Diethylene glycol functionalized self-assembling peptide nanofibers and their hydrophobic drug delivery potential

Acta Biomaterialia 8 (2012) 3241–3250

Contents lists available at SciVerse ScienceDirect

Acta Biomaterialia journal homepage: www.elsevier.com/locate/actabiomat

Diethylene glycol functionalized self-assembling peptide nanofibers and their hydrophobic drug delivery potential Parisa Sadatmousavi a, Tewodros Mamo b, P. Chen a,⇑ a b

Department of Chemical Engineering, University of Waterloo, Waterloo, Ontario, Canada N2L 3G1 Nanotechnology Engineering, University of Waterloo, Waterloo, Ontario, Canada N2L 3G1

a r t i c l e

i n f o

Article history: Received 20 December 2011 Received in revised form 8 May 2012 Accepted 18 May 2012 Available online 27 May 2012 Keywords: Drug delivery Self-assembling peptide Amino acid pairing Diethylene glycol Functionalization

a b s t r a c t Self-assembling peptide nanofibers have emerged as important nanobiomaterials, with such applications as delivery of therapeutic agents and vaccines, nanofabrication and biomineralization, tissue engineering and regenerative medicine. Recently a new class of self-assembling peptides has been introduced, which takes into consideration amino acid pairing (AAP) strategies in the peptide sequence design. Even though these peptides have shown promising potential in the design of novel functional biomaterials, they have a propensity to initiate uncontrollable aggregation and be degraded by proteolytic enzymes. These present the most significant challenge in advancing self-assembling peptides for in vitro and in vivo applications. Functionalizing biomaterials with polyethylene glycol (PEG) has been shown to surmount such problems. Here the results of conjugating diethylene glycol (DEG), a short segment of PEG, to one of the AAP peptides, AAP8, with eight amino acids in sequence, are reported. The results indicate that incorporation of DEG into the peptide sequence modulates fiber self-assembly through creating more aligned and uniform nanostructures. This is associated with increasing solubility, stability, and secondary structure b-sheet content of the peptide. The DEG conjugate of AAP8 also shows reduced cellular cytotoxicity. Functionalization of AAP8 improves the capability of the peptide to stabilize and deliver a hydrophobic anticancer compound, ellipticine, in aqueous solution, consequently inducing greater cytotoxicity to lung carcinoma cells over a relatively long time, compared with non-functionalized AAP8. The presented functionalized peptide and its drug delivery application indicate a potentially useful design strategy for novel selfassembling peptide biomaterials for biotechnology and nanomedicine. Ó 2012 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

1. Introduction Self-assembling peptides are a class of biomaterials showing promising results for various biomedical applications, including delivery of therapeutic agents and vaccines, nanofabrication, biomineralization, tissue engineering, and regenerative medicine [1–3]. Recently a new class of self-assembling peptides based on the design principle amino acid pairing (AAP) has been introduced, with the model peptide AAP8 showing great promise in creating b-sheet-rich nanofibers, and stabilizing and delivering a hydrophobic anticancer agent [4]. The design of AAP self-assembling peptides is unique as it is based on the combination of several side-chain interactions, including hydrophobic interactions, electrostatic interactions, and hydrogen bonding. While these newly designed systems and other self-assembling peptides have shown immense potential, issues remain in optimizing the self-assembled structures and making them more robust for in vivo applications [5]. A major issue to be addressed in improving the self-assembled ⇑ Corresponding author. Tel.: +1 519 888 4567x35586; fax: +1 519 746 4979. E-mail address: [email protected] (P. Chen).

fibril nanostructure is avoiding uncontrollable aggregation of the b-sheet structures while improving the capability of the fibers to form predictable nanostructures [6,7]. The surface of the nanofibers must be modified for effective biomedical applications in vivo. Conjugating the peptide units with a short segment of polyethylene glycol (PEG) polymer is here proposed to address these issues. Surface functionalization of nanoparticles (NPs) with PEG has become a standard strategy to increase the NP half-life in the bloodstream, as the functionalization reduces protein opsonization and macrophage uptake [5,8,9]. By reducing non-specific interactions with proteins through its hydrophilicity and steric repulsion effects, PEG results in long lasting circulating drug delivery systems with reduced opsonization and complement activation [5]. PEG has been approved by the US Food and Drug Administration (FDA) for clinical use due to its low toxicity and lack of immunogenicity. A number of clinically approved therapeutics rely on PEG for improved in vivo profiles, including liposomes (Doxil), PEG–drug conjugates (Oncaspar) and polymeric NPs (Genexol-PM) [10]. Various researchers have also shown an effect of PEGylationon iron oxide NPs for in vivo cancer imaging due to the enhanced permeability and retention

1742-7061/$ - see front matter Ó 2012 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.actbio.2012.05.021

