Construction and Application of Elastin Like Polypeptide Containing IL-4 Receptor Targeting Peptide

Various human solid tumors highly express IL-4 receptors which amplify the expression of some of anti-apoptotic proteins, preventing drug-induced cancer cell death. Thus, IL-4 receptor targeted drug delivery can possibly increase the therapeutic efficacy in cancer treatment. Macromolecular carriers with multivalent targeting moieties offered great advantages in cancer therapy as they not only increase the plasma half-life of the drug but also allow delivery of therapeutic drugs to the cancer cells with higher specificity, minimizing the deleterious effects of the drug on normal cells. In this study we designed a library of elastin like polypeptide (ELP) polymers containing tumor targeting AP1 peptide using recursive directional ligation method. AP1 was previously discovered as an atherosclerotic plaque and breast tumor tissue homing peptide using phage display screening method, and it can selectively bind to the interleukin 4 receptor (IL-4R). The fluorescently labeled [AP1-V12]6, an ELP polymer containing six AP1 enhanced tumor-specific targeting ability and uptake efficiency in H226 and MDA-MB-231 cancer cell lines in vitro. Surface plasmon resonance analysis showed that multivalent presentation of the targeting ligand in the ELP polymer increased the binding affinity towards IL-4 receptor compared to free peptide. The binding of [AP1-V12]6 to cancer cells was remarkably reduced when IL-4 receptors were blocked by antibody against IL-4 receptor further confirmed its binding. Importantly, the Cy5.5-labeled [AP1-V12]6 demonstrated excellent homing and longer retention in tumor tissues in MDA-MB-231 xenograft mouse model. Immunohistological studies of tumor tissues further validated the targeting efficiency of [AP1-V12]6 to tumor tissue. These results indicate that designed [AP1-V12]6 can serve as a novel carrier for selective delivery of therapeutic drugs to tumors.


Introduction
Targeted macromolecular polymer carriers offer the potential of effective drug delivery by virtue of their ability to decrease the rate of drug clearance after systematic administration and improve the plasma half-life of drugs [1]. One such engineered biopolymer is elastin-like polypeptide (ELP), which is an emerging drug carrier under development for cancer therapy. In several cancer models, ELPs have already been used for targeted delivery of small molecules drugs (doxorubicin) [2,3], therapeutic peptides (c-Myc inhibitory peptide) [4] and proteins [5]. They have unique characteristics compared to other polymeric drug delivery systems, including low toxicity, good biodegradability, and biocompatibility [6,7]. ELP biopolymers are derived from the structural motif found in mammalian elastin protein, which consists of the pentapeptide repeat, Val-Pro-Gly-Xaa-Gly (VPGXG), where Xaa (''guest'' residue) can be any amino acid except proline [8,9]. ELP polymers can be synthesized at the genetic level using recombinant DNA methods; thus, their sequence, composition, and molecular weight can be controlled. Accordingly, the hydrophobicity and degree of ionization can be precisely tuned for proper tissue distribution and subcellular uptake [10,11]. Moreover, the numbers of targeting peptides or specific reactive sites for drug conjugation can be incorporated at the genetic level along with the ELP sequence, which is one of the difficulties to accomplish using synthetic polymers. ELPs are soluble in aqueous solutions below their transition temperature (T t ), but hydrophobically collapse and aggregate at temperatures greater than T t [12,13]. This inverse transition temperature is fully reversible. The transition temperature of ELP can be controlled by varying the guest residue, molecular weight and concentration [10]. Notably, ELPs can be easily expressed and purified at high yield simply by exploiting inverse temperature cycling (ITC) method [14,15].
Using phage display screening, we previously discovered an atherosclerotic plaque and breast tumor tissue homing peptide, CRKRLDRN, termed AP1 peptide that selectively binds to the interleukin -4 receptors (IL-4Rs) [16][17][18]. IL-4Rs are highly expressed in a wide variety of human tumors, including renal cell carcinoma, squamous cell carcinoma of the head and neck, malignant glioma, AIDS-associated Kaposi's sarcoma and breast cancer cell lines [19][20][21][22]. It has been reported that IL-4/IL-4R interactions amplify the expression of some anti-apoptotic proteins, including PED/PEA15 (15 kDa phosphoprotein enriched in astrocytes), cFLIP/CFLAR (CASP8 and FADD-like apoptosis regulator), and the BCL family proteins Bcl-xL and Bcl, thereby preventing drug-induced cancer cell death [23,24]. Thus, targeting the IL-4R could possibly increase the therapeutic efficacy of anticancer drugs. Even though AP1 peptide possesses high target-specificity and efficient tumor-homing, it has limitation due to easy elimination from the circulation and short half-life.
The aim of this study is to increase the stability and binding affinity of the AP1 peptide in vivo, and thus increase its utility, through multivalent presentation of the AP1 peptide in an ELP polymer backbone. Many studies have demonstrated that the presentation of multiple targeting moieties on a polymer backbone can improve binding avidity and specificity compared to monovalent ligand presentation [25]. Accordingly, we designed and prepared a series of an AP1 peptide containing ELP polymer using recursive directional ligation (RDL) method and analyzed the characteristics and the efficacy in vitro and in vivo. The binding specificity of [AP1-V 12 ] 6 polymer towards cancer cells was analyzed using confocal microscopy and flow cytometry, and in vivo optical imaging in a tumor xenograft mouse model. The studies have revealed that the incorporation of multivalent targeting peptide ligand into ELP polymer facilitated the greater targeting to IL-4R expressing cancer cells.

