Rare-earth-incorporated polymeric vector for enhanced gene delivery

Qiwen Wang a,1, Weihong Jin b,1, Guosong Wub, Ying Zhao b, Xue Jin a,b, Xiurong Hu a,Jun Zhou a, Guping Tang a,b,**, Paul K. Chu b,*

a Institute of Chemical Biology and Pharmaceutical Chemistry, Zhejiang University, Hangzhou 310028, PR ChinabDepartment of Physics & Materials Science, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong, PR China

a r t i c l e i n f o

Article history:Received 16 July 2013Accepted 20 September 2013Available online 5 October 2013

Keywords:PEI-CyDNeodymiumEnergetic ionsPlasma treatmentCellular pathways

a b s t r a c t

Cationic polymer PEI-CyD is doped with Nd by plasma technology to produce the gene vector: Nd@PEI-CyD. Luciferase expression and EGFP transfection experiments performed in vitro reveal that Nd@PEI-CyD has significantly higher transfection efficiency than lipofectamine 2000 and PEI-CyD and themechanism is studied and proposed. The rare-earth element, Nd, stimulates the energy metabolism ofcells, enhances cell uptake of complexes/pDNA, and regulates the cellular pathways. These special fea-tures suggest a new strategy involving metal-incorporated non-viral gene vectors.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

In recent years, two kinds of gene delivery systems for genetherapies: viral and non-viral [1] have been widely reported [2,3].Viral gene vectors have high efficiency, but serious safety risks suchas immunogenicity, carcinogenicity and inflammation limit theirclinical implementation [4]. In comparison, non-viral gene vectorswith several advantages over viral ones such as low immunoge-nicity and toxicity, large DNA loading capacity, and tissue-specifictargeting are typically less effective [2]. Therefore, it is critical todevelop more efficient non-viral vectors for clinical applications[2,5].

Among the various types of non-viral vectors, polyethylenimine(PEI), especially high-molecular weight (HMW) PEI (～25 kDa), is awidely used cationic polymer known as the gold standard [6].Although HMW PEI (～25 kDa) has high gene transfection effi-ciency, the high toxicity remains a drawback for application in vivo[7]. Our previous studies have shown that b-cyclodextrin-poly-ethylenimine(PEI-CyD), in which b-CyD is crosslinked by low

molecular weight PEI (600Da), has lower cytotoxicity than PEI25 kDa and similar transfection efficiency as PEI 25 kDa [8]. Ourpreliminary studies suggest that the gene transfection efficiency ofPEI-CyD can be improved by surface modification of the polymer[9,10].

Metallic elements play vital roles in biological systems and ac-tivities, for example, bones and teeth [11], functional componentsof proteins [12], activators of enzymes [13], essential parts in theelectron transport of the respiratory chain [14], and maintenance ofnormal functions in cell membranes [15]. In particular, rare earthelements (REEs) can change the mitochondrial metabolic activity[16,17], promote DNA synthesis [18], regulate the activity ofcalmodulin [19,20], and participate in a variety of physiological andbiochemical processes [21]. Although REEs have been used topromote the growth and development of plants and animals[22,23], their use in gene vectors have seldom been explored.

In this work, the benefits of metallic elements in biologicalprocesses such as DNA transcription, mRNA translation [24e26]and surface modification of polymeric materials by ion beamsand plasmas [27e29] are combined. The cationic polymer PEI-CyDis doped with a rare earth element, neodymium (Nd), to produceNd@PEI-CyD complexes. To determine the effects of the plasmatreatments, the chemical, physical, and biological characteristics ofthe Nd@PEI-CyD are investigated. The transfection efficiency ofNd@PEI-CyD is compared to that of PEI-CyD and the commercialtransfection agent Lipofectamine 2000. To illustrate the underlyingmechanism (Fig. 1), the ATP assay is performed to evaluate the

* Corresponding author. Tel.: þ86 852 34427724; fax: þ86 852 34420542.** Corresponding author. Institute of Chemical Biology and PharmaceuticalChemistry, Zhejiang University, Hangzhou 310028, PR China. Tel./fax: þ86 57188273284.

E-mail addresses: tangguping@zju.edu.cn (G. Tang), paul.chu@cityu.edu.hk(P.K. Chu).

