A Heparan Sulfate-Binding Cell Penetrating Peptide for Tumor Targeting and Migration Inhibition

As heparan sulfate proteoglycans (HSPGs) are known as co-receptors to interact with numerous growth factors and then modulate downstream biological activities, overexpression of HS/HSPG on cell surface acts as an increasingly reliable prognostic factor in tumor progression. Cell penetrating peptides (CPPs) are short-chain peptides developed as functionalized vectors for delivery approaches of impermeable agents. On cell surface negatively charged HS provides the initial attachment of basic CPPs by electrostatic interaction, leading to multiple cellular effects. Here a functional peptide (CPPecp) has been identified from critical HS binding region in hRNase3, a unique RNase family member with in vitro antitumor activity. In this study we analyze a set of HS-binding CPPs derived from natural proteins including CPPecp. In addition to cellular binding and internalization, CPPecp demonstrated multiple functions including strong binding activity to tumor cell surface with higher HS expression, significant inhibitory effects on cancer cell migration, and suppression of angiogenesis in vitro and in vivo. Moreover, different from conventional highly basic CPPs, CPPecp facilitated magnetic nanoparticle to selectively target tumor site in vivo. Therefore, CPPecp could engage its capacity to be developed as biomaterials for diagnostic imaging agent, therapeutic supplement, or functionalized vector for drug delivery.


Introduction
Carcinoma is a malignant cancer originating in the ectodermal and endodermal epithelial cells. Interaction between cell surface and microenvironment plays a crucial role in malignant tumor progression. Alterations of cell surface receptor, coreceptor, and adhesive protein expression are reported in various cancer types in vitro and in vivo [1][2][3]. Abnormal expression of cell surface molecules notably contributes to enhance tumor cell growth, survival, migration, and invasiveness [4]. Characterization of such alterations and development of novel agent for specific targeting are unmet medical need for early cancer diagnosis.
Glycosaminoglycans (GAGs) including heparan sulfate (HS), chondroitin sulfate (CS), keratan sulfate (KS), or dermatan sulfate (DS) are covalently attached to their core proteins to form proteoglycans. HS proteoglycan (HSPG) present in the extracellular matrix (ECM) provides structural frameworks to mediate cell-cell communication and function in growth factor-receptor binding [5,6]. HSPGs are key players in modulating tumor progression processes including metastasis, angiogenesis, proliferation, and malignant transformation [4]. Thus, upregulation of cell surface HS may play an active and crucial role in directing malignant phenotype of cancer during different developmental stages.
Cell penetrating peptides (CPPs) are short-chain cationic and/or amphipathic peptides which may be internalized into living cells [7]. CPPs are able to mediate translocation of a conjugated cargo (e.g., anticancer therapeutics) across plasma membrane, providing an effective and nontoxic mechanism 2 BioMed Research International for drug delivery [8]. Most CPPs are rich in positively charged Arg and Lys residues and are internalized after initially interacting with cell surface negatively charged GAGs which cluster CPPs on outer membrane surfaces [9,10].
CPPs might be potentially used in clinical procedures such as gene therapy and cancer therapy [8,11]. However, most CPPs are unfeasible for in vivo researches due to nonspecificity of their highly cationic characteristics. Cell surface negatively charged HS initializes the contact of CPPs, so particular HS binding CPPs might own mysterious sequence to exert multiple functions including HS binding, cellular binding, lipid binding, and in vivo tissue targeting activities. CPPecp is a recently identified CPP not only binding to negatively charged molecules including GAGs and lipids on cell surface in vitro but also targeting mucosal tissues in vivo [12][13][14]. In this study, we aim to collect and analyze the characteristics of HS-binding cell penetrating peptides derived from natural proteins. Besides, CPPecp itself falling in this classification has demonstrated multiple functions including in vitro tumor binding, tumor migration inhibition and angiogenesis inhibition activities, and in vivo cargo delivery to tumor site. Here, we provide more clues for the design of peptide therapeutics or intratumor delivery strategy by linking of a tumor targeting CPP. Furthermore, CPPecp might be a unique HS probe for cancer diagnosis to facilitate the quality of therapeutic index and molecular imaging in translational medicine. [32][33][34][35][36][37][38][39][40][41] or CPPecp with N-terminally conjugated fluorescein isothiocyanate (FITC) or tetramethylrhodamine (TMR) were synthesized by Genemed Synthesis Inc. and their purities (>90%) were assessed by analytical high-performance liquid chromatography. Peptide sequences were confirmed by matrix-assisted laser desorption/ionisation time-of-flight mass spectrometry in Genemed Synthesis Inc.