Designing of monomer gene and oligomerization
Synthetic oligonucleotides encoding genes of [VGVPG] 14 and VGRKRLDRNG[VGVPG] 12 referred as V 14 and AP1-V 12 were designed to contain BamH I and HinD III compatible cohesive ends upon annealing, which allows the annealed product to be directly ligated into a BamH I and HinD III-cleaved pRSET B+ vector (Invitrogen, CA, USA) [26,27]

Thermal characterization
Transition temperature (T t ) of [V 14 ] 6 and [AP1-V 12 ] 6 were determined by monitoring the turbidity profile of protein solutions at wavelength 350 nm as a function of temperature using Cary UV-visible spectrophotometer equipped with temperature controller (Agilent Technologies). The absorbance was monitored from 20uC to 45uC in 1uC/min increments. The T t of [V 14 ] 6 and [AP1-V 12 ] 6 protein were determined at a concentration of 10 mM.  for 1 h at 4uC. After washing with PBS for two times, cells were suspended with 300 ml of PBS and analyzed by flow cytometry. 10,000 events were analyzed for each sample.

Confocal Microscopy
H226, MDA-MB-231 and H460 cells were seeded on four chambered slide and grown to 80% confluence. Cells were then incubated with 10 mM Alexa 488 labeled [V 14 ] 6 and [AP1-V 12 ] 6 proteins and AP1 peptides for 1 h at 4uC and 37uC. Unbound peptides were washed out with PBS, and cells were fixed with 4% paraformaldehyde (Sigma Aldrich). Cell nuclei were stained with 49,6-diamidino-2-phenylindole (DAPI; Sigma Aldrich), and chamber slides were mounted with anti-fade reagent (Invitrogen). Images were captured and analyzed in sequential scanning mode using a Zeiss LSM-510 Meta confocal microscope.

Surface Plasmon Resonance (SPR) analysis
Interactions of AP1, [V 14 ] 6 and [AP1-V 12 ] 6 with the IL-4R were analyzed at 25uC using a surface plasmon resonance instrument (Reichert Life Sciences, NY, USA). IL-4Ra (CD124) (Sino Biological, Beijing, China) was immobilized by activating the carboxy methyl group on dextran-coated chips (Reichert Life Sciences) through a reaction with N-hydroxysuccinimide (Sigma Aldrich), followed by covalent bonding of the ligands to the chip surface via amide linkages and blocking of excess activated carboxyls with ethanolamine. Different concentrations of AP1 peptide (3.125 to 50 mM) and [AP1-V 12 ] 6 (125 nM to 1 mM) in binding buffer were allowed to flow over surfaces containing immobilized IL-4Ra (20006400 RU) for 3 min at a rate of 25 mL/min. The sensor surface was regenerated after each association and dissociation cycle by injecting 10 mM HCl for 10 s. The interaction of ligand and analyte was analyzed using Scrabber 2 Biologic software. Binding kinetics was assessed by determining association (k on ), dissociation (k off ), and equilibrium (K D ) constants.