3242

P. Sadatmousavi et al. / Acta Biomaterialia 8 (2012) 3241–3250

(EPR) [11]. Liu et al. reported the properties of synthesized PEGylated NPs with low cytotoxicity and high resistance to phagocytosis by macrophages in vitro, as well as low accumulation in liver and spleen in vivo biodistribution tests. These results make these NPs highly preferable in tumor imaging. The authors claim that these outstanding characteristics are dependent on the significant shielding effect of the PEG coating, providing a neutral surface charge under physiological conditions [11]. The exceptional properties of PEG make this polymer a special candidate to resolve the issues involved in NP formation. NPs can have problems of insolubility, toxicity, bio-incompatibility and a low half-life in the bloodstream for in vitro, in vivo, and pre-clinical applications. The PEGylation of b-sheet-forming self-assembling peptides has been a subject of study since the first report by Lynn et al. on conjugation of PEG to the amyloid peptide (10-35 A b) [12,13]. This early work showed that PEGylation prevented the lateral association of fibrils, inhibiting irreversible steps of fibrillogenesis. Elastin-like polypeptides based on (VPGVG)4 and (VPAVG)4 were also conjugated with PEG, and enhanced the self-assembly properties [14]. Moreover, PEGylation of the GAGA peptide, which is based on Bombyxmori silk, enhanced peptide self-assembly through the formation of both parallel and anti-parallel b-sheets [15]. Another interesting work studied the effect of PEGylation on the Q11 self-assembling peptide through conjugation of different PEG chains at various points on the backbone of the peptide [7]. This work showed that PEG conjugation strongly prevented lateral aggregation of fibril-forming peptides, resulting in well-ordered peptide matrices. In the current work we focus on an amino acid pairing peptide, AAP8, with an eight amino acid sequence. AAP8 contains four hydrophobic phenylalanine (F) side-chains, a hydrogen bonding pair (QN), and an ionic pair (EK) in the sequence FEFQFNFK. This peptide previously showed great promise for various applications because of its unique self-assembling properties [4]. However, a lack of solubility and the possibility of uncontrollable aggregation led to the formation of non-uniform nanostructures, making this peptide unsuitable for further application. To overcome this issue and enhance the nanostructure configuration diethylene glycol (DEG), a short segment of the PEG chain, was conjugated to the model peptide AAP8. Since AAP8 is a short eight amino acid peptide it is hypothesized that conjugating a shorter segment of the PEG molecule should be enough to obtain the desired effects. We believe that the longer the PEG chain the more likely it is to interfere with self-assembly of the peptide. To investigate the various configurations of DEG conjugation AAP8 was conjugated with DEG on one or both terminals of the molecule, forming AAP8– DEG and DEG–AAP8–DEG. We utilized spectroscopy and microscopy techniques to determine the effect of DEG conjugation on peptide self-assembly, the secondary structure b-sheet content, and the formation of peptide nanostructures. To further investigate the drug delivery potential of AAP8 modified with diethylene glycol we used an anticancer agent, ellipticine, as a model hydrophobic drug. Lung carcinoma cells (A549) were used to evaluate the anticancer activity of ellipticine encapsulated with either AAP8 or modified AAP8.

2. Experimental 2.1. Materials The peptide AAP8 (Ac-FEFQFNFK-NH2) and the two diethylene glycol (DEG) (–NH–CH2–CH2–O–CH2–CH2–O–CH2–CO–) conjugated amino acid pairing (AAP8) peptides, Ac–FEFQFNFK–DEG– NH2 and Ac–DEG–FEFQFNFK–DEG–NH2, were purchased from CanPeptide Inc. (Quebec, Canada). They were characterized for their

self-assembling properties in aqueous solution at 0.5 mg ml1. The synthesis method for all three peptides was based on solid phase peptide synthesis. Matrix-assisted laser desorption ionization time of flight mass spectroscopy (MALDI-TOF-MS) gave a molecular weight of 1147.31 g mol1 for AAP8, 1293 g mol1 for Ac–AAP8– DEG–NH2, and 1437 g mol1 for Ac–DEG–AAP8–DEG–NH2. The purity of the peptides were verified by CanPeptide as >95% by liquid chromatography–mass spectrometry. The peptide aqueous solutions were prepared with pure water (Milli-Q, 18.2 MX). The anticancer agent ellipticine (99.8% pure) and fluorescent marker 1anilino-8-naphthalene sulfonate (ANS) (>97%) were purchased from Sigma-Aldrich (Oakville, Canada). Cell culture agents, including low glucose Dulbecco’s modified Eagle’s medium (DMEM), and phosphate-buffered saline (PBS) were purchased from HyClone (Ontario, Canada). Fetal bovine serum (FBS) and trypsin–EDTA were purchased from Invitrogen Canada Inc. (Burlington, Canada). A free comprehensive chemical drawing package, ACD/ChemSketch Freeware (Toronto, Canada), was used to draw the chemical structures of the peptide molecules. This software optimizes all bond lengths and angles based on energy minimization. 2.2. Peptide solution preparation Solutions of the APP8 peptide and DEG-conjugated AAP8 peptides were prepared by dissolving the peptide powder in pure water (18.2 MX, Millipore Milli-Q system) at specified concentrations for various experiments, including nanostructure visualization by atomic force microscopy (AFM), hydrophobicity testing by fluorescence spectroscopy, secondary structure evaluation and anticancer activity tests in vitro of the peptides in combination with an anticancer drug. 2.3. Atomic force microscopy (AFM) The self-assembled peptide nanostructures were imaged by AFM (PicoScan, Molecular Imaging, Phoenix, AZ) on a mica substrate. The samples were prepared as follows. A 10 ll sample of the peptide solution (15 min after preparation) was placed on a freshly cleaved mica substrate that was fixed on an AFM sample plate. The solution was incubated on the mica surface for 10 min, to allow the peptide nanostructures to adhere to the surface, and then rinsed with water twice. The extra water (moisture) on the surface was removed with tissue paper. The sample plate was then covered with a Petridish to avoid possible contamination and left to dry overnight. AFM images were taken at room temperature in tapping mode. All images were acquired using a 225 lm silicon single crystal cantilever (type NCL-16, Molecular Imaging, AZ) with a typical tip radius of 10 nm and frequency of 175 kHz. A scanner with a maximum scan size of 6  6 lm was used [16]. Data sets were subjected to third order flattening. The AFM images underwent further dimension analysis with Gwyddion software (free SPM data analysis software). 2.4. Steady-state fluorescence measurements The hydrophobicity of the peptide assemblies was investigated using the fluorescent probe ANS. ANS with a molecular weight of 299.3 g mol1 was used to characterize the hydrophobicity of the prepared AAP8 assemblies. ANS is a fluorescent dye that binds with high affinity to the hydrophobic surfaces of protein assemblies. The emission maximum of ANS undergoes a blue shift and the fluorescence intensity increases significantly upon ANS binding to low polarity regions of the protein surface. These properties mean that ANS is well suited to determining the hydrophobicity of peptide nanostructures. ANS can be excited at 360 nm in pure water and emits light at 520 nm, whereas in hydrophobic solution the

P. Sadatmousavi et al. / Acta Biomaterialia 8 (2012) 3241–3250

emission wavelength moves toward a lower value. A solution of 10 lM ANS was prepared in 10 mM phosphate buffer, pH 6. The fresh peptide solution was mixed with the same volume of ANS solution on a vortex mixer for 10 s. LS-100 spectrofluorometer (Photon Technology International, London, Canada) was used in this study. The light source used was a pulsed xenon lamp. For solution samples spectra were obtained using a quartz cell (1  1 cm). The ANS-containing samples were excited at 360 nm and the emission spectra collected from 420 to 670 nm. The excitation and emission slit widths were set at 0.5 mm and 1.25 mm, respectively [17].

