ARTICLE pubs.acs.org/Langmuir
The Effects of Substituent Grafting on the Interaction of pH-Responsive Polymers with Phospholipid Monolayers Shengwen Zhang, Andrew Nelson,* Zachary Coldrick, and Rongjun Chen Centre for Molecular Nanoscience (CMNS), School of Chemistry, University of Leeds, Leeds LS2 9JT, United Kingdom ABSTRACT: pH-responsive amphiphilic polymers with suitable graftings have demonstrated highly efficient cell membrane activity and hence are promising applicants for drug-delivery. Grafting the hydrophobic amino acid L-phenylalanine and the hydrophilic methoxy poly(ethylene glycol) amine onto the pendant carboxylic acid moieties of a linear polyamide, poly(L-lysine isophthalamide), can effectively modify the amphiphilicity and conformation of the amphiphilic polymers. Here, the interactions of these polymers with phospholipid monolayers adsorbed on mercury (Hg) electrodes have been studied. AC voltammetry (ACV), rapid cyclic voltammetry (RCV), and electrochemical impedance spectroscopy (EIS) have been applied to monitor phospholipid monolayer associations with different polymer concentrations under different pH values. The polymers interact reversibly with the monolayer shown by altering the monolayer capacitance and inhibiting the phospholipid reorientation in electric field. Polymer grafting enhances the pH-mediated conformational change of the polymers which in turn increases their phospholipid monolayer activity. The most significant monolayer interactions have been observed with the polymer grafted with hydrophobic L-phenylalanine. A low level of PEGylation of the backbone also increases the monolayer activity. The polymer/DOPC interactions have been represented with an impedance model, which takes account of the interaction giving rise to an increase in monolayer capacitance and inhomogeneity and a Debye type dielectric relaxation. The extent of penetration of the polymers into the monolayer is inversely related to the electrical resistance they give rise to during the Debye relaxation. The cell membrane activities of these amphiphilic polymers have been successfully mirrored in this supported DOPC monolayer system, isolating the key parameters for biomembrane activities and giving insight into the mechanism of the interactions. The conclusions from this study provide strategic directions in material design catering to different requirements in biomedical applications.
1. INTRODUCTION Synthetic amphiphilic polymers with alternating hydrophobic segments and ionizable carboxyl groups can mimic the celltransduction function of some fusogenic viral peptides.16 These polymers exhibit a change of conformation from extended charged chains to globular hydrophobic structures upon a decrease of solution pH below their pKa ranges. Hydrophobic interactions between the resulting compact hydrophobic structures and phospholipid cell membranes facilitate the membrane-binding of the polymers, which leads to subsequent membrane disruption.1,3,5,7 The pH-mediated conformational changes are attributed to the balance between hydrophobic association, originating from the hydrophobic segments, and the electrostatic repulsion from the charged carboxyl groups.8 Hence, modulating the balance between the hydrophobic and hydrophilic components of the polymers can manipulate the conformational change and, consequently, the membrane disruption within the pH range characteristic of endosomes. By disrupting the endosomal membrane, cytoplasmic delivery to subcellular targets can be achieved, thus avoiding the lysosomal trafficking of biomacromolecular drugs (e.g., proteins and nucleic acids), which will be degraded by lysosomal enzymes.9,10 Poly(L-lysine isophthalamide) (PLP), a biodegradable, metabolite-derived polyamide, has attracted r 2011 American Chemical Society