1 Both authors contributed equally to this work.

Contents lists available at ScienceDirect

Biomaterials

journal homepage: www.elsevier .com/locate/biomater ia ls

0142-9612/$ e see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.biomaterials.2013.09.069

Biomaterials 35 (2014) 479e488

energy metabolism of cells and the reverse transcription-polymerase chain reaction (RT-PCR) and western blot analysis areconducted to assess several mRNA and proteins.

2. Materials and methods

2.1. Materials

Branched polyethylenimine (PEI; average 25 000 Da), linear polyethylenimine(average MW 600), b-cyclodextrin (b-CyD, MW 1135), 1,10-carbonyldiimidazole(CDI, MW 162.15), dimethyl sulfoxide (DMSO, �99.5%), N,N-dimethylformanide(DMF, �99%), [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide](MTT, MW 218.1), phosphate buffered saline (PBS), and triethylamine (Et3N, 99%)were obtained from Sigma (St. Louis, MO, USA). DMF and DMSO were dried by

refluxing over CaH2 and distilled before use. Calcium hydride (CaH2) was purchasedfrom Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China).

The plasmid DNA (pDNA) encoding firefly luciferase (pGL3-Luc) or green fluo-rescent protein (pEGFP) were purchased from Promega Corporation (Madison, WI,USA) and Guangzhou Jetway Biotech CO., Ltd. (Guangzhou, Guangdong, China). TheBCATM protein assay kit and luciferase activity assay kit were purchased from PierceBiotechnology, Inc. (Rockford, IL) and Promega Corporation (Madison, WI, USA),respectively. The ATP assay kit was purchased from Nanjing Jiancheng Bioengi-neering Institute (Nanjing, Jiangsu, China).

2.2. Production of the Nd@PEI-CyD

The PEI-CyD was synthesized as described previously [8]. The PEI-CyD filmswere prepared by dropping the solution containing 10 mg of PEI-CyD dissolved in200 mL of deionized water onto a 1 cm � 1 cm silicon chip and dried in air overnight.

Fig. 1. Illustration of the synthesis of Nd@PEI-CyD and process of gene delivery mediated by Nd@PEI-CyD.

Table 1Primers used in RT-PCR of selected gene transcripts.

Gene Primer sequences Annealing temperature (�C)

calmodulin Sense primer: 50-TGAGATAGGGTTCCTGGTTTG-30

Antisense primer: 50-GAGGGTGTAGGGTTTCTGGTT-3056

caveolin Sense primer: 50-TAGGATGCTCCCTTGTCGC-30

Antisense primer: 50-TGCTTCTCGCTCAGCTCGT-3056

metallothionein Sense primer: 50-CTCAACTTCTTGCTTGGGATC-30

Antisense primer: 50-AATGGGTCAGGGTTGTATGG-3056

GAPDH Sense primer: 50-AAGGTCGGAGTCAACGGATTT-30

Antisense primer: 50-AGATGATGACCCTTTTGGCTC-3056

Q. Wang et al. / Biomaterials 35 (2014) 479e488480

Fig. 2. (A) Tomoscan of Nd@PEI-CyD films; (B) XPS spectra acquired from Nd@PEI-CyD at different depths; (C) XPS spectrum of O 1s at different depths; (D) Theoretical fitting of O1s at 50 nm; (E) XPS spectrum of Nd 3d at different depths; (F) Theoretical fitting of Nd 3d at 50 nm.

Q. Wang et al. / Biomaterials 35 (2014) 479e488 481

Nd plasma immersion ion implantation (Nd-PIII) was conducted on a metal plasmaion implanter equipped with a neodymium cathodic arc source for 10 min at apulsed voltage of �20 kV.

2.3. Polymer characterization

The structure of the PEI-CyD was characterized by 1H nuclear magnet-icresonance (1H NMR) spectra. The 1H NMR spectra were recorded on a Bruker AV-400NMR spectrometer at 400 MHz and room temperature.

