lable at ScienceDirect

Biomaterials 34 (2013) 9688e9699

Contents lists avai

Biomaterials

journal homepage: www.elsevier .com/locate/biomateria ls

Poly(ethylene glycol)-block-cationic polylactide nanocomplexesof differing charge density for gene delivery

Chih-Kuang Chen a,1, Charles H. Jones a,1, Panagiotis Mistriotis a, Yun Yu a, Xiaoni Ma a,Anitha Ravikrishnan a, Ming Jiang a, Stelios T. Andreadis a,b,c, Blaine A. Pfeifer a,**,Chong Cheng a,*

aDepartment of Chemical and Biological Engineering, University at Buffalo, The State University of New York, Amherst, NY 14260, USAbDepartment of Biomedical Engineering, University at Buffalo, The State University of New York, Amherst, NY 14260-4200, USAcCenter of Excellence in Bioinformatics and Life Sciences, Buffalo, NY 14203, USA

a r t i c l e i n f o

Article history:Received 17 July 2013Accepted 20 August 2013Available online 10 September 2013

Keywords:Cationic polymerGene deliveryNanocomplexPEGylationPolylactide

* Corresponding author. Department of ChemicalUniversity at Buffalo, The State University of New YUSA. Tel.: þ1 716 645 1193; fax: þ1 716 645 3822.** Corresponding author. Tel.: þ1 716 645 1198.

E-mail addresses: blainepf@buffalo.edu (B.A. P(C. Cheng).

1 These authors contributed equally.

0142-9612/$ e see front matter � 2013 Elsevier Ltd.http://dx.doi.org/10.1016/j.biomaterials.2013.08.063

a b s t r a c t

Representing a new type of biodegradable cationic block copolymer, well-defined poly(ethylene glycol)-block-cationic polylactides (PEG-b-CPLAs) with tertiary amine-based cationic groups were synthesized bythiol-ene functionalization of an allyl-functionalized diblock precursor. Subsequently the application ofPEG-b-CPLAs as biodegradable vectors for the delivery of plasmid DNAs (pDNAs) was investigated. Viathe formation of PEG-b-CPLA:pDNA nanocomplexes by spontaneous electrostatic interaction, pDNAsencoding luciferase or enhanced green fluorescent protein were successfully delivered to four physio-logically distinct cell lines (including macrophage, fibroblast, epithelial, and stem cell). Formulatednanocomplexes demonstrated high levels of transfection with low levels of cytotoxicity and hemolysiswhen compared to a positive control. Biophysical characterization of charge densities of nanocomplexesat various polymer:pDNA weight ratios revealed a positive correlation between surface charge and genedelivery. Nanocomplexes with high surface charge densities were utilized in an in vitro serum genedelivery inhibition assay, and effective gene delivery was observed despite high levels of serum. Overall,these results help to elucidate the influence of charge, size, and PEGylation of nanocomplexes upon thedelivery of nucleic acids in physiologically relevant conditions.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

With high delivery efficacy, viral vectors have been extensivelyused in gene therapy research over the previous decades [1]. Spe-cifically, viral-mediated carriers have been employed in more than70% of gene therapy clinical trials [2]. However, there are appre-ciable concerns about viral vectors regarding their biosafety, cyto-toxicity, immunogenicity, tumorigenicity, gene capacity, andtargeting efficiency [3,4]. These limitations have led to increasedinterest in the development of safe and efficacious non-viralvectors.

Accordingly, two major categories of non-viral vectors havebeen developed: cationic lipids (CLs) and cationic polymers (CPs).

and Biological Engineering,ork, Buffalo, NY 14260-4200,

feifer), ccheng8@buffalo.edu

All rights reserved.

Since the initial studies by Felgner and co-workers [5], CLs repre-sent the most studied systems for non-viral gene delivery [6]. Incontrast, CPs have gained increasing attention due to their inherentflexibility in design and formulation, which allows precise structureand surface modification for specific biomedical applications [7e10]. Both CLs and CPs can spontaneously form nanocomplexeswith negatively charged nucleic acids (RNA and DNA) throughelectrostatic interactions. Within the nanocomplexes, CLs and CPsserve as a protective barrier to prevent nucleic acids from degra-dation via ubiquitous nucleases. However, CL and CP-based nano-complexes differ significantly in endosomal/phagosomal escapemechanisms. CL-based systems operate on the basis of lipid mixing,which may involve membrane fusion and formation of transientand local perturbations, thus allowing the release of nucleic acidinto the cytosol. In contrast, CP-based systems most likely functionthrough a proton spongemechanism, inwhich amine groups of CPs(pKa 5w7) become protonated during endosome (or phagolysome)acidification due to the inflow of Hþ by the activity of H-ATPase andcounter ion Cl� influx to restore charge neutrality. It is believed thatwater intake occurs in endosomes to compensate for the increased

Fig. 1. Conceptual diagram of pDNA gene delivery using PEG-b-CPLA.

C.-K. Chen et al. / Biomaterials 34 (2013) 9688e9699 9689

ion inflow, causing osmotic swelling followed by endosomalrupture. However, reports have questioned the validity of theproton sponge effect [11,12], and recent findings by Rehman and co-workers indicate that cellular administration of CLs and CPs did notresult in complete endosomal rupture or the release of nano-complexes into the cytosol, suggesting the presence of alternativedelivery mechanisms [13].

Along with the development of polymer chemistry, the appli-cability of CPs in gene delivery has been enhanced by incorporatingbiodegradability within their macromolecular structures. A varietyof biodegradable CPs such as poly(b-amino ester)s [14e17], poly[a-(4-aminobutyl)-L-glycolic acid] [18], poly(4-hydroxy-L-prolineester) [19], poly(D-glucaramidoamine) [20], cationic poly(a-hy-droxy acid) [21], and cationic cyclodextrin [22] have been suc-cessfully synthesized and used in gene delivery studies. In additionto protecting therapeutic genes from nuclease degradation, syn-thetic design of CPs can be directed to optimize their biodegrad-ability and improve their biocompatibility for repeatedadministration of gene-based therapies [23,24]. In particular, theintracellular cleavage of the biodegradable polymer backbone aidsnanocomplexes in unpacking gene payloads, and effective cytosolicrelease of nucleic acids from nanocomplexes was reported to bepositively correlated with enhanced gene transfection [25,26].

Although biodegradable CPs may be used as safe gene deliveryscaffolds, most of the resulting nanocomplexes are unstable underphysiological conditions due to undesired interactions with serumproteins and salts, resulting in breakdown or aggregation and sub-sequent clearance [27]. To address this issue, certain studies havesuggested the use of anionic polymers [28,29]; however, mostpolymer delivery systems require excess cationic charge to remainstable. On the other hand, as observed in polyethyleneimine (PEI)-based systems, excess cationic charge of nanocomplexes can desta-bilize plasmamembranes of red blood cells and cause aggravated celldamage, resulting in necrosis, apoptosis, and autophagy [30]. Toameliorate such concerns, PEGylation has been used to stericallyshield CPs, thereby hindering andminimizing undesired interactionsbetween the cationic nanocomplex and serum albumin [31e33].Accordingly, colloidal stability and transfection efficiency of PEGy-lated nanocomplexes under physiological conditions can improve ascompared with their unPEGylated analogs. Recently, several studiesusing PEGylated CPs for enhanced gene transfection efficiency inserum have been reported. For example, Won and co-workerssynthesized poly(ethylene glycol)-poly(n-butyl acrylate)-poly(2-

(dimethylamino)ethyl methacrylate) (PEG-PnBA-PDMAEMAs) andformulated them into micelles for siRNA delivery [34]. The micelle/siRNA nanocomplexes, i.e. micelleplexes, demonstrated better genesilencing efficiency than PDMAEMA/siRNA nanocomplexes. More-over, improved accumulation of micelleplexes in tumor tissues wasobserved and credited to PEG shielding and size effects. However, thebackbone of PEG-PnBA-PDMAEMAs is unable to degrade underphysiological aqueous conditions, which may significantly hamperpotential clinical applications. Kwon and co-workers have preparedPEG-conjugated poly(ketalized serine) (PEG-poly(kSer)) for DNAcomplexation [35]. Possessing ketal-linkages, such nanocomplexescan effectively release DNA payloads into the cytoplasm throughacid-hydrolysis. PEG-poly(kSer)/DNA nanocomplexes exhibitedthree times higher transfection efficiency than poly-L-lysine (PLL) inNIH3T3 cells but were not as effective as typical commercial poly-meric vectors (such as 25 kDa branched PEI), presumably due to theirinsufficient buffering capacity. Similarly, poly(ethylene glycol)-block-poly(L-lysine) (PEG-b-PLL) has been synthesized by Jin and co-workers and utilized to form nanocomplexes with siRNA for genesilencing [36]. However, gene knockdown efficiency via such nano-complexes was lower than commercially available Lipofectamin�2000, presumably because of low dissociation and release efficiencyof siRNA from the nanocomplexes. Although PEGylation can beimportant for the rational design of CP vectors, it remains a challengeto successfully synthesize PEGylated degradable CPs that possesshigh complexation ability, desirable safety profile, and high trans-fection efficiency.