3243

resuspended in fresh DMEM at a concentration of 2.5  104 cells ml1. A 200 ll volume of the cell suspension was added to each well of a clear, flat-bottomed, 96-well plate (Costar) and incubated for 24 h. The culture medium was replaced with 150 ll of fresh culture medium, followed by the addition of 50 ll of the complexes, their dilutions, or control samples to each well. Each treatment was replicated in four wells. The plates were incubated for 24 h prior to performing a cell viability assay. The MTT assay (TOX1. Sigma-Aldrich, Oakville, Canada) was used to determine cell viability after each treatment. The detailed procedure has been described in previous publications [4,18].

2.5. Fourier transform infrared (FTIR) spectroscopy The secondary structure of the peptides was determined by FTIR spectroscopy (Tensor37, Bruker Optics Ltd., Canada), using a BioATR II cell. A 30 ll of sample solution was injected into an ATR crystal (silicon) chamber. Scanning was between 1550 and 1700 cm1, with 256 runs for each measurement. The obtained spectra were analyzed using OPUS software (v.6.5, Bruker Optics Ltd, Canada). 2.6. Circular dichroism (CD) spectroscopy The secondary structures of the peptides were evaluated using another standard method, CD spectrometry (J-815, JASCO Inc., Easton, MD). A 150 ll aliquot of the sample solution at a concentration of 0.5 mg ml1 was transferred to a quartz cuvette with a path length of 1 mm. CD spectra were collected by a continuous scanning method from 190 to 260 nm at a scanning speed of 50 nm min1. Data were analyzed with JASCO software (J-810). 2.7. Dynamic light scattering (DLS) The zeta potentials of the peptide–ellipticine complexes were investigated in a Zetasizer Nano ZS (Malvern Instruments, Malvern, UK) with suitable viscosity and refractive index settings at 25 °C. Samples were injected into a disposable cell (folded capillary DTS-1060, Malvern Instruments, Malvern, UK) with a volume of 1 ml. The zeta potential distribution was directly calculated from the electrophoretic mobility distribution based on the Smoluchowski formula. For each solution the zeta potential determination was repeated three times. The values reported here correspond to the average peak zeta potential value distribution from replicates. 2.8. Peptide and ellipticine complex preparation Complexes were formed of ellipticine (0.1 mg ml1, 406 lM) with freshly prepared AAP8 (0.5 mg ml1 435 lM), AAP8–DEG (0.5 mg ml1 386 lM), and DEG–AAP8–DEG (0.5 mg ml1 347 lM). Ellipticine (1 mg ml1) crystals were dissolved in tetrahydrofuran (THF). Aliquots of ellipticine–THF were transferred to a 1.5 ml centrifuge tube and dried under flowing air. Fresh peptide solution was then added to the tube, followed by continuous probe sonication at 6 W power for 10 min. A control of ellipticine in pure water (without peptide) at the same concentration was prepared for comparison. 2.9. Cellular toxicity of the peptide–drug complexes A cancer cell line, non-small lung cancer cells A549, was used to determine the cellular toxicity of the peptide–ellipticine complexes. The cells were cultured in low glucose DMEM containing 10% FBS and 1% penicillin–streptomycin in an environmentally controlled incubator (37 °C, 5% CO2). After reaching 95% confluence the cells were trypsinized and suspended in DMEM, followed by centrifugation at 500 r.p.m. for 5 min. The cell pellets were

3. Results and discussion Self-assembly of short peptides into b-sheet fibril nanostructures paves the way for the synthesis of novel biomaterials for a wide range of applications (e.g. drug delivery and tissue engineering). To develop peptide-based nanostructures it is necessary to ensure that their configuration is controllable and uniform (e.g. of either fibrillar or spherical shape). One of the main drawbacks of b-sheet fibrillar structures is their tendency to aggregate in a tangled form, which are unusable in both in vitro and in vivo applications. We previously introduced the de novo design principle of amino acid pairing (AAP) and created a novel class of self-assembling peptides [4]. This design strategy focuses on amino acid pair interactions, including hydrogen bonding, ionic pairing and hydrophobic interactions. The current work concentrates on an AAP peptide with all three of the mentioned pairing interactions. This self-assembling peptide, AAP8, consists of eight amino acids: four phenylalanine (F) with hydrophobic side chains; a lysine (K) and a glutamic acid (E) forming an ionic pair; a glutamine (Q) and an asparagine (N) forming a hydrogen bonding pair. This peptide is phenylalanine-rich, which may generate b-sheet-rich fibrils but possibly reduce the solubility of the peptide in aqueous media. Since b-sheet formation is mainly due to intermolecular hydrogen bonding and hydrophobic interactions it is hypothesized that introducing organic molecules, such as polymers, via conjugation could regulate b-sheet formation and lead to a more oriented, uniform nanostructure. This hypothesis is supported by earlier work by Burkoth et al. [13] and Collier et al. [7], who showed that PEG-modified A b and Q11 peptides did not aggregate laterally. In addition to improvements in peptide selfassembly due to PEG conjugation, there are also numerous biological advantages with respect to drug delivery, such as solubility, long-term circulation in the bloodstream and protection of the NPs from enzymatic degradation [7,12–15]. In this work a short segment of PEG, DEG, with a molecular weight of 145 Da, was selected to functionalize AAP8 and forma peptide–DEG hybrid molecule. Since AAP8 is a short peptide, conjugating a short ethylene glycol would be appropriate, so that it does not interfere with peptide self-assembly to any large extent. Two hybrids were studied here: one a conjugate of DEG to the Cterminal of AAP8, the other a conjugate of DEG to both the Cand N-termini of AAP8. A schematic of the molecular formuli is presented in Fig. 1 (ACD/ChemSketch Freeware, Toronto, Canada). Hydrophilic amino acid residues (lysine and glutamic acid) are positioned on one side of the molecule, with hydrophobic amino acid residues (phenylalanine rings) located on the other, creating an amphiphilic side-to-side configuration. 3.1. Effect of DEG functionalization on self-assembly of the nanofibers 3.1.1. Nanostructure morphology AFM images of self-assembled peptide nanostructures on a mica surface showed significant differences in assembly morphologies between AAP8 and functionalized AAP8 (Fig. 2). AFM images

