Designing Peptide Bunches on Nanocage for Bispecific or

Feb 22, 2016 - Department of Biochemistry and Cell Biology, Cell and Matrix Research Institute, School of Medicine, Kyungpook National University, Dae...
0 downloads 0 Views 6MB Size
Download PDF
Article pubs.acs.org/Biomac

Designing Peptide Bunches on Nanocage for Bispecific or Superaffinity Targeting Sooji Kim,† Jae-Ok Jeon,† Eunsung Jun,† JunGoo Jee,‡ Hyun-Kyung Jung,† Byung-Heon Lee,† In-San Kim,*,§,∥ and Soyoun Kim*,† †

Department of Biochemistry and Cell Biology, Cell and Matrix Research Institute, School of Medicine, Kyungpook National University, Daegu 700-422, Republic of Korea ‡ College of Pharmacy, Research Institute of Pharmaceutical Sciences, Kyungpook National University, Daegu 702-701, Republic of Korea § Biomedical Research Institute, Korea Institute of Science and Technology, Seoul 136-791, Republic of Korea ∥ KU-KIST School, Korea University, Seoul 136-701, Republic of Korea S Supporting Information *

ABSTRACT: Ferritin cage nanoparticles are promising platforms for targeted delivery of imaging and therapeutic agents because their cage structure can accommodate small molecules and their surfaces can be decorated with multiple functionalities. However, selective targeting is still a challenge for translating ferritin-based nanomedicines into the clinic, especially for heterogeneous diseases such as cancer. Targeting peptides can be genetically fused onto the surface of a ferritin cage, forming peptide bunches on nanocages (PBNCs) that offer synergistic increases in binding avidity. Here, we utilized two sites of the ferritin monomer, the Nterminus and the loop between the fourth and fifth helices, which are exposed on the surface of the assembled 24-subunit ferritin cage, to ligate one or two types of peptides to achieve “super affinity” and bispecificity, respectively. PBNCs formed by ligation of the IL-4 receptor-targeting peptide, AP1, to both sites (48AP1-PBNCs) tethered IL-4R, expressing tumor cells with greater affinity than did PBNCs with AP1 ligated to a single site (24AP1-PBNCs). Moreover, bispecific PBNCs containing 24 RGD peptides and 24 AP1 peptides (24RGD/24AP1-PBNCs) were capable of independently targeting cells expressing the corresponding receptors. Bispecific and superaffinity PBNCs could be useful for efficient targeting of ferritin-based therapeutic/ diagnostic agents in a clinical setting.

■

INTRODUCTION Protein-based supramolecular assemblies, such as ferritins and heat-shock proteins, with their well-characterized cage-like structural features, are promising nanoscale particles for several reasons: they are readily produced in large quantities, have defined interior cavities, are usually monodispersed in solution, and are amenable to chemical and biological modification.1,2 Ferritin is ubiquitous in nature and acts as an intracellular iron storage system, accommodating up to 4500 iron atoms inside of its cavity. Ferritin has been the subject of research for various applications in bionanotechnology that offers a number of unique advantages. First, the ferritin cage itself is highly symmetrical, comprising 24 subunits arranged in an octahedral (432) symmetry with remarkable thermal and chemical stability. Second, it is possible to reconstitute the ferritin structure through controlled disassembly and reassembly, a feature that can be used to configure ferritin to carry small molecules inside of the cage.3,4 Third, the selectivity of loaded metals in ferritin is broad.5,6 Fourth, it is possible to modify the © XXXX American Chemical Society

surface of the ferritin cage through the addition of peptide and protein tags.7,8 Finally, it is biocompatible, biodegradable, less toxic, and less immunogenic compared with synthetic polymers.9 These characteristics have made ferritins attractive vectors for the delivery of drug molecules10−12 and as scaffolds for vaccine design.13−15 Furthermore, the plasticity of in vitro mineralization makes ferritin an ideal tool for cellular and medical imaging, because labeled heavy atoms and heavy atom complexes can be readily sequestered within its core.16−18 Decorating the cage with multiple ligands using genetic and chemical methods allows the development of various nanoparticle platform technologies for diagnostic and therapeutic purposes.7,8,19,20 Extensive research on ferritin-based nanoparticles carries the promise of advanced, next-generation therapeutic and diagReceived: December 28, 2015 Revised: February 18, 2016

A

DOI: 10.1021/acs.biomac.5b01753 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 1. Designed PBNCs with one- or two-site peptide display for bispecific and super affinity targeting. (A) 3D structure models of PBNCs with peptides inserted at two sites were calculated based on wild-type ferritin light chain crystal structure (PDB 2FG4). The calculated 24 subunits were assembled into PBNCs by computer simulation based on the wild-type ferritin cage structure (PDB 3A68). The peptide inserted at the N-terminus was highlighted blue and peptide, at position 157, magenta. For super affinity targeting PBNC, peptides were highlighted with magenta and pink. (B) Schematic diagrams of the PBNCs with one- or two-site peptide display. RGD, GRGDSP, and AP1 (RKRLDRN) peptides were inserted into two different positions, N-terminal and position before 157 between the fourth and firth helices. A flexible linker (GGGSG) was inserted between Nterminal peptide ligand and ferritin. (C) The purified PBNCs were subjected to SDS-PAGE showing more than 95% homogeneity.

tumors through the enhanced permeation and retention (EPR) effect, but more than 95% of administered nanoparticles actually end up at sites other than the targeted tumor.21 Untargeted nanoparticles are retained in the kidney or liver and

nostic nanomedicines; however, translating these nanomedicines into the clinic requires a greater effort to overcome current limitations, such as efficient targeting in complex clinical situations. Nanoparticles can be passively targeted to B