Recently, we synthesized biodegradable cationic polylactides(CPLAs) through ring-opening polymerization (ROP) and thiol-enefunctionalization, and further utilized them for delivery of siRNA toprostate cancer cells and pDNA to macrophage and fibroblast cells[37,38]. As comparedwith commercial gene delivery vectors (Fugene6), CPLAs provided enhanced transfection efficiencies due in part tohigh complexation ability and degradability at physiological condi-tions. However, despite possessing promising transfection and low-toxicity properties, preliminary studies demonstrated that CPLA-based nanocomplexes suffer from problems of low stability anddecreasing transfection efficiency with increased serum concentra-tions. Thus, to further enhance the potential clinical applicability ofCPLAs for gene delivery, PEGylation of CPLAs is required to limitnonspecific interactions at physiological conditions.

In this study, we report the synthesis of well-defined poly(-ethylene glycol)-block-cationic polylactides (PEG-b-CPLAs) with

C.-K. Chen et al. / Biomaterials 34 (2013) 9688e96999690

different charge densities and their applications as gene deliverycarriers for four physiologically distinct cell lines (Fig. 1). Effects ofcharge density and environmental pH on degradation rates of PEG-b-CPLAs were systematically studied. Subsequently, using PEG-b-CPLAs to form nanocomplexes with pDNA for delivery, genetransfection magnitude and efficiency were investigated. Relativeto positive controls, comparable or better transfection magnitudeand efficiency via PEG-b-CPLAs were achieved, while exhibiting notoxicity and minimal hemolysis. PEGylated nanocomplexes werefurther characterized by dynamic light scattering (DLS) and trans-mission electron microscopy (TEM) to measure biophysical prop-erties. In order to further demonstrate their potential applicabilitytowards clinical use, the influence of serum levels on transfectionefficiency was studied. Enhanced resistance to serum for PEGylatednanocomplexes when compared to unPEGylated versions andpositive controls was observed, highlighting the advantages andpotential of these carriers for nucleic acid delivery applications.

2. Materials and methods

2.1. Measurements

All 1H NMR spectra were measured at 500 MHz in CDCl3 using a Varian INOVA-500 spectrometer maintained at 25 �C with tetramethylsilane (TMS) as an internalreference standard. Gel permeation chromatography (GPC) data were acquired froma Viscotek GPC system equipped with a VE-3580 refractive index (RI) detector, a VE1122 pump, and two mixed-bed organic columns (PAS-103M and PAS-105M).Dimethylformamide (DMF; HPLC) containing 0.01 M LiBr was used as mobilephase with a flow rate of 0.5 mL/min at 55 �C. The GPC instrument was calibratedthrough narrowly-dispersed linear polystyrene standards purchased from Varian.Average hydrodynamic diameters (Dh) and zeta potential of nanocomplexes wereobtained using dynamic light scattering (DLS) on a Zetasizer nano-ZS90 (Malvern,Inc.) in water at 25 �C. All experiments were conducted using a 4 mW 633 nm HeeNe laser as the light source at a fixed measuring angle of 90� to the incident laserbeam. The correlation decay functions were analyzed by cumulants method coupledwith Mie theory to obtain volume distribution. Transmission electron microscopy(TEM) images were acquired using a JEOL 2010 microscope. TEM samples wereprepared by dip coating a 300 mesh carbon-coated copper grid with a dilute samplesolution, followed by staining using ruthenium tetroxide for 1 day. UV irradiationwas carried out using a UVGL-58 handheld UV lamp (UVP ultraviolet lamps; 6 W,0.12 A) with lmax of 365 nm. Dialysis procedures were conducted against acetoneusing molecularporous membrane tubing (Spectra/Por Dialysis Membrane, Spec-trum Laboratories Inc.) with an approximate molecular weight cut off (MWCO) at3500 Da.

2.2. Materials

4-Dimethylaminopyridine (DMAP; 99þ%) and L-lactide (L-LA, 98%) were pur-chased from SigmaeAldrich. 2,20-Dimethoxy-2-phenylacetophenone (DMPA; 98%)was purchased from Acros Organics. Dichloromethane (DCM; HPLC), acetone(HPLC), ethyl acetate (HPLC), hexane (HPLC), DMF (HPLC), and diethyl ether (HPLC)were purchased from Fisher Chemical. a-Methoxy-uehydroxyl PEG (mPEG-OH;MW: 2000 Da) was purchased from RAPP Polymere. 2-(Diethylamino)ethanethiolhydrochloride (DEAET, >98%) was purchased from Amfinecom Inc. DCM, DMF andethyl acetate were dried by distillation over CaH2. LA was recrystallized from dryethyl acetate four times prior to use. mPEG-OH was dried as follows prior to use:mPEG-OH was dissolved in 1 mL dried DCM, followed by complete solvent removal,and the cycle was repeated five times; toluenewas used as solvent to treat mPEG-OHfor another five cycles. Allyl-functionalized LA monomer 1 was prepared throughthe method we reported previously [37]. All other chemicals were used withoutfurther purification.

2.3. Synthesis of PEG-b-allyl-functionalized PLA (2)

In a 10 mL reaction flask with a magnetic stirring bar under nitrogen atmo-sphere, 1 (544 mg; 3.78 mmol), L-LA (643 mg; 3.78 mmol), and dried mPEG-OH(189 mg; 0.09 mmol) were added with dried DCM (4 mL). The solution was heat-ed to 35 �C using an oil bath for 1 h, followed by adding DMAP (44mg; 0.36mmol) in0.5 mL of dried DCM. After incubation for 1 week at 35 �C, the reactionwas stoppedat comonomer conversion of w80%, as determined by 1H NMR spectroscopy of analiquot of polymerization solution, based on the resonance intensities of the CH3

protons of remaining comonomers at 1.67e1.71 ppm relative to the CH3 protons ofthe resulting polymer at 1.49e1.59 ppm. The reaction mixture was precipitated bycold diethyl ether three times. Then the precipitate was collected and dried in avacuum to give 2 as a white solid powder in 30% isolated yield. 1H NMR (500 MHz,

CDCl3, ppm): d 1.49e1.59 (br m, CH3 of units from LA and 1), 2.67e2.73(br m,CH2CH]CH2 of units from 1), 3.38 (s, terminal CH3O of mPEG-OH), 3.54e3.68 (br m,CH2O of mPEG-OH), 5.14e5.30 (br m, CHCH3 of units from LA, CHCH3, CHCH2CH]CH2; and CH2CH]CH2 of units from 1), 5.77e5.79 (m, CH2CH]CH2 of units from 1).MNMR

n ¼ 5:5 kDa, MGPCn ¼ 14:0 kDa, PDIGPC ¼ 1.05. The mole fraction of 1 in PLA-

based block was 50% based upon the 1H NMR resonance intensities of 1H fromunits of 1 at 5.77e5.79 ppm relative to 4H from units of 1 and 2H from units of LA at5.14e5.30 ppm.

2.4. Synthesis of PEG-b-CPLAs

For the synthesis of PEG-b-CPLA-20, in a 10 mL flask, 2 (100 mg), DEAET(16.7 mg), and photoinitiator DMPA (10.15 mg) were dissolved in CDCl3 (5 mL) withmolar ratio of [allyl of 2]0:[SH of DEAET]0:[DMPA]0 ¼ 1:0.5:0.2. To remove oxygen inthe reactant solution, a freeze-pump-thaw procedure was carried out for three cy-cles. Subsequently, the thiol-ene reaction was induced by UV irradiation(lmax ¼ 365 nm) for 30 min. The reactant solution was dialyzed against acetone forfive days to completely remove unreacted DEAET and catalysts and then dried invacuum to give PEG-b-CPLA-20 with 87% yield. 1H NMR (500 MHz, CDCl3, ppm) ofPEG-b-CPLA-20 : d 1.28e1.36 (br m, (CH3CH2)2NHþCl� from amine-functionalizedunits), 1.51e1.63 (br m, CH3 of units from LA, 1 and amine-functionalized units),1.76e1.81 (br m, CH2CH2CH2SCH2 from amine-functionalized units), 2.00e2.04 (brm, CH2CH2CH2SCH2 from amine-functionalized units), 2.64e3.20 (br m, CH2CH]CH2 of units from 1 and CH2CH2CH2SCH2; SCH2CH2NHþCl�(CH2CH3)2 from amine-functionalized units), 3.40 (s, terminal CH3O of mPEG-OH), 3.53e3.68 (br m, CH2OofmPEG-OH), 5.13e5.30 (br m, CHCH3 of units from LA; CHCH3, CHCH2CH]CH2, andCH2CH]CH2 of units from 1; CHCH3 and CHCH2CH2CH2S from amine-functionalizedunits), 5.79e5.83 (br m, CH2CH]CH2 of units from 1). MNMR