3244

P. Sadatmousavi et al. / Acta Biomaterialia 8 (2012) 3241–3250

Fig. 1. Molecular structure of (A) diethylene glycol (DEG), (B) amino acid pairing peptide AAP8 with monomer length 3.53 nm, (C) AAP8–DEG with monomer length 4.52 nm, (D) DEG–AAP8–DEG with monomer length 5.51 nm. The width of all molecular structures was 0.2–0.8 nm. All dimensions were estimated using ACD/ChemSketch Freeware (Toronto, Canada).

of AAP8 revealed that this peptide forms extremely tangled fibers and large aggregates (Fig. 2A). In contrast, the modified AAP8 showed a greatly improved morphology, revealing more uniform fibrillar structures, with little evidence of aggregation (Fig. 2B and C). Images were taken to observe the peptide assembly at two time points. AAP8 formed larger aggregates over time and in most cases did not attach to the mica surface (Fig. 2D). This latter observation is due to the hydrophobicity of AAP8, which reduces the likelihood of stabilization on mica. However, AAP8–DEG formed more structured fibers, with bundles consisting of a few fibers of increased length (Fig. 2E). The most interesting morphology observed was from DEG–AAP8–DEG, which produced highly uniform, long fibers aligned in parallel, with almost no evidence of aggregation or tangled fibers (Fig. 2F). In the analysis of the dimensions of the peptide assemblies the widths were deconvoluted using the method reported by Hong et al., to eliminate the convolution effect arising from the finite size of the AFM tip [16,19]. The observed dimensions have to be corrected. For spherical shapes with a radius Rm the observed width of the sample follows the relationship: Wobs = 4(RtRm)0.5, where Rt is the radius of the AFM tip, which is 10 nm in the present case. If the sample is a sheet the real width of the sheet can be calculated using the equation W = Wobs  2(2RtH  H2)0.5, where H is observed height. Since we later show a significant amount of b-sheet in the peptide secondary structure, use of the sheet formula for the dimension analysis is acceptable. The corrected width and height were calculated from an analysis of at least 50 independent fibers in each image, and are given in Table 1. The heights of assembled nano fibers were 0.7–0.8 nm for all three peptides in the freshly prepared samples, in agreement with the corresponding molecular determinations using the ACD/

ChemSketch drawing software, which determined the width to be 0.5–0.8 nm, from the top-most and bottom-most atoms in the van der Waals-based structures. The height of the AAP8 fibers increased, to almost twice the initial height, over the 10 day period, which is likely due to aggregation of the peptide, which did not attach to the mica surface. AAP8–DEG increased in height to some extent, but the height of the DEG–AAP8–DEG assemblies did not show any significant increase. Since DEG–AAP8–DEG is more hydrophilic and stable in aqueous solution the molecules tend to attach to the mica surface. AFM images of freshly made AAP8 demonstrated 16.86 ± 1.07 nm wide, short tangled fibers, with aggregates 76.08 ± 2.98 nm in diameter. The aggregates seemed to grow over 10 days, to 109.4 ± 5.6 nm in width. The ACD/ChemSketch software determined the length of the individual AAP8 molecules to be 3.35 nm. Thus the fiber width observed by AFM corresponds to more than one peptide molecule length. The dimension analysis of AAP8–DEG found thinner fibers, 10.2 ± 0.61 nm in width, even though the molecule is longer, 4.52 nm, than AAP8, as predicted by the ACD/ChemSketch software. Each fiber consists of approximately two or three DEG-conjugated AAP8 molecules stacked together width-wise. Over 10 days the fibers only grew to about 13.53 ± 0.87 nm in width, however some of the fibers associated together and created bundles 58.95 ± 3.99 nm in width. DEG–AAP8–DEG presented the most uniform fibrillar structures, whether the samples were fresh or had been aged for 10 days. The initial fibers were 8.21 ± 0.79 nm wide with no trace of aggregation. After 10 days the average fiber width was 10.97 ± 0.87 nm and the fibers showed an oriented and parallel pattern. The DEG–AAP8–DEG molecules were about 5.7 nm in length, as calculated using ACD/ChemSketch, longer than those of AAP8 and DEG–AAP8, but the resulting thinner fibers indicate a

3245

P. Sadatmousavi et al. / Acta Biomaterialia 8 (2012) 3241–3250

Fig. 2. Nanostructure of freshly prepared (A) AAP8, (B) AAP8–DEG, (C) DEG–AAP8–DEG, (D)10-day-old AAP8, (E) 10-day-old AAP8–DEG, and (F) 10-day-old DEG–AAP8–DEG determined by tapping mode atomic force microscopy (AFM) on a mica surface. All concentrations were 0.5 mg ml1.

Table 1 The dimensions of the AAP8, AAP8–DEG, and DEG–AAP8–DEG assemblies at two time points. Fresh

10 Days

Shapea

Width (nm)

Height (nm)

Shapea

Width (nm)

Height (nm)

AAP8–DEG

F A F

16.86 ± 1.07 76.08 ± 2.98 10.2 ± 0.61

0.69 ± 0.06 0.92 ± 0.03 0.74 ± 0.06

DEG–AAP8–DEG

F

8.21 ± 0.79

0.81 ± 0.081

F A F B F

17.06 ± 2.02 109.4 ± 5.65 13.53 ± 0.87 58.95 ± 3.99 10.97 ± 0.87

1.21 ± 0.21 3.15 ± 0.45 1.03 ± 0.09 2.16 ± 0.35 0.86 ± 0.09

AAP8

The heights and widths were obtained using Gwyddion software (free SPM data analysis software), with the widths corrected using the above deconvolution method [16]. a F, fibrillar assemblies; A, aggregates; B, bundles.

lower probability of aggregation. In comparison with AAP8, the DEG-conjugated AAP8 peptides were shown to spread more widely on the mica surface and assembled in a more uniform fashion. Conjugating DEG to the peptide significantly changes the nanostructure, which can be interpreted as due to steric shielding of the fibrils by DEG and an increase in stability in aqueous solution [7].