particular interests.6,11 With a hydrophobic backbone and pendant carboxyl groups, PLP displays a typical coil-to-globule conformational change.12 However, its cell membrane disruptive capacity is rather limited and only occurs at lysosomal pH values.1,11 Further studies have shown that, by grafting hydrophobic amino acids (including L-phenylalanine) or hydrophilic methoxy poly(ethylene glycol) amine (mPEG-NH2) on to the pendant carboxylic acid moieties of PLP, the amphiphilicity and structure of the polymer can be modified, facilitating dramatically enhanced conformational changes and biomembrane disruptive activity.1,35 Effective biomembrane activity of the grafted polymers, compared to the parent polymer PLP, has been demonstrated at endosomal pH values with the hemolysis of red blood cell and release of lactate dehydrogenase of HeLa and Chinese hamster ovary cells.1,3,5,11 In the sheep erythrocyte models of endosomes, the phenylalanine-grafted polymer can be 35-fold more effective in membrane disruption on a molar basis than the highly lytic bee sting peptide melittin at pH 6.5, within the pH range characteristic of early endosomes. Successful cytoplasmic delivery of the Received: January 4, 2011 Revised: April 19, 2011 Published: June 09, 2011 8530
dx.doi.org/10.1021/la105125d | Langmuir 2011, 27, 8530–8539
Langmuir model drug calcein3 and the therapeutic protein MBP-Apoptin2 has also been achieved with these pH-responsive polymers. Together with their low cytotoxicity,6,11 these pH-responsive polymers have potential applications in intracellular drug delivery. Although much work has been done regarding the polymer synthesis, conformation, aggregation, cell membrane disruptive activity, in vitro cytotoxicity, intracellular trafficking of the polymers, and their drug delivery applications,16,1012 the fundamental mechanism of the polymer/biomembrane interaction remains unexplored. Therefore, it is the aim of this work to examine the effect of the grafting, particularly the hydrophobicity of the grafting, on the pH-response and the subsequent membrane activity of the amphiphilic parent polymer PLP using a biomembrane model. Among all polymer derivatives grafted with hydrophobic amino acids, PP-75 (degree of grafting with L-phenylalanine = 63.2 mol %), exhibits the strongest cell membrane interaction.3,5 For the PLP derivatives grafted with hydrophilic (mPEG-NH2), PA-1 (degree of PEGylation = 0.9 mol %) and PA-1.5 (degree of PEGylation =1.3 mol %) have been proven to be the most effective.1,11 Therefore, PP-75 and PA-1, together with the parent polymer PLP (Mn = 17.9 kDa, polydispersity = 1.99), have been chosen in this work to study the effect of grafting. Investigations of biomembrane interactions can be carried out using a biological membrane model. The phospholipid monolayer on a mercury (Hg) electrode is such a classic supported biomembrane model.1318 Phospholipids, for example, dioleoyl phosphatidylcholine (DOPC), form fluid, yet impermeable monolayers on the Hg electrode and as such resemble the outer leaflet of phospholipid bilayer vesicles.19,20 Therefore, the modifications of the monolayer caused by the active compounds can be simultaneously monitored electrochemically in a well controlled manner. This system has been successfully employed in the study of protein binding,21 antibiotic/biomembrane interaction,2224 ion channel function,25 and peptide/biomembrane interaction,2628 with exceptional sensitivity, accuracy, and reproducibility. Recently, this supported phospholipid/Hg system has been improved by replacing the traditional hanging Hg drop electrode (HMDE) with a modernized chip based device.20,29 This Hg electrode which is formed by electrodepositing Hg on to a Pt electrode fabricated on to a silicon wafer substrate shows much improved mechanical stability without compromising the advantages of the HMDE.20 In addition, this wafer-based Hg film electrode (MFE) can be electrochemically cleaned in situ, and hence it uses substantially less Hg during the measurements.20 As a result, this device has been chosen in part of this work to investigate the polymer/monolayer interaction.
2. EXPERIMENTAL SECTION 2.1. Materials. Poly(L-lysine isophthalamide) (PLP, Mn = 17.9 kDa, polydispersity = 1.99) was synthesized by hydrolyzing poly(L-lysine methyl ester isophthalamide) and converting to the sodium salt form, as reported previously.6 Its derivative, PA-1, was synthesized by grafting methoxy poly(ethylene glycol) amine (mPEG-NH2, Mn = 4.4 kDa, degree of PEGylation = 0.9 mol %) onto the pendant carboxylic acid of PLP.10 Grafting L-phenylalanine onto the pendant carboxylic acid moieties along the backbone of PLP produced another derivative PP-75 (degree of grafting with L-phenylalanine = 63.2 mol %).3,4 The numbers 1 and 75 in the polymer derivatives PA-1 and PP-75 indicate the stoichiometric molar percentages of the substituent of mPEG-NH2 and L-phenylalanine, respectively, relative to the pendant carboxyl groups in PLP. The unit structures of these polymers are presented in Scheme 1.