n ¼ 6:2 kDa,MGPC

n ¼ 14:3 kDa, PDIGPC ¼ 1.06.PEG-b-CPLA-50 was prepared ([allyl of 2]0:[SH of DEAET]0:[DMPA]0 ¼ 1:3:0.4)

using the same method applied to PEG-b-CPLA-20. 1H NMR (500 MHz, CDCl3, ppm) ofPEG-b-CPLA-50: d 1.28e1.37 (br m, (CH3CH2)2NHþCl� from amine-functionalizedunits), 1.53e1.63 (br m, CH3 of units from LA, amine-functionalized units), 1.76e1.82(br m, CH2CH2CH2SCH2 from amine-functionalized units), 2.00e2.18 (br m,CH2CH2CH2SCH2 from amine-functionalized units), 2.60e2.72 (br m, CH2CH2CH2SCH2

from amine-functionalized units), 2.80e3.20 (SCH2CH2NHþCl�(CH2CH3)2 from amine-functionalized units), 3.40 (s, terminal CH3O of mPEG-OH), 3.53e3.68 (br m, CH2O ofmPEG-OH), 5.12e5.25 (br m, CHCH3 of units from LA, CHCH3 and CHCH2CH2CH2S fromamine-functionalized units). MNMR

n ¼ 7:3 kDa, MGPCn ¼ 14:0 kDa, PDIGPC ¼ 1.06.

2.5. Degradation studies

PEG-b-CPLA aqueous solutions with concentrations of 1 mg/mL were preparedfor degradation studies. RPMI-1640 (pH ¼ 7.4) or 25 mM sodium acetate buffer(pH ¼ 5.5) were used as the aqueous media. Solutions were sealed in vials andincubated at 37 �C with mild shaking. At predefined time intervals, aliquots of 1 mLof polymer solution were withdrawn and fully dried in vacuum. The resultingpolymer was dissolved in DMF for GPC analysis.

2.6. Cell lines and reporter plasmids

A RAW264.7 (murine macrophage) cell line was provided by Dr. Terry Connell(Department of Microbiology and Immunology, University at Buffalo, SUNY). The cellline was maintained in medium prepared as follows: 50 mL of fetal bovine serum(heat inactivated), 5 mL of 100 mM MEM sodium pyruvate, 5 mL of 1 M HEPES buffer,5 mL of penicillin/streptomycin solution, and 1.25 g of D(þ)-glucose added to 500 mLRPMI-1640 and filter sterilized. The 293T/17 (human embryonic kidney cells) cellline was obtained from ATCC (Manassas, VA) and cultured in Dulbecco’s ModifiedEagle Medium (DMEM; Gibco BRL, Grand Island, NY) supplemented with 10% (v/v)Fetal Bovine Serum (FBS; GIBCO) and 1% (v/v) Antibiotic-Antimycotic (A/A; Gibco);whereas, NIH-3T3 mouse fibroblasts (ATCC) were cultured in DMEM supplementedwith 10% (v/v) Bovine Serum (Gibco) and 1% (v/v) A/A. Human hair follicle derivedmesenchymal stem cells were isolated as described previously [39,40]. Briefly, skintissue from the scalp of a 45 year old male was obtained by the Cooperative HumanTissue Network (CHTN, Philadelphia, PA). For dissociation of the hair follicle fromsurrounding dermis, the sample was treated with 0.5% collagenase Type I (Invi-trogen, Carlsbad, CA) for 4 h at 37 �C. Next, single hair follicles were unplugged andplated on a 24-well plate with 200 mL of culture medium to promote attachment ofthe hair on the plate. A week later, hHF-MSCs started to migrate out of the hairfollicle. Subsequently, HF-MSCs were expanded and cultured in DMEM supple-mented with 10% MSC qualified FBS (GIBCO), 1 ng/mL bFGF, (BD Biosciences, SanJose, CA), and 1% (v/v) A/A. All cell lines were housed in T75 flasks and cultured at37 �C/5% CO2.

To determine the transfection efficacy of our DNA delivery system, we employedtwo reporter genes, luciferase and enhanced green fluorescent protein (EGFP), eachdriven by a cytomegalovirus promoter within plasmids pCMV-Luc (Elim Bio-pharmaceuticals) and pCMV-EGFP (Addgene). Plasmids were transformed into andprepared from an Escherichia coli cloning host (GeneHogs, Invitrogen) prior to beingused in the experiments outlined below.

Fig. 2. Synthesis of well-defined PEG-b-CPLA via ROP and thiol-ene functionalization.

C.-K. Chen et al. / Biomaterials 34 (2013) 9688e9699 9691

2.7. DNA gel-shift assay

pDNA (900 ng/sample) was mixed by gentle vortexing with PEG-b-CPLA-20 andPEG-b-CPLA-50 at various ratios in sterile water. After 15 min incubation at roomtemperature, mixtures were loaded into a 0.8% agarose gel containing SYBR� SafeDNA Gel Stain (Invitrogen) and electrophoresed at 80 V for 30 min. Gels werevisualized under UV after electrophoresis.

2.8. Preparation of Polymer:pDNA nanocomplexes

Nanocomplexeswere prepared by electrostatic interaction between PEG-b-CPLAand pCMV-Luc/pCMV-EGFP by mixing various weight ratios of polymer to pDNA.First, working dilutions of PEG-b-CPLA (PEG-b-CPLA-20, PEG-b-CPLA-50) wereprepared in water to a final concentration of 4 mg/mL. Depending on the desiredweight ratio, differing amounts of PEG-b-CPLA stock solution were mixed withpDNA in water, gently vortexed, and incubated 30 min to form the PEG-b-CPLA:pDNA complexes. Taking 512:1 PEG-b-CPLA-50 as an example, 512 mL of PEG-b-CPLA-50 stock solution was mixed with pCMV-Luc (final concentration of 4 mg/mL) to a final volume of 1 mL.

2.9. Transfection studies

For gene delivery experiments, RAW264.7, NIH3T3, 293T, and HFMSC cell lineswere seeded at densities of 30 � 104, 1.4 � 104, 2.9 � 104, and 2.9 � 103 cells/well,respectively, in two different types of 96-well plates (four plates in total) in 100 mLmedium/well and allowed 24 h for attachment. A tissue culture-treated, flat-bottom,sterile, white, polystyrene 96-well plate was used for luciferase assays; whereas,BCA assays were conducted in tissue culture-treated, sterile, polystyrene 96-wellplates. Each plate was carried out as a duplicate of the other.

Following the 24 h incubation, the medium was removed and replaced with70 mL growth medium plus 30 mL of PEG-b-CPLA:pDNA nanocomplexes (total vol-ume of 100 mL) and incubated for an additional 24 h. Fugene 6 (Promega) wasincluded as a positive control and complexed to pDNA (100 ng/well) according to themanufacturer’s instructions. After a second 24 h incubation (48 h after initialseeding), plates were analyzed for luciferase expression using the Bright Glo assay(Promega) and protein content using the Micro BCA Protein Assay Kit (Pierce) ac-cording to each manufacturer’s instructions. Gene delivery was calculated bynormalizing luciferase expression by protein content for each well/plate. Resultsderive from six replicates and two independent experiments.

To measure the transfection efficiency of cells seeded and transfected as describedabove, flow cytometry was performed. Two days after transfection, cells were washedwith PBS and detached from the surface using 0.25% Trypsin/1 mM EDTA. Trypsin wasinactivated using equal volume of DMEM supplementedwith 10% FBS. Next, cells werecentrifuged (250 g, 5min) and resuspended in PBS. The fraction of GFP expressing cellswas quantified using a FACS Calibur flow cytometer (Becton Dickinson, Franklin Lakes,NJ). For proper gating, cells transfected with water were used as a negative control;whereas, cells transfectedwith Fugene:pDNA complex were used as a positive control.

2.10. MTT assay for cell viability

Cytotoxicity of PEG-b-CPLA:pDNA nanocomplexes was determined by the MTT(3-(4,5-dimethylthiazol-2-yl)-diphenyl tetrazolium bromide) colorimetric assay.RAW264.7, NIH3T3, 293T, and HFMSC cells were seeded in tissue culture-treated,

sterile, polystyrene 96-well plates at densities indicated above. After incubationfor 24 h, mediumwas replaced with that containing various concentrations of PEG-b-CPLA:pDNA nanocomplexes (100 ng DNA/well). Following an additional 48 h in-cubation, cells were incubated with MTT solution (5 mg/mL), added at 10% v/v, for3 h at 37 �C/5% CO2. The medium containing MTT solution was then aspirated andformazan reaction products were dissolved in DMSO and shaken for 1 h. The opticaldensity of the formazan solution was quantified on a Synergy 4 Multi-ModeMicroplate Reader (BioTek Instruments, Inc.) at 570 nm with 630 nm serving asthe reference wavelength. Results are shown as a percentage of untreated cells with100% viability for three independent experiments.