DOI: 10.1021/acs.biomac.5b01753 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules can be causes of toxicity.22 Furthermore, the heterogeneity of tumor cells and temporal expression of biomarkers are problems that may not easily be solved by a simple targeting strategy. The main direction for research on the translation of nanomedicines is the development of strategies to deliver a diagnostic or therapeutic agent selectively to a target site for enhanced efficacy with reduced side effects. To achieve this goal, researchers have developed active-targeting, high-affinity nanoparticles with multifunctional targeting moieties. We previously reported the design of peptide bunches on nanocages (PBNCs), in which genetic fusion of specific peptides onto the surface of ferritin cage and subsequent symmetrical assembly of ferritin forms clusters of the peptides in bunches.8 We showed that this clustering synergistically increases the affinity of the peptide ligands and substantially enhances their therapeutic efficacy, suggesting possible applications of various PBNCs for targeting different diseases. Using the interleukin-4 receptor (IL-4R) targeting peptide, AP1, identified from phage-display libraries,23 we demonstrated that PBNCs are capable of overcoming problems associated with peptide drugs, such as weak affinity and short half-life.24 Here, we generated a “super-affinity” PBNC that displays 48 targeting peptides (24 subunits × 2 sites/subunit) and tethers tumor cells with remarkable affinity by forming a stable complex. We also generated a bispecific targeting PBNC consisting of two different targeting peptides, one for blood vessels and the other for tumor cells. Different peptides bunches, simultaneously ligated onto two exposed sites, were independently fully functional, targeting cells with high affinity without impeding each other. Our findings indicate that ferritin can be decorated with various combinations of peptides and suggest that the resulting multifunctional PBNCs could achieve selective targeting with minimal side effects in a complex clinical setting.

■

1 h. Afterward, the denatured protein was loaded onto a nickel ion chelate affinity column, washed with wash buffer (20 mM Tris-HCl pH 8.0, 500 mM NaCl, 20 mM imidazole, 8 M urea), and refolded with a gradient addition of 8−0 M urea containing refolding buffer (20 mM Tris-HCl pH 8.0, 500 mM NaCl, 20 mM imidazole). The renatured protein was eluted with elution buffer (20 mM Tris-HCl pH 8.0, 150 mM NaCl, 180 mM histidine). Cell Culture. Human cancer cell lines (U87MG (glioblastoma), HT1080 (fibrosarcoma) and A549 (lung carcinoma)) were respectively maintained in high glucose Dulbecco’s modified Eagle’s Media (DMEM), low glucose DMEM, RPMI 1640 containing 10% fetal bovine serum (FBS), 100 units/mL penicillin, and 100 μg/mL streptomycin in a 37 °C incubator with 5% CO2. Size Characterization. After purification, each protein was analyzed by SEC (Superdex 200 10/300 GL column). Oligomeric states were judged from the elution volume compared with molecular weight standard. Protein elution profiles in SEC were monitored by measuring absorbance at 280 nm. TEM pictures were recorded using an FEI Tecnai (the Korea Basic Science Institute, KBSI). Surface Plasmon Resonance Analysis. Interactions of RGDPBNCs with Integrin αvβ3 and Interactions of AP1-PBNCs with IL4R were analyzed at 25 °C using a surface plasmon resonance instrument (SR7500 DC, Reichert Inc., NY, U.S.A.). Human Integrin αvβ3 was purchased at Millipore (Cat. No. CC1018) and IL-4R was expressed and purified in Sf21 cells, as described previously.8 Integrin αvβ3 was immobilized by activating the carboxymethyl group on dextran-coated chips through a reaction with a mixture of N-(3(dimethylamino)propyl)-N′-ethylcarbodiimide hydrochloride and Nhydroxysuccinimide (Sigma-Aldrich). Different concentrations of RGD-PBNCs (1.3−83 nM) in binding buffer (20 mM Hepes pH 7.4, 150 mM NaCl, 180 mM histidine, 1 mM MgCl2, 0.005% Tween 20) were allowed to flow over the surface containing immobilized Integrin αvβ3 (approximately 1800 RU) at a rate of 30 μL/min. In the same way, IL-4R was immobilized on dextran-coated chips (approximately 2000 RU) and AP1-PBNCs flow over at a rate of 25 μL/min. The sensor surface was regenerated after each association and dissociation cycle by injecting 2 M NaCl for 1 min. Sensorgrams were fit to a simple 1:1 Langmuir interaction model (A + B ⇌ AB) using data analysis program Scrubber 2.0 (BioLogic Software, Australia, and KaleidaGraph Software, Australia). FACS Analysis. The expression levels of integrin (αvβ3 and αvβ5) on the surface of human cancer cell lines (U87MG, HT1080, A549) were determined by flow cytometry. Each type of cell was detached by 1 mM EDTA and washed with phosphate buffered saline (PBS) containing 1% bovine serum albumin (BSA; 1% BSA/PBS). A total of 2 × 105 cells were incubated with respective human integrin αvβ3 antibody (1:500 dilution, MAB1976, Millipore), human integrin αvβ5 antibody-FITC conjugated (1:200 dilution, MAB1961F, Millipore), human IL-4R antibody (1:500 dilution, MAB230, R&D systems) at 4 °C for 1 h and washed two times with 1% BSA/PBS, followed by incubation with Alexa Fluor 488 goat antimouse IgG (H+L) antibody (1:200 dilution, Invitrogen) at 4 °C for 20 min. After washing two times with 1% BSA/PBS, the stained cells were resuspended in 350 μL of PBS for analysis. Flow cytometry was performed by using BD Accuri C6 cytometry (BD Biosciences, San Jose, CA, U.S.A.). Cell Adhesion Assay. The cellular adhesion activities of integrin expressing cells were tested as described before.25 U87MG and HT1080 cells (3 × 104 cells/100 μL media) were added to each well of 96-well ELISA plates (Costar, Corning, Inc. NY), which were coated for 1 h at room temperature with wt FTL, 1RGD-PBNCs, 157 AP1-PBNCs, and 1RGD157AP1-PBNCs (2.5 to 40 μg/mL) and blocked for 1 h at room temperature with PBS containing 2% BSA. After incubation for 30 min at 37 °C, unattached cells were removed by rinsing with PBS. Attached cells were incubated for 1 h at 37 °C in 50 mM citrate buffer (pH 5.0) containing 3.75 mM ρ-nitrophenyl-Nacetyl-β-D-glucosaminide (hexosaminidase substrate) and 0.25% Triton X-100. Enzyme activity was blocked by adding 5 mM glycine buffer (pH 10.4) containing 5 mM EDTA, and the absorbance was measured at 405 nm using a Sunrise Basic Microplate Reader (Tecan).