Animal Model
This study strictly followed the recommendations of National Institute of Health (NIH) for the Care and Use of Laboratory Animals. Animal experiments were reviewed and approved by the Committee on the Ethics of Animal Experiments of the Kyungpook National University (Permit Number: KNU 2012-124). All efforts were made for minimizing animal suffering. Female nude mice (BALB/c nude; body weight, 2063 g; n = 5) were housed in a specific pathogen-free environment at 2262uC, 5565% relative humidity with light. Tumors were generated by subcutaneously injecting MDA-MB-231 cells (5610 6 cells) into the right flank of 5-wk-old BALB/c nude female mice and allowing them to grow for 10days [17,29]. For in vivo analysis of tumor targeting, [V 14 ] 6 and [AP1-V 12 ] 6 were labeled with Alexa 680 (Cy5.5) at the C-terminal Cys residue of the protein. Tumorbearing mice were anesthetized under inhalational anesthesia (1%, w/v, isofurane in 2 L oxygen), then injected with Cy5.5-labeled [AP1-V 12 ] 6 (n = 5) or, [V 14 ] 6 control (n = 5) of approximately 8 mg/kg via the tail vein. In vivo fluorescence images were taken at different time intervals after anesthetization (10 min, 1 h, 6 h, 12 h, and 24 h) using an eXplore Optix system (ART Advanced research technologies Inc., Montreal, Canada).

Tissue preparation
At 6 h time point, animals were euthanized with CO 2 , tumors and organs were removed from a subset of animals, and ex vivo fluorescence images were collected. The tumor tissues were then fixed by incubating with 4% paraformaldehyde overnight and frozen for cryosectioning. Tissue slices (8-mm thick) were incubated overnight at 4uC with anti-IL-4R antibody (R&D Systems; 1:100) and then incubated for 30 min at room temperature with Alexa 488-labeled goat anti-mouse IgG secondary antibody (1:200). After staining nuclei with DAPI, sections were slide-mounted and observed under a confocal microscope.

Statistical analysis
The statistical significance was determined using Student's t-test for two groups and one-way ANOVA for comparing multiples groups. ***P,0.0001, **P,0.001, and *P,0.05 were considered as statistically significant and denoted by asterisks in the Figures.

Design of [AP1-V 12 ] n genes and protein expression
We successfully constructed [V 12 ] n and [AP1-V 12 ] n gene libraries using the RDL method [11]. The ELP libraries consisted of the monomer genes repeat Val-Pro-Gly-Val-Gly with Valine at the guest residue of ELP pentapeptide (Fig. 1A), whereas the [AP1-V 12 ] n library contained modified AP1 sequence, in the N-terminal region of the ELP coding sequences (Fig. 1B). The monomer gene of [AP1-V 12 ] n was designed such that at each round of gene oligomerization by RDL, the AP1 sequence was repeated along with the ELP sequence. This iterative process yielded [V 12 ] n (  14 ] 6 WPC proteins were purified using the ITC method; the total yield was ,20 mg/ L. After four rounds of ITC, SDS-PAGE followed by Coomassie blue staining showed that [AP1-V 12 ] 6 and [V 14 ] 6 were approximately ,37 kDa and ,35 kDa in size, respectively, with minimal contamination (Fig. 2A).

Thermal characterization
The T t of [AP1-V 12 ] 6 and [V 14 ] 6 proteins were characterized at a concentration of 10 mM in PBS. The turbidity profiles were determined by measuring optical density at 350 nm (OD 350 ) as function of temperature in 1uC min 21 increments. The T t , defined as the temperature at which turbidity in the protein solution reached 50%, was in the range of 37uC to 39uC for [AP1-V 12 ] 6 and 26uC to 28uC for and [V 14 ] 6 ( Fig. 2B). Incorporation of the hydrophilic AP1 sequence into the ELP gene increased the T t compared to the ELP control [30]. We observed that the inverse transition of the polymer [AP1-V 12 ] 6 is just above the physiological body temperature, which is preferable for clinical applications.