2.11. Characterization of nanocomplexes

Using PEG-b-CPLA:pDNA nanocomplexes prepared as described above, zetapotential and effective diameter were measured by dynamic light scattering (DLS)on a Zetasizer nano-ZS90 (Malvern, Inc.). All data points result from measurementsof three independently formulated nanocomplexes measured five times. Addition-ally, nanocomplex samples were prepared by the drop-wise addition of 10 mLaqueous PEG-b-CPLA-50:pDNA (weight ratio 512:1) solution on a carbon-coatedcopper grid, followed by drying in a vacuum oven. Dried nanocomplexes weretreated with volatile ruthenium tetroxide vapor overnight prior to conductingtransmission electron microscopy (TEM, JEOL 2010 microscope).

2.12. Hemolytic activity assay

The hemolysis assay was modified from that previously described [41]. Briefly, a5% RBC/PBS solutionwas prepared by washing sheep blood (HemoStat Laboratories)with PBS until the supernatant became clear of red color. Next, 100 mL of purified 5%RBC solution was incubated with 900 mL of PEG-b-CPLA-20 and PEG-b-CPLA-50 inPBS at various concentrations for 1 h at 37 �C. Triton-X 100 (1% solution) was used toconstruct a % blood lysis standard curve by altering the amount of blood added toeach respective sample. For example, for 50 and 100% lysis, 50 and 100 mL of purified5% RBC solution was mixed with 1% Triton-X (to 1 mL). PBS was used to generatenegative controls. Samples were centrifuged and hemolysis quantified bymeasuringsupernatant at 541 nm and comparing to the % blood lysis standard curve. Dataweretaken from three independent experiments.

2.13. Serum inhibition to gene delivery

To determine if transfection was affected by the presence of increasing levels ofserum, Fugene 6, CPLA-50, and PEG-b-CPLA-50 nanocomplexes were prepared atoptimal conditions and vector:pDNA ratios as described above and before [38]. Vectorswere then incubatedwithRAW264.7cells (100ng/well) inRPMI-1640mediumwith10,20, 30, 40, 50, and 60% FBS for 24 h. Gene delivery was determined as described above.

3. Results and discussion

3.1. Synthesis and characterization of PEG-b-CPLAs

There is a pressing need for the creation of tunable well-definedpolymeric structures for controlled delivery and release of gene

Fig. 3. (a) GPC curves of mPEG-OH, precursor polymer 2 and PEG-b-CPLAs, and (b) 500 MHz 1H NMR spectra of PEG-b-CPLA-20 (top) and PEG-b-CPLA-50 (bottom) in CDCl3.

C.-K. Chen et al. / Biomaterials 34 (2013) 9688e96999692

therapeutics. By precise control of polymer structure, the influenceof molecular architecture, charge density, and pendant function-ality on gene transfection results can be elucidated, therebyincreasing screening efficacy in finding optimal polymer/genenanocomplex formulations [38]. In this context, a synthetic routefor the preparation of PEG-b-CPLAs was designed via living ROP andclick functionalization (Fig. 2). PEGwas covalently linked with allyl-functionalized PLA to form a diblock copolymer by using mPEG-OHas the initiator in the organocatalyzed ROP synthesis of the func-tional PLA block. Subsequently, allyl functionalities of the diblockcopolymer were converted into tertiary amine groups in acontrolled manner via thiol-ene chemistry, resulting in a series ofPEG-b-CPLAs with various charge densities. Thus, PEG-b-CPLAscould be synthesized from small molecules via a two-step reactionsequence under mild conditions without the use of metal catalysts.In addition to the hydrolytically degradable ester bonds of the CPLAblock, the introduction of tertiary amine cationic groups is expectedto aid endosomal escape via osmotic swelling and a possible“proton sponge” effect [42].

In this study, PEG-b-CPLA-20 and PEG-b-CPLA-50 (numberindicating amine mol% relative to the repeat units of CPLA block)were synthesized. ROP of allyl-functionalized LA monomer 1 withL-LA was conducted using mPEG-OH (MW ¼ 2.0 kDa; number-

average degree of polymerization (DPn) ¼ 45) as the initiator and4-dimethylaminopyridine (DMAP) as the organocatalyst ([1]0:[L-LA]0:[mPEG-OH]0:[DMAP]0 ¼ 40:40:1:4; 35 �C, 168 h, in DCM;w80% overall conversion of comonomers) [43]. Well-definedchemical structure of the resulting PEG45-b-poly(10.5-co-L-LA0.5)22(i.e., 2) was verified by 1H NMR and gel permeation chromatog-raphy (GPC) (Fig. 3a). Based on resonance intensities of allyl CHprotons (CH]CH2) from units of 1 at 5.77e5.79 ppm and the CHprotons from units of both comonomers and the allyl CH2 protons(CH]CH2) from units of 1 at 5.14e5.23 ppm, a 50% mole fraction ofunits of 1 in the PLA-based block of 2 was deduced. The DPn of 22for the PLA-based block of 2 was calculated based upon the com-parison of resonance intensities of CH protons from the units ofboth comonomers and allyl CH2 protons (CH ¼ CH2) from units of 1at 5.14e5.23 ppm with the resonance intensities of 180 protonsfrom mPEG-OH at 3.54e3.68 ppm. Accordingly, the number-average molecular weight (MNMR

n ) of 5.5 kDa was determined for2. GPC analysis further indicated that 2 had a Mn of 14.0 kDa and apolydispersity index (PDI) of 1.05, relative to linear polystyrenestandards. As shown in Fig. 3a, 2 exhibited a very narrow andmonomodal GPC peak with smaller elution volume (at 17.6 mL)than that of the initiatormPEG-OH (at 19.0mL), indicating thewell-controlled formation of the PLA-based block of 2 due to the living

Table 1Synthesis of PEG-b-CPLAs via thiol-ene functionalization of PEG45-b-poly(10.5-co-L-LA0.5)22, 2.a

CP [ene]0:[SH]0:[DMPA]0 Aminemol%b

MNMRn

(kDa)PDIGPCc

PEG-b-CPLA-20

1:0.5:0.2 20 6.2 1.06

PEG-b-CPLA-50

1:3:0.4 50 7.3 1.06

a Reaction conditions: UV irritation for 30 min in CDCl3; 2: MNMRn ¼ 5.5 kDa,

PDIGPC ¼ 1.05.b Determined by 1H NMR spectroscopy and relative to repeat units of the CPLA

block.c Relative to linear polystyrenes.

Fig. 4. Degradation profiles at 37 �C: (a) PEG-b-CPLA-20 in RPMI-1640 (pH ¼ 7.4), (b)PEG-b-CPLA-50 in RPMI-1640 (pH ¼ 7.4), (c) PEG-b-CPLA-20 in sodium acetate buffer(pH ¼ 5.5), (d) PEG-b-CPLA-50 in sodium acetate buffer (pH ¼ 5.5).

C.-K. Chen et al. / Biomaterials 34 (2013) 9688e9699 9693

ROP process. The significant difference between the MNMRn and

MGPCn of 2 can be ascribed to the different MW-hydrodynamic

volume relationship for 2 and polystyrene. It should be notedthat PLA-based oligomer was also formed along with 2 in the ROPprocess, presumably due to trace amounts of hydroxyl-containingimpurities in the reaction system. However, the oligomer wasreadily removed through work-up procedure.

To partly or fully convert units of 1 into tertiary amine-basedcationic LA (CLA) units, UV-induced thiol-ene functionalization ofthe allyl-functionalized PLA block of 2was conducted, using DEAETas the thiol-containing functionalization agent and DMPA as thephotoinitiator in CDCl3 under UV irradiation (l ¼ 365 nm) for30 min at ambient temperature. PEG-b-CPLA-20 and PEG-b-CPLA-50 (i.e., PEG45-b-poly(10.3-co-CLA0.2-co-L-LA0.5)22 and PEG45-b-pol-y(CLA0.5-co-L-LA0.5)22) were successfully obtained by varying the[ene]0:[SH]0:[DMPA]0 feed ratios (Table 1). Because of the overlapof the characteristic resonances of protons from CLA units withother resonances, the mole fraction of CLA units in the CPLA blockof PEG-b-CPLA-20 was deduced from the remaining mole fractionof units of 1, which was obtained by comparing the resonance in-tensities of allyl CH protons (CH]CH2) from the monomer unit of 1(at 5.83 ppm) to the methyl CH2CH3 protons of CLA units (at 1.28e1.38 ppm) (Fig. 3b). For PEG-b-CPLA-50, the complete disappear-ance of resonances of allyl CH protons (CH]CH2) from units of 1 (at5.83 ppm) indicated that all the allyl functionalities of 2 wereconverted into tertiary amine groups. Based on the mole fraction ofamine-functionalized units in CPLA blocks, MNMR

n values of PEG-b-CPLA-20 and PEG-b-CPLA-50 were determined to be 6.2 and7.3 kDa, respectively. GPC analysis showed that their elution peakpositions at 17.5e17.6 mL and PDI values of 1.06 were very close tothose of precursor 2 (Fig. 3a), indicating that the side group func-tionalization did not considerably affect hydrodynamic volume ofthe polymer and that side reactions in click functionalization weresuccessful suppressed [37]. The high degree of molecular homo-geneity exhibited by PEG-b-CPLAs is beneficial for screeningoptimal cationic polymer:gene nanocomplex formulations becauseconcerns of artifacts attributed to broad molecular weight distri-butions can be excluded.