There is no clear peak observed for a-helix or b-turn. The intensity of these peaks increased from AAP8 to AAP8–DEG to DEG–AAP8– DEG (Fig. 3), which showed that the b-sheet content increased when DEG was conjugated to AAP8. This evidence indicates that functionalization of the peptide with DEG has a significant impact on its secondary structure and modulates its b-sheet content.

3.1.2. Secondary structure To further investigate the effect of DEG conjugation on the peptide assemblies the secondary structures of the peptide–DEG hybrids were determined by FTIR spectroscopy and CD. In all cases the b-sheet content increased when DEG was conjugated to the peptide sequence. The FTIR results indicated a strong peak at 1614–1622 cm1, and the CD spectra had a strong negative peak at 210–215 nm. The three main peaks in the amide I region are for b-sheet (1614–1622 cm1), a-helix (1650–1658 cm1) and bturn (1680 cm1) structures. In the FTIR spectra obtained for AAP8, AAP8–DEG and DEG–AAP8–DEG the dominant peak is at 1620 cm1, which corresponds to b-sheet secondary structure.

3.1.3. Hydrophilicity and the critical aggregation concentration (CAC) Since PEG is a hydrophilic polymer we hypothesized that conjugation of a short segment of PEG, such as DEG, to our relatively hydrophobic peptide, AAP8 with four phenol rings, would reduce the hydrophobicity of the peptide. In order to determine the hydrophobicity we utilized a fluorescent probe, which can bind to hydrophobic compounds and, as a consequence, emit light. We used the fluorescence probe ANS, which emits light at 470 nm in a hydrophobic environment, whereas its emission wavelength in pure water is 520 nm. Fig. 4A shows the fluorescence spectra of the ANS probe in the three different peptide solutions compared with that in pure water. All spectra were normalized to the light scattering in air at

3246

P. Sadatmousavi et al. / Acta Biomaterialia 8 (2012) 3241–3250

Fig. 3. Secondary structure determination by (A) FTIR, (B) circular dichroism. —, AAP8; —, AAP8–DEG; —, DEG–AAP8–DEG. All concentrations were 0.5 mg ml1 in Milli-Q water.

360 nm, to correct for fluctuations in the light source. The normalized fluorescence intensity of ANS shows the following trend: AAP8  AAP8–DEG  DEG–AAP8–DEG  H2O. The peak position of ANS in pure water (inset) is observed at 520 nm, but shifts to 470 nm for AAP8 and its DEG conjugates. In addition, the fluorescence intensity of the ANS probe decreased upon DEG functionalization of AAP8. Among AAP8, AAP8–DEG, and DEG–AAP8–DEG the latter showed the lowest fluorescence intensity, and was thus the least hydrophobic. As expected, conjugation of DEG on both ends enhanced the hydrophilicity of AAP8 compared with conjugation at one end, which explains the significant reduction in aggregation, likely caused by hydrophobic residues. The same method was used to determine the CAC. The CAC is a parameter evaluating the strength of self-assembly. The CAC is defined as the concentration of peptides above which peptide aggregates form spontaneously. The lower the CAC the higher the tendency to self-associate. In this study the CACs of the three peptides were determined by (ANS) fluorescence assay. The fluorescence intensity of the ANS probe did not change on increasing the peptide concentration at low concentrations of peptide in pure water, however above a certain concentration, which we call the CAC, the fluorescence intensity of ANS increased significantly. AAP8 showed the lowest CAC amongst the DEG-conjugated AAP8 peptides, and DEG–AAP8–DEG the highest. The observed values were 0.015, 0.06, and 0.15 mg ml1 for AAP8, DEG–AAP8, and DEG–AAP8–DEG, respectively (Fig. 4B). Thus the CAC increases with increasing DEG conjugation. On DEG conjugation the peptide becomes more hydrophilic, which would require more peptide chains to form self-assembled nanostructures, increasing the minimum peptide concentration required (i.e. a higher CAC). DEG functionalization facilitates peptide self-assembly in more organized ways and avoids aggregation at low concentrations. As observed

Fig. 4. (A) Hydrophobicity of AAP8 (—), AAP8–DEG (—), DEG–AAP8–DEG (—), and pure water (—). The inset shows the fluorescence control ANS probe in water without the peptide. The peptide concentration was 0.5 mg ml1 and the ANS concentration 10 lM. (B) CAC determination using an ANS fluorescence probe.

in the AFM images, DEG–AAP8–DEG does not show lateral aggregation, even at a high concentration (0.5 mg ml1), however, both AAP8 and AAP8–DEG showed aggregation and tangled fibers. Thus DEG conjugation to both ends of AAP8 provides amore controllable and stable structure. 3.2. The effect of DEG functionalization on encapsulation of a hydrophobic drug and cytotoxicity 3.2.1. Ellipticine encapsulation As a next step we investigated the potential use of AAP8 and its DEG conjugates as carriers for hydrophobic anticancer drugs. Ellipticine, with anticancer activity, has been used as a hydrophobic model drug in past drug delivery research [4,18,20,21]. Ellipticine, a cytotoxic plant alkaloid, is a polycyclic molecule that intercalates between DNA base pairs, inhibits topoisomerase II, and induces G2/M phase cell cycle arrest [22]. The fluorescence properties of ellipticine make it easy to characterize encapsulation and release of the drug. Despite these advantages, clinical trials of ellipticine were suspended due to an inability to target the drug, and accumulation at undesired sites, which can be explained by the insolubility of ellipticine in aqueous media [23]. The challenges regarding delivery of hydrophobic compounds highlight a strong need for more effective delivery systems. We previously showed the great potential of AAP8 to deliver ellipticineto cancer cells [4]. However,

3247

P. Sadatmousavi et al. / Acta Biomaterialia 8 (2012) 3241–3250

AAP8 without modification might be challenging for further in vitro and in vivo tests due to its tendency to aggregate and its cytotoxicity. The therapeutic efficiency of drug carriers may be limited by particle recognition by macrophages of the mononuclear phagocyte system (MPS) and the possibility of rapid elimination from the bloodstream after intravenous injection. Thus the modification of AAP8 with DEG is a means to overcome aggregation and may improve its drug delivery potential [8,24,25]. Preliminary characterization has been performed on the peptide and drug complexes to investigate the differences between complexes formed from modified and unmodified AAP8. Table 2 summarizes the data for the three complexes. Ellipticine emission spectra of collected with a spectrofluorometer showed ellipticineto be in the protonated state (peak at 520 nm) for all AAP8–ellipticine, AAP8–DEG–ellipticine and DEG–AAP8–DEG–ellipticinecomplexes at a peptide concentration of 0.5 mg ml1 and ellipticine concentration of 0.1 mg ml1. The differences between the current study and Fung et al. [4] regarding the form of ellipticine in solution is due to the age of the complexes in the two experiments. In Fung et al. [4]AAP8 was added to ellipticine and mechanically stirred at 900 r.p.m. for 24 h. However, in the current study AAP8 was added to ellipticine and the complexes formed by a probe sonication method at a power of 6–8 W for 5–10 min and immediately analyzed by fluorescence spectroscopy and studied for cytotoxicity. In this case the complex product is a transparent yellow color, similar to that of the EAK–ellipticine complexes reported in Fung et al. [18]. As discussed in Fung et al.’s work [18] and shown in figure 4 of that report, whether ellipticine is protonated or crystalline is time dependent. In the case of freshly prepared complexes protonated ellipticine is dominant, whereas after 10 h the crystalline form of ellipticine dominates in solution. Since in the current study the focus was not on studying the state of ellipticine over time in solution and only freshly prepared samples were used in the experiments the protonated state of is reported for AAP8–ellipticine complexes. Note that the protonation of ellipticine usually occurs at higher peptide concentrations and relatively low pH (i.e. 65), lower than the pKa of ellipticine (6.0) [17]. The pH of freshly prepared AAP8 was 4.3, which provides an acidic environment for the protonation of ellipticine. Since ellipticine is stable in solution in the protonated form it is expected to be positively charged. This was confirmed by zeta potential measurements for the complexes, with no significant differences being observed for DEG-conjugated and non-conjugated AAP8 in complex with ellipticine. Since the zeta potential of all three complexes was more positive than +30 mV they are considered to be stable in solution. The interaction between the hydrophobic drug ellipticine and the peptide AAP8 is based on hydrophobic interactions between phenylalanine side-chains and ellipticine. The four Phe residues are strongly hydrophobic and, assisted by p–p stacking to stabilize the nanostructure, consequently interact with ellipticine. This hydrophobic interaction between the peptide and ellipticine remains constant on modification of AAP8 with DEG, however, DEG helps to stabilize the complex in aqueous solution. Unmodified AAP8 tends to aggregate overtime when in complex with ellipticine in an aqueous environment, whereas DEG-modified AAP8 provides a more hydrophilic environment resulting in stable fibers in

Table 2 Complex characterization.

AAP8–ellipticine AAP8–DEG–ellipticine DEG–AAP8–DEG–ellipticine

Zeta potential (mV)

pH

Ellipticine state

36.6 ± 0.41 30.46 ± 0.64 30.86 ± 0.38

5.0 ± 0.1 5.1 ± 0.1 5.36 ± 0.03

Protonated Protonated Protonated

Peptides concentration 0.5 mg ml–1; ellipticine concentration 0.1 mg ml–1.

complex with ellipticine in aqueous solution. AFM images illustrated the above hypothesis, showing that DEG-modified AAP8 suppressed aggregation of AAP8 in complex with ellipticine and more efficiently stabilized ellipticine in aqueous solution. AAP8 in complex with ellipticine formed cylindrical particles with a hydrodynamic diameter of 212 nm, which is comparable with the DLS results (Table 3). However, both DEG-modified AAP8 peptides in complex with ellipticine formed fiber structures (Fig. 5). The nanostructure of the peptide–ellipticine complexes observed in the AFM images showed wider fibers compared with those in the peptide nanostructure, as shown in Fig. 2. The calculated width, height and length of observed fibers are summarized in Table 3. The mechanism of complexation between the peptide and ellipticine is not fully understood. As discussed in Fung et al. [4], a model has been proposed for AAP8 nanofibers via amino acid pairing, where peptide molecules are aligned in an anti-parallel fashion to form nanofibers with a width equal to the length or a multiple of the length of the peptide chain. This may be seen in Fig. 2 and Table 1. However, in peptide–ellipticine complexes this alignment is disrupted and as a result wider nanofibers may form, perhaps because of potential multiple interactions between the peptide and ellipticine molecules (e.g. ionic, hydrogen bonding and hydrophobic interactions). It can be observed from the AFM images that the height of the nanofibers does not increase significantly. One may postulate that fiber growth in peptide–ellipticine complexes is not isotropic, but the reason is unclear and requires more investigation. The average hydrodynamic diameter of the particles was calculated based on the following equation for cylindrical particles length L and diameter (width) w observed in AFM images. This equation is applied for AAP8 and ellipticine complex particles, since they formed cylinders, whereas this equation cannot be applied to fibrillar structures [26]

pffiffiffiffiffiffiffiffiffiffiffiffiffiffi DH ¼ L 1  x2

, ln

1þ

pffiffiffiffiffiffiffiffiffiffiffiffiffiffi! 1  x2 ; x

x¼

w ½1 þ 0:37ðL  wÞ=L: L

3.2.2. Cytotoxicity The cytotoxicity of unmodified AAP8 and DEG-modified AAP8 peptises was evaluated by MTT assay on lung carcinoma cells, i.e. cell line A549, after incubation times of 24 and 48 h. Cells treated with AAP8 showed 30–40% cytotoxicity. However, cells treated with AAP8 modified with DEG on one end or both ends showed lower cytotoxicity (0–5%). DEG functionalization of AAP8 has a clear beneficial effect on the cytotoxicity of the peptide, rendering

Table 3 Dimensions of ellipticine complexes with AAP8, AAP8–DEG, and DEG–AAP8–DEG assemblies.

AAP8–ellipticine AAP8–DEG–ellipticine DEG–AAP8–DEG–ellipticine

Width (nm)

Length (nm)

Height (nm)

(DH)

126.11 ± 4.94 98.79 ± 3.369 74.07 ± 2.044

339.36 ± 8.57 283.891 ± 1.374 472.17 ± 27

5.398 ± 0.325 1.21 ± 0.106 0.819 ± 0.103

212.91

Heights and widths were obtained using Gwyddion software (free SPM data analysis software), and the widths were corrected using the above deconvolution method [16]. The hydrodynamic diameter (DH) was calculated based on L and w [26].

3248

P. Sadatmousavi et al. / Acta Biomaterialia 8 (2012) 3241–3250

Fig. 5. Nanostructure of freshly prepared ellipticine complexes with (A) AAP8, (B) AAP8–DEG, and (C) DEG–AAP8–DEG determined by tapping mode atomic force microscopy (AFM) on a mica surface. Concentrations were 0.5 mg ml1peptides and 0.1 mg ml1ellipticine.

Fig. 6. The cytotoxicity (A) controls (negative control, AAP8, AAP8–DEG, and DEG–AAP8–DEG (1.07–109 lM) incubated for 24 h; (B) ellipticine complexed with AAP8, AAP8– DEG, and DEG–AAP8–DEG and in pure water incubated for 24 h; (C) ellipticine complexed with AAP8, AAP8–DEG, and DEG–AAP8–DEG and in pure water incubated for 48 h. The maximum ellipticine concentration was 100 lM serially diluted to 1.56 lM. The peptide concentration rangeg from 1 to 100 lM. Note that the molar ratio of ellipticine to peptide remained constant at 1:1. DMEM corresponds to untreated cells. Error bars represent the standard error of the means for the 95% confidence interval.

it completely non-toxic at all concentrations evaluated (up to 100 lM) (Fig. 6A). The anticancer activity of ellipticine encapsulated with unmodified AAP8 and modified AAP8 peptides was examined on cell line A549 after incubation for 24 and 48 h by MTT assay. The results show that the cytotoxicity of ellipticine in complex with modified

AAP8 peptides was lower than that of the corresponding unmodified AAP8, which can be understood by our two main observations: first, the cytotoxicity of AAP8 is higher compared with that of modified AAP8s, which consequently may induce greater cytotoxicity together with ellipticine; second, the modified AAP8 peptides, i.e. AAP8–DEG and DEG–AAP8–DEG, stabilize ellipticine which

P. Sadatmousavi et al. / Acta Biomaterialia 8 (2012) 3241–3250

remains suspended longer in aqueous media, resulting in slower release of ellipticine from the complex and a delay in drug contact with adhesive cells, compared with AAP8–ellipticine complexes. This was observed after 24 h incubation (Fig. 6B). Despite the fact that complexes of modified AAP8 peptides and ellipticine showed lower toxicity to cancer cells than the corresponding unmodified AAP8 complexes after 24 h treatment the cytotoxicity of the complexes is higher than the ellipticine control at the highest concentration (100 lM). This indicates that both modified and unmodified AAP8 carriers are more effective in delivering ellipticine and causing cytotoxicity than the drug control. This is primarily because ellipticine is stabilized in AAP8 and DEG-conjugated AAP8 peptides in aqueous media. However, the cells treated with the complexes for 48 h showed the reverse trend in cytotoxicity for modified and unmodified AAP8 in complex with ellipticine, up to 50–100 lM of both peptide and ellipticine. The cytotoxicity of the complexes followed the trend: DEG–AAP8–DEG–ellipticine > AAP8–DEG–ellipticine > AAP8–ellipticine. This can be explained by aggregation of particles in the AAP8 and ellipticine solution overtime and settling in the cellcontaining wells without penetration of the cell membranes, whereas at shorter incubation times, within 24 h, the possibility of aggregation is lower, which allows penetration of the cell membrane, resulting in greater toxicity. On the other hand, the cytotoxicity of the modified AAP8 peptides and ellipticine complexes improved overtime as ellipticine stabilized in solution will eventually reach the cells, causing greater cytotoxicity (Fig. 6C). This phenomenon happened only for higher concentrations of complexes of DEG-modified AAP8 and ellipticine (50–100 lM peptide and ellipticine), i.e. up to 4 times dilution, and was expected, as the difference in aggregation between the modified and unmodified peptides is significant only at relatively high concentrations.

4. Conclusions Nanostructural characterization of the peptide AAP8 and DEGconjugated AAP8 indicates that functionalizing the short selfassembling peptide with DEG is an effective means to enhance the b-sheet secondary structure content and modulate the nanostructure to an aligned and uniform fibrillar configuration, minimizing random aggregation. Unmodified AAP8 showed significant aggregation over time, whereas DEG-conjugated AAP8 peptides, particularly DEG–AAP8–DEG, overcame this problem. We also observed higher hydrophilicity, in particular when DEG was conjugated to both ends of AAP8. The improvement in peptide self-assembly upon DEG conjugation supports the hypothesis that conjugating a short chain of PEG to peptides produces a more uniform and organized nanostructure, increased b-sheet content and enhance solubility in aqueous media, which all lead to low lateral aggregation. These peptides showed reduced cytotoxicity to lung carcinoma cells compared with unmodified AAP8, which is considered an advantage for drug carriers. Both the modified and unmodified AAP8 peptides showed an impressive ability to stabilize the hydrophobic anticancer drug ellipticine in aqueous solution. All three complexes of AAP8, AAP8–DEG and DEG–AAP8–DEG with ellipticine exhibited greater cytotoxicity to A549 cells than an ellipticine control, indicating a greater ability of peptide-based carriers to stabilizing and deliver ellipticine. The viability of cells treated with the two DEG-modified AAP8 peptides in complex with ellipticine was higher than the corresponding AAP8 and ellipticine complexes after 24 h incubation. This may be due to delayed contact between ellipticine, when complexed with DEG–modified AAP8 peptides, and adherent cells. After 48 h incubation the DEGmodified AAP8 peptides were more effective than unmodified AAP8 complexed with ellipticine, possibly due to aggregation

3249

associated with AAP8 and ellipticine in aqueous solution over a relatively long time. The results reported here provide a means to overcome critical issues in aggregation and inherent cytotoxicity associated with the use of biomaterials, and pave the way to the further design of functionalized biomaterials for improved drug delivery applications. Acknowledgements This research was financially supported by the Natural Sciences and Engineering Research Council of Canada, the Canadian Foundation for Innovation and the Canada Research Chairs Program to one of the co-authors (P.C.) and the Alexander Graham Bell Doctoral Scholarship for another co-author (P.S.). We also thank Dr E. Meiering for letting us use the FTIR system and Donghan Sohn for some cellular studies at the University of Waterloo. Appendix A. Figures with essential color discrimination Certain figures in this article, particularly Figs. 2–6, are difficult to interpret in black and white. The full color images can be found in the on-line version, at http://dx.doi.org/10.1016/j.actbio.2012. 05.021. References [1] Collins L, Parker AL, Gehman JD, Eckley L, Perugini MA, Separovic F, et al. Selfassembly of peptides into spherical nanoparticles for delivery of hydrophilic moieties to the cytosol. ACS Nano 2010;4:2856–64. [2] Rudra JS, Tian YF, Jung JP, Collier JH. A self-assembling peptide acting as an immune adjuvant. Proc Natl Acad Sci USA 2010;107:622–7. [3] Zhao X, Pan F, Xu H, Yaseen M, Shan H, Hauser CAE, et al. Molecular selfassembly and applications of designer peptide amphiphiles. Chem Soc Rev 2010;39:3480–98. [4] Fung S, Yang H, Sadatmousavi P, Sheng Y, Mamo T, Nazarian R, et al. Amino acid pairing for de novo design of self-assembling peptides and their drug delivery potential. Adv Funct Mater 2011;21:2456–64. [5] Alexis F, Pridgen E, Molnar LK, Farokhzad OC. Factors affecting the clearance and biodistribution of polymeric nanoparticles. Mol Pharm 2008;5:505–15. [6] Koenig HM, Kilbinger AFM. Learning from nature: beta-sheet-mimicking copolymers get organized. Angew Chem Int Ed 2007;46:8334–40. [7] Collier JH, Messersmith PB. Self-assembling polymer–peptide conjugates: nanostructural tailoring. Adv Mater 2004;16:907–10. [8] Owens DE, Peppas NA. Opsonization, biodistribution, and pharmacokinetics of polymeric nanoparticles. Int J Pharm 2006;307:93–102. [9] Langer R. Drug delivery and targeting. Nature 1998;392:5–10. [10] Zhang L, Gu FX, Chan JM, Wang AZ, Langer RS, Farokhzad OC. Nanoparticles in medicine: therapeutic applications and developments. Clin Pharmacol Ther 2008;83:761–9. [11] Liu D, Wu W, Ling J, Wen S, Gu N, Zhang X. Effective PEGylation of iron oxide nanoparticles for high performance in vivo cancer imaging. Adv Funct Mater 2011;21:1498–504. [12] Lynn DG, Meredith SC. Review. Model peptides and the physicochemical approach to beta-amyloids. J Struct Biol 2000;130:153–73. [13] Burkoth TS, Benzinger TLS, Jones DNM, Hallenga K, Meredith SC, Lynn DG. Cterminal PEG blocks the irreversible step in beta-amyloid (10–35) fibrillogenesis. J Am Chem Soc 1998;120:7655–6. [14] Pechar M, Brus J, Kostka L, Konak C, Urbanova M, Slouf M. Thermoresponsive self-assembly of short elastin-like polypentapeptides and their poly(ethylene glycol) derivatives. Macromol Biosci 2007;7:56–69. [15] Hamley IW, Ansari A, Castelletto V, Nuhn H, Rosler A, Klok HA. Solution selfassembly of hybrid block copolymers containing poly(ethylene glycol) and amphiphilic beta-strand peptide sequences. Biomacromolecules 2005;6:1310–5. [16] Hong YS, Legge RL, Zhang S, Chen P. Effect of amino acid sequence and pH on nanofiber formation of self-assembling peptides EAK16-II and EAK16-IV. Biomacromolecules 2003;4:1433–42. [17] Fung SY, Yang H, Chen P. Sequence effect of self-assembling peptides on the complexation and in vitro delivery of the hydrophobic anticancer drug ellipticine. PLoS ONE 2008;3:e1956. [18] Fung SY, Yang H, Bhola PT, Sadatmousavi P, Muzar E, Liu M, et al. Selfassembling peptide as a potential carrier for hydrophobic anticancer drug ellipticine: complexation, release and in vitro delivery. Adv Funct Mater 2009;19:74–83. [19] Yang H, Fung S, Pritzker M, Chen P. Surface-assisted assembly of an ioniccomplementary peptide: controllable growth of nanofibers. J Am Chem Soc 2007;129:12200–10.

3250

P. Sadatmousavi et al. / Acta Biomaterialia 8 (2012) 3241–3250

[20] Fung SY, Yang H, Chen P. Formation of colloidal suspension of hydrophobic compounds with an amphiphilic self-assembling peptide. Colloids Surf B 2007;55:200–11. [21] Fung SY, Duhamel J, Chen P. Solvent effect on the photophysical properties of the anticancer agent ellipticine. J Phys Chem A 2006;110:11446–54. [22] Tian E, Landowski TH, Stephens OW, Yaccoby S, Barlogie B, Shaughnessy JD. Ellipticine derivative NSC 338258 represents a potential new antineoplastic agent for the treatment of multiple myeloma. Mol Cancer Ther 2008;7(3):500–9. [23] Cavalcanti LP, Torriani IL. Thermotropic phase behavior of DPPC liposome systems in the presence of the anti-cancer agent ‘Ellipticine’. Eur Biophys J 2006;36:67–71.

[24] Peracchia MT, Vauthier C, Desmaele D, Gulik A, Dedieu JC, Demoy M, et al. Pegylated nanoparticles from a novel methoxypolyethylene glycol cyanoacrylate hexadecyl cyanoacrylate amphiphilic copolymer. Pharm Res 1998;15:550–6. [25] Fant K, Eshjoerner EK, Jenkins A, Grossel MC, Lincoln P, Norden B. Effects of PEGylation and acetylation of PAMAM dendrimers on DNA binding, cytotoxicity and in vitro transfection efficiency. Mol Pharm 2010;7:1734–46. [26] Lomakin A, Chung DS, Benedek GB, Kirschner DA, Teplow DB. On the nucleation and growth of amyloid beta-protein fibrils: detection of nuclei and quantitation of rate constants. Proc Natl Acad Sci USA 1996;93:1125–9.