ARTICLE
Scheme 1. Unit Structures of PLP, PA, and PP Polymers
Polymer additions to buffered electrolyte with concentrations of 1 and 10 μg cm3 were prepared.
2.2. Microfabricated Wafer Device with Mercury Film Electrode. The fabrication of the wafer based MFE has been described previously.20,29,30 Pt discs with a radius of 0.480 mm and contact pads were developed on a silicon wafer substrate by a wet etching process.20 Electrodeposition of Hg onto Pt discs was carried out in a standard three electrode cell in 0.05 mol dm3 Hg(NO3)2 (Sigma-Aldrich Co.) at 0.4 V with a Pt counter electrode (Metrohm UK Ltd.) and a double junction reference electrode (3.5 mol dm3 KCl, Ag/AgCl inner filling, 0.1 mol dm3 perchloric acid outer filler, Metrohm UK Ltd.). One coulomb of charge was passed during the deposition so forming a Hg film with surface area of ∼0.75 mm2 on the Pt disk.20 Initially, Hg forms droplets on the Pt substrate during electrodeposition. However, once a sufficient amount of Hg has been deposited on the Pt, the Hg coalesces to form a semi-oblate spheroid structure. Light microscope images of the Pt electrode with Hg droplets during the initial deposition and the subsequent Hg film formation have been displayed in a previous publication.20 In addition, with reference to the geometry of the Hg film deposited after the passage one coulomb of charge, a surface area of 0.744 mm2 has been calculated for the semi-oblate spheroid Pt/Hg electrode.20 The surface area of the Pt disk electrode has been optically measured as 0.724 mm2.20 The calculated surface area for the Pt/Hg electrode is slightly higher than that for the disk because the semi-oblate spheroid surface area lies between the bounds of a perfectly flat film and a hemisphere.20 The MFEs were then rinsed with 18.2 MΩ Milli-Q water and dried under nitrogen gas before use. A special cell with a volume of ∼1.5 cm3 has been designed to host the MFE wafer.20 The cell employed a Ag/AgCl:3.5 mol dm3 KCl reference electrode separated from the electrolyte by a porous glass frit. All potentials in this paper are quoted versus the potential of this reference electrode. A rectangular Pt electrode of ∼14.9 mm2 on the wafer is used as the auxiliary electrode. A DOPC (Avanti Polar Lipids) aqueous suspension (2.54 mmol dm3) is introduced into 0.1 mol dm3 KCl electrolyte, and repetitive rapid cyclic voltammetry (RCV) at scan rate of 40 V s1 is applied to the MFE for the duration of 1 min (860 cycles) between 0.2 and 1.6 V to form a stable DOPC monolayer on the Hg. Inspection with RCV between 0.2 and 1.6 at 40 V s1 had confirmed the integrity of the monolayer.20 In addition, this MFE can be electrochemically cleaned in situ by RCV to potentials between ca. 1.8 and 2.3 V, which progressively destabilizes adsorbed organic matter on the electrode.20 The pH of the 0.1 mol dm3 KCl electrolyte for the cleaning procedure ranges from 5.5 to 8.2 using added buffers as described in section 2.3. No hydrogen evolution has been observed during the cleaning procedure using RCV between 0.40 V and 2.3 V. 2.3. Electrochemical Measurements. Voltammetric techniques have been employed in this work to study (a) the monolayer capacitative elements with out-of-phase alternating current voltammetry (ACV),17,28,31 8531
dx.doi.org/10.1021/la105125d |Langmuir 2011, 27, 8530–8539
Langmuir
ARTICLE
(b) the recovery of the phospholipid monolayer with rapid cyclic voltammetry (RCV), and (c) the initial polymer/phospholipid interactions with electrochemical impedance spectroscopy (EIS). Within the capacitance minimum region, that is, at the potential of zero charge (PZC), the capacitance of this DOPC monolayer is directly related to its structure and dielectric constant, as C¼
εε0 d
ð1Þ