3.2. Degradation of PEG-b-CPLAs

The unpackaging of genes from nanocomplexes requiresthoughtful design considerations because the release of gene pay-loads from carrier in cytoplasm is a necessary step for effective genedelivery [44]. A myriad of strategies have been developed to pro-mote the cytoplasmic unpacking of genetic cargo from CP-basednanocomplexes, including decreasing MW and reducing chargedensity of CPs [45e47]. Among these options, decreasing MWof CPvia polymer degradation can be considered as the most usefulapproach to address this issue because degraded polymer species

can more readily unload gene payloads in addition to limitingcytotoxic interactions with the targeted cells [48]. However,enhanced degradability can sacrifice nanocomplex stability andmay lead to the unwanted occurrence of gene release in intercel-lular space, thereby resulting in decreased transfection efficiency.Thus, identifying factors that affect polymer degradation rates iscritical for tailoring an ideal polymeric gene delivery scaffold.

GPC was used as an analytic tool to investigate hydrolytic de-gradability of PEG-b-CPLAs. Degradation of the polymer backbonecan be probed via the increases in elution volumes of degradedspecies as compared with the initial polymers. Degradation ex-periments were conducted by incubating 1 mg/mL of PEG-b-CPLAsin RPMI-1640 (pH ¼ 7.4) and 25 mM sodium acetate buffer(pH¼ 5.5) at 37 �C, followed bywithdrawing aliquots at 1, 3, 9, 24 h,and 1 week for GPC assessment. According to GPC curves (Fig. 4),PEG-b-CPLA-20 exhibited significant degradation after 24 h in

Fig. 6. Agarose gel electrophoresis of polymer:pDNA nanocomplexes. Lane assign-ments are as follows: (1)e(7) PEG-b-CPLA-20:pCMV-Luc nanocomplexes at poly-mer:pDNA ratios of (1) 8:1, (2) 16:1, (3) 32:1, (4) 64:1, (5) 128:1, (6) 256:1, and (7)512:1; (8) molecular weight marker; (9)e(15) PEG-b-CPLA-50:pCMV-Luc nano-complexes at polymer:DNA ratios of (9) 8:1, (10) 16:1, (11) 32:1, (12) 64:1, (13) 128:1,(14) 256:1, and (15) 512:1.

C.-K. Chen et al. / Biomaterials 34 (2013) 9688e96999694

RPMI-1640 at 37 �C, with an obvious intensity reduction of theinitial elution peak at 17.5 mL associated with the appearance of amajor elution peak attributed to PEG-2000 at 18.7 mL and a minorelution peak at 20.7 mL resulting from oligomeric residuals of CPLAblocks. Moreover, the initial elution peak disappeared after 1 weekwith only detached PEG-2000 chains and oligomeric residue beingobserved after this time, indicating that the CPLA block was thor-oughly cleaved into oligomeric residuals via hydrolytic degradation.PEG-b-CPLA-50 demonstrated faster degradation than PEG-b-CPLA-20, as indicated by the appearance of the PEG-2000 elutionpeak with significant intensity at 18.7 mL after 3 h as well as thecomplete disappearance of the initial elution peak after 24 h. Theincubationmediawere further changed from RPMI-1640 (pH¼ 7.4)to 25 mM sodium acetate buffer (pH ¼ 5.5), in order to study pHeffects on degradation behavior of PEG-b-CPLAs. For PEG-b-CPLA-20, the initial elution peak at 17.5 mL remained essentially un-changed after 24 h in sodium acetate buffer. In the new medium,PEG-b-CPLA-50 also demonstrated a reduction in degradation,although the breakdown rate was still faster than PEG-b-CPLA-20.As a result, acidic conditions led to a decreased degradation rate forboth PEG-b-CPLAs when compared to neutral conditions. Overall,the CPLA blocks of PEG-b-CPLAs were readily degraded undernormal physiological pH conditions. In contrast, the PEG block wasnot degraded, but completely detached from PEG-b-CPLAs. Theobserved degradation rate of CPLA blocks in PEG-b-CPLAs wassignificantly enhanced by increasing their amine mol%; however, itwas decreased by changing medium pH from 7.4 to 5.5.

The degradation rate of the CPLA blocks of PEG-b-CPLAs isinfluenced by polymer amine concentration and environmental pHcondition. Because a tertiary amine is generally considered toosterically hindered to act as nucleophile, it is expected that thetertiary amine groups in PEG-b-CPLA catalyze the hydrolysis ofester groups on the polymer main chains through a general basecatalysis mechanism (Fig. 5), in which a neutral tertiary aminemoiety partially abstracts a proton from a nucleophilic watermolecule in the transition state [49]. In support of this mechanism,degradation rate increased with a respective increase in amine mol%. Moreover, because under acidic conditions protonationedepro-tonation equilibrium favors the formation of a protonated aminewhich cannot function as a general base, reduced degradation rateat pH 5.5 relative to pH 7.4 was observed. As such, the aminecontent of PEG-b-CPLAs is the key determinant in pH-dependentdegradation mechanism.

3.3. Nanocomplex formation and gene delivery

To determine the efficacy of gene delivery of the newly syn-thesized PEG-b-CPLA variants, nanocomplexes of PEG-b-CPLAswith luciferase- and enhanced green fluorescent protein (EGFP)-

Fig. 5. Proposed degradation m

encoding pDNAs were formed via electrostatic interaction andsubsequently utilized to transfect four physiologically distinct celllines in serum-containingmedium. Following 24 h incubation, cellstransfected with luciferase-encoding pDNA were measured forluciferase activity and total protein levels using Bright Glo andmicro BCA assay kits coupled with quantification via 96-well platereader. Luciferase gene delivery was calculated by dividing lucif-erase luminescence by total protein content of wells and data werenormalized relative to gene delivery results obtained with thecommercially-available Fugene 6 vector. In contrast, following 24 hincubation, cells transfected with EGFP-encoding pDNA wereevaluated using flow cytometry to quantify the percentage ofpositively expressing cells. EGFP data was not normalized and waspresented as raw data of representative experiments.

To form nanocomplexes, polymer:pDNA weight ratios wereevaluated by using an agarose gel retardation assay to determinethe minimal dosage of each respective PEG-b-CPLA to completelycomplex pCMV-Luc pDNA (Fig. 6). PEG-b-CPLA-20 moderatelyretarded pCMV-Luc progression at polymer:pDNA of 512:1;whereas, PEG-b-CPLA-50 interacted completely with pDNA atpolymer:pDNA of 256:1. PEG-b-CPLA-20 could only partially bindpDNA and required an excessive weight ratio of 1024:1 (data notshown) to completely retard pDNA progression. In contrast, thehigher charge density PEG-b-CPLA-50 was able to retard a majorityof pDNA progression at a weight ratio of 64:1. For each polymer, thehighest two polymer:pDNA weight ratios (256:1, 512:1) wereselected and screened for gene delivery.

echanism of PEG-b-CPLA.

C.-K. Chen et al. / Biomaterials 34 (2013) 9688e9699 9695

To further understand gene delivery properties of PEG-b-CPLAs,four different cell line types (macrophage, fibroblast, epithelial,stem cell) were selected on the basis of being physiologicallydistinct from one another. In addition, in order to observe the effectof increasing charge density and weight ratios, PEG-b-CPLA-20 and-50 were used in duplicate experiments at two separate poly-mer:pDNA weight ratios (as identified above). Gene delivery ofpCMV-Luc using PEG-b-CPLAs (Fig. 7aeb, top frames), revealed apositive correlation of gene delivery with both charge density andweight ratios. PEG-b-CPLA-20 performed comparable to Fugene 6at higher weight ratios. In contrast, PEG-b-CPLA-50 performedcomparable to Fugene 6 at lower weight ratios and better at higherweight ratios. The greatest improvement in luciferase expression,w2.5 times the expression levels produced with Fugene 6, wasobserved with PEG-b-CPLA-50 nanocomplexes and hair folliclederived mesenchymal stem cells (hHF-MSC). Interestingly, despitethe significant improvement when compared to a well-establishedcommercial vector, the raw expression levels for this cell line weresignificantly lower (data not shown) than that of any other cell linetested in this study, highlighting the challenges associatedwith thishistorically difficult to transfect cell lineage. Similarly, EGFP genedelivery into the same four cell lines was evaluated by flowcytometry (Fig. 7aeb, bottom frames). As before, gene delivery ef-ficiency was positively correlated with charge density and weightratio, and the results were comparable (in most cases) to thoseproduced with Fugene 6. Generally, % positive cells were lowerin macrophage and stem cell lines, as compared to epithelialand fibroblast lines. Additionally, despite doubling of polymer

Fig. 7. Gene delivery studies of (a) PEG-b-CPLA-20 and (b) PEG-b-CPLA-50 at various polymeluminescence units normalized by total protein and reported relative to gene delivery valdeviation values resulting from three independent experiments. EGFP gene delivery (bottomreported by percentage of positive cells in 10,000 cell samples. *Statistically significant (95

concentration and maintenance of pDNA concentration, a highercharge density resulted in minimal changes in % positivelyexpressing cells, suggesting a potential limit to which uptake andsubsequent endosomal/phagosomal release are influenced.

