUSSION The preparation of surface-functionalized nanocontainers starting from the supramolecular building blocks adamantaneterminated poly(acrylic acid) (Ad-PAA) and cyclodextrin vesicles (CDV) is depicted in Scheme 1. An important advantage of CDV in contrast to conventional liposomes is the fact that these dynamic platforms can selectively bind hydrophobic guest molecules, considerably simplifying the preparation of polymer nanocontainers.26 In a first step, AdPAA was self-assembled on the surface of bare CDV (hydrodynamic diameter dH = 120−150 nm) templates via a highly efficient host−guest recognition of β-cyclodextrin and adamantane-conjugated polymers (Ka ∼ 103 M−1)27 delivering polymer decorated vesicles (PDV). The subsequent 1-ethyl-3(3-dimethylaminopropyl)carbodiimide (EDC) mediated cross-

linking of the poly(acrylic acid) shell with cystamine linkers results in stable PSVSS and simultaneously allows for a facile decoration of the polymer shell with primary amine functionalized small molecules via amide bond formation. The ratio of diamine linker versus functionalized amine used for amide formation will allow for varying the density of the functionalities in the polymer layer. In comparison to physical methods for adsorbing ligands on nanocarrier surfaces, this covalent functionalization approach allows for a more robust and irreversible attachment of the ligands.10 Applying this single-step procedure, the nanocontainers were covalently functionalized with biotin-NH2 (PSVSSBiotin), mannose-NH2 (PSVSSMan), PEG-NH2 (Mn ≈ 750 g mol−1) (PSVSSPEG), or combinations thereof (PSVSSBiotin,PEG and PSVSSMan,PEG). 41762


Research Article


Figure 2. (a) Time-dependent OD600 and (b) DLS measurements monitoring the aggregation of PSVSSLigand in the presence of tetravalent binding proteins. In OD600 measurements, ConA or STA was added at t = 5 min. Concentrations of PSVSSLigand (2% maximum functionalization of carboxylic acid groups by biotin-NH2 or mannose-NH2) correspond to 100 μM β-cyclodextrin amphiphiles in HBS (20 mM HEPES, 150 mM NaCl, pH 7.4) supplemented with 1 mM CaCl2 and 1 mM MnCl2 for ConA recognition experiments, [STA] = [ConA] = 0.1 mg mL−1, measurements were performed at 25 °C.

of numerous noncovalent cross-links between the nanocontainers, thereby causing the aggregation and an increase in turbidity. Indeed, a spontaneous time-dependent increase in optical density at λ = 600 nm (OD600) was observed upon the addition of proteins to the biotin-conjugated or mannoseconjugated nanocontainers (Figure 2a). When PSVSS without the surface functionalization were treated with STA or ConA, there was no change in OD600 verifying the specificity of multivalent interactions. The results of optical density measurements were corroborated by DLS measurements. Upon the addition of ConA to PSVSSMan, the average hydrodynamic diameter increased drastically from about 135 nm to average aggregate sizes >1000 nm, whereas the treatment of unfunctionalized PSVSS did not cause a change in size distributions (Figure 2b). Similarly, the aggregation of PSVSSBiotin in the presence of STA resulted in highly polydispersed samples with large aggregates (Figure S2). Furthermore, agglutination of PSVSSMan was immediately reversed by the addition of a large excess of mannose, displaying the highly dynamic nature of the noncovalent recognition of mannose ligands at the nanocontainers’ surfaces (Figures 2b and S3). In contrast, the initial OD600 could not be recovered for aggregated PSVSSBiotin in the presence of an excess of biotin ligand. This can be explained by the remarkably strong binding and slow dissociation (half-life time of days to weeks)21 of biotin and STA inhibiting a dynamic ligand exchange and a redispersion of the aggregates. To evaluate if the functionalization with PEG-NH2 still facilitates the recognition of surface-conjugated ligands, we first prepared PSVSSBiotin,PEG functionalized with biotin-NH2 (2% stoichiometric maximum functionalization of carboxylic acid groups by biotin-NH2 ) and PEG-NH 2 simultaneously. Interestingly, PSVSSBiotin,PEG in the presence of STA show a much slower aggregation and a lower OD600 compared to PSVSSBiotin, and the size distribution according to DLS indicates the copresence of aggregates with sizes of ca. 1000 nm and remaining, nonaggregated nanocontainers (Figure S4). For PSVSSMan,PEG functionalized with mannose-NH2 and PEG-NH2, the agglutination with the weaker binding ConA was fully inhibited and OD600 remained low for a stoichiometric maximum of 2% functionalized carboxylic acid groups with mannose-NH2 (Figure S4). Again, these results imply that the functionalization with PEG groups sterically stabilizes the nanocontainer surface, diminishing the recognition and