The Nd@PEI-CyD samples were analyzed by attenuated total-reflection Fouriertransformed infrared spectroscopy (ATR-FTIR, Varian, ExcaliburTM, USA). Eachsample was scanned 16 times in the spectral region of 4000e600 cm�1 with aresolution of 4 cm�1 and averaged to produce each spectrum. The surface chemical

structure of Nd@PEI-CyD was determined by X-ray photoelectron spectroscopy(XPS, Physical electronics PHI 5802). The overlapping peaks were resolved into in-dividual components by the peak analysis software, XPSPEAK version 4.1 adoptingthe LorentzianeGaussian sum function [30]. The tomoscan of Nd@PEI-CyD wasexamined by field-emission scanning electron microscopy (FE-SEM, JEOL JSM-7400F, Japan).

2.4. Preparation and characterization of Nd@PEI-CyD/pDNA

The Nd@PEI-CyD/pDNA complexes were prepared by mixing Nd@PEI-CyD withpDNA at different N/P ratios. Briefly, the Nd@PEI-CyD solutions were added to thepDNA solutions with the same volume, mixed, and incubated for 30 min at roomtemperature to form the Nd@PEI-CyD/pDNA complexes. The images of the Nd@PEI-

Fig. 3. ATR-FTIR spectra of PEI-CyD, Nd-PEI-CyD, and Nd@PEI-CyD.

Fig. 4. (A) Particle size and zeta potential of Nd@PEI-CyD/pDNA complexes (mean � SD, n ¼ 3) for various N/P ratios and (B) TEM image of Nd@PEI-CyD/DNA complexes for an N/Pratio of 30.

Q. Wang et al. / Biomaterials 35 (2014) 479e488482

CyD/pDNA complexes in the 293T cell lines were acquired by field-emission scan-ning electron microscopy (FE-SEM, JEOL JSM-7400F, Japan). The energy-dispersiveX-ray spectroscopy (EDS) spectra were taken on the SEM to determine the chemi-cal composition of the complexes in the 293T cell lines.

The particle size and zeta potentials of the complexes were measured at 25 �C bydynamic light scattering (DLS) on the Zetasizer 3000 (Malvern Instruments, Wor-cestershire, UK). Complex solutions (20 mL) containing 1 mg of pDNA were preparedat various N/P ratios. The average size (nm) and zeta potential (mV) were used in thedata analysis. The morphology of the complexes was examined by transmissionelectron microscope ( TEM, Philips CM20, Holland).

2.5. Cell culture

The 293T cells, HEK-293 cells, and 3T3 cells were cultured in Dulbecco’s modi-fied eagle medium (DMEM) containing 10% FCS. The MCF-7 cells were cultured inDMEM containing 10% FCS and the SGC-7901 cells were cultured in RPMI-1640containing 10% FCS. The cells were grown under humidified air containing 5% CO2

at 37 �C.

2.6. In vitro transfection efficiency assay

The 293Tcells were seeded at a density of 2�104 cells/well on a 48-well plate in400 mL of DMEM medium containing 10% FCS and incubated for 18 h before trans-fection. The medium was replaced by 400 mL of fresh serum-free DMEM containingcomplexes (of Nd@PEI-CyD with pEGFP or Nd@PEI-CyD with pGL3-Luc) with 1 mgpDNA at different N/P ratios. After incubation for 6 h at 37 �C, the transfectionmediumwas replaced with a fresh growth medium and the cells were incubated foradditional 36 h.

In the green fluorescent protein assay, the cells expressing the green fluorescentprotein were observed and the pictures were captured by fluorescence microscopy(Nikon Eclipse TE200, Nikon Corporation, Tokyo, Japan) at 36 h transfection. TheEGFP fluorescence was excited at 488 nm and emissionwas collected using a 515 nmfilter. The EGFP transfection efficiency was also determined by flow cytometry. Aftertransfection, the cells were trypsinized, washed, and re-suspended in PBS. The FACSanalysis of the EGFP-expressing cells was performed by flow cytometry (BD LSR,USA). The untransfected cells were used to establish the background. The Cell Quest3.3 software (BD Biosciences, San Jose, CA) was used for data acquisition and dataanalysis was performed with the FCS Express V3 (De Novo Software, Thornhill,Canada).

The luciferase activity was measured according to standard protocols of theluciferase assay system (Promega). The measurements were conducted in a single-tube luminometer (Berthold lumat LB9507, Germany) for 10 s. The relative lightunits (RLU) were normalized against protein concentration in the cell extracts,which was measured using a BCA protein assay kit (Pierce, Rockford, USA). Theluciferase expression efficiency was reported in terms of RLU/mg cellular protein.