In our previous studies of delivering pDNA via CPLAs [38], weconcluded that macrophages possessed superior gene delivery ca-pabilities because of an additional uptake mechanism (phagocy-tosis), resulting in increased luciferase expression as compared toNIH3T3. Our non-normalized luciferase data (not shown)confirmed our previous observations, while presenting new ques-tions pertaining to gene expression magnitudes versus expressionprevalence. Recent observations by Rehman and coworkers indi-cate that gene expression is influenced by uptake mechanisms[13,50,51]. Non-phagocytic cells uptake and encapsulate a smallnumber of nanocomplexes into low-pH compartments via endo-cytosis, resulting in few efficient gene release events (1e3) per cell[13]. Phagocytosis is believed to operate independently of pureclathrin-mediated and clathrin-independent endocytosis; howev-er, following internalization, phagosomes readily fuse to early andlate endosomes [52]. Thus, we propose the additional uptakemechanism of macrophages is the cause of the discrepancies be-tween our flow cytometry and luciferase experiments. Since releaseof genetic payloads and not uptake is the proposed limiting step innanocomplex gene delivery, an additional uptake mechanismwould work synergistically to enrich late endosomal/phagosomalcompartments with pDNA. The following nanocomplex-inducedcompartmental escape would be governed by universal cell-specific endocytic release kinetics; however, release following

r:pDNAweight ratios in different cell lines. Luciferase gene delivery (top) is measured inues obtained with Fugene 6. Error bars of luciferase experiments represent standard) data are presented as representative experiments measured by flow cytometry and

% confidence) when compared to respective Fugene 6 control.

C.-K. Chen et al. / Biomaterials 34 (2013) 9688e96999696

enhanced uptake (in the case of phagocytic cells) would allowhigher concentrations of pDNA to enter the cytosol and subse-quently the nucleus. In contrast, non-phagocytic cells demon-strated similar % positive cells and non-normalized gene deliverytrends, suggesting that all cells were governed by similar trends ofendocytosis and subsequent release and expression. In conclusion,distinctive difference between respective lines can be observedwith analysis focused on general gene delivery efficacy (i.e., non-normalized luciferase and % EGFP expression) in the order293T > RAW264.7 > NIH3T3 > hHF-MSC.

3.4. Cytotoxicity and hemolysis

For eventual applicability, a polymer must possess a safe cyto-toxicity and hemolysis profile. Thus, PEG-b-CPLAswere examined fortheir cytotoxicity in all cell types using an MTT assay (Fig. 8). Theviability after treatment using both PEG-b-CPLAs at three concentra-tions (1, 2, and3mg/mL) resulted innoobvious toxicity. Conversely, inourprevious studies, CPLA-50demonstrated low tomoderate toxicitytoward RAW264.7 and NIH3T3 cell lines (w80% or greater viability).This supports previous claims that PEGylation further diminishesexcess charge and results in decreased toxicity [47].

To further evaluate biocompatibility, hemolysis assays aretypically performed to assess a polymer’s potential interactionwiththe cell membrane of red blood cells (RBCs). It is proposed thatexcessive charges of CPs can interact with the negatively chargedmembranes of RBCs to form nanosized pores in cell membranes,resulting in an influx of small solutes and leading to osmotic lysis[53,54]. Hemolysis is related to the release of hemoglobin fromRBCs and measurement by a plate reader of pelleted RBC super-natant after sample treatment. For analysis, a standard curve ofhemolysis from 1 to 100% was constructed using Triton X-100 inconjugation with a phosphate-buffered-saline (PBS) negative con-trol. As before, nanocomplexes obtained by mixing PEG-b-CPLAs of1, 2, 3 mg/mL with 100 ng/sample pDNA, as well as Fugene 6 pre-pared to match respective concentrations, were incubated withRBCs and subsequently evaluated for their hemolytic activity(Fig. 9). Upon PEGylation of CPLAs, hemolysis was significantlyreduced regardless of charge density. PEG-b-CPLA-20 hemolysisresulted in no significant hemolysis when compared to the PBScontrol. PEG-b-CPLA-50, however, resulted in low hemolysis butonly at the highest dosages tested (3 mg/mL is w6 times themaximum gene delivery dosage used). Thus, PEG-b-CPLAs offerenhanced gene delivery with minimal cytotoxicity and hemolysis.

Fig. 8. MTT assay: (a) PEG-b-CPLA-20 and (b) PEG-b-CPLA-50 incubated at three dosages fofrom three independent experiments.

3.5. Biophysical characterization

Efficient gene delivery by cationic nanocomplexes is stronglyinfluenced by net surface charge and effective particle diameter. Toinvestigate such biophysical properties, nanocomplexes were pre-pared as described above at polymer:pDNA weight ratios wherepDNA was at least partially retarded by PEG-b-CPLAs and thenmeasured by DLS. Zeta potential (Fig. 10) values of both PEG-b-CPLA-20 and -50 possessed net cationic values with a positive trendin relation to increasing polymer:pDNA weight ratios. At 512:1polymer:pDNA weight ratio (the ratio associated with the mosteffective gene delivery), nanocomplexes of PEG-b-CPLA-20 andPEG-b-CPLA-50 possessed zeta potential values of 21.9 � 1.4 mVand 28.3 � 1.4 mV, respectively. Previous studies have proposedthat net cationic surface charges can prevent nanocomplex aggre-gation in addition to being required for efficient transfection by thespontaneous formation of small, positively charged complexes[55e57]. However, arbitrarily increasing charge density can resultin undesired cytotoxicity by continued destabilization of mito-chondrial, nuclear, and cell membranes [58,59]. Thus, CPLA-basedvariants, especially those altered by PEGylation, offer capabilitiesto finely balance surface charge, gene delivery, and cytotoxicity.

To be considered for clinical studies, cationic polymers need theability to self-assemble with pDNA (without use of harmful sol-vents) into particles that are small enough to be taken up byreceptor-mediated endocytosis and pinocytosis by non-phagocyticcells. The size limit for these types of uptake mechanisms variesfrom cell to cell, but is generally on the order of 200 nm for receptor-mediated endocytosis and greater than 500 nm for pinocytosis [60].Additionally, phagocytic cells (macrophages, dendritic cells, neu-trophils) can accommodate and process particles of substantiallylarger sizes of 2e10 mm [61]. To this end, we evaluated PEG-b-CPLA-based nanocomplexes for effective particle diameter measurementsby DLS (Fig. 10b). A positive correlation is observed with particleeffective diameter and both increasing polymer:pDNA weight ratioand charge density. At 64:1 and 256:1 ratios, PEG-b-CPLA-20-basednanocomplexes are equal to or slightly larger in effective diameterthan PEG-b-CPLA-50-based nanocomplexes but fail to reach this sizeat other ratios. Thus, observed diameters are generally smaller thanpreviously reported values for non-PEGylated CPLA-based nano-complexes [38], and can be explained by PEGylation of CPLAslimiting aggregations due to shielding effects. The reduction in sizecan also explainwhy non-phagocytic cell lines demonstrated similartransfection results as RAW264.7 despite lacking additional uptake

r 24 h with various cell types. Error bars represent standard deviation values resulting

Fig. 9. Hemolytic activity of CPLA and PEG-b-CPLA:pDNA nanocomplexes. RBCs were incubated with various concentrations of nanocomplexes. Results are expressed as percentageof total RBC lysis as determined by the Triton-X 100 control. Error bars represent standard deviation values resulting from three independent experiments. *Statistically significant(95% confidence) when compared to respective Fugene 6 control. #Statistically significant (95% confidence) when compared to respective non-PEGylated sample.

C.-K. Chen et al. / Biomaterials 34 (2013) 9688e9699 9697

mechanisms. Drastic distinction of gene delivery between chargedensities for RAW264.7 also confirms previous findings thatincreased nanocomplex charge heightens gene delivery to nega-tively charged macrophages [61e63].