Flow
Cytometry. Cells (3.0 × 10 5 /well) were added into six-well plates and cultured in the indicated medium. After 24 h, 5 M FITC-CPPecp dissolved in medium was added into a well and the samples were incubated for 1 h. Cells were then harvested, washed, and suspended in PBS. The fluorescent intensities of the cell samples were measured using a FACSCalibur flow cytometer (BD Biosciences, Franklin Lakes, NJ) and excitation and emission wavelengths of 488 nm and 515-545 nm, respectively. The relative internalization of FITC-CPPecp was reported as the mean fluorescent signal for 10,000 cells.

Fluorescence Microscopy.
CT-26 cells were cultured on coverslips (5.0 × 10 3 /coverslip) in RPMI-1640. After 24 h, cell samples were incubated with FITC or FITC-CPPecp at 37 ∘ C for 10 min. Alternatively, CT-26 cells were pretreated with heparinase II ( 2.5 mU/mL) (Sigma-Aldrich, Missouri, USA) at 37 ∘ C for 2 h followed by treatment with 5 M TMR-CPPecp at 37 ∘ C for 10 min. The cells were then washed twice with PBS, fixed with 4% (w/v) paraformaldehyde, and rinsed twice with PBS. The coverslips were mounted in a Vectashield antifade mounting medium with DAPI (Vector Labs). Inverted fluorescent microscopy was performed using Axiovert 135 (Carl Zeiss, Göttingen, Germany) to assess the distribution of the FITC-CPPecp or TMR-ECPecp in the cells.

In Vitro Cell Migration Assay.
Effect of CPPecp on cell migration was assessed using a 24-well transwell plate inserted with incorporating polyethylene terephthalate filter membrane with 8 m pores (BD FalconTM Cell Culture Insert System).
Approximately 5 × 10 4 human umbilical vein endothelial cells (HUVECs) (obtained from BCRC, number: H-UV001) were suspended in 200 L complete EC medium (Gibco) containing 0, 5, or 12.5 M CPPecp and then seeded on the upper compartment of filter. The lower compartment of filter contains 500 L complete EC medium with 20 ng/mL VEGF (R&D) as stimulator. After incubating at 37 ∘ C and 5% CO 2 for 4 h, the migrated cells on the lower surface of filter were fixed with 4% formaldehyde at RT for 15 min and stained with Hocechst at RT for 15 min. The nonmigrated cells on the upper surface of filter were removed by cotton swab. Filter membrane of transwell insert was cut down and mounted with Fluoromount mounting medium (Sigma Aldrich, Missouri, USA). Numbers of migrated cells were counted in five randomly selected microscopic fields at magnification 100x using inverted fluorescent microscope (TE2000E, Nikon, Kanagawa, Japan) with a cooled CCD (Evolution VF, MediaCybernetics, Bethesda, MD).
The result was represented as mean ± SD (standard deviation) of three independent experiments. Statistically significant differences were analyzed using unpaired Student's -test. Asterisks showed level of statistical significance: * < 0.05; < 0.01; * * * < 0.001 compared with control.

Zebrafish Angiogenesis
Model. Tg(kdr:EGFP) zebrafish, a well-studied model for vascular embryogenesis [15], was performed to assess the effects of CPPecp on angiogenesis. The Tg(kdr:EGFP) (kindly provided by Dr. Yung-Jen Chuang's lab at NTHU) is a transgenic zebrafish line that expresses eGFP driven by the kdr promoter in vasculature endothelial cells during zebrafish embryogenesis, which can serve as an in vivo angiogenesis model for drug screening [16]. Fertilized eggs were generated from adult mating pairs and incubated at 28.5 ∘ C in a recirculating aquaculture system.  [17]. CT-26 tumorbearing mouse was utilized to investigate biodistribution of MNP-CPPecp and Prussian blue staining was employed to demonstrate ferric iron in mouse tissues. The CT-26 tumorbearing mouse was intravenously injected with 150 L MNP-CPPecp (0.06 emu/g) and sacrificed by CO 2 narcosis at a time point of 3, 6, 12, and 24 h after administration. The kidney, heart, liver, spleen, stomach, pancreas, small intestine, large intestine, trachea, lung, and tumor of mice were taken, fixed with paraformaldehyde, embedded in paraffin, and sectioned into 5 m thick tissue slides, following by deparaffinizing in xylene solution (J. T. Baker Phillipsburg, NJ, USA) and serially rehydrating with 100%, 95%, 85%, 75%, and 50% alcohol. The slides were continuously immersed in working solution (20% hydrochloric acid and 10% potassium ferrocyanide (Sigma, MO, USA) solution mixture, 1 : 1 volume ratio) at room temperature for 30 min and then counterstained with fast nuclear red (Sigma, MO, USA) at RT for 5 min. After dehydration through 95% and 100% alcohol and clearing with xylene, each slide was finally covered with coverslip. Tissue images were digitized using light microscope (Eclipse E400, Nikon) with digital microscopy camera (AxioCam ICc 5, ZEISS).

Heparan Sulfate Binding Cell Penetrating Peptides
Derived from Natural Proteins. Heparan sulfate (HS) serves as the initial anchoring site for many CPPs through electrostatic interactions between negatively charged sulfates or carboxyl groups and basic amino acids Arg as well as Lys [18]. Till now 27 CPPs from natural proteins including 14 viral proteinderived peptides, 7 animal homeostatic modulator-derived peptides, 3 antimicrobial peptides, and 3 toxin-derived peptides have been demonstrated or predicted to be able to interact with cell surface HS and penetrate cross the plasma membrane. In silico secondary structures of all 27 HS-binding CPPs were predicted by Network Protein Sequence Analysis [19]. As shown in Table 1 [20]. All 14 viral protein-derived CPPs are highly cationic (high pI values) with 10 peptides forming helix and 4 existing as random coil, penetrating cells through direct translocation [21][22][23][24] and lipid raft-mediated endocytosis [25][26][27][28][29]. Most of the 7 animal homeostatic modulator-derived CPPs may be internalized into cytosol through HS-mediated and energydependent endocytosis, among which 5 animal proteinderived peptides are demonstrated to possess either helix or sheet to interact with the plasma membrane, while our CPPecp and apolipoprotein B binding domain are unique such that they hold random coil structures in this category. As for 3 antimicrobial peptides, all of them are suggested to interact with cell surface HS and penetrate membrane barrier via energy-dependent endocytosis. LL-37 holds high level of helix, SynB1 possesses sheet, and SynB3 retains random coil structures [30][31][32]. For the last category toxinderived CPPs, bovine prion-derived bPrPp forming helix and mixed helix with strand are distributed in the internal region of venom-derived crotamine, and scorpion toxinderived maurocalcine [33][34][35][36].
Previous researches have shown that the interactions between the positively charged peptide and highly negatively charged membrane components, such as the GAG moieties of cell surface proteoglycans, play a crucial role in the overall process of cellular permeability of highly basic or amphipathic CPPs [37]. Although this investigation may also reflect nonspecific electrostatic interactions between these The confidence of the prediction is denoted by scaling the predictions from week (lower-case letter) to strong (upper-case letter). "H, " "E, " and "C" refer to -helical, -strand, and random coil propensities, respectively.

6
BioMed Research International basic peptides and HS, it has been characterized that negatively charged heparin more effectively blocks uptake of CPPs than other soluble GAGs such as CS and hyaluronic acid [38], likely suggesting that there might be some structural requirements involved in the strong interaction between CPP and HS. In Table 1, 19 of these 27 HS-binding CPPs generally possess conventional heparin binding sequences such as XBBXB and XBBBXXBX where B is a basic amino acid and X represents a random amino acid, and they can also be divided into cationic and amphipathic groups. Most viral factorderived peptides are basic amino acid-rich. For example, cationic TAT is an extensively used CPP rich in Arg and can interact with sulfated proteoglycans and negatively charged phospholipids on the cell membrane [25]. It should be noted that although 10-amino acid CPPecp is almost equal size to 9residue TAT and 10-residue SynB3, the features of TAT and SynB3 are quite different from CPPecp. Both TAT derived from viral protein and SynB3 belonging to antimicrobial peptide are highly cationic peptides with high pI values above 12, while our newly identified CPPecp containing only 2 Arg and 1 Lys in a total of 10 amino acids is amphipathic with a pI value of 10.05. Interestingly, the proportion of basic residues in amphipathic crotamine (26%) is close to CPPecp (30%). "RWRCK" motif of CPPecp was previously predicted as a unique functional pattern in all 13 hRNaseA family members employing Reinforced Merging for Unique Segments system (ReMUS) [39]. Another peptide CyLoP-1 (CRWRWKCCKK) derived from crotamine also exhibited efficient intracellular delivery activity. In both cases positively charged residues conducting electrostatic interaction and aromatic Trp exerting transient membrane destabilization were essential to maintain CPP functionality [40,41]. Taken together, a similar motif "RWRXK" shown on the loop, where X might be a random amino acid, is present in both CPPecp and crotamine, suggesting that combination of positively charged residues and nonpolar aromatic residues, especially Trp, might provide a design rationale for novel amphipathic cell penetrating peptides. [42,43] Figure 1(a) showed significant FITC-CPPecp binding activity to CT-26 cells, which correlated well with significantly higher HS expression (Figure 1(b)). In addition, 5 M FITC-CPPecp rapidly and efficiently internalized into CT-26 cells within 10 min as analyzed by fluorescent microscopy (Figure 1(c)).

Cellular Binding of CPPecp to Tumor Cell with Higher HS Expression Level . Heparan sulfate (HS) is reported to be overexpressed in several tumors
To further address the importance of HS for CPPecp anchor in the absence of autofluorescence background, removal of cell surface HS by heparinase was carried out along with CPPecp labeled with tetramethylrhodamine (TMR). CT-26 cells were incubated in medium with (+) or without (−) heparinase II for 2 h and then treated with 5 M TMR-CPPecp for 10 min. TMR-CPPecp rapidly and efficiently bound to CT-26 cell surface (Figure 1(d), upper panel), while removal of cellular HS led to significant reduction in CPPecp attachment (Figure 1(d), lower panel). Taken together, our HS-binding CPPecp possessed strong binding activity to tumor cell surface with higher HS expression, while depletion of cell surface HS abolished such highly selective binding activity of CPPecp to tumor cells.

Effect of CPPecp on Migration of Mouse Colon Carcinoma
Cell. It has been shown that HSPGs may modulate cell migration by interacting with growth factors or chemokines and drives cell migrate toward specific stimuli [44]. Since CPPecp with a novel heparin-binding motif in ECP has already been identified to possess high recognition activity to cellular surface HSPG and penetration activity into cells [12], here whether CPPecp might modulate cancer cell migration through interaction with HSPG was further investigated using in vitro transwell migration assay, while EDN 32-41 , a 10amino acid peptide derived from comparable sequence motif of human RNase2 (EDN), possessing a conventional heparinbinding motif was also analyzed as a control. Figure 2 (Figure 2, gray bar). These results indicated that CPPecp containing core RWRCK motif, rather than containing known heparin-binding motif, inhibited CT-26 cell migration across the membrane in vitro. It has been reported that cancer migration was inhibited by antagonism of HS side chains. For example, A5G27 peptide derived from laminin 5 globular domain recognizes HS side-chains of CD44 variant 3 and blocks bioactivity of fibroblast growth factor-2 (FGF-2). It significantly inhibits FGF-2-induced WiDr colon cancer cell migration and invasion [45]. Collectively, inhibitory effect of CPPecp on cancer cell migration is possibly arisen from interaction with cell surface HS.

Effects of CPPecp on Migration of Vascular Endothelial
Cell. Cell surface HS proteoglycan (HSPG) serves as a coreceptor to coordinate binding of vascular endothelial growth factor (VEGF) toward its receptor. It has been reported to be associated with angiogenesis [46,47]. However, vascular endothelial cell migration is a crucial step in formation of new blood vessel and tumor angiogenesis [48]. To test the hypothesis that CPPecp interacting with cell surface HSPGs also affected angiogenesis, a common model cell line human umbilical vein endothelial cell (HUVEC) was used for in vitro transwell migration assay. Figure 3 indicated that VEGFinduced HUVEC migration was restored by cotreatment with 5 or 12.5 M CPPecp, leading to, respectively, 77% and 64% migration activity. This result indicated that CPPecp could inhibit VEGF-induced HUVEC migration. Likewise, the CD44-binding peptide A5G27 derived from laminin 5 globular domain inhibits FGF-induced angiogenesis in Chick CAM Assay [49]. Moreover, an HS-binding peptide 6a-P, corresponding to the HSPG binding domain of VEGF, binds to HSPG and affects interaction between VEGF and HSPG [50]. It interferes with angiogenesis by inhibiting VEGF-induced HUVEC migration and binding of VEGF to HUVEC. As a result, involvement of our CPPecp in angiogenesis may be attributed to interaction with cell surface HSPG.

Effects of CPPecp on Angiogenesis during Embryonic
Development of Zebrafish. Cell surface HSPGs serve as a coreceptor to coordinate binding of VEGF toward its receptor and have been reported to be associated with angiogenesis [46,47]. Tg(kdr:EGFP) zebrafish, a well-studied model for vascular embryogenesis, has been used as model for drug screening and angiogenesis studies [51,52]. It was thus utilized to investigate CPPecp effects on in vivo embryonic angiogenesis by injecting 4.6 nL of 4.56 or 22.8 pmol CPPecp or PBS (control) into yolk sac of zebrafish at 60 h postfertilization (hpf), and the development of subintestinal vessel (SIV) pattern (Figure 4(a), SIV networks are indicated with red rectangle) at 24 h postinjection (hpi) was monitored with images by inverted fluorescent microscope. Here 16-20 zebrafish were tested for each treatment group. The observed SIV patterns of zebrafish were divided into three groups according to growth level of SIV: normal, mildly inhibited, and severely inhibited phenotypes (Figure 4(b)). In the normal phenotype SIV developed as smooth basketlike pattern with 5-6 arcades. Both mild and severe inhibition phenotypes could be further classified as ectopic SIV pattern, in which SIV exhibited tortuous network and was unable to demonstrate complete basket-like pattern that normal phenotype developed. However, severe inhibition phenotype displayed more incomplete SIV network than mild inhibition phenotype did. In contrast, the zebrafish injected with CPPecp appeared to be tortuous, in which SIV pattern shrank significantly as compared with that of PBS control (Figure 4(c)). Figure 4( (Figure 4(e)). These data revealed that our CPPecp possessed antiangiogenesis activity in inhibiting SIV growth of zebrafish. As a result, involvement of CPPecp in angiogenesis may be attributed to interaction with cell surface HS. CPPecp is the first antiangiogenic peptide deciphered in embryonic development of zebrafish.

Time-Dependent Biodistribution of MNP-CPPecp in CT-26 Tumor-Bearing Mouse.
To better understand biodistribution of our HS-binding CPPecp in vivo, CPPecp was conjugated with well-dispersed Fe 3 O 4 magnetic beads (59.3 nm for diameter) to form magnetic nanoparticle-conjugated CPPecp (MNP-CPPecp). CT-26 tumor-bearing mice were intravenously injected with MNP-CPPecp (0.06 emu/g) and sacrificed at different time point after administration ( Figure 5(a)). MNP-CPPecp signal was detected using Prussian blue staining to indicate ferric iron in tissue section (blue color). Figure 5(b) indicated that stainable ferric iron (blue color as indicated by yellow arrow) was barely detectable in trachea, heart, and large intestine at all indicated time points, and so did other tissues including stomach, pancreas and kidney (data not shown). The MNP-CPPecp mainly accumulated in liver tissues from 3 h up to 24 h owing to uptake and removal by macrophages of reticuloendothelial system, which played a role in clearance of external substance in liver [53,54]. Interestingly, Prussian blue staining signals in CT-26 tumor section suggested MNP-CPPecp accumulation from 12 h to 24 h, whereas MNP signal was only detected in liver at 24 h. One recent report showed that exendin-4 peptide-conjugated superparamagnetic iron oxide nanoparticles were inevitably accumulated in liver tissue, suggesting that a nanoparticle might unavoidably be captured by this metabolic organ [55]. However, it is worth noting that CPPecp has potential to target colon carcinoma in vivo, suggesting that CPPecp might be applied for a potent carrier for drug delivery.

Heparan Sulfate-Binding Cell Penetrating Peptide for
Tumor Targeted Strategy. Although CPPs as noninvasive agents have promising biomedical potential for molecular delivery, they are mostly unfeasible for in vivo researches due to nonspecificity of their highly cationic characteristic such as TAT peptide. Due to high uptake rates in vitro and relatively low specificity in vivo of most CPPs, conventional CPPs would be designed for topical applications in CPP-based delivery (Table 1). Further analysis of natural protein-derived CPPs revealed that 5 CPPs exerted in vitro tumor suppression as well as cell internalization activities (Table 2). Although TAT peptide (46)(47)(48)(49)(50)(51)(52)(53)(54)(55)(56)(57) demonstrated antiangiogenesis and apoptosis-inducing activities, TAT peptide was proved to show low target specificity in vivo [56]. Distinct from conventional highly cationic CPPs, 4 amphipathic CPPs including CPPecp, crotamine, NFL-TBS , and p28 peptides demonstrated unique tumor targeting activity in vivo. Even though specific protein receptors for CPPecp and crotamine remain to be investigated, HSPG acting as coreceptor is indispensable for the translocation of CPPecp and crotamine [12,34]. In addition, both CPPecp and crotamine targeted highly proliferating cells such as tumor tissues [14,57]. Interestingly, a motif decorating a hydrophilic aromatic amino acid participating in membrane permeation between two arginines (RWR) appeared to be conserved in both CPPecp and crotamine, leading to similar characteristics of these 2 multifunctional HS-binding CPPs. Therefore, amphipathic CPPs might own promising potential to be designed as peptide-based drugs. In particular, HS-binding CPPs are suitable drug carriers for in vivo application in delivery of functional therapeutics.

Conclusions
CT-26 colon tumor cells revealed high CPPecp binding activity due to high HSPG expression on cell surface. CPPecp displays not only significantly inhibitory effects on CT-26 cancer migration and angiogenesis in vitro but also antiangiogenesis activity during zebrafish embryogenesis in vivo. Moreover, covalent linkage of CPPecp to magnetic nanoparticle shows potential for in vivo targeting to a subcutaneous CT-26      [75] tumor site. Moreover, CPPecp containing a core RWRXK sequence demonstrates both cell penetrating and epithelial tumor targeting activities. Taken together, our HS-binding CPPecp might be feasible for further application in molecular imaging for tumor homing and selectively targeting drug delivery system.