EXPERIMENTAL SECTION

Construction of Peptide Conjugated Ferritin. The recombinant plasmid for the expression of wt FTL proteins was described previously.8 The RGD peptides (RGD, GRGDSP) and AP1 (RKRLDRN) peptides were inserted at the 1 and 157 positions of wt FTL, respectively. To insert RGD and AP1 peptides at the 157 position, BamHI and ApaI restriction sites were introduced into wt FTL by using PCR-mediated site-directed mutagenesis. The primers encoding RGD and AP1 peptides were annealed and inserted into the BamHI and ApaI site of the plasmid. The insertion of the peptide resulted in the change of the original amino acid sequence of wt FTL from 1 5 5 LGGPE to 1 5 5 GSRGDGGPE ( 1 5 7 RGD-PBNC), 155 GSGRGDSPGGPE (157GRGDSP-PBNC), and 155 GSRKRLDRNGGPE (157AP1-PBNC). The nucleotides encoding RGD, GRGDSP, and AP1 were inserted into the 1 position with flexible linker (GGGSG) for 1RGD-PBNC, 1GRGDSP-PBNC, and 1 AP1-PBNC, respectively (Figure 1B). Protein Expression and Purification. The peptide conjugated ferritin proteins were overexpressed in E. coli BL21 (DE3) cells and purified as previously reported.8 Briefly, cells were grown at 37 °C to an OD600 of 0.5 in LB medium containing 50 μg/mL of kanamycin, and protein expression was induced by 1 mM IPTG at 20 °C for 18 h. After induction, cells were harvested by centrifugation (2,265 g), and the pellets were suspended in lysis buffer (20 mM Tris-HCl pH 8.0, 100 mM NaCl, 1 mM EDTA, 1% Triton X-100, 1 mM PMSF, 0.5 mM DTT) and homogenized with an ultrasonic processor. After sonication, the cell lysate was centrifuged at 10400 rpm (12930g) for 30 min, and the inclusion bodies from cell lysates were solubilized by incubating in binding buffer (20 mM Tris-HCl pH 8.0, 500 mM NaCl, 5 mM imidazole) containing 8 M urea at room temperature for C

DOI: 10.1021/acs.biomac.5b01753 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 2. Physicochemical characterization of bisite decorating PBNCs. (A) Size exclusion elution profiles of wt FTL and peptide conjugated PBNCs. (B) Transmission electron microscopy (TEM) images of 2% uracyl acetate stained peptide conjugated PBNCs. Cell Binding Assay. To perform cell binding analysis, 2 × 105 A549 cells in suspension were incubated with 1% bovine serum albumin at 37 °C for 30 min for blocking and incubated with 400 nM 157 AP1-PBNCs or 1RGD157AP1-PBNCs at 4 °C for 20 min. The cells were incubated with the same amount of wt FTL and 1RGD-PBNCs as control. After each protein incubated with cells, cells were incubated with human ferritin light chain antibody (1:400 dilutions, SC-14420, Santa Cruz Biotechnology Inc.), followed by incubation with Alexa Fluor 488-conjugated donkey antigoat IgG (H+L) antibody (Invitrogen), and analyzed with FACS calibur cytometry (BD Biosciences, San Jose, CA, U.S.A.). For comparison of cell binding ability between 157AP1-PBNC and 1 AP1157AP1-PBNC, series of concentrations (4.2, 8.3, and 16.6 nM) of 157 AP1-PBNC or 1AP1157AP1-PBNC were incubated with A549 cells. To assess competing ability of AP1-PBNCs (157AP1-PBNC, 1 AP1157AP1-PBNC) with IL-4 against IL-4R expressing cells, A549 cells (1 × 105/100 μL media) were incubated with mixture of biotinylated IL-4 (Cat. No. NF400, R&D systems) and increasing concentrations of AP1-PBNCs (2.6−20.8 nM) for 30 min at 4 °C. After incubation, the cell suspensions were treated with avidin-FITC reagent for 30 min at 4 °C in the dark, followed by washing with RDF1 cell wash buffer (Cat. No. NF400, R&D systems) twice to remove unreacted avidin-fluorescein. Cells were resuspended in 250 μL of 1× RDF1 and analyzed with FACS calibur cytometry (BD Biosciences, San Jose, CA, U.S.A.).