Dynamic light-scattering (DLS) measurements show that the self-assembly of the thin polymer shell increases the hydrodynamic diameter slightly by ca. 10 nm, but during the crosslinking and ligand conjugation, no change in particle diameter was observed. This clearly indicates that no interparticle crosslinking occurred, which is most probably a consequence of a high local concentration of preorganized carboxylic acid groups at the vesicle surface reacting with free amine groups of the cross-linker in combination with the steric and electrostatic stabilization resulting from the polymer shell (Figure 1a and Table S1). To assess the surface properties of the nanocontainers, ζ-potential measurements were performed (Figure 1b and Table S1). At pH 7.4, the poly(acrylic acid) shell of PDV creates a strongly decreased ζ-potential of −21.3 mV for PDV. The cross-linking and attachment of biotin-NH2 or mannose-NH2 increases the ζ-potential by ca. 6 mV as a consequence of amide bond formation. Because only a small number of ligands were added in the amide-formation step to limit the functionalization of carboxylic acid groups with ligands to a stoichiometric maximum of 2%, the ζ-potentials of PSVSS, PSVSSBiotin, and PSVSSMan do not differ significantly. PEG-NH2 was added in higher concentration to achieve a maximum functionalization of 40% of carboxylic acid groups. In this case, a ζ-potential of −8.2 mV was observed, which is typical for PEG-coated nanoparticles and indicates a successful functionalization. As a protective layer of PEG is a common method to sterically stabilize nanoparticles in the biological media and to prevent unspecific protein absorption (stealth effect),8,28,29 we tested the colloidal stability in a buffered saline containing 10% fetal bovine serum. It can be seen from the DLS data that the average hydrodynamic diameter of both PSVSS and PSVSSPEG did not change over 4 days, and no aggregation could be observed (Figure S1), documenting a remarkable colloidal stability of the polymer nanocontainers in a biological environment. An agglutination assay was performed to prove the functionalization of the polymer shell with biotin or mannose ligands. In this experiment, the STA was added to PSVSSBiotin, whereas the lectin Concanavalin A (ConA), which has a high affinity to derivatives of α-D-mannose (Ka ∼ 104 M−1),12,13 was added to PSVSSMan. Each of these tetrameric proteins can bind four ligands and the specific and multivalent interaction of these proteins with the corresponding ligands conjugated to the surface of PSVSSBiotin or PSVSSMan should result in the formation 41763


Research Article


Figure 3. (a) Fluorometric quantification of biotin in the dispersions of PSVSSBiotin with varied stoichiometric maximum functionalization of carboxylic acid groups by biotin-NH2 (performed as duplicates). (b) Time-dependent aggregation PSVSSBiotin with different maximum functionalization by biotin-NH2 upon the addition of STA at t = 5 min. (c) Time-dependent aggregation PSVSSMan with different maximum functionalization by mannose-NH2 upon the addition of ConA at t = 5 min. (d) Initial rates of aggregation of PSVSSMan + ConA or PSVSSBiotin + STA calculated by a linear approximation of the increase in OD600 within the first 60 s of aggregation. Concentrations of PSVSSLigand correspond to 100 μM β-cyclodextrin amphiphiles in HBS (20 mM HEPES, 150 mM NaCl, pH 7.4) supplemented with 1 mM CaCl2 and 1 mM MnCl2 for ConA recognition experiments, [STA] = [ConA] = 0.1 mg mL−1, measurements were performed at 25 °C.