In vitro cell binding of [AP1-V 12 ] 6 polymer
To study binding efficiency, we labeled the C-terminal Cys residue of [AP1-V 12 ] 6 and [V 14 ] 6 proteins with Alexa 488. FACS analysis confirmed that the cell-binding capacity of the targeted [AP1-V 12 ] 6 polymer in H226 (Fig. 3A, B) and MDA-MB-231 cells (Fig. 3C, D) after a 1-h incubation at 4uC was higher than that of the non-targeted [V 14 ] 6 polymer (P,0.0001). Both non-targeted [V 14 ] 6 and targeted [AP1-V 12 ] 6 polymers showed significantly lower binding towards H460 cells (Fig. 3E, F). For the [AP1-V 12 ] 6 polymer, cell attachment was 5.961.6 fold higher in H226 cells and 5.661.3 fold higher in MDA-MB-231 cells compared to H460 cells. In addition, the AP1 peptide showed 4.261.8 and 3.461.1 folds higher binding to H226 and MDA-MB-231 cells than to H460 cells. Thus, the [AP1-V 12 ] 6 polymer exhibited 1.7 and 2.2 fold higher cell-binding capacity in H226 (P,0.0001) and MDA-MB-231(P,0.05) cells than the targeting AP1 peptide alone. Our results suggested that multivalent presentation of the AP1 sequence in the [AP1-V 12 ] 6 polymer allowed greater accumulation on cells compared to the AP1 peptide.
The IL-4R dependence of [AP1-V 12 ] 6 polymer binding to tumor cells was further confirmed using competition assays. The binding of tumor cells by [AP1-V 12 ] 6 polymer was remarkably reduced in a concentration-dependent manner by pre-incubation with different concentrations (1, 5 and 10 mg/mL) of anti-IL-4R antibody (Fig. 3 G, H).

Cellular localization of [AP1-V 12 ] 6 polymer and affinity for IL-4R
The cellular localization of [AP1-V 12 ] 6 polymer at different temperatures were further confirmed by confocal microscopic imaging, which clearly revealed that [AP1-V 12 ] 6 polymer and AP1 peptide accumulated more efficiently to the cell surface of H226 (Fig. 4A) and MDA-MB-231 (Fig. S2A) cells at 4uC compared to H460 cells (Fig. S3A). [AP1-V 12 ] 6 polymer was well internalized when incubated at 37uC whereas no uptake was seen in case of [V 14 ] 6 polymer in both H226 (Fig. 4B) and MDA-MB-231 cells (Fig. S2B). Both non-targeted [V 14 ] 6 and targeted [AP1-V 12 ] 6 polymers have shown lower cell binding and uptake in case of IL-4R negative H460 cells (Fig. S3A, B). Thus, these results indicated that cellular uptake of [AP1-V 12 ] 6 polymers were not due to  nonspecific effects of increased temperature. Further evidence for cellular localization of [AP1-V 12 ] 6 polymer on the cells were obtained by exploiting confocal microscopic Z-series images captured in sequential scanning mode, showed that the [AP1-V 12 ] 6 polymer was well distributed on the plasma membrane of H226 cells at 4uC (Fig. 4A, right panel) and internalized into cytoplasm when temperature was increased to 37uC (Fig. 4B, right  panel).

Surface plasmon resonance analysis to check [AP1-V 12 ] 6 protein affinity for IL-4R
The kinetics of AP1 peptide and [AP1-V 12 ] 6 protein binding to the IL-4R were analyzed using surface plasmon resonance [31,32]. The [AP1-V 12 ] 6 polymer (Fig. 6A) showed higher binding affinity than the AP1 peptide (Fig. 6B) and [V 14 ] 6 (Fig. 6C). The equilibrium constants (K D ) of free AP1 peptide and [AP1-V 12 ] 6 protein were found to be 5.5460.3610 23 and 2.0760.3610 27 , respectively (Table 1). Overall, the affinity of the [AP1-V 12 ] 6 polymer for the IL-4R was ,10000-fold higher than that of free AP1 peptide. No binding of [V 14 ] 6 was detected, even at concentrations as high as 1 mM. Thus, multivalent presentation of the targeting ligand AP1 on the ELP polymer backbone indeed increased the binding affinity towards the IL-4R.