endothelial cells of tumor and the increase of interleukin-4 (IL4) level in tumor environment have been long observed.27 RGD-modified nanoparticles have been applied for efficient drug delivery to tumors and the RGD-modified ferritin also demonstrated that it can efficiently home to tumors through RGD-integrin αvβ3 interaction.28−30 IL-4R targeting peptide is also used for tumor homing peptide for drug delivery. Recently, IL-4R-peptide decorated liposomes were reported to efficiently deliver doxorubicin into lung tumor cells.31 Blockade of signals from integrin αvβ3 or IL-4R by competition against their natural ligands, vitronectin, or IL-4, respectively, gives additional therapeutic effects.32 In a previous study, we constructed IL-4 receptor-targeting PBNCs using human ferritin light (FTL) chain.8 The IL-4 targeting peptide, AP1, was genetically ligated to the exposed loop region (155LGGPE159) between the fourth and fifth helices of FTL. The FTL monomer, comprising five helical bundles, assembles into 24-subunit cage structures in which the Nterminal end and the loop between the fourth and fifth helices are positioned toward to the surface. The N-terminus of the ferritin monomer is frequently used for genetic and chemical modification because this exposed end can be easily modified and places ligands in a dispersed pattern over the surface.14,20 The second exposed site, the short loop between the fourth and the fifth helices, is 4-fold symmetry axis. Therefore, peptides ligated on the loop are clustered, producing synergistic binding effects8,33 (Figure 1A). Only a single site has previously been used for decorating ferritin with peptides or proteins since bisite ligation can negatively impact protein expression or cause steric hindrance among peptides at different sites. To examine the possibility of simultaneously modifying two sites, we constructed one- or two-site-ligated PBNCs using the RGD

■

RESULTS AND DISCUSSION Construction of Bisite-Decorated, Tumor Targeting PBNCs. The RGD peptide is a well-known tumor homing peptide via interaction with integrin αvβ3 that is highly expressed on activated endothelial cells, new-born vessels as well as some tumor cells, but is not detected in resting endothelial cells and most normal organs.26 The up-regulation of IL-4 receptor (IL-4R) on tumor cells as well as in vascular D

DOI: 10.1021/acs.biomac.5b01753 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 3. Surface plasmon resonance studies for the binding kinetics of IL-4R and AP1 conjugated PBNCs. (A) The concentrations of AP1 conjugated PBNCs injected are indicated. (B) Plots of RU values of AP1 conjugated PBNCs vs different injection concentrations.

Table 1. Binding Kinetics between IL-4 Receptor and AP1 Conjugated PBNC ligand

kona (M−1 S1−)

AP1-PBNC AP1-PBNC 1 AP1 157AP1-PBNC

(1.22 ± 0.64) × 10 (3.50 ± 2.56) × 106 (5.25 ± 3.78) × 105

157 1

a

koffa (S1−)

KDb (M) −3

(1.25 ± 0.4) × 10 (1.42 ± 1.91) × 10−2 (1.55 ± 2.04) × 10−4

6

(1.44 ± 0.9) × 10−9 (5.29 ± 3.84) × 10−9 (2.99 ± 2.69) × 10−10

Obtained by saturated binding responses averaging at least three independent runs of SPR measurements. bKD = koff/kon.

peptides, RGD, and GRGDSP, and the AP1 peptide, RKRNDRN (Figure 1B). For N-terminal ligation, a short linker (GGGSG) was introduced between the peptide and FTL to improve peptide flexibility. For two-site display and bispecific targeting, the RGD peptides (RGD and GRGDSP) were ligated to the N-terminus (inserted before residue number 1) and AP1 peptides were ligated to the loop (inserted before residue number 157) or vice versa. We also generated a superaffinity targeted PBNC by ligating AP1 peptides to both the N-terminal end and the loop. The two-site-ligated PBNCs were expressed in Escherichia coli to a level comparable to that of one-siteligated PBNCs and were purified to >95% homogeneity (Figure 1C). Physicochemical Characterization of Bisite-Decorated PBNCs. Bispecific and superaffinity PBNCs were characterized by size-exclusion chromatography (SEC) and transmission electron microscopy (TEM) and compared with wild-type ferritin cage proteins (wt FTL) lacking displayed peptides. All two-site-ligated PBNCs eluted at a position similar to that of wt FTL in SEC, suggesting that simultaneous peptide modification of the N-terminus and the loop did not hamper formation of authentic cage architectures (Figure 2A). TEM images also confirmed that two-site-ligated PBNCs formed an intact, cage architecture with a uniform size distribution (Figure 2B). These results indicate that different combinations of bispecific PBNCs form intact, cage nanoparticle architectures without significant

deviation from structures formed by wt FTL. Only the AP1- 157 GRGDSP PBNC was precipitated by time at concentration of >0.3 mg/mL during SEC. Binding of Superaffinity PBNCs. Binding Kinetics of AP1-PBNCs with One-Site and Two-Site Modifications. We expected that one-site ligation of AP1 to the N-terminus (124AP1-PBNC) or C-terminal loop region (15724AP1-PBNC) on the surface of FTL nanoparticles would enhance binding avidity toward the designated target, IL-4R, as previously reported.8 The N-termini of ferritin monomers constitute a 3fold local symmetry with a distance of 5 nm, whereas the Ctermini loops constitute a 4-fold local symmetry with distance of 1 nm. Thus, the ligands that are fused to the N- or C- termini of FTL monomers are displayed in proximity of each other, leading to synergistically enhanced binding avidity.34 We further investigated the binding kinetics of 124AP1-PBNC and 15724AP1-PBNC because three-dimensional models of Nor C-termini-modified PBNCs predicted different patterns of peptide display, namely, a dispersed pattern of N-terminally ligated 124AP1-PBNC and a bunched pattern of the 4-fold loop-ligated 15724AP1-PBNC. To this end, we immobilized a component of the extracellular region of the IL-4R, applied a series of different concentrations of 24AP1-PBNCs, and examined binding kinetics using surface plasmon resonance (SPR) analysis (Figure 3). The binding responses similarly increased and saturated upon addition of increasing concen1