biotin ligands within a densely functionalized polymer shell, reducing the number of free-biotin-binding sites and thereby the probability of an interparticle cross-linking. This hypothesis is corroborated by a theoretical estimation of the number of biotin-binding sites necessary to fully saturate a shell of STA on the surface of PSVSSBiotin (see the SI). In this simple geometric consideration, we assumed that STA covers the volume of a thin shell around a spherical nanocontainer core. Based on this estimation, the biotin surface density for a starting saturation of STA is in the same order of magnitude as for 3−4% biotinylation, where a reduced interparticle cross-linking was observed. For PSVSSMan, the mannose density on the surface was varied in the same manner. Here, the initial rate of increase in OD600 was increased almost 10-fold from 0.03 min−1 for 1% to 0.24 min−1 for 2% maximum functionalization. For samples with higher ligand densities, only a slightly stronger agglutination was observed (Figure 3c and 3d). This experiment demonstrates two vital points: a fast and extensive agglutination is a result of high mannose density and a most likely multivalent recognition by ConA (i.e., cluster glycoside effect)30,31 and the dynamic nature of interaction at high ligand densities still allows for a dynamic exchange of ligands and interparticle cross-linking. In contrast, the slow dynamics of biotin−STA binding and saturation of binding sites reduce interparticle recognition at high biotin ligand densities. Finally, we applied the rational surface design of the redoxresponsive nanocontainers to immobilize them on solid

aggregation process. However, the strong biotin−STA binding still facilitates the interparticle cross-linking. We next investigated if the ligand density at the nanocontainers’ surface can be varied and how this influences the multivalent recognition by proteins. Therefore, increasing concentrations of biotin-NH2 were added during the preparation of PSVSSBiotin, resulting in a theoretical maximum biotinylation of 1, 2, 3, or 4% of carboxylic acid groups at the nanocontainer surface. After dialysis to remove free biotin-NH2, the biotin concentration was measured by a fluorometric method based on the quenching of the natural fluorescence of STA upon binding to biotin (see Experimental Section and Figure S5).25 Figure 3a shows that the measured biotin concentration correlates with the theoretical degree of functionalization, demonstrating that the concentration of surface-attached ligands can be easily varied by adapting the stoichiometry during synthesis. Upon addition of STA to PSVSSBiotin with a low surface density of biotin (1 or 2% maximum biotinylation), a fast and extensive increase in OD600 was monitored. Interestingly, nanocontainers with higher biotin surface densities (3 and 4%) showed only a slow and moderate increase in OD600 in the presence of STA (Figure 3b). Similarly, the DLS measurements show aggregated and nonaggregated nanocontainers for 3 and 4% functionalized nanocontainers, which indicates a strongly reduced interparticle cross-linking (Figure S6). This phenomenon most probably can be attributed to a starting saturation of STA binding sites with 41764


Research Article


Figure 4. (a) Illustration of the biotin−STA binding motif for the immobilization of PSVSSBiotin. Fluorescence microscopy images of (b) rhodamine B-labeled and (c) CF-loaded PSVSSBiotin immobilized on the STA-patterned surfaces. (d) Time-dependent release of CF from immobilized redoxresponsive PSVSSBiotin.

immobilized PSVSSBiotin loaded with CF and measured the fluorescence of the supernatant solution. In the absence of a reducing agent, only a moderate increase in fluorescence intensity was monitored, indicating a slow passive release of CF from the aqueous lumen of the immobilized nanocontainers. Strikingly, upon the addition of the reducing agent TCEP, a highly accelerated release of cargo was observed (Figures 4d and S9). Taken together, the immobilization experiments indicate that the PSVSSBiotin stay intact and maintain their redoxresponsive release properties upon attachment to a solid substrate.