2.7. MTT cytotoxicity assay

The cytotoxicity of the complexes was evaluated using theMTTassay using 293T,3T3, and MCF-7 cells. Generally, the cells were seeded on 96-well plates at 1 � 104

cells/well in 200 mL of DMEM medium containing 10% FCS and allowed to grow for18 h. The medium was replaced by 200 mL of serum-free culture media containingserial dilutions of Nd@PEI-CyD/pDNA solutions at a series of N/P ratios for 4 h. Afterreplacing the mediumwith 90 mL of fresh medium and 10 mL of MTT solution (5 mg/mL in PBS buffer), the cells were incubated for another 4 h. Finally, each well wasreplaced with 100 mL of DMSO and measured spectrophotometrically on an ELISAplate reader (Model 550, Bio-Rad) at a wavelength 570 nm. The relative cell growth(%) related to control cells cultured in themedia without the polymer was calculatedby the following relationship: ([A]test�[A]blank)/([A]control�[A]blank) � 100.

2.8. ATP assay

In the ATP assay, the cells were transfected as described in section 2.6. After celldisruption, each well was treated with the ATP Assay Kit. The samples were incu-bated for 5 min at room temperature. Each sample was analyzed by UV-Vis spec-trophotometry at a wavelength 636 nm. The concentration of ATP was calculated by([A]test�[A] control)/([A]standard�[A]blank) � dilution factor � 1 mmol/L. The concen-tration of ATP was normalized against the protein concentration in the cell extracts.It was measured using a BCA protein assay kit (Pierce, Rockford, USA). The con-centration of ATP was reported in terms of mmol/g cellular protein.

2.9. In vitro analysis of calmodulin, caveolin, and metallothionein expressions

The HEK-293 cells and SGC-7901 cells were seeded on 6-well culture plates at adensity of 5.0 � 105 cells per well and incubated at 37 �C in 5% CO2 for 18 h to reach70% confluence. The medium was replaced by 3 mL of fresh serum-free DMEM orRPMI-1640 containing different complexes (PEI-CyD/pDNA or Nd@PEI-CyD/pDNA)with 4 mg of pDNA with different N/P ratios, PBS, or Nd@PEI-CyD, respectively. Af-ter incubation for 6 h at 37 �C, the mediumwas replaced by a fresh growth medium

and the cells were incubated for additional 48 h. The cellular levels of calmodulin,caveolin, and metallothionein mRNA as well as protein were assessed using theRT-PCR and Western blot analysis.

In the Western blot analysis, the cells were transfected as described above andthe cell proteins were extracted after transfection for 48 h. The total protein wasquantified by the BCA protein assay kit (Pierce, Rockford, USA). An equal amount ofprotein was separated on the SDS-PAGE, transferred onto the nitro-cellulosemembrane, and blocked and incubated overnight with monoclonal antibodiesagainst calmodulin (1:400), caveolin (1:400), and metallothionein (1:400). Afterwashing, the membrane was incubated with the HRP-conjugated secondary anti-body (1:5000) for 2 h at room temperature. The bands were visualized using theWestzol enhanced chemiluminescence kit (Intron, Sungnam, Korea) and theexpression was normalized to the housekeeping gene (b-actin) expression.

In the RT-PCR analysis, mRNA was extracted from a total of 1 � 106 cells 48 hpost-transfection and then reverse-transcribed into cDNA using the Qiagen RNEasyMini Kit (Cat. No.74104). Quantification of the PCR product was performed in the 1�TE buffer using absorption values at 260 nm and 280 nm (c(RNA)¼ OD260 � dilutionfactor � 0.04 mg/mL). The RNA was reverse-transcribed with OligodT (Fermentas)according to the protocol. The relative gene expression values were determinedusing the BandScan4.3 software and the data representing the calmodulin, caveolin,and metallothionein expression were normalized to the housekeeping geneglycer-aldehyde-3-phosphate dehydrogenase (GAPDH) as the endogenous reference. Threeselected genes and housekeeping gene (GAPDH) used in this study are listed inTable 1. The PCR parameters consisted of 5min of Taq activation at 95 �C, followed by32 cycles of 95 �C 30 s, 56 �C 45 s, and 72 �C 45 s.