E

DOI: 10.1021/acs.biomac.5b01753 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 4. Cellular binding of super affinity PBNCs. (A) FACS analysis of A549 cells incubated with indicated concentrations of wt FTL and AP1 conjugated PBNCs. The green and red lines indicate cells incubated with 157AP1-PBNCs and 1AP1157AP1-PBNCs, respectively. The black line indicates cells incubated with wild type FTL. (B) A549 cells were incubated with IL-4 in the absence (green) or presence of AP1 conjugated PBNCs (blue, 2.6 nM; yellow, 5.2 nM; cyan, 10.4 nM; magenta, 20.8 nM). The black line indicated untreated cell control. (C) Bar charts show relative IL-4 binding (%) to A549 cells in the presence of different concentrations of 157AP1-PBNCs (white) or 1AP1157AP1-PBNCs (black). The binding activity of IL-4 alone is normalized as 100%. The results are means + s.d. (n = 3 independent examinations); *P < 0.05, unpaired Student’s t test.

as evidenced by their significantly higher binding responses compared with 24-peptide conjugated PBNCs.5 Cellular Binding of Super Affinity PBNCs. We previously reported that 15724AP1-PBNCs specifically targeted A549 cells expressing IL-4R on their surface.8 The specific interaction was mediated by IL-4R and blocked IL-4R-mediated signaling. We further examined the binding capacity of the superaffinity PBNC, 124AP1/15724AP1-PBNC, for A549 cells compared with that of one-site-ligated 15724AP1-PBNCs using fluorescence-activated cell sorting (FACS) analysis (Figure 4A). The binding capacity of 124AP1/15724AP1-PBNCs to A549 cells was enhanced compared with that of one-site−ligated 157 24AP1-PBNCs, indicating that superaffinity PBNCs are capable of binding IL-4R-expressing cells more efficiently. Next, A549 cells were preincubated with the authentic ligand, IL-4, and the formed complex was challenged with 15724AP1PBNCs or 124AP1/15724AP1-PBNCs (Figure 4B,C). The superaffinity AP1-PBNCs more effectively competed with IL4 for the IL-4R, suggesting that superaffinity AP1-PBNCs block IL-4R-mediated signaling and exert biological activity more efficiently than single-site-ligated AP1-PBNCs. Bispecific Targeting of 124RGD/15724AP1-PBNCs. Binding Kinetics of Bispecific-Targeting PBNCs. The RGDmodified ferritin platform has been extensively studied and reported for tumor imaging and therapy.28−30 In most previous studies, the RGD containing peptides are fused to the Ntermini of ferritin monomers for targeting and delivering therapeutic agents to the tumor region. However, systematic binding kinetics of RGD peptides ligated for bispecific-targeting has not been investigated. The affinity of bispecific-targeting PBNCs for the corresponding receptor was examined by applying 1 24AP1-PBNCs, 1 24AP1 157 24RGD-PBNCs, or 1 24AP115724GRGDSP-PBNCs to an IL-4R-coated surface and

trations of both one-site-conjugated 24AP1-PBNCs, resulting in an equilibrium dissociation constant (KD) of ∼10−9 M, indicating synergistically enhanced avidity (Table 1). However, the real-time binding kinetics of the two 24AP1-PBNCs were different; the association of 15724AP1-PBNC was slower than that of 124AP1-PBNC, but its dissociation was also slower, resulting in a stable complex with an overall affinity that was stronger than that of 124AP1-PBNC. The different peptide display patterns might be responsible for the different association and dissociation rates. For example, the closer proximity of peptides that are fused to C-termini loops may hinder accessibility and decrease association rate, but their peptide bunches also decrease dissociation rates and form stable complex, in comparison with N-termini modified PBNCs. We also examined the binding kinetics of the twosite-ligated PBNC, 124AP1/15724AP1-PBNC. Conjugation of AP1 peptides to both the N-terminal end and the loop, yielding an FTL displaying 48 peptides on its surface, resulted in an association rate that was even slower than that of 15724AP1PBNC, but it also slowed the dissociation rate 10-fold. The KD of 124AP1/15724AP1-PBNC for immobilized IL-4R was calculated to be ∼10−10 M and was termed a superaffinity PBNC. Binding responses increased continuously with addition of increasing concentrations of 124AP1/15724AP1-PBNCs, reflecting the remarkable stability and minimal dissociation of the resulting complex, whereas responses of one-site−ligated PBNCs became saturated (Figure 3B). The multivalent binding of nanoparticles is described by avidity, defined as the ability to form a stable complex with a target, which is dependent on both intrinsic affinity and number of binding sites.35 Thus, the superaffinity PBNCs, displaying 48 AP1 peptides, exhibited superior binding ability through formation of stable complexes, F

DOI: 10.1021/acs.biomac.5b01753 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 5. Surface plasmon resonance studies for the binding kinetics of RGD-conjugated bispecific PBNCs. (A) Binding responses of bispecific PBNCs against IL-4R coated surface at different injection concentrations. (B) Binding responses of bispecific PBNCs against αvβ3-coated surface at different injection concentrations. Wild-type FTL is control.

measuring binding responses using SPR analysis (Figure 5A). A broad range of AP1-PBNC concentrations was applied to assess overall binding avidity. Binding avidity was similar among AP1PBNCs, indicating that RGD peptide bunches on the surface did not hamper complex formation between AP1 and IL-4R. Next, human integrin αvβ3 was immobilized and binding responses were monitored upon application of a series of different concentrations of PBNCs surface-decorated with RGD alone or with RGD and AP1 peptides. Binding of RGDdecorated PBNCs to αvβ3 protein was dependent on displayed RGD because wt FTL showed only nonspecific, minimal binding responses. Multivalent conjugation of RGD peptides (RGD or GRGDSP) to the surface of FTL nanoparticles enhanced binding avidity by ∼103−104-fold (Table 2). Binding avidities of 124RGD-PBNCs and 124RGD/15724AP1-PBNCs were similar, implying that peptides in different bunches did not interfere with each other for binding to designated receptors. 124RGD-PBNCs showed a trend toward greater avidity of binding to αvβ3 compared with 15724RGD-PBNCs (∼3-fold), and 15724AP1-PBNCs showed a similar trend toward binding IL-4R more tightly than 124AP1-PBNCs; however, neither difference reached statistical significance. On the basis