substrates, creating a surface coating that can release its cargo upon a reductive trigger. We decided to exploit the biotin−STA receptor−ligand pair to immobilize PSVSSBiotin (Figure 4a) due to its high association constant, which should inhibit a dynamic dissociation of the PSVSSBiotin from the surface.22 Applying a recently published procedure,22 the surface of silicon substrates was patterned by microcontact printing (μCP)32 of a thiolfunctionalized biotin derivative (biotin-SH) reacting with a silicon-supported self-assembled monolayer (SAM) of alkenes. To generate a hydrophilic surface with a static contact angle of 39.5° (Figure S7), unreacted alkenes were saturated with mercaptoethanol and to minimize unspecific protein adsorption, the surfaces were blocked with bovine serum albumin (BSA) prior to the application of STA. Finally, these patterned substrates were incubated with PSVSSBiotin, which were fluorescently labeled by including rhodamine B-conjugated amphiphilic β-cyclodextrin33 in the template vesicles. Fluorescence microscopy analysis clearly shows the specific immobilization of PSVSSBiotin because the signal of rhodamine B resembles the printed patterns of 10 μm dots that are spaced at 10 μm (Figure 4b). To show that PSVSSBiotin remained intact upon immobilization, they were loaded with the hydrophilic fluorescent dye carboxyfluorescein (CF) before immobilization.24 Here, the immobilization in a stripe pattern nicely demonstrates the recognition of intact CF-loaded PSVSSBiotin (Figure 4c). In the negative control experiments, dye-loaded or rhodamine B-labeled PSVSS without biotin functionalization were incubated with STA-patterned surfaces. No immobilization was observed, verifying the specific deposition of nanocontainers on the surface (Figure S8). In previous studies, we demonstrated the redox-responsive release of hydrophilic payloads from PSVSS in solution.24 To investigate the influence of surface immobilization on the redox-responsive release of a payload from PSVSS, we

■

CONCLUSIONS In summary, we have introduced the rational surface design of self-assembled polymer-shelled vesicles and employed this methodology for their immobilization in a noncovalent biomimetic fashion. The conjugation of biotin or mannose ligands to the polymer shell proved to be appropriate for the recognition and binding of biological macromolecules and indicates that this complex process can be controlled by the variation in ligand densities on the nanocontainer surface and a proper choice of the ligand−receptor pair. Moreover, the patterned immobilization of these redox-responsive polymer nanocontainers provides a new strategy to construct tethered nanocapsules, which can release their payload upon a reductive trigger. Such materials may have promise for use in targeted drug delivery applications and may be considered in the design of microarrays for sensing applications.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.7b15516. 41765