Fig. 10. Biophysical characterization of PEG-b-CPLA:pDNA nanocomplexes: (a) zeta potentpolymer:pDNA weight ratios, and (c) TEM image of PEG-b-CPLA-50:pDNA nanocomplexes

Transmission electron microscopy was further utilized to verifythe size of the PEG-b-CPLA nanocomplexes. The samplewith chargedensity and weight ratio corresponding to the most effective genedelivery was selected for observation (PEG-b-CPLA-50:pDNA

ial measurements, (b) effective diameter measurements of nanocomplexes at variousat the 512:1 weight ratio.

Fig. 11. Serum inhibition study of gene delivery of CPLA-50 and PEG-b-CPLA-50 nanocomplexes compared to Fugene 6. Error bars represent standard deviation values resultingfrom three independent experiments. Statistically significant (95% confidence) when compared to respective Fugene 6 control.

C.-K. Chen et al. / Biomaterials 34 (2013) 9688e96999698

weight ratio of 512:1). The diameter was observed to be approxi-mately 200 nm (Fig. 10c), which is consistent with the DLS data.

3.6. In vitro serum inhibition to gene delivery

For PEG-b-CPLA-based nanocomplexes to be relevant for in vivoapplications, gene delivery must remain effective in the presence ofheightened levels of extracellular protein. Generally, negativelycharged serum proteins bind positively charged complexes,resulting in aggregation and clearance and ultimately hinderinggene delivery. As such, Fugene 6, CPLA-50, and PEG-b-CPLA-50nanocomplexes were prepared as described above and before [38],and incubated with RAW264.7 cells with varying percentages offetal bovine serum (FBS). Fig. 11 shows that Fugene 6 and CPLA-50gene delivery is negatively correlated with increasing levels of FBS.A large decrease in gene delivery occurs with each moderate in-crease (10%) of serum. However, the rate of decrease is significantlyreduced when using PEG-b-CPLA-50. It is generally accepted thatphysiological serum levels range from 45 to 60% of volume [64].Under such condition, PEG-b-CPLA-50 is able to provide compara-ble gene delivery to the Fugene 6 positive control at 10% FBS con-ditions. At the highest percentage of FBS, PEG-b-CPLA-50 performsslightly below the level of the Fugene 6 control measured at 10%FBS and 3e5 times better than the positive control and un-PEGylated counterparts at the same conditions. This result sug-gests that PEG-b-CPLA-based nanocomplexes when compared withnon-PEGylated CPLA offer structural characteristics, such as refinedpolymer-pDNA chargeecharge interactions and PEG-shielding,which can withstand drastic reductions of gene delivery thatwould normally accompany increased levels of serum.

4. Conclusions

PEGylation has beneficially improved a new class of well-definedcationic polylactides with tunable charge densities. We chemicallyand biophysically characterized PEG-b-CPLA variants and investi-gated transfection magnitude and efficiency using four physiologi-cally distinct cell lines. Results indicate that this class of PEG-b-CPLApolymers exhibits impressive gene delivery capabilities when

compared to an effective commercially-available control and resistsserum inhibition of gene delivery while possessing high degrad-ability, low cytotoxicity, and minimal hemolysis.

Acknowledgment

The authors thank the NSF grants (CBET-1019227 and DMR-1133737) and a SUNY-Buffalo Schomburg fellowship (CHJ) forfinancial support.

References

[1] Giacca M, Zacchigna S. Virus-mediated gene delivery for human gene therapy.J Control Release 2012;161:377e88.

[2] Vannucci L, Lai M, Chiuppesi F, Ceccherini-Nelli L, Pistello M. Viral vectors: a lookback and ahead on gene transfer technology. New Microbiol 2013;36:1e22.

[3] Thomas CE, Ehrhardt A, Kay MA. Progress and problems with the use of viralvectors for gene therapy. Nat Rev Genet 2003;4:346e58.

[4] Bouard D, Alazard-Dany D, Cosset FL. Viral vectors: from virology to transgeneexpression. Br J Pharmacol 2009;157:153e65.

[5] Felgner PL, Gadek TR, HolmM, Roman R, Chan HW, Wenz M, et al. Lipofection:a highly efficient, lipid-mediated DNA-transfection procedure. Proc Natl AcadSci U S A 1987;84:7413e7.

[6] Chen Y, Bathula SR, Yang Q, Huang L. Targeted nanoparticles deliver siRNA tomelanoma. J Invest Dermatol 2010;130:2790e8.

[7] Zheng M, Pavan GM, Neeb M, Schaper AK, Danani A, Klebe G, et al. Targetingthe blind spot of polycationic nanocarrier-based siRNA delivery. ACS Nano2012;6:9447e54.

[8] Patil ML, Zhang M, Minko T. Multifunctional triblock nanocarrier (PAMAM-PEG-PLL) for the efficient intracellular siRNA delivery and gene silencing. ACSNano 2011;5:1877e87.

[9] Samal SK, Dash M, Van Vlierberghe S, Kaplan DL, Chiellini E, van Blitterswijk C,et al. Cationic polymers and their therapeutic potential. Chem Soc Rev2012;41:7147e94.

[10] Mintzer MA, Simanek EE. Nonviral vectors for gene delivery. Chem Rev2009;109:259e302.

[11] Benjaminsen RV, Mattebjerg MA, Henriksen JR, Moghimi SM, Andresen TL.The possible “proton sponge” effect of polyethylenimine (PEI) does notinclude change in lysosomal pH. Mol Ther 2013;21:149e57.

[12] Funhoff AM, van Nostrum CF, Koning GA, Schuurmans-Nieuwenbroek NM,Crommelin DJ, Hennink WE. Endosomal escape of polymeric gene deliverycomplexes is not always enhanced by polymers buffering at low pH. Bio-macromolecules 2004;5:32e9.

[13] Rehman ZU, Hoekstra D, Zuhorn IS. Mechanism of polyplex- and lipoplex-mediated delivery of nucleic acids: real-time visualization of transientmembrane destabilization without endosomal lysis. ACS Nano 2013;7:3767e77.

C.-K. Chen et al. / Biomaterials 34 (2013) 9688e9699 9699

[14] Lynn DM, Langer R. Degradable poly(b-amino esters): synthesis, character-ization, and self-assembly with plasmid DNA. J Am Chem Soc 2000;122:10761e8.

[15] Anderson DG, Lynn DM, Langer R. Semi-automated synthesis and screening ofa large library of degradable cationic polymers for gene delivery. Angew ChemInt Ed 2003;42:3153e8.

[16] Zugates GT, Anderson DG, Little SR, Lawhorn IEB, Langer R. Synthesis ofpoly(b-amino ester)s with thiol-reactive side chains for DNA delivery. J AmChem Soc 2006;128:12726e34.

[17] Tzeng SY, Green JJ. Subtle changes to polymer structure and degradationmechanism enable highly effective nanoparticles for siRNA and DNA deliveryto human brain cancer. Adv Healthcare Mater 2013;2:468e80.

[18] Lim Y-b, Choi YH, Park J-s. A self-destroying polycationic polymer: biode-gradable poly(4-hydroxy-L-proline ester). J Am Chem Soc 1999;121:5633e9.

[19] Putnam D, Langer R. Poly(4-hydroxy-L-proline ester): low-temperaturepolycondensation and plasmid DNA complexation. Macromolecules1999;32:3658e62.

[20] Liu Y, Wenning L, Lynch M, Reineke TM. New poly(D-glucaramidoamine)sinduce DNA nanoparticle formation and efficient gene delivery intomammalian cells. J Am Chem Soc 2004;126:7422e3.

[21] Zhang Z, Yin L, Xu Y, Tong R, Lu Y, Ren J, et al. Facile functionalization ofpolyesters through thiol-yne chemistry for the design of degradable, cell-penetrating and gene delivery dual-functional agents. Biomacromolecules2012;13:3456e62.

[22] Li J, Yang C, Li H, Wang X, Goh SH, Ding JL, et al. Cationic supramoleculescomposed of multiple oligoethylenimine-grafted b-cyclodextrins threaded ona polymer chain for efficient gene delivery. Adv Mater 2006;18:2969e74.

[23] Merkel OM, Kissel T. Nonviral pulmonary delivery of siRNA. Acc Chem Res2012;45:961e70.

[24] Kulkarni A, DeFrees K, Schuldt RA, Hyun S-H, Wright KJ, Yerneni CK, et al.Cationic a-cyclodextrin:poly(ethylene glycol) polyrotaxanes for siRNA de-livery. Mol Pharm 2013;10:1299e305.

[25] Liu Y, Reineke TM. Degradation of poly(glycoamidoamine) DNA delivery ve-hicles: polyamide hydrolysis at physiological conditions promotes DNArelease. Biomacromolecules 2010;11:316e25.