Table 2. Quantified Binding Constants of the RGD-PBNCs to αvβ3 Measured by SPR

a b

ligand

KDa (M)

RGD peptide 157 RGD-PBNC 157 GRGDS-PBNC 1 RGD-PBNC 1 RGD 157AP1-PBNC 1 GRGDS 157AP1-PBNC

4.0 × 10−4 13.8 × 10−8 22.5 × 10−8 5.2 × 10−8 2.5 × 10−8 3.2 × 10−8

βb 2.9 1.8 7.6 1.6 1.2

× × × × ×

103 103 103 104 104

Obtained by saturated binding responses of SPR measurements. Multivalency parameter β = KD(free)/KD(multivalent).

of these observations, we selected 124RGD/15724AP1-PBNC as the bispecific PBNC for subsequent cell-based experiments. 1 24GRGDSP/15724AP1-PBNCs showed avidity similar to that of 124RGD/15724AP1-PBNCs, but they were less stable. Dual Functionalities of Bispecific-Targeting PBNCs. To further test bispecific-targeting PBNCs, we employed three cell lines that express αvβ3 and IL-4R: U87MG cells, which express αvβ3;36,37 A549 cells, which express IL-4R; and HT1080 cells, which express both αvβ3 and IL-4R.37 Receptor expression was G

DOI: 10.1021/acs.biomac.5b01753 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 6. Specific cellular binding. (A, B) FACS analysis of expression of integrin αvβ3, αvβ5, IL-4 receptor in U87MG, HT1080, and A549 cell lines. The expression level of receptors are plotted against each cell line (B). (C) Cell adhesion levels of integrin-expressing cells (U87MG, HT1080) onto PBNC or BSA coated wells were measured. Triplets of indicated amounts of WT-FTL, 157AP1-FTL, 1RGD-FTL, and 1RGD157AP1-FTL were coated and cells were incubated for 30 min at 37 °C. Bar charts show mean of adherent cells from three independent experiments (error bars; s.d.; **P < 0.01, ***P < 0.001, unpaired Student’s t test). (D) Cellular binding of A549 cells was detected by flow cytometry. A total of 400 nM of WTFTL, 157AP1-PBNC, 1RGD-PBNC, and 1RGD157AP1-PBNC were incubated for 20 min at 4 °C.

confirmed by FACS analysis (Figure 6A,B). Since the RGD peptide sequence is responsible for integrin-mediated adhesion

of cells to collagen, we tested the functionality of RGDdisplaying PBNCs by examining cell adhesion. To this end, we H

DOI: 10.1021/acs.biomac.5b01753 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

*Tel.: +82-53-420-4829. Fax: +82-53-422-1466. E-mail: [email protected].

coated 96-well plates with 124RGD-PBNCs, 1 24RGD/15724AP1-PBNCs, 15724AP1-PBNCs or wt FTL, and applied αvβ3-expressing U87MG or HT1080 cells for 30 min. RGD-decorated PBNCs specifically mediated cell adhesion in a dose-dependent manner, whereas 15724AP1-PBNCs and wt FTL did not. The cell adhesion activities of 124RGD-PBNCs and 124RGD/15724AP1-PBNCs were not significantly different (Figure 6C). HT1080 cells, which express both IL-4R and αvβ3, adhered to 15724AP1-PBNC-coated plates after a longer incubation (50 min) owing to interactions between AP1 and IL-4R (Supporting Information, Figure 1). AP1-mediated cell binding was examined using FACS analysis, which showed that 157 24AP1-PBNCs and 124RGD/15724AP1-PBNCs bound IL4R-expressing A549 cells, whereas 124RGD-PBNCs and wt FTL did not (Figure 6D). These results indicate that bispecifictargeting PBNCs have dual functionality, with the capacity for targeting two different sites in complex diseases.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (Ministry of Education; NRF-2015R1D1A3A01019018); by the National Research Foundation of Korea (NRF) grant funded by the Korea government (2014R1A5A2009242); by the KIST Institutional Program (Project No. 2E25270); by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIP, Ministry of Science, ICT and Future Planning; No. 2015R1A2A1A15056039).

■

■

CONCLUSIONS We have demonstrated that ferritin cage nanoparticles can be engineered for use as superaffinity and bispecific targeting agents through genetic incorporation of peptides onto two different sites. Upon assembly of the cage architecture, IL-4Rtargeting PBNCs containing 48 AP1 peptides, displayed in bunches on the nanocage surface, exhibit 10× higher binding avidity than 24-peptide AP1-PBNCs. The remarkable complex stability demonstrated here would be advantageous in therapeutic applications, reflecting the high selectivity and diminished toxicity. Bispecific-targeting PBNCs were developed by genetic incorporation of RGD and AP1 peptides on the same surface of ferritin nanocages. Peptides in each bunch were fully capable of binding their corresponding receptors, implying that ferritin-based nanocages can be simultaneously decorated by multifunctionalities at exposed sites of the N-terminal end and 4-fold symmetry loop. We also demonstrated that the display patterns of peptide bunches on the N-terminal end and loop are different: more dispersed in N-terminal-ligated bunches and more clustered in loop-ligated bunches. Binding kinetics and preference of peptides tended to differ depending on where the peptides were ligated. Although these specific differences were not significant, they suggest an avenue for future research that might fine-tune the development of peptide-conjugated nanocages. More extensions of this technology such as developing other bispecific targeting ferritins or loading small molecules into the PBNCs are opened for future research. Various combinations of superaffinity and multitargeting PBNCs could aid in the translation of many valuable ferritin-based nanomedicines to a clinical setting.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.biomac.5b01753. Additional data of cell adhesion assay are available (PDF).