Research Article


■

(13) Lis, H.; Sharon, N. Lectins: Carbohydrate-Specific Proteins That Mediate Cellular Recognition. Chem. Rev. 1998, 98, 637−674. (14) Chen, Y.; Espeel, P.; Reinicke, S.; Du Prez, F. E.; Stenzel, M. H. Control of Glycopolymer Nanoparticle Morphology by a One-Pot, Double Modification Procedure Using Thiolactones. Macromol. Rapid Commun. 2014, 35, 1128−1134. (15) Sun, P.; He, Y.; Lin, M.; Zhao, Y.; Ding, Y.; Chen, G.; Jiang, M. Glyco-Regioisomerism Effect on Lectin-Binding and Cell-Uptake Pathway of Glycopolymer-Containing Nanoparticles. ACS Macro Lett. 2014, 3, 96−101. (16) Dhaware, V.; Shaikh, A. Y.; Kar, M.; Hotha, S.; Gupta, S. S. Synthesis and Self-Assembly of Amphiphilic Homoglycopolypeptide. Langmuir 2013, 29, 5659−5667. (17) Li, X.; Chen, G. Glycopolymer-Based Nanoparticles: Synthesis and Application. Polym. Chem. 2015, 6, 1417−1430. (18) Parry, A. L.; Clemson, N. A.; Ellis, J.; Bernhard, S. S. R.; Davis, B. G.; Cameron, N. R. “Multicopy Multivalent” GlycopolymerStabilized Gold Nanoparticles as Potential Synthetic Cancer Vaccines. J. Am. Chem. Soc. 2013, 135, 9362−9365. (19) Le Guével, X.; Perez-Perrino, M.; Fernández, T. D.; Palomares, F.; Torres, M.-J.; Blanca, M.; Rojo, J.; Mayorga, C. Multivalent Glycosylation of Fluorescent Gold Nanoclusters Promotes Increased Human Dendritic Cell Targeting via Multiple Endocytic Pathways. ACS Appl. Mater. Interfaces 2015, 7, 20945−20956. (20) Vermette, P.; Griesser, H. J.; Kambouris, P.; Meagher, L. Characterization of Surface-Immobilized Layers of Intact Liposomes. Biomacromolecules 2004, 5, 1496−1502. (21) Liu, W.; Samanta, S. K.; Smith, B. D.; Isaacs, L. Synthetic Mimics of Biotin/(strept)avidin. Chem. Soc. Rev. 2017, 46, 2391−2403. (22) Roling, O.; Wendeln, C.; Kauscher, U.; Seelheim, P.; Galla, H.J.; Ravoo, B. J. Layer-by-Layer Deposition of Vesicles Mediated by Supramolecular Interactions. Langmuir 2013, 29, 10174−10182. (23) Grzelakowski, M.; Onaca, O.; Rigler, P.; Kumar, M.; Meier, W. Immobilized Protein−Polymer Nanoreactors. Small 2009, 5, 2545− 2548. (24) de Vries, W. C.; Grill, D.; Tesch, M.; Ricker, A.; Nüsse, H.; Klingauf, J.; Studer, A.; Gerke, V.; Ravoo, B. J. Reversible Stabilization of Vesicles: Redox-Responsive Polymer Nanocontainers for Intracellular Delivery. Angew. Chem., Int. Ed. 2017, 56, 9603−9607. (25) Der-Balian, G. P.; Gomez, B.; Masino, R. S.; Parce, J. W. A Fluorometric Method for Determining the Degree of Biotinylation of Proteins. J. Immunol. Methods 1990, 126, 281−285. (26) Samanta, A.; Tesch, M.; Keller, U.; Klingauf, J.; Studer, A.; Ravoo, B. J. Fabrication of Hydrophilic Polymer Nanocontainers by Use of Supramolecular Templates. J. Am. Chem. Soc. 2015, 137, 1967− 1971. (27) Ravoo, B. J.; Jacquier, J.-C.; Wenz, G. Molecular Recognition of Polymers by Cyclodextrin Vesicles. Angew. Chem., Int. Ed. 2003, 42, 2066−2070. (28) Neoh, K. G.; Kang, E. T. Functionalization of Inorganic Nanoparticles with Polymers for Stealth Biomedical Applications. Polym. Chem. 2011, 2, 747−759. (29) Torchilin, V. P. Recent Advances with Liposomes as Pharmaceutical Carriers. Nat. Rev. Drug Discovery 2005, 4, 145−160. (30) Voskuhl, J.; Stuart, M. C. A.; Ravoo, B. J. Sugar-Decorated Sugar Vesicles: Lectin−Carbohydrate Recognition at the Surface of Cyclodextrin Vesicles. Chem. − Eur. J. 2010, 16, 2790−2796. (31) Lundquist, J. J.; Toone, E. J. The Cluster Glycoside Effect. Chem. Rev. 2002, 102, 555−578. (32) Perl, A.; Reinhoudt, D. N.; Huskens, J. Microcontact Printing: Limitations and Achievements. Adv. Mater. 2009, 21, 2257−2268. (33) Voskuhl, J.; Fenske, T.; Stuart, M. C. A.; Wibbeling, B.; Schmuck, C.; Ravoo, B. J. Molecular Recognition of Vesicles: Host− Guest Interactions Combined with Specific Dimerization of Zwitterions. Chem. − Eur. J. 2010, 16, 8300−8306.