In vivo imaging of [AP1-V 12 ] 6 polymer tumor targeting
The in vivo tumor-targeting efficiency of the [AP1-V 12 ] 6 polymer was studied using a near-infrared fluorescence (NIRF) live optical imaging system. In vivo fluorescence images taken at different time intervals showed that the [AP1-V 12 ] 6 polymer rapidly accumulated at tumor tissue (within a little as 10 min of injection) in MDA-MB-231 (Fig. 7A, B) breast cancer-bearing mice and was retained in tumor tissues for 24 h after injection. This specific targeting and sustained localization within tumors could provide sufficient time for [AP1-V 12 ] 6 polymer-conjugated drugs to exert their effects. In contrast, [V 14 ] 6 treated mice showed minimal fluorescence intensity at the tumors tissue and more nonspecific tissue localization. Six hours after injection, mice were sacrificed and ex vivo fluorescence images of tumor and excised organs were collected to evaluate the biodistribution of [AP1-V 12 ] 6 and [V 14 ] 6 . [AP1-V 12 ] 6 fluorescence intensity was higher in the target tumor compared with that of [V 14 ] 6 , confirming the increased tumortargeting specificity and retention time of the AP1 peptidemodified ELP polymer. However, fluorescence intensity was also found to be strong in the liver and kidney of both targeted [AP1-V 12 ] 6 and non-targeted [V 14 ] 6 injected mice, indicating rapid metabolism and secretion in urine (Fig. 7C). An immunohistological examination of tumor tissue showed that abundant staining for [AP1-V 12 ] 6 was confined to the tumor tissue, consistent with our in vivo and ex vivo imaging results (Fig. 7D).