■

REFERENCES

(1) Doll, T. A.; Raman, S.; Dey, R.; Burkhard, P. Nanoscale assemblies and their biomedical applications. J. R. Soc., Interface 2013, 10 (80), 20120740. (2) Maham, A.; Tang, Z.; Wu, H.; Wang, J.; Lin, Y. Protein-based nanomedicine platforms for drug delivery. Small 2009, 5 (15), 1706− 21. (3) Uchida, M.; Kang, S.; Reichhardt, C.; Harlen, K.; Douglas, T. The ferritin superfamily: Supramolecular templates for materials synthesis. Biochim. Biophys. Acta, Gen. Subj. 2010, 1800 (8), 834−45. (4) Dominguez-Vera, J. M.; Colacio, E. Nanoparticles of Prussian blue ferritin: a new route for obtaining nanomaterials. Inorg. Chem. 2003, 42 (22), 6983−5. (5) Pead, S.; Durrant, E.; Webb, B.; Larsen, C.; Heaton, D.; Johnson, J.; Watt, G. D. Metal ion binding to apo, holo, and reconstituted horse spleen ferritin. J. Inorg. Biochem. 1995, 59 (1), 15−27. (6) Haikarainen, T.; Thanassoulas, A.; Stavros, P.; Nounesis, G.; Haataja, S.; Papageorgiou, A. C. Structural and thermodynamic characterization of metal ion binding in Streptococcus suis Dpr. J. Mol. Biol. 2011, 405 (2), 448−60. (7) Lee, J. H.; Seo, H. S.; Song, J. A.; Kwon, K. C.; Lee, E. J.; Kim, H. J.; Lee, E. B.; Cha, Y. J.; Lee, J. Proteinticle engineering for accurate 3D diagnosis. ACS Nano 2013, 7 (12), 10879−86. (8) Jeon, J. O.; Kim, S.; Choi, E.; Shin, K.; Cha, K.; So, I. S.; Kim, S. J.; Jun, E.; Kim, D.; Ahn, H. J.; Lee, B. H.; Lee, S. H.; Kim, I. S. Designed nanocage displaying ligand-specific Peptide bunches for high affinity and biological activity. ACS Nano 2013, 7 (9), 7462−71. (9) He, D.; Marles-Wright, J. Ferritin family proteins and their use in bionanotechnology. New Biotechnol. 2015, 32 (6), 651−7. (10) Schoonen, L.; van Hest, J. C. Functionalization of protein-based nanocages for drug delivery applications. Nanoscale 2014, 6 (13), 7124−41. (11) Zhen, Z.; Tang, W.; Chuang, Y. J.; Todd, T.; Zhang, W.; Lin, X.; Niu, G.; Liu, G.; Wang, L.; Pan, Z.; Chen, X.; Xie, J. Tumor vasculature targeted photodynamic therapy for enhanced delivery of nanoparticles. ACS Nano 2014, 8 (6), 6004−13. (12) Liang, M.; Fan, K.; Zhou, M.; Duan, D.; Zheng, J.; Yang, D.; Feng, J.; Yan, X. H-ferritin-nanocaged doxorubicin nanoparticles specifically target and kill tumors with a single-dose injection. Proc. Natl. Acad. Sci. U. S. A. 2014, 111 (41), 14900−5. (13) Zhao, L.; Seth, A.; Wibowo, N.; Zhao, C. X.; Mitter, N.; Yu, C.; Middelberg, A. P. Nanoparticle vaccines. Vaccine 2014, 32 (3), 327− 37. (14) Kanekiyo, M.; Wei, C. J.; Yassine, H. M.; McTamney, P. M.; Boyington, J. C.; Whittle, J. R.; Rao, S. S.; Kong, W. P.; Wang, L.; Nabel, G. J. Self-assembling influenza nanoparticle vaccines elicit broadly neutralizing H1N1 antibodies. Nature 2013, 499 (7456), 102− 6. (15) Han, J. A.; Kang, Y. J.; Shin, C.; Ra, J. S.; Shin, H. H.; Hong, S. Y.; Do, Y.; Kang, S. Ferritin protein cage nanoparticles as versatile

AUTHOR INFORMATION

Corresponding Authors

*Tel.: +82-2-958-5935. Fax: +82-2-958-5478. E-mail: [email protected]. I

DOI: 10.1021/acs.biomac.5b01753 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

recognition drug modular nanoplatform. Biomaterials 2012, 33 (21), 5423−30. (34) Ghisaidoobe, A. B.; Chung, S. J. Functionalized protein nanocages as a platform of targeted therapy and immunodetection. Nanomedicine (London, U. K.) 2015, 10 (24), 3579−95. (35) Hong, S.; Leroueil, P. R.; Majoros, I. J.; Orr, B. G.; Baker, J. R., Jr.; Banaszak Holl, M. M. The binding avidity of a nanoparticle-based multivalent targeted drug delivery platform. Chem. Biol. 2007, 14 (1), 107−15. (36) Zhou, Y.; Kim, Y. S.; Chakraborty, S.; Shi, J.; Gao, H.; Liu, S. 99mTc-labeled cyclic RGD peptides for noninvasive monitoring of tumor integrin alphaVbeta3 expression. Mol. Imaging 2011, 10 (5), 386−97. (37) Sutherland, M.; Gordon, A.; Shnyder, S. D.; Patterson, L. H.; Sheldrake, H. M. RGD-Binding Integrins in Prostate Cancer: Expression Patterns and Therapeutic Prospects against Bone Metastasis. Cancers 2012, 4 (4), 1106−45.