Experimental details, synthesis of Ad-PAA, amphiphilic β-cyclodextrin and ligands, DLS, OD600, and contact angle data, release profiles, fluorescence microscopy images, control experiments, and calculations (PDF)

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Armido Studer: 0000-0002-1706-513X Bart Jan Ravoo: 0000-0003-2202-7485 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS W.C.d.V. acknowledges a fellowship of the Fonds der Chemischen Industrie. We sincerely thank the Deutsche Forschungsgemeinschaft (DFG SFB858 and EXC 1003) for funding. The authors thank Dr. Christian Wendeln for providing mannose-NH2 and Dr. Oliver Roling for supplying biotin-SH.

■

REFERENCES

(1) Kamaly, N.; Xiao, Z.; Valencia, P. M.; Radovic-Moreno, A. F.; Farokhzad, O. C. Targeted Polymeric Therapeutic Nanoparticles: Design, Development and Clinical Translation. Chem. Soc. Rev. 2012, 41, 2971−3010. (2) Pawar, P. V.; Gohil, S. V.; Jain, J. P.; Kumar, N. Functionalized Polymersomes for Biomedical Applications. Polym. Chem. 2013, 4, 3160−3176. (3) Zhang, S.; Geryak, R.; Geldmeier, J.; Kim, S.; Tsukruk, V. V. Synthesis, Assembly, and Applications of Hybrid Nanostructures for Biosensing. Chem. Rev. 2017, 117, 12942−13038. (4) Elsabahy, M.; Heo, G. S.; Lim, S.-M.; Sun, G.; Wooley, K. L. Polymeric Nanostructures for Imaging and Therapy. Chem. Rev. 2015, 115, 10967−11011. (5) Qi, K.; Ma, Q.; Remsen, E. E.; Clark, C. G.; Wooley, K. L. Determination of the Bioavailability of Biotin Conjugated onto Shell Cross-Linked (SCK) Nanoparticles. J. Am. Chem. Soc. 2004, 126, 6599−6607. (6) Najer, A.; Wu, D.; Bieri, A.; Brand, F.; Palivan, C. G.; Beck, H.-P.; Meier, W. Nanomimics of Host Cell Membranes Block Invasion and Expose Invasive Malaria Parasites. ACS Nano 2014, 8, 12560−12571. (7) Culver, H. R.; Steichen, S. D.; Herrera-Alonso, M.; Peppas, N. A. Versatile Route to Colloidal Stability and Surface Functionalization of Hydrophobic Nanomaterials. Langmuir 2016, 32, 5629−5636. (8) Wu, J.; Zhao, L.; Xu, X.; Bertrand, N.; Choi, W. I.; Yameen, B.; Shi, J.; Shah, V.; Mulvale, M.; MacLean, J. L.; Farokhzad, O. C. Hydrophobic Cysteine Poly(disulfide)-Based Redox-Hypersensitive Nanoparticle Platform for Cancer Theranostics. Angew. Chem., Int. Ed. 2015, 54, 9218−9223. (9) Dou, W. T.; Zeng, Y.-L.; Lv, Y.; Wu, J.; He, X.-P.; Chen, G.-R.; Tan, C. Supramolecular Ensembles Formed between Charged Conjugated Polymers and Glycoprobes for the Fluorogenic Recognition of Receptor Proteins. ACS Appl. Mater. Interfaces 2016, 8, 13601−13606. (10) Yang, W. J.; Zhao, T.; Zhou, P.; Chen, S.; Gao, Y.; Liang, L.; Wang, X.; Wang, L. “Click” Functionalization of Dual StimuliResponsive Polymer Nanocapsules for Drug Delivery Systems. Polym. Chem. 2017, 8, 3056−3065. (11) Lee, S.-M.; Chen, H.; O’Halloran, T. V.; Nguyen, S. T. “Clickable” Polymer-Caged Nanobins as a Modular Drug Delivery Platform. J. Am. Chem. Soc. 2009, 131, 9311−9320. (12) Miura, Y.; Hoshino, Y.; Seto, H. Glycopolymer Nanobiotechnology. Chem. Rev. 2016, 116, 1673−1692. 41766


Molecular Recognition and Immobilization of Ligand-Conjugated

Recommend Documents