2.10. Statistical analysis

All the experiments were repeated at least three times and the data are pre-sented as means � standard deviation. The statistical significance (p < 0.01) wasevaluated by the Student t-test when only two groups were compared. If more thantwo groupswere compared, evaluation of significancewas performed using the one-way analysis of variance (ANOVA) followed by Bonferroni’s post hoc test. In all thetests, the statistical significance was set at p < 0.01.

3. Results and discussion

Fig. S1 depicts 1H NMR spectrum of PEI-CyD in D2O. Fig. 2Ashows the tomoscan of the Nd@PEI-CyD samples. The modifiedsurface darkens after the plasma treatment. The X-ray

Fig. 5. SEM micrographs of 293T cells incubated with Nd@PEI-CyD/pDNA complexesand corresponding EDS spectra.

Q. Wang et al. / Biomaterials 35 (2014) 479e488 483

Fig. 6. Fluorescence microscopy images of 293T cells incubated with (A) PEI-CyD/pEGFP complexes, (B) Lipofectamine/pEGFP complexes 2000, (C) Nd@PEI-CyD/pEGFP complexes,and (D) Nd-PEI-CyD/pEGFP complexes; Quantified expressions of EGFP by flow cytometry: (E) PEI-CyD/pEGFP, (F) Lipofectamine/pEGFP, and (G) Nd@PEI-CyD/pEGFP, and (H) Nd-PEI-CyD/pEGFP, (I) In vitro gene transfection - Luciferase expression levels in 293T cells transfected with PEI-CyD/pDNA, Lipofectamine/pDNA, Nd@PEI-CyD/pDNA, and Nd-PEI-CyD/pDNA.

Q. Wang et al. / Biomaterials 35 (2014) 479e488484

photoelectron spectroscopy (XPS) spectra of the Nd@PEI-CyD inFig. 2B with different colors representing various depths show theNd3d and O1s peaks. The O1s peak remains at about 532 eV at adepth of 100 nm (Fig. 2C), indicating the existence of CeOeH [31]and that the OeH bonds of PEI-CyD are unreacted. The O1s peakin the near surface shifts to 530 eV as a result of the NdeO bond[32] disclosing the formation of Nd@PEI-CyD and complete reac-tion of CeOeH. The O1s peak at 50 nm in Fig. 2D is Gaussian fitted tolocate the CeOeH and NdeO peaks. By integrating the peaks, 86%of OeH is found to be converted into NdeO. The Nd3d peaks at adepth of 50 nm (Fig. 2E and F) consist of two subpeaks at 1003 eV(3d3/2) and 979 eV (3d5/2) indicating the formation of NdeO bond(Nd3þ) [33,34]. According to Gaussian fitting, the satellite peaks onthe lower binding energy side of the 3d5/2 and 3d3/2 peaks areseparated by about 5 eV. The broad peaks may not arise fromdifferent chemical states but be attributed to the special nature ofrare earth atoms [35].

The ATR-FTIR spectra in Fig. 3 further confirm the formation ofthe NdeO bond. Compared to PEI-CyD, Nd@PEI-CyD exhibits asharp peak at 654 cm�1 which is the characteristic peak of NdeO. Inaddition, the characteristic peak of OeH at 3291 cm�1 diminishessubstantially and the other peaks remain at about the same posi-tion as in PEI-CyD, confirming that the plasma treatment does notdamage the polymer skeleton. However, after the Nd@PEI-CyD isdissolved and freeze-dried, the characteristic peak of NdeO at654 cm�1 almost shift back to 629 cm�1, which is part of theoriginal PEI-CyD, as shown in Fig. 3. The red (inweb version) profileand characteristic OeH peak at 3291 cm�1 reappear showing agingwhich is common to polymeric materials. The aged materials aretermed Nd-PEI-CyD to distinguish them from the fresh sampledesignated as Nd@PEI-CyD.

Cationic polymers bind with DNA and form nanoparticles viaelectrostatic interactions. The particle size and zeta potential of theNd@PEI-CyD/pDNA complexes with various N/P ratios are deter-mined by dynamic light scattering and the results are displayed inFig. 4A. The PEI-CyD/pDNA and Nd@PEI-CyD/pