Arginine-tagging of polymeric nanoparticles via histidine to improve cellular uptake

Accepted Manuscript Arginine-tagging of Polymeric Nanoparticles Via Histidine to Improve Cellular Uptake Jasper Z. S. Chiu, Ian G. Tucker, Bernie J. McLeod, Arlene McDowell PII: DOI: Reference:

S0939-6411(14)00341-5 http://dx.doi.org/10.1016/j.ejpb.2014.11.014 EJPB 11751

To appear in:

European Journal of Pharmaceutics and Biopharmaceutics

Received Date: Accepted Date:

1 April 2014 19 November 2014

Please cite this article as: J.Z. S. Chiu, I.G. Tucker, B.J. McLeod, A. McDowell, Arginine-tagging of Polymeric Nanoparticles Via Histidine to Improve Cellular Uptake, European Journal of Pharmaceutics and Biopharmaceutics (2014), doi: http://dx.doi.org/10.1016/j.ejpb.2014.11.014

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

Manuscript for submission to European Journal of Pharmaceutics and Biopharmaceutics

Arginine-tagging of Polymeric Nanoparticles Via Histidine to Improve Cellular Uptake Jasper Z. S. Chiu, Ian G. Tucker, Bernie J. McLeod, Arlene McDowell* School of Pharmacy, University of Otago, Dunedin 9054, New Zealand

*Corresponding author. Tel.: +64 3 479 7145; Fax.: + 64 3 479 7034 Email address: [email protected] (A. McDowell)

Abbreviations

PECA, Poly(ethyl-cyanoacrylate); ECA, Ethyl-2-cyanoacrylate; FACS, Flourescence-associated cell sorting; R, Arginine; MALDI-TOF, Matrix assisted laser desorption ionization – time of flight; FITC, Fluorescein isothiocyanate; HEPES, 2-[4-(2-hydroxyethyl)piperazine-1-yl]ethyanesulfonic acid; HBSS, Hanks balanced salt solution; TFA, Trifluoroacetic acid

1

Abstract Polyarginine, a cell-penetrating peptide, has been shown to aid cellular penetration of bioactives into cells. We utilized a novel approach of using a histidine linker to produce poly(ethyl-cyanoacrylate) (PECA) nanoparticles tagged with oligoarginine and investigated cellular uptake. MALDI TOF/TOF (tandem) analysis revealed that di-arginine-histidine (RRH) covalently bound to PECA nanoparticles to form cationic particles (+18 mV), while longer oligoarginine peptides did not co-polymerize with PECA nanoparticles. Although RRH-tagged nanoparticles had similar size and FITC-dextran entrapment efficiency compared to unmodified nanoparticles, RRH-tagging of nanoparticles resulted in a greater release of FITC-dextran. As the nanoparticles were found to aggregate in Hanks Balanced Salt Solution (HBSS), the effect of phosphate on the zeta-potential of nanoparticles was studied. Treating the nanoparticles with poloxamer-407 prevented aggregation. RRH-tagged PECA nanoparticles increased cellular uptake by a further 30% compared to unmodified PECA nanoparticles and was concentration dependent. We suggest that enhanced cell uptake can be achieved using a di-arginine histidine construct as opposed to the previously published findings that a minimum of hexa-arginine is necessary. Further, the cationic zeta-potential of the cell-penetrating peptide may not be needed to enhance uptake.

Keywords Poly(ethyl-cyanoacrylate)

nanoparticles;

polyarginine;

MALDI-TOF;

cationic;

cell-

penetrating peptides

2

1.

Introduction Cell-penetrating peptides (CPPs) are short cationic peptides of 10 to 30 amino acids

that consist of mainly basic residues, such as arginine and lysine, and can be either derived from natural sources or produced synthetically [1, 2]. These peptides are able to improve cellular uptake [3] and enhance permeation [4] with minimal toxicity [5], thus making them potentially valuable tools for drug delivery. CPPs have been investigated for the delivery of therapeutic proteins [6], nucleic acids [7], and small molecules [8]. An example of a widely studied CPP is oligoarginine, which has been documented to improve insulin absorption across the gut in rats [9]. Although the greatest improvement in cell uptake has been shown with 15 arginine residues, octa-arginine (R8) is more extensively studied due to the costeffectiveness in oligoarginine synthesis [10]. CPPs have to be associated with bioactive molecules, either covalently or noncovalently, to enhance cell penetration [11]. Current covalent conjugation methods involve fusion proteins produced synthetically in bacteria or attachment to the bioactive through covalent side-chain linkages, such as a peptide bond or a disulfide bridge. Non-covalent linkages include simple electrostatic interaction and “piggy-back” attachment, where the CPP is attached covalently to a smaller molecule that is linked non-covalently to the bioactive [11]. The major limitation of direct covalent linkage is alteration of the activity and/or efficacy of the bioactive [3], whereas non-covalent linking requires a substantially higher amount of CPP for effective electrostatic association with the bioactive [9]. An alternative strategy to overcome the disadvantages of both covalent and noncovalent attachment in linking CPPs to a bioactive molecule is to covalently associate CPPs to nanoparticles containing the bioactive, thus leaving the encapsulated bioactive molecule in its native form. Also, colloidal carriers, such as polymeric nanoparticles, could increase the oral bioavailability of the bioactive by protecting the encapsulated bioactive from the hostile gastrointestinal environment [12].

Polymeric nanoparticles, such as poly(alkyl-

cyanoacrylate) (PACA) nanoparticles, prepared using water-in-oil microemulsions can offer additional advantages. These polymeric nanoparticles do not require purification and can be administered orally in the microemulsion template as the template is biodegradable, biocompatible [13] and exhibits permeation enhancing effects [14]. CPPs have been conjugated to polymeric carriers through epoxy conjugation [15] and cross-linking with dextran [16-18].

However, both techniques are time-consuming and

require multiple steps. Therefore, a simple and fast conjugation method is desirable. Kafka 3

et al. [19, 20] discovered that the histidine residue in the gonadotropin-releasing hormone analogue D-Lys6-GnRH covalently conjugates to poly(ethyl-cyanoacrylate) (PECA) nanoparticles both during and after nanoparticle formation. These authors suggested that the unprotonated histidine side chain (pKa 6.04) in D-Lys6-GnRH acts as a nucleophile [20], instead of only hydroxyl ions, to initiate the anionic interfacial polymerization of ethylcyanoacrylate (ECA) subunits [21]. However, the usefulness of histidine as a linker for deliberate covalent attachment of functional peptides such as oligoarginine to PECA nanoparticles in a single step polymerization has yet to be evaluated. The aim of this study was to prepare arginine-tagged (R-tagged) PECA nanoparticles via histidine anchoring, to characterize these particles and to evaluate the cellular uptake of PECA nanoparticles with arginine-tagging.

2.

Experimental sections

2.1

Reagents Arginine-histidine (RH), di-arginine-histidine (RRH), tetra-arginine-aminocaproic

acid-histidine (R4-aca-H),octa-arginine-histidine (R8H), octa-arginine-di-histidine (R8H2) and octa-arginine (R8) (all ≥ 95% purity) from GLS (Shanghai, China) were used as supplied. For preparation of the microemulsion, ethyloleate (GPRTM) was supplied by BDH Laboratory Supplies (Poole, England), while the surfactants sorbitan monolaurate (Crill 1) and ethoxy 20 sorbitan mono-oleate (Crillet 4) were kindly provided by BTB chemicals (Auckland, New Zealand).

The monomer ethyl-2-cyanoacrylate (ECA), fluorescein

isothiocyanate (FITC)-dextran (MW = 70 kDa, 500 kDa and 2,000 kDa), 2-[4-(2hydroxyethyl)piperazin-1-yl]ethanesulfonic acid (HEPES) (≥ 99.5% purity), sodium bicarbonate (≥ 99.5% purity), Hanks Balanced Salt Solution (HBSS; without sodium bicarbonate and phenol) and trifluoroacetic acid (TFA) were purchased from Sigma-Aldrich (St. Louis, MO, USA).

Poloxamer-407 (Lutrol-F127®) was purchased from BASF

(Ludwigshafen, Germany). Methanol (HPLC grade), acetonitrile (HPLC grade), chloroform were supplied by BDH Laboratory Supplies (Poole, England) and hydrochloric acid (fuming 37%, for analysis) and D(+)-glucose were sourced from Merck (Darmstadt, Germany). Sodium chloride (NaCl), di-sodium-hydrogen phosphate (Na2HPO4) and sodium-di-hydrogen phosphate (NaH2PO4) were purchased from Univar, Asia Pacific Specialty Chemicals Limited (Sydney, Australia) and absolute ethanol was supplied by Anchor Ethanol (Auckland, New Zealand). Distilled, ultra-pure water was produced using a Milli-Q® water 4

Millipore Purification SystemTM (Billerica, USA). BD PharmingenTM Propidium iodide 50 µg/mL was purchased from BD Biosciences (California, USA).

2.2

Preparation of PECA nanoparticles from a microemulsion template The microemulsion was prepared based on the method described by Watnasirichaikul

et al. [22]. Briefly, a mixture of Crill 1 and Crillet 4 were prepared in a weight ratio of 4 : 6. Then the surfactant mix, ethyloleate oil and water were mixed together at a weight ratio of 5.4 : 3.6 : 1, respectively, to produce the microemulsion template [20]. ECA monomer (200 µL ECA dissolved in 600 µL of chloroform) was added drop-wise to 10 g microemulsion template under constant stirring at 700 rpm at 4°C. The polymerization process was allowed to progress for a minimum of 4 h.

2.3

R-tagging and FITC-dextran encapsulation To produce R-tagged nanoparticles, either RH (10 µmol), RRH (10 µmol), R4-aca-H

(5.6 µmol), R8H (4 µmol), R8H2 (5.2 µmol) or R8 (4 µmol) was added to the aqueous phase of the microemulsion, prior to polymerization. To load the nanoparticles with FITC-dextran, 1 mg of FITC attached to dextran of either MW 70 kDa, 500 kDa and 2,000 kDa was dissolved in the aqueous phase, prior to polymerization.

2.4

Isolation of nanoparticles from microemulsion Either 0.1 or 0.2 g microemulsion containing the nanoparticles was gently mixed with

600 µL of dilute HCl (pH 2.5) and 600 µL of methanol 80% (v/v). Then the sample was centrifuged at 20,800 g (Eppendorf Centrifuge 5417C) for 30 min to isolate the nanoparticles from the microemulsion. The pellet of nanoparticles was re-suspended via brief sonication in absolute ethanol to wash the isolated nanoparticles and then spun at 20,800 g for 30 min. This washing procedure was repeated twice. Bulk isolation of 1 or 2 g microemulsion was performed by mixing with 6 mL of dilute HCl (pH 2.5) and 6 mL of methanol 80% (v/v). The sample was centrifuged at 40,100 g at 25°C for 35 min in Beckman OptimaTM L-80 Ultracentrifuge. The pellet of nanoparticles was re-suspended via brief sonication in absolute ethanol to wash the isolated nanoparticles and then spun at 40,100 g for 30 min. This cleaning procedure was repeated twice.

5

2.5

Characterization of nanoparticles 2.5.1

Screening for covalent association using MALDI-TOF mass spectrometry

Detailed sample preparation and operation protocols used were as described by Kafka et al. [20]. Briefly, the nanoparticles were isolated from 0.1 g microemulsion and dissolved with sonication in 20 to 30 µL acetonitrile. An aliquot (0.5 µL) of the dissolved nanoparticle suspension was added to 9.5 µL matrix (10 mg/mL α-cyano-4-hydroxycinnamic acid dissolved in aqueous acetonitrile 60% (v/v) with TFA 0.1% (v/v)). An aliquot (0.8 µL) was then spotted onto the MALDI-plate (Opti-TOF 384 well plate, Applied Biosystems, Framingham, MA, USA) and air-dried prior to analysis. All mass spectrometry (MS) spectra were obtained in positive-ion reflector mode with 1000 laser pulses per sample spot. Relevant precursor ions were chosen for collision induced dissociation tandem mass spectrometry (CID-MS/MS).

Laser pulses were then set at 2000 to 4000 per selected

precursor using 2 kV mode and air as the collision gas at a pressure of 1 x 10-6Torr. Nanoparticles with oligoarginine covalently bound to the PECA polymer were selected for further characterization. 2.5.2

Size and zeta-potential

Isolated nanoparticles from 0.1 g microemulsion were either re-suspended in absolute ethanol (containing 0.2% polysorbate 80) for size measurements, or 1mM NaCl solution for zeta-potential measurements (Zetasizer ZEN3600, Malvern instruments Ltd, UK).

The

arginine-tagged nanoparticle formulation that bore the highest zeta-potential was selected for in vitro release studies. 2.5.3

Entrapment efficiency

Isolated nanoparticles from 0.1 g microemulsion were dissolved in 1 mL aqueous acetonitrile 80% (v/v) and the FITC-dextran released was determined using Hitachi F-7000 Spectrofluorometer (excitation 485 nm and emission 516 nm). Entrapment efficiency was calculated as the percentage FITC-dextran recovered upon dissolving the nanoparticles as a proportion of the total FITC-dextran added to the microemulsion. 2.5.4

Quantifying RRH associated with PECA nanoparticles

Reversed-phase high-performance liquid chromatography (RP-HPLC) was used to detect and quantify the amount of unassociated RRH in the microemulsion through an indirect method. Nanoparticles were isolated from 0.2 g microemulsion and the supernatant (1 mL) was drawn for HPLC analysis. Agilent

TM

RP-HPLC analysis was carried out using an

series 1200 HPLC system (CA, USA) with a micro-vacuum degasser (G1379B), a 6

quaternary pump, a temperature-controlled auto-sampler (G1329A with G1330B thermostat), a temperature-controlled column compartment (G1316A) coupled to a variable wavelength UV detector (G1314B). Samples in the auto-sampler were kept at 4ºC and the column (HiChrom Ultrasphere 5 ODS, 250 x 4.6 mm i.d. 5μm particle size, 80 Å pore size, Berkshire, UK) was heated to 40ºC.

The column was protected by a widepore C18

Phenomenex Analytical Guard cartridge column (KJ0-4282, 4.0 x 3.0 mm i.d.).

The

injection volume for all samples was 100 µL with detection wavelength set at 198 nm for RRH. The mobile phase consisted of solvent A: TFA 0.1% (v/v) in ultra-pure water and solvent B: acetonitrile. An isocratic elution was used namely; 1% solvent A in 99% solvent B with a flow rate of 1 mL/min over 10 min. A washing step was added to flush out residual formulation components such as oil and surfactants by increasing the concentration of solvent B to 80% over an additional 3 min at a flow rate of 1.1 mL/min. Then the system was reequilibrated by reducing the concentration of solvent B to 1% and the flow rate to 1 mL/min. The system was allowed to re-equilibrate for a further 5 min prior to the next run. The validated isocratic method (ICH guidelines) [23] for RRH concentration ranging from 4 to 29 µg/mL was linear (r2 = 0.997). All 3 concentration levels (4, 13 and 29 µg/mL) had greater than 95% accuracy with both intraday and interday variability of less than 2%. Limits of quantification (LoQ) and limits of detection (LoD) were 2.3 µg/mL and 0.8 µg/mL, respectively.

2.6

Effects of phosphate on nanoparticles R-tagged and unmodified nanoparticles (without FITC-dextran loading) were isolated

from 2 g microemulsions (Section 2.4). The isolated nanoparticles were then re-dispersed in 10 mL ultrapure water. The suspension of nanoparticles was titrated with phosphate buffer (5 mM; pH 7) at room temperature under constant stirring at 1,000 rpm. A sample aliquot (0.5 to 1 mL) was withdrawn for zeta-potential determination after each volume addition of phosphate buffer. To investigate the effect of different pHs, the experiment was repeated with phosphate buffers (5 mM) of pH 6 and 8.

2.7

Treating nanoparticles with poloxamer-407 R-tagged and unmodified nanoparticles were isolated from 0.1 g microemulsions and

then re-dispersed in ultrapure water containing poloxamer-407, in mass ratios (poloxamer407 : dry nanoparticles) of 1 : 8.3, 1 : 4.2, 1 : 1.4, 1 : 0.8, prior to adding equal amounts of 7

HBSS (2X concentrated, pH 7.1). The zeta-potential and size of the nanoparticles were measured.

The stability of poloxamer-407 treated nanoparticles in HBSS at 37°C was

investigated over 2 h.

2.8

Cell culture and cellular uptake in a Caco-2 cell model Caco-2 cells were grown in 75 cm2 culture flasks. Cells were fed with Dulbecco’s

Modified Eagle Medium (supplemented with 10% fetal bovine serum, 1% non-essential amino acid and 1% penicillin-streptomycin) every other day. Upon 70% cell confluency, cells were trypsinized and seeded on 12-well plate at 100,000 cells/well for 2 days. Cells were washed twice with 2 mL HBSS (supplemented with 4.2 mmol sodium bicarbonate, 25 mmol D(+)-glucose and 25 mmol HEPES), before replacing with 2 mL of FITC-dextran loaded nanoparticle formulations (12.5 µg dry weight/mL) for incubation at 37°C. After 2 h, the formulations were discarded and cells were washed twice with ice cold HBSS. Then the cells were trypsinized and FACS (florescence-activated cell sorting) buffer (1% w/v bovine serum albumin and 0.01% w/v sodium azide in phosphate buffered saline) was added to stop the trypsinization process. The cells were centrifuged at 210 g for 5 min and washed again with FACS buffer, before re-suspending each sample in 100 µL FACS buffer with 50 ng propidium iodide. Samples were placed on ice in the dark until analysis. Cells were analysed with FACSCantoTM II flow cytometery (BD Biosciences, California, USA) and data was analyzed using FlowJo 7.6 analysis software (Tree Star, Oregon, USA).

2.9

FITC-dextran release study in Hanks Balanced Salt Solution (HBSS) FITC-dextran (MW = 2,000 kDa) encapsulated nanoparticles (either unmodified or R-

tagged) were extracted from 1 g microemulsions. The nanoparticles were re-suspended in 10 mL HBSS (pH 7.4). The suspension was stirred at 300 rpm at 37ºC and samples (1 mL) were drawn at 15, 45, 75, 105, 135 and 180 min. Samples were spun at 20,800 g for 30 min and the supernatant was analyzed for FITC-dextran released by spectrofluorometery (excitation 485 nm and emission 516 nm).

2.10

Statistical analysis Comparisons of size, zeta-potential and cellular uptake between the different

nanoparticle formulations were analyzed with General Linear Model ANOVA using 8

Minitab® ver. 16.1.0.0.

Post ANOVA pair-wise comparisons were assessed using the

Bonferroni method. P< 0.05 was considered significantly different.

3.

Results

3.1

Characterization of nanoparticles Using MALDI-TOF spectrometry, covalent bonding of oligoarginine to the PECA

nanoparticles can be determined by the presence of prominent peaks of the parent peptide mass (Mpeptide) plus 125 mass unit increments due to sequential addition of ECA monomer units (n), resulting in a mass pattern of [Mpeptide + (125)n + H]+ in the mass spectrum. By selection of a prominent peak in the mass spectrum, collision-induced-dissociation (CID) fragmentation can be used to identify the location of covalent association. The mass spectrum of unmodified PECA nanoparticles had characteristic peaks in 125 mass unit increments that represent sequential ECA monomer subunit additions (data not shown). The mass spectrum of RH-tagged nanoparticles had prominent peaks at parent RH mass with 125 mass unit increments [MRH (311) + (125)n + H]+ at m/z 812, 937, 1062, 1187 and 1312 corresponding to parent RH conjugated to 4, 5, 6, 7 and 8 monomer units respectively (Fig. 1(a)). Subsequent fragmentation of parent ion 937 (Fig. 1(b)) resulted in peaks of parent mass [MRH+ H]+ at m/z 312with subsequent ECA additions at m/z 562, 687 and 812. Similarly, the mass spectrum of RRH-tagged nanoparticles showed peaks, [MRRH (467) + (125)n + H]+, at m/z 718, 843, 968, 1093 and 1218 corresponding to parent RRH conjugated to 2, 3, 4, 5 and 6 monomer units (Fig. 1(c)) with fragmentation of precursor 968 bearing RRH mass peak [MRRH+ H]+ at m/z 468 and subsequent ECA additions at m/z 718 and 843 (Fig.1(d)). The mass spectrum intensity of RRH conjugated to ECA was lower compared to RH conjugated to ECA. Low mass immonium ions (i) are generated through multiple fragmentation of a precursor and can be used to confirm the presence of certain amino acids [24]. Fragmentation of RH and RRH-tagged nanoparticles generated iHistidine ions (Mihistidine = 110) conjugated to 5 ECA subunits (m/z 735 in Fig. 1(b)) and 4 ECA subunits (m/z 610 in Fig. 1(d)), respectively. MALDI-TOF spectra did not reveal any evidence of covalent bonding of longer oligoarginine (R8H, R8H2 and R4-aca-H) to PECA nanoparticles (Table 1). As RH and RRH were found to covalently bind to the PECA nanoparticles, they were used to produce Rtagged nanoparticles for further study. 9

R-tagged and unmodified nanoparticles had a similar size of approximately 200 nm (P> 0.05) with polydispersity index less than 0.3 (Table 2). However, there were significant differences (P< 0.001) in zeta-potential of nanoparticle formulations, being -13.7 mV for unmodified nanoparticles, -3.5 mV for RH-tagged nanoparticles and +18.4 mV for RRHtagged nanoparticles. FITC-dextran loading did not affect the size (P> 0.05) or zeta-potential (P> 0.05) of the nanoparticles. FITC-dextran entrapment efficiency was higher with higher molecular weight FITC-dextran, with similar entrapment in both unmodified and RRH-tagged nanoparticles (Table 2). As the RRH-tagged nanoparticles had the most positive zeta-potential, this formulation was selected for further investigation. The amount of RRH recovered from the microemulsion after preparing RRH-tagged nanoparticles, and equivalent to amount not associated with the nanoparticles, was 11.9% ± 3.5 (n = 7).

3.2

Effects of phosphate on nanoparticles Both unmodified and RRH-tagged nanoparticles formed aggregates upon re-

dispersion in HBSS; therefore the effect of phosphate on aggregation of nanoparticles was explored to better understand the behaviour of these polymeric nanoparticles in an in vitro system.

Unmodified and RRH-tagged nanoparticles were stable in ultrapure water and

maintained their respective zeta-potentials of -29 mV and +18 mV over 4 h under constant stirring at room temperature (data not shown). Upon titration with phosphate buffer, the zetapotential of unmodified nanoparticles increased, eventually reaching a plateau at 3 mM phosphate ion concentration (Fig. 2(a)).

Conversely, the zeta-potential of RRH-tagged

nanoparticles decreased with increasing concentration of phosphate, ultimately reaching a stable zeta-potential (Fig. 2(b)). When titrated with HBSS (pH 7.4), the zeta-potential of unmodified nanoparticles increased to -23 mV (Fig. 2), while RRH-tagged nanoparticles decreased to -7 mV and further addition of HBSS resulted a plateau at approximately -12 mV for both unmodified nanoparticles and RRH-tagged nanoparticles (Fig. 2).

3.3

Effect of poloxamer-407 treatment Treating the nanoparticles with poloxamer-407 at a mass ratio of at least 1 : 4.2

(poloxamer : dry nanoparticles) inhibited aggregation (size > 1000 nm) upon re-suspension of particles in HBSS (Fig. 3(a)). However size variability of the particles was increased with 10

decreasing concentrations of poloxamer-407. Treating the nanoparticles with poloxamer-407 at 1 : 4.2 mass ratios maintained the size of unmodified and RRH-tagged nanoparticles at approximately 200 nm over 2 h at 37°C (Fig. 3(b)).

3.4

Cellular uptake in a Caco-2 cell model The addition of RRH to a solution of FITC-dextran did not improve cellular uptake of

FITC-dextran (Fig. 4). Encapsulation FITC-dextran in PECA nanoparticles increased the cellular uptake to approximately 16%, significantly higher than FITC-dextran in solution (P< 0.001). Associating the nanoparticles with RRH, either by a physical mix or RRH-tagging did not further improve the cellular uptake (Fig. 4). However, increasing the amount of RRH used for tagging by 10-fold (to 100 μmol) significantly increased the cellular uptake of the FITC-dextran to 45% (P< 0.001) compared to unmodified nanoparticles (Fig. 4).

3.5

Release profile of FITC-dextran in HBSS Both unmodified and RRH (100 µmol)-tagged nanoparticles showed a burst release of

FITC-dextran (MW = 2,000 kDa) over the first 30 min. Unmodified nanoparticles had an incomplete release over 3 h, while RRH (100 µmol)-tagged nanoparticles released about 90% of its contents within 3 h (Fig. 5).

4.

Discussion The mass spectra of RH-tagged nanoparticles (Fig. 1(a))and RRH-tagged

nanoparticles (Fig. 1(c)) with their respective fragmented precursors of 937 and 968 is evidence that oligoarginine covalently bound to the PECA polymer, consistent with the findings of Kafka et al. [20]. Fragmentation of these parent precursors generated i[histidine] peak at m/z 110, with the–COOH group cleaved. Therefore, the presence of i[histidine] attached to 5-ECA and 4-ECA subunits (m/z 735 in Fig. 1(b) and 610 in Fig. 1(d)), indicated that the histidine residue was the side chain that associated covalently with the PECA polymer. Although RRH was found to be covalently bound to the PECA, the occurrence of this binding as seen with the peak intensity in the mass spectrum was less evident compared to the peak intensity for RH. Furthermore, longer oligoarginines (R8H2, R8H and R4-aca-H) did not covalently associate with the PECA nanoparticles. These observations are suggestive of a trend in reducing capability of histidine to act as a nucleophile with increasing arginine residues. An electron inductive effect due to the arginine is unlikely as the inclusion of a 11

spacer in R4-aca-H oligoarginine did not result in covalent binding. This phenomenon can perhaps be better reasoned by the rigidity of arginine on the peptide backbone. Rothbard et al. [25] discussed the importance of spacing and rigidity of arginine residues in CPPs to allow interaction between the guanidine headgroup of arginine with the cell surface. Therefore, increasing arginine residues in oligoarginine could decrease the rigidity of the peptide backbone and allow the peptide to fold and shield the imidazole ring of histidine, thus preventing histidine from participating in the nucleophilic reaction during the polymer polymerization. Although RH appeared to bind more readily to the PECA nanoparticles compared to RRH, nanoparticles tagged with RH were less cationic (-3.5 mV) than particles tagged with RRH (+18.4 mV). This finding suggests that most of RH peptide is covalently bound to the interior of the nanoparticles, rather than the external surface. The terminal arginine of the longer RRH peptide could also protrude through the meshwork of the polymer, allowing the arginine to be presented closer to the surface of nanoparticles, giving the particles a more cationic zeta-potential. Arginine-tagging of PECA nanoparticles did not alter the size of nanoparticles, all being approximately 200 nm in diameter. The surface of PECA nanoparticles are normally negatively charged (-13.7 mV) due to the presence of acrylic head groups [26], however the zeta-potential was shifted to +18 mV and -3.5 mV with RRH and RH–tagging, respectively. This alteration in zeta-potential could be explained by covalent association of cationic arginine residues on the surface of the nanoparticles. Analysis of the supernatant from the nanoparticle microemulsion after polymerization revealed that approximately 90% of the RRH peptide was associated with the nanoparticles. Interaction of cationic chitosan particulates with electrolytes and proteins in media has been previously reported [27]. A rise in pH from 6 to 7.4 in HBSS increased the size of chitosan nanoparticles from 25 nm to 333 nm and slightly shifted the zeta-potential from +5.3 mV to +3.3 mV.[27] An increase in pH of HBSS buffer reduces the ionization of the amine group of chitosan (pKa = 6.31 - 6.51) [28] and alters the composition of phosphate species in the buffer. Both decreased ionization of chitosan and increased surface adsorption HPO42lead to a reduction in zeta-potential, forming a less stable particle suspension that is prone to aggregation. In contrast to chitosan particles, where a pH rise decreases ionization, the rise in pH from 6 to 8 caused a drop in zeta-potential for RRH-tagged PECA nanoparticles (Fig. 2(b)), presumably due to the ionization of the acrylic acid head groups (pKa = 4.35) in PECA 12

nanoparticles as discussed by Müller et al. [29] as well as surface adsorption of HPO42-. At pH 8, HPO42- would be the major ion present and therefore, would displace H2PO4- adsorbed on the surface of the RRH-tagged nanoparticles, resulting in a more negative zeta-potential. For unmodified PECA nanoparticles, the increase in ionic strength of buffer is the likely explanation for the less negative zeta-potential (Fig. 2(a)). When the titration was performed with HBSS buffer, the unexpected drop in zetapotential with the initial addition of 200 µL HBSS (containing 0.015 mM phosphate) could be attributed to the presence of 2-[4-(2-hydroxyethyl)piperazin-1-yl]ethanesulfonic acid (HEPES), a zwitterionic buffering agent, in HBSS (25 mM HEPES). Addition of 0.3 mM HEPES (pH 7.4) was sufficient to negate the cationic surface charge, giving a zeta-potential of -3 mV (data not shown). The negative charge from HEPES carboxy groups could aid in the neutralization of the cationic charge on the surface of the RRH-tagged nanoparticles. Although phosphate and HEPES were found to be contributing factors in charge alteration in RRH-tagged nanoparticles, the impact of other components in HBSS has yet to be identified. Stabilizing agents such as poloxamer have been used to prevent aggregation of particulates [30]. Here, we have demonstrated that treating both the unmodified and RRHtagged nanoparticles with poloxamer-407 of at least 1 : 4.2 (poloxamer : dry nanoparticles) mass ratio prevented aggregation upon re-dispersion in HBSS, with all particles maintaining their size of approximately 200 nm after incubation for 2 h at 37°C. However, treating the nanoparticles with non-ionic poloxamer-407 also altered the zeta-potential of both unmodified and RRH-tagged nanoparticles to 0 mV. Cellular uptake of RRH-tagged nanoparticles was studied in a Caco-2 cell model. The oligoarginie RRH alone did not enhance cell uptake of FITC-dextran. Incorporation of FITC-dextran within a nanoparticulate formulation improved cell uptake. A physical mixture of PECA nanoparticles with RRH or RRH-tagging at 10 µmol RRH did not further improve the uptake of PECA nanoparticles. However tagging with a higher RRH concentration (100 µmol) doubled the cellular uptake compared to unmodified PECA nanoparticles (Fig. 4). Tagging the nanoparticles with 100 µmol also resulted in a more cationic zeta-potential of 34.8 ± 2.8 mV indicating increased presence of the RRH on the surface of the nanoparticle. However the zeta-potential was 0 mV when the particles were treated with poloxamer-407 to prevent aggregation of the nanoparticles in the cell culture buffer used. Contrary to common findings that neutral particles either do not interact with phospholipids [31] or interact at a low level [32, 33], our results suggest cellular interactions between the surface oligoarginine 13

of the RHH-tagged nanoparticles and the cell membrane. Nel et al. [34] provide an excellent review on behavior of nanoparticles at the nano-bio interface. These authors state that special consideration of van der Waals, electrostatic and other forces (such as solvation) is required for nanoscale events. Electrostatic forces are only present within a few nanometers from the surface of the particle in a typical biological fluid [34]. Although RRH-tagged nanoparticles in the present study had zeta-potential of 0 mV after poloxamer treatment, electrostatic forces between the oligoarginine and cell membrane could still be in effect provided the nanoparticles are in close proximity to the cell membrane. This attractive force, reinforced by van der Waals forces, would therefore cause accumulation of RRH-nanoparticles on the cellular surface. Thus, allowing oligoarginine to interact with the cell surface leading to greater cellular uptake. The release of fluorescent material (FITC-dextran) was greater and occurred more rapidly from the RRH-tagged nanoparticles compared to the FITC-dextran release from the unmodified nanoparticles (Fig. 5), suggesting that R-tagging the nanoparticles may have interfered with the polymerization of PECA, affecting the integrity of the polymer wall formed [35]. RRH (100 µmol)-tagged nanoparticles lost almost all of the FITC-dextran encapsulated within the first 30 minutes (Fig. 5). This is a major limitation to our cell uptake study, where our uptake result would very likely be an underestimate of the true uptake of the nanoparticles.

Nevertheless, the proportion of cells that took up the RRH-tagged

nanoparticles was still significantly increased compared to unmodified nanoparticles. Physical mixture of FITC-dextran and RRH solution did not increase the uptake of the FITC-dextran. Although FITC-dextran is negatively charged and is expected to bind noncovalently with RRH through electrostatic interaction, the uptake of FITC-dextran was not increased (Fig. 4). The physical mixture failed to increase penetration of FITC-dextran could be due to the oligoarginine being RRH, rather than octa-arginine which is more efficient at penetration and the RRH concentration added was insufficient for penetration as effective penetration through electrostatic linking requires a substantially high oligoarginine concentration [9]. Increased cellular penetration with oligoarginine conjugated to a polymer backbone have also been shown by Sakuma et al. [36], where physically mixed insulin and polymer backbone conjugated with octa-arginine (R8) increased nasal absorption of insulin. The group also demonstrated increased penetration of fluorescent in Caco-2 cells and Calu-3 cells when the fluorescent probe was mixed with polymer conjugated with R8, while physical 14

mixture with R8 alone did not improve the uptake of the fluorescent probe into the cells. In our experiment, we have shown that RRH-tagged nanoparticles can also increase cell penetration. This is contrary to the finding that at least hexa-arginine (R6) is required for effective penetration [10]. The presence of multiple arginine molecules on the surface of the nanoparticles may have allowed arginines to be perceived as oligoarginine when interacting with cells for enhanced penetration. We have prepared and characterized cationic RRH-tagged nanoparticles and shown the effect of phosphates on PECA nanoparticles. As size is a critical factor in determining the particulate uptake in cells [37], poloxamer-407 is a potential candidate to prevent aggregation of PECA nanoparticles in HBSS, but at the expense of negating the positive zetapotential of the nanoparticles.

RRH (100 µmol)-tagged nanoparticles, treated with

poloxamer-407, increased cellular uptake of PECA nanoparticles suggesting that arginine could still interact with the cell membrane at close proximity, although the particles have a zeta-potential of 0 mV. Additionally, it is important to consider the influence of the media on nanoparticles when performing in vitro experiments as cellular response would depend on the surface characteristics of the particulates presented at the nano-bio interface.

Acknowledgements This study was funded by a University of Otago Research Grant, New Zealand. We thank Dr Torsten Kleffmann, Department of Biochemistry, University of Otago for assistance with MALDI analysis and Prof. Thomas Rades, Prof. Sarah Hook and Mr David Schmierer, School of Pharmacy, University of Otago for valuable discussions.

References [1] C. Foged, H.M. Nielsen, Cell-penetrating peptides for drug delivery across membrane barriers, Expert Opin. Drug Del., 5 (2008) 105-117. [2] V. Kersemans, K. Kersemans, B. Cornelissen, Cell penetrating peptides for in vivo molecular imaging applications, Curr. Pharm. Desgin, 14 (2008) 2415-2427. [3] S.B. Fonseca, M.P. Pereira, S.O. Kelley, Recent advances in the use of cell-penetrating peptides for medical and biological applications, Adv. Drug Deliver. Rev., 61 (2009) 953964. [4] R. Fischer, M. Fotin-Mieczek, H. Hufnagel, R. Brock, Break on through to the other sidebiophysics and cell biology shed light on cell-penetrating peptides, Chembiochem, 6 (2005) 2126-2142. 15

[5] H. Nagahara, A.M. Vocero-Akbani, E.L. Snyder, A. Ho, D.G. Latham, N.A. Lissy, M. Becker-Hapak, S.A. Ezhevsky, S.F. Dowdy, Transduction of full-length TAT fusion proteins into mammalian cells: TAT-p27Kip1 induces cell migration, Nat. Med., 4 (1998) 1449-1452. [6] S. Fawell, J. Seery, Y. Daikh, C. Moore, L.L. Chen, B. Pepinsky, J. Barsoum, TATmediated delivery of heterologous proteins into cells, P. Natl. Acad. Sci. USA, 91 (1994) 664-668. [7] Y.L. Chiu, A. Ali, H. Cao, T.M. Rana, Visualizing a correlation between siRNA localization, cellular uptake and RNAi in living cells, Chem. Biol., 11 (2004) 1165-1175. [8] J.B. Rothbard, S. Garlington, Q. Lin, T. Kirschberg, E. Kreider, P.L. McGrane, P.A. Wender, P.A. Khavari, Conjugation of arginine oligomers to cyclosporin A facilitates topical delivery and inhibition of inflammation, Nat. Med., 6 (2000) 1253-1257. [9] N. Kamei, M. Morishita, Y. Eda, N. Ida, R. Nishio, K. Takayama, Usefulness of cellpenetrating peptides to improve intestinal insulin absorption, J. Control. Release, 132 (2008) 21-25. [10] P.A. Wender, W.C. Galliher, E.A. Goun, L.R. Jones, T.H. Pillow, The design of guanidium-rich transporters and their internalization mechanisms, Adv. Drug Deliver. Rev., 60 (2008) 452-472. [11] K.M. Wagstaff, D.A. Jans, Protein transduction: Cell penetrating peptides and their therapeutic applications, Curr. Med. Chem., 13 (2006) 1371-1387. [12] S.A. Galindo-Rodriguez, E. Allemann, H. Fessi, E. Doelker, Polymeric nanoparticles for oral delivery of drugs and vaccines: A critical evaluation of in vivo studies, Crit. Rev. Ther. Drug, 22 (2005) 419-464. [13] S. Watnasirichaikul, T. Rades, I.G. Tucker, N.M. Davies, Effects of formulation variables on characteristics of poly(ethylcyanoacrylate) nanocapsules prepared from w/o microemulsions, Int. J. Pharm., 235 (2002) 237-246. [14] P.P. Constantinides, C.M. Lancaster, J. Marcello, D.C. Chiossone, D. Orner, L. Hidalgo, P.L. Smith , A.B. Sarkahian, S.H. Yib, A.J. Owen, Enhanced intestinal absorption of an RGD peptide from water-in-oil microemulsions of different composition and particle size, J. Control. Release, 34 (1995) 109-116. [15] K.S. Rao, M.K. Reddy, J.L. Horning, V. Labhasetwar, TAT-conjugated nanoparticles for the CNS delivery of anti-HIV drugs, Biomaterials, 29 (2008) 4429-4438. [16] A.M. Koch, F. Reynolds, H.P. Merkle, R. Weissleder, L. Josephson, Transport of surface-modified nanoparticles through cell monolayers, Chembiochem, 6 (2005) 337-345. [17] J.M. de la Fuente, C.C. Berry, TAT peptide as an efficient molecule to translocate gold nanoparticles into the cell nucleus, Bioconjugate Chem., 16 (2005) 1176-1180. [18] C. Chauvierre, D. Labarre, P. Couvreur, C. Vauthier, Novel polysaccharide-decorated poly(isobutyl cyanoacrylate) nanoparticles, Pharm. Res., 20 (2003) 1786-1793. [19] A.P. Kafka, T. Kleffmann, T. Rades, A. McDowell, Characterization of peptide polymer interactions in poly(alkylcyanoacrylate) nanoparticles: A mass spectrometric approach, Curr. Drug Del., 7 (2010) 208-215. [20] A.P. Kafka, T. Kleffmann, T. Rades, A. McDowell, Histidine residues in the peptide DLys6-GnRH: Potential for copolymerization in polymeric nanoparticles, Mol. Pharm., 6 (2009) 1483-1491. [21] C. Vauthier, C. Dubernet, E. Fattal, H. Pinto-Alphandary, P. Couvreur, Poly(alkylcyanoacrylates) as biodegradable materials for biomedical applications, Adv. Drug Deliver. Rev., 55 (2003) 519-548. [22] S. Watnasirichaikul, N.M. Davies, T. Rades, I.G. Tucker, Preparation of biodegradable insulin nanocapsules from biocompatible microemulsions, Pharm. Res., 17 (2000) 684-689. 16

[23] ICH, Guidance for industry - Q2B Validation of analytical procedures: Methodology, in, Food and Drug Administration (FDA), 1996. [24] K. Biemann, Nomenclature for peptide fragment ions (positive-ions), Method. in Enzymol., 193 (1990) 886-887. [25] J.B. Rothbard, E. Kreider, C.L. VanDeusen, L. Wright, B.L. Wylie, P.A. Wender, Arginine-rich molecular transporters for drug delivery: Role of backbone spacing in cellular uptake, J. Med. Chem., 45 (2002) 3612-3618. [26] T. Pitaksuteepong, N.M. Davies, I.G. Tucker, T. Rades, Factors influencing the entrapment of hydrophilic compounds in nanocapsules prepared by interfacial polymerisation of water-in-oil microemulsions, Eur. J. Pharm. Biopharm., 53 (2002) 335-342. [27] J.W. Loh, G. Yeoh, M. Saunders, L.-Y. Lim, Uptake and cytotoxicity of chitosan nanoparticles in human liver cells, Toxicol. Appl. Pharm., 249 (2010) 148-157. [28] Q.Z. Wang, X.G. Chen, N. Liu, S.X. Wang, C.S. Liu, X.H. Meng, C.G. Liu, Protonation constants of chitosan with different molecular weight and degree of deacetylation, Carbohyd. Polym., 65 (2006) 194-201. [29] R.H. Müller, C. Lherm, J. Herbort, T. Blunk, P. Couvreur, Alkylcyanoacrylate drug carriers: I. Physicochemical characterization of nanoparticles with different alkyl chain length, Int. J. Pharm., 84 (1992) 1-11. [30] P.J. White, P. Hickey, L. Sze Tu, S.J. Headey, M.J. Scanlon, B.J. Boyd, C.W. Pouton, Colloidal characteristics and formulation of pure protein particulate vaccines, J. Pharm. Pharmacol., 64 (2012) 1386-1393. [31] L. Leserman, P. Machy, J. Barbet, Cell-specific drug transfer from liposomes bearing monoclonal antibodies, Nature, 292 (1981) 226-228. [32] R.M. Miller, B. Bondurant, S.D. McLean, K.A. McGovern, D.F. O'Brien, Liposome-cell interactions in vitro: Effect of liposome surface charge on the binding and endocytosis of conventional and sterically stabilized liposomes, Biochemistry, 37 (1998) 12875-12883. [33] K.-D. Lee, S. Nir, D. Papahadjopoulos, Quantitative analysis of liposome-cell interaction in vitro: Rate constants of binding and endocytosis with suspension and adherent J774 cells and human monocytes, Biochemistry, 32 (1993) 889-899. [34] A.E. Nel, L. Madler, D. Velegol, T. Xia, E.M.V. Hoek, P. Somasundaran, F. Klaessig, V. Castranova, M. Thompson, Understanding biophysicochemical interactions at the nanobio interface, Nat. Mater., 8 (2009) 543-557. [35] A. Graf, A. McDowell, T. Rades, Poly(alkycyanoacrylate) nanoparticles for enhanced delivery of therapeutics - Is there real potential?, Expert Opin. Drug Del., 6 (2009) 371-387. [36] S. Sakuma, M. Suita, Y. Masaoka, M. Kataoka, N. Nakajima, N. Shinkai, H. Yamauchi, K.-i. Hiwatari, H. Tachikawa, R. Kimura, S. Yamashita, Oligoarginine-linked polymers as a new class of penetration enhancers, J. Control. Release, 148 (2010) 187-196. [37] A. des Rieux, V. Fievez, M. Garinot, Y.-J. Schneider, V. Preat, Nanoparticles as potential oral delivery systems of proteins and vaccines: A mechanistic approach, J. Control. Release, 116 (2006) 1-27.