where C is the specific capacitance, ε is the relative permittivity of the low dielectric region of the DOPC monolayer, ε0 is the dielectric constant of vacuum, and d is the dielectric spacing. All the electrochemical measurements were conducted in a Faraday cage. The KCl (0.1 mol dm3) electrolyte was prepared from Analar KCl (Fisher Chemicals Ltd.) calcined at 600 °C for 5 h and dissolved in 18.2 MΩ Milli-Q water with added 0.01 mol dm3 phosphate buffer (pH 6.5, pH 7.4, and pH 8.2) or citrate buffer (pH 5.5). A 3.5 mol dm3 Ag/AgCl reference electrode with a porous sintered glass frit and a platinum bar as counter electrode have been used throughout. All potentials in this paper are quoted versus the aforesaid reference electrode. ACV measurements were carried out on the MFE in the cell device. RCV and EIS measurements were performed on a HMDE in an argon deaerated system. DOPC monolayers on HMDE were prepared by spreading 13 μL of 2.54 mmol dm3 DOPC in pentane at the argon/ electrolyte interface, followed by slowly lowering the HMDE through the phospholipid on the electrolyte interface, as described before.17,22,23,27,28 Before, after, and between experiments, all compartments in the cell were cleaned in a hot solution of 98% H2SO4 and 30% H2O2 mixture in the volume ratio of approximately 1:3 and then thoroughly rinsed with 18 MΩ Milli-Q water. ACV was used (Autolab, Ecochemie, Utrecht, Netherlands) to obtain the capacitancepotential profiles by measuring the out-of-phase current (I00 ) at negatively going potentials from 0.4 to 1.6 V in response to an applied voltage wave of 75 Hz and an ac amplitude of ΔV = 0.005. Specific capacitance was calculated from17 C¼
I 00 ΔV ωA
ð2Þ
where ω is the angular frequency (ω = 2πf) and A is the area of the electrode. RCV measurements were carried out with a PowerLab 4/25 signal generator (AD Instruments Ltd.) which was controlled by Scope software (AD Instruments Ltd.). Potential scans between 0.2 and 1.5 V at the scan rate (ν) of 40 V s1 were performed on the HMDE to investigate the polymer/DOPC interaction. The DOPC monolayer recovery scans after 15 min interactions were conducted between 0.2 and 1.7 V at the same scan rate. RCV was used because it is a very effective way to obtain a snapshot or fingerprint of the state of the interface since it is almost instantaneous. For example, at the scan rate of 40 V s1, the potential scan from 0.2 to 1.5 V is completed in ∼0.03 s. For the Hg supported monolayer, the current from RCV is linearly related to scan rate in the absence of faradaic reactions so a higher scan rate is preferable for enhancing the sensitivity of measurement.20 However, at scan rates higher than 40 V s1, a gradual broadening of the current peaks representing the phospholipid monolayer phase transitions is observed and a decrease in current peak height at scan rates higher than 100 V s1 has been observed previously.13 As a result, 40 V s1 is chosen as the optimum scan rate to use maximizing sensitivity and minimizing resolution degradation. A further reason for using RCV is that when a slow scan is applied to a phospholipid layer, the layer is exposed to substantial electric fields of increasing magnitude. The electric field can induce irreversible changes in the phospholipid monolayer, especially in the region of the desorption potential of the monolayer.31 Rapid potential
scanning enables the layer not to be exposed to any field for a significant time period, avoiding irreversible changes taking place. In particular, phospholipid which desorbs from the electrode during the RCV cathodic sweep has no time to diffuse away from the interface and is fully recovered on the RCV anodic sweep. RCV at 40 V s1 has a small influence on the actual potential applied to the electrode in the context of the experiments carried out. The height of the capacitance peak of DOPC on Hg at ∼ 0.96 V is ∼80 μF cm2.22 The solution resistance of the electrochemical cells used is