antigen delivery nanoplatforms for dendritic cell (DC)-based vaccine development. Nanomedicine 2014, 10 (3), 561−9. (16) Jin, R.; Lin, B.; Li, D.; Ai, H. Superparamagnetic iron oxide nanoparticles for MR imaging and therapy: design considerations and clinical applications. Curr. Opin. Pharmacol. 2014, 18, 18−27. (17) Geninatti Crich, S.; Bussolati, B.; Tei, L.; Grange, C.; Esposito, G.; Lanzardo, S.; Camussi, G.; Aime, S. Magnetic resonance visualization of tumor angiogenesis by targeting neural cell adhesion molecules with the highly sensitive gadolinium-loaded apoferritin probe. Cancer Res. 2006, 66 (18), 9196−201. (18) Cao, C.; Wang, X.; Cai, Y.; Sun, L.; Tian, L.; Wu, H.; He, X.; Lei, H.; Liu, W.; Chen, G.; Zhu, R.; Pan, Y. Targeted in vivo imaging of microscopic tumors with ferritin-based nanoprobes across biological barriers. Adv. Mater. 2014, 26 (16), 2566−71. (19) Hwang, M. P.; Lee, J. W.; Lee, K. E.; Lee, K. H. Think modular: a simple apoferritin-based platform for the multifaceted detection of pancreatic cancer. ACS Nano 2013, 7 (9), 8167−74. (20) Fantechi, E.; Innocenti, C.; Zanardelli, M.; Fittipaldi, M.; Falvo, E.; Carbo, M.; Shullani, V.; Di Cesare Mannelli, L.; Ghelardini, C.; Ferretti, A. M.; Ponti, A.; Sangregorio, C.; Ceci, P. A smart platform for hyperthermia application in cancer treatment: cobalt-doped ferrite nanoparticles mineralized in human ferritin cages. ACS Nano 2014, 8 (5), 4705−19. (21) Park, K. Facing the truth about nanotechnology in drug delivery. ACS Nano 2013, 7 (9), 7442−7. (22) Kell, D. B.; Pretorius, E. Serum ferritin is an important inflammatory disease marker, as it is mainly a leakage product from damaged cells. Metallomics 2014, 6 (4), 748−73. (23) Hong, H. Y.; Lee, H. Y.; Kwak, W.; Yoo, J.; Na, M. H.; So, I. S.; Kwon, T. H.; Park, H. S.; Huh, S.; Oh, G. T.; Kwon, I. C.; Kim, I. S.; Lee, B. H. Phage display selection of peptides that home to atherosclerotic plaques: IL-4 receptor as a candidate target in atherosclerosis. J. Cell. Mol. Med. 2008, 12 (5B), 2003−14. (24) Edwards, C. M.; Cohen, M. A.; Bloom, S. R. Peptides as drugs. Qjm 1999, 92 (1), 1−4. (25) Son, H. N.; Nam, J. O.; Kim, S.; Kim, I. S. Multiple FAS1 domains and the RGD motif of TGFBI act cooperatively to bind alphavbeta3 integrin, leading to anti-angiogenic and anti-tumor effects. Biochim. Biophys. Acta, Mol. Cell Res. 2013, 1833 (10), 2378−88. (26) Liu, Z.; Wang, F.; Chen, X. Integrin alpha(v)beta(3)-Targeted Cancer Therapy. Drug Dev. Res. 2008, 69 (6), 329−339. (27) Li, Z.; Jiang, J.; Wang, Z.; Zhang, J.; Xiao, M.; Wang, C.; Lu, Y.; Qin, Z. Endogenous interleukin-4 promotes tumor development by increasing tumor cell resistance to apoptosis. Cancer Res. 2008, 68 (21), 8687−94. (28) Zhen, Z.; Tang, W.; Chen, H.; Lin, X.; Todd, T.; Wang, G.; Cowger, T.; Chen, X.; Xie, J. RGD-modified apoferritin nanoparticles for efficient drug delivery to tumors. ACS Nano 2013, 7 (6), 4830−7. (29) Uchida, M.; Flenniken, M. L.; Allen, M.; Willits, D. A.; Crowley, B. E.; Brumfield, S.; Willis, A. F.; Jackiw, L.; Jutila, M.; Young, M. J.; Douglas, T. Targeting of cancer cells with ferrimagnetic ferritin cage nanoparticles. J. Am. Chem. Soc. 2006, 128 (51), 16626−33. (30) Lin, X.; Xie, J.; Niu, G.; Zhang, F.; Gao, H.; Yang, M.; Quan, Q.; Aronova, M. A.; Zhang, G.; Lee, S.; Leapman, R.; Chen, X. Chimeric ferritin nanocages for multiple function loading and multimodal imaging. Nano Lett. 2011, 11 (2), 814−9. (31) Chi, L.; Na, M. H.; Jung, H. K.; Vadevoo, S. M.; Kim, C. W.; Padmanaban, G.; Park, T. I.; Park, J. Y.; Hwang, I.; Park, K. U.; Liang, F.; Lu, M.; Park, J.; Kim, I. S.; Lee, B. H. Enhanced delivery of liposomes to lung tumor through targeting interleukin-4 receptor on both tumor cells and tumor endothelial cells. J. Controlled Release 2015, 209, 327−36. (32) Kwon, M.; Kim, J. W.; Roh, J. L.; Park, Y.; Cho, K. J.; Choi, S. H.; Nam, S. Y.; Kim, S. Y.; Lee, B. H. Recurrence and cancer-specific survival according to the expression of IL-4Ralpha and IL-13Ralpha1 in patients with oral cavity cancer. Eur. J. Cancer 2015, 51 (2), 177−85. (33) Kang, H. J.; Kang, Y. J.; Lee, Y. M.; Shin, H. H.; Chung, S. J.; Kang, S. Developing an antibody-binding protein cage as a molecular J

DOI: 10.1021/acs.biomac.5b01753 Biomacromolecules XXXX, XXX, XXX−XXX