17

Figure 1. Mass spectra of PECA nanoparticles tagged with (a) RH and (c) RRH. CID fragmentation of selected precursor ion of (b) 937 from RH-tagged nanoparticles and (d) 968 from RRH-tagged nanoparticles. [MRH+ H]+ = 312 and RH/ECA copolymers at m/z 812.36, 937.41, 1062.46, 1187.50; [MRRH+ H]+= 467 and RRH/ECA copolymers at m/z 718.36, 843.43, 968.47, 1093.53, 1218.59.

Figure 2. Zeta-potential of (a) unmodified nanoparticles (open symbols) and (b) R-tagged nanoparticles (closed symbols) titrated with phosphate buffer of varying pHs (□ pH 6, ◊ pH 7, ○ pH 8) and HBSS (∆pH 7.4). Data are means ± SD (n = 3).

Figure 3. (a) Size of unmodified nanoparticles (unfilled) and RRH-tagged nanoparticles (filled) treated with different poloxamer-407 concentrations before re-dispersion in HBSS. (b) Size of unmodified nanoparticles (△) and RRH-tagged nanoparticles (▲) coated with poloxamer-407 at 1 : 4.2 (poloxamer : dry nanoparticles) mass ratio, incubated at 37°C over 2 h. Data are means ± SD (n = 3).

Figure 4. Cellular uptake of different nanoparticle formulations in Caco-2 cells at 37°C over 2 h. Data are means ± SD (n ≥ 6).

Figure 5. Release of FITC-dextran (MW = 2,000 kDa) from either unmodified nanoparticles (△) or RRH (100 µmol)-tagged nanoparticles (▲) in HBSS, pH 7.4 with constant stirring at 300 rpm at 37°C. Data are means ± SD (n = 3).

18

Table 1. Covalent bonding of oligoarginine to the poly(ethyl-cyanoacrylate) (PECA) nanoparticles as determined by MALDI-TOF spectrometry. Peptide Arginine-histidine (RH) Di-arginine-histidine (RRH) Tetra-arginine-amino caproic acid-histidine (R4-aca-H) Octa-arginine (R8) Octa-arginine-histidine (R8H) Octa-arginine-di-histidine (R8H2)

Molecular weight 311.3 467.5 893.7

Covalent association Yes Yes No

1267.5 1404.6 1541.8

No No No

Table 2. Characterization of unmodified and R-tagged nanoparticles (either tagged with RH or RRH), with or without entrapment of FITC-dextran of different molecular weights and their entrapment efficiencies. Data are means ± SD (two independent batches). Values sharing the same letter are not significantly different, while values with different letters are significantly different. Nanoparticles

Size (d.nm) 186 ± 6a 206 ± 24a 196 ± 26a 198 ± 8a

Polydispersity Index 0.02 ± 0.03 0.08 ± 0.07 0.02 ± 0.00 0.17 ± 0.12

Zeta-potential (mV) -13.7 ± 2.9a -3.5 ± 0.2b 18.4 ± 1.9c -12.4 ± 1.0a

FITC-dextran loaded (kDa) Nil Nil Nil 70

Entrapment efficiency (%) 42.9 ± 3.3

Unmodified RH-tagged RRH-tagged Unmodified70 RRH-tagged70 Unmodified500 RRH-tagged500 Unmodified2000 RRH-tagged2000

195 ± 15a

0.09 ± 0.01

18.2 ± 0.4c

70

52.4 ± 4.4

188 ± 5a

0.07 ± 0.02

-11.9 ± 0.2a

500

80.5 ± 11.7

208 ± 15a

0.04 ± 0.02

18.4 ± 1.9c

500

77.7 ± 4.6

204 ± 2a

0.10 ± 0.08

-11.9 ± 2.0a

2,000

89.1 ± 0.8

186 ± 5a

0.03 ± 0.03

18.1 ± 1.6c

2,000

92.9 ± 1.3

19

Figure 1

(a)

(c)

(b)

Figure 1.

(d)

Figure 2.

Phosphate concentration (mM)

-30

-40

Phosphate concentration (mM)

4

-20

3

0

2

-10

Zeta potential (mV)

(a)

0. 00 0. 05 0. 10 0. 15 0. 20 0. 25 0. 30 0. 35 0. 40 1

4

3

2

0. 40 1

0. 35

0. 30

0. 25

0. 20

0. 15

0. 10

0. 05

0. 00

Zeta potential (mV)

Figure 2

(b)

40

30

20

10 0

-10

-20

-30

Figure 3

(a)

S iz e ( d .n m )

400 300 200 100

4 .2 8 .3 1 .4 0 .8 : : : : 1 1 1 1 P o lo x a m e r : N P s ( w e ig h t r a tio )

(b)

Size (d.nm)

300 250 200 150 100

Figure 3.

0

50 100 Time (min)

150

Cellular uptake (% of viable cell population)

Figure 4

100

FITC-Dextran solution FITC-Dextran + RRH solution Unmodified NPs Unmodified NPs + RRH RRH (10 µmol)-tagged NPs RRH (100 µmol)-tagged NPs

80 60 40 20 0 Formulations

Figure 4.

Figure 5

FITC-dextran released (%)

120 100 80 60 40 20 0 0

Figure 5.

30

60

90 120 150 180 210 Time (min)

Graphical abstract

Highlights Chiu et al. EJPB • • •

We surface decorated PECA nanoparticles with arginine using a novel histidine anchoring method Arginine-tagging increased cellular uptake by 30% compared to unmodified nanoparticles Di-arginine was sufficient to enhance cell uptake

20