[26] Luo X-h, Liu C-W, Li Z-y, Qin S-y, Feng J, Zhang X-z, et al. OEI800 poly-conjugates linked with ketalized glycolic acid for use as gene vectors. J MaterChem 2011;21:15305e15.

[27] Tros de Ilarduya C, Sun Y, Duzgunes N. Gene delivery by lipoplexes and pol-yplexes. Eur J Pharm Sci 2010;40:159e70.

[28] Nicol F, Wong M, MacLaughlin FC, Perrard J, Wilson E, Nordstrom JL, et al.Poly-L-glutamate, an anionic polymer, enhances transgene expression forplasmids delivered by intramuscular injection with in vivo electroporation.Gene Ther 2002;9:1351e8.

[29] Schlegel A, Largeau C, Bigey P, Bessodes M, Lebozec K, Scherman D, et al.Anionic polymers for decreased toxicity and enhanced in vivo deliveryof siRNA complexed with cationic liposomes. J Control Release 2011;152:393e401.

[30] Gao X, Yao L, Song Q, Zhu L, Xia Z, Xia H, et al. The association of autophagywith polyethylenimine-induced cytotoxicity in nephritic and hepatic celllines. Biomaterials 2011;32:8613e25.

[31] Glodde M, Sirsi SR, Lutz GJ. Physiochemical properties of low and high molec-ular weight poly(ethylene glycol)-grafted poly(ethylene imine) copolymers andtheir complexes with oligonucleotides. Biomacromolecules 2005;7:347e56.

[32] Yang J, Wang H-Y, Yi W-J, Gong Y-H, Zhou X, Zhuo R-X, et al. PEGylatedpeptide based reductive polycations as efficient nonviral gene vectors. AdvHealthcare Mater 2013;2:481e9.

[33] Yu Y, Chen C-K, Law W-C, Mok J, Zou J, Prasad PN, et al. Well-defineddegradable brush polymer-drug conjugates for sustained delivery of Pacli-taxel. Mol Pharm 2013;10:867e74.

[34] Gary DJ, Lee H, Sharma R, Lee JS, Kim Y, Cui ZY, et al. Influence of nano-carrierarchitecture on in vitro siRNA delivery performance and in vivo bio-distribution: polyplexes vs micelleplexes. ACS Nano 2011;5:3493e505.

[35] Shim MS, Kwon YJ. Acid-transforming polypeptide micelles for targetednonviral gene delivery. Biomaterials 2010;31:3404e13.

[36] Qi RG, Liu S, Chen J, Xiao HH, Yan LS, Huang YB, et al. Biodegradable co-polymers with identical cationic segments and their performance in siRNAdelivery. J Control Release 2012;159:251e60.

[37] Chen C-K, Law W-C, Aalinkeel R, Nair B, Kopwitthaya A, Mahajan SD, et al.Well-defined degradable cationic polylactide as nanocarrier for the delivery ofsiRNA to silence angiogenesis in prostate cancer. Adv Healthcare Mater2012;1:751e61.

[38] Jones CH, Chen C-K, Jiang M, Fang L, Cheng C, Pfeifer BA. Synthesis of cationicpolylactides with tunable charge densities as nanocarriers for effective genedelivery. Mol Pharm 2013;10:1138e45.

[39] Bajpai VK, Mistriotis P, Andreadis ST. Clonal multipotency and effect of long-term in vitro expansion on differentiation potential of human hair folliclederived mesenchymal stem cells. Stem Cell Res 2012;8:74e84.

[40] Liu JY, Peng HF, Gopinath S, Tian J, Andreadis ST. Derivation of functionalsmooth muscle cells from multipotent human hair follicle mesenchymal stemcells. Tissue Eng Part A 2010;16:2553e64.

[41] Higgins DE, Shastri N, Portnoy DA. Delivery of protein to the cytosol ofmacrophages using Escherichia coli K-12. Mol Microbiol 1999;31:1631e41.

[42] Y-b Lim, Kim S-M, Lee Y, Lee W-k, Yang T-g, Lee M-j, et al. Cationic hyper-branched poly(amino ester): a novel class of DNA condensing molecule withcationic surface, biodegradable three-dimensional structure, and tertiaryamine groups in the Interior. J Am Chem Soc 2001;123:2460e1.

[43] Zou J, Hew CC, Themistou E, Li Y, Chen CK, Alexandridis P, et al. Clicking well-defined biodegradable nanoparticles and nanocapsules by UV-induced thiol-ene cross-linking in transparent miniemulsions. Adv Mater 2011;23:4274e7.

[44] Jeong JH, Kim SW, Park TG. Molecular design of functional polymers for genetherapy. Prog Polym Sci 2007;32:1239e74.

[45] Jeong JH, Christensen LV, Yockman JW, Zhong ZY, Engbersen JFJ, Kim WJ, et al.Reducible poly(amido ethylenimine) directed to enhance RNA interference.Biomaterials 2007;28:1912e7.

[46] Erbacher P, Roche AC, Monsigny M, Midoux P. The reduction of the positivecharges of polylysine by partial gluconoylation increases the transfection ef-ficiency of polylysine/DNA complexes. Biochim Biophys Acta Biomembr1997;1324:27e36.

[47] Sharma R, Lee JS, Bettencourt RC, Xiao C, Konieczny SF, Won YY. Effects of theincorporation of a hydrophobic middle block into a PEG-polycation diblockcopolymer on the physicochemical and cell interaction properties of thepolymer-DNA complexes. Biomacromolecules 2008;9:3294e307.

[48] Kunath K, von Harpe A, Fischer D, Peterson H, Bickel U, Voigt K, et al. Low-molecular-weight polyethylenimine as a non-viral vector for DNA delivery:comparison of physicochemical properties, transfection efficiency and in vivodistribution with high-molecular-weight polyethylenimine. J Control Release2003;89:113e25.

[49] Fife TH, Singh R, Bembi R. Intramolecular general base catalyzed ester hy-drolysis. The hydrolysis of 2-aminobenzoate esters. J Org Chem 2002;67:3179e83.

[50] Rehman Zu, Sjollema KA, Kuipers J, Hoekstra D, Zuhorn IS. Nonviral genedelivery vectors use syndecan-dependent transport mechanisms in filopodiato reach the cell surface. ACS Nano 2012;6:7521e32.

[51] Rehman Zu, Hoekstra D, Zuhorn IS. Protein kinase A inhibition modulates theintracellular routing of gene delivery vehicles in HeLa cells, leading to pro-ductive transfection. J Control Release 2011;156:76e84.

[52] El-Sayed A, Harashima H. Endocytosis of gene delivery vectors: from clathrin-dependent to lipid raft-mediated endocytosis. Mol Ther 2013;21:1118e30.

[53] Sovadinova I, Palermo EF, Huang R, Thoma LM, Kuroda K. Mechanism ofpolymer-induced hemolysis: nanosized pore formation and osmotic lysis.Biomacromolecules 2011;12:260e8.

[54] Fischer D, Li Y, Ahlemeyer B, Krieglstein J, Kissel T. In vitro cytotoxicity testingof polycations: influence of polymer structure on cell viability and hemolysis.Biomaterials 2003;24:1121e31.

[55] Pochard I, Boisvert J-P, Persello J, Foissy A. Surface charge, effective chargeand dispersion/aggregation properties of nanoparticles. Polym Int 2003;52:619e24.

[56] Putnam D. Polymers for gene delivery across length scales. Nat Mater 2006;5:439e51.

[57] Shmueli RB, Anderson DG, Green JJ. Electrostatic surface modifications toimprove gene delivery. Expert Opin Drug Delivery 2010;7:535e50.

[58] Grandinetti G, Smith AE, Reineke TM. Membrane and nuclear permeabiliza-tion by polymeric pDNA vehicles: efficient method for gene delivery ormechanism of cytotoxicity? Mol Pharm 2012;9:523e38.

[59] Lv H, Zhang S, Wang B, Cui S, Yan J. Toxicity of cationic lipids and cationicpolymers in gene delivery. J Control Release 2006;114:100e9.

[60] Rice-Ficht AC, Arenas-Gamboa AM, Kahl-McDonagh MM, Ficht TA. Polymericparticles in vaccine delivery. Curr Opin Microbiol 2010;13:106e12.

[61] Doshi N, Mitragotri S. Macrophages recognize size and shape of their targets.PLoS One 2010;5:e10051.

[62] Thiele L, Merkle HP, Walter E. Phagocytosis and phagosomal fate of surface-modified microparticles in dendritic cells and macrophages. Pharm Res2003;20:221e8.

[63] Gomez JMM, Csaba N, Fischer S, Sichelstiel A, Kundig TM, Gander B, et al.Surface coating of PLGA microparticles with protamine enhances theirimmunological performance through facilitated phagocytosis. J ControlRelease 2008;130:161e7.

[64] He XM, Carter DC. Atomic structure and chemistry of human serum albumin.Nature 1992;358:209e15.