Discussion
Cancer cells express a number of cell surface and matrix proteins that mediate tumor growth, migration, invasion, and metastasis. Screening for ligands that specifically target these receptors represents an excellent cancer therapy strategy [33,34]. Initial studies using phage display technology discovered an IL-4R binding peptide (AP1) that was highly expressed on atherosclerotic plaques and cancer cells [16,17]. The AP1 peptide was found to bind the IL-4R with low affinity (i.e., micromolar dissociation constant), which means that more targeting peptide is necessary to achieve high receptor occupancy. To improve the binding affinity of AP1 as well as its avidity towards the IL-4 receptor, we incorporated multiple AP1 peptides into the ELP polymer.
The genetically encoded synthesis of [AP1-V 12 ] 6 allows easy synthesis of ligand-presenting ELP polymers and makes it possible to control their molecular weight and T t to conform the requirements of physiological conditions. The ELP polymer backbone is capable of accommodating virtually any target specific ligand and can support multivalent presentation of target ligands without a change in its physical properties [35,36]. Many researchers have investigated the potential of ELP based incorporation of peptide as targeting moieties and cell penetrating peptide to improve the uptake by cancer cells [37,38]. Moreover there are many reports on application of diblock ELPs polymers consisting of a hydrophilic block and hydrophobic block capable of forming monodisperse spherical micelles above critical micelle temperature (CMT) for presentation of targeting ligands at their N-terminus site which enables multivalent presentation on their selective targets [36,39].
In this study for the first time we attempted to design ELP based multivalent targeting polymer by accommodating tandem repeat of highly specific IL-4R binding ligands, AP1 along the ELP polymer to improve its presentation towards IL-4R highly expressed cancer cells. Incorporation of AP1 on ELP elevated the T t compared to ELP alone, likely due to the charged and polar surface residues present on AP1 [30]. Trabbic-Carlson et al. 2004 [30] already reported that the T t of ELP fusion protein was negatively correlated with the fraction of hydrophobic area  6 and AP1 for 1 h at 4uC. Cell binding was determined using flow cytometry. Histograms are representative of three independent experiments. Graphical bars (on right) represent the percent of Alexa 488 labeled polymer bound to cells as mean 6SD of data obtained from three separate experiments performed in triplicates. ***P,0.0001, **P,0.001, and *P,0.05, one-way ANOVA; n = 3. (G, H) H226 cells (3610 5 cells) were pre-incubated with different concentrations (1, 5 and 10 mg/mL) of anti-IL-4 receptor antibody followed by 1 h incubation with 10 mM Alexa-labeled [AP1-V 12 ] 6 at 4uC. The cells were further suspended in 300 mL of PBS after washing and analyzed using flow cytometry. Histograms are representative of three independent experiments. Graphical bars represent the percent of Alexa 488 labeled polymer bound to cells as mean 6SD of data obtained from three separate experiments performed in triplicates. ***P,0.0001, One way ANOVA; n = 3. doi:10.1371/journal.pone.0081891.g003 AP1-ELP Based Tumor Targeting PLOS ONE | www.plosone.org presented on the surface of the fused folded protein. Proteins with relatively high hydrophobic solvent-accessible surface area depressed the inverse transition temperature of the fused ELP whereas proteins with more hydrophilic surfaces slightly increase elevated the transition temperature of the ELP fusion protein relative to that of the ELP.
The multivalent presentation of the targeting ligand AP1 on the ELP polymer backbone increased binding affinity ,10000-fold compared to the free AP1 peptide clearly denoted its augmentation in affinity and IL-4R interaction which is important for tumor targeting approach. In vitro studies further showed significantly higher binding of the [AP1-V 12 ] 6 polymer to IL-4R highly expressed H226 lung cancer and MDA-MB-231 human breast cancer cells than to IL-4R negative H460 cells. The increased cellular localization of the [AP1-V 12 ] 6 polymer over free AP1 peptide thus verifies the effectiveness of the multivalent, ELPdelivery strategy in improving tumor-targeting activity. Competitive inhibition of [AP1-V 12 ] 6 binding by different concentrations of anti-IL-4R antibody confirmed the specificity towards IL-4R.
Additionally confocal imaging results revealed the increased accumulation and uptake of [AP1-V 12 ] 6 polymer than [V 14 ] 6 control by IL-4R highly expressed tumor cells. Consistent with other findings [4,40], we believed that the significant enhancement in cellular uptake of thermally responsive [AP1-V 12 ] 6 polymer was due to increased temperature mediated phase transition which could possibly increase the exposure of IL-4R binding site to the cells thereby enhanced localization on cell surface and mediate cellular uptake.   [18]. Notably, our genetically engineered ELP-based AP1 polymers further enhanced tumor targeting ability, showing high tumor accumulation that was evident as early as 10 min after intravenous injection, peaked after about 6 h, and was maintained up to 24 h (Fig. 7). With the exception of kidney and liver, fluorescence levels were negligible in off-target organs, including the lung, spleen, and heart. Even though our designed the [AP1-V 12 ] 6 polymer has T t just above the physiological body temperature, which is preferable for clinical applications [42], T t of [V 14 ] 6 control is found to be relatively lower than the [AP1-V 12 ] 6 . However, the difference in T t was not an emphasis due to application of lesser concentration of nontargeting and targeting polymer during in vivo experiments. Besides, [V 14 ] 6 control was found to accumulate in various organs post intravenous injection except in the tumor tissue unlike the case of [AP1-V 12 ] 6 injected groups with precise localization to tumor tissue. Collectively, these findings highlight the promise of our [AP1-V 12 ] 6 polymer as a candidate macromolecular drug delivery system. Although our ELP-based polymers can target only IL-4R overexpressed cancer cells, many tumor of different origin have been reported to highly expressed IL-4R [19][20][21][22] and thus our designed polymer can be used as a potent drug carrier for various targeted cancer therapies.
The strategy of designing macromolecular carrier systems with multivalent ligand presentation for targeting surface molecules displayed on solid tumors has clear value in clinical applications. Because of relatively simple and novel synthetic process and the ability to use multivalent peptides and targeting moieties, we proposed a method which is distinctive and explore a new arena of multivalent cancer targeting. Although our multivalent targeting system exhibited effective tumor accumulation, further study is needed to realize the promise of ELP-based targeting polymers as carriers for delivering chemotherapeutic agents and therapeutic peptides.  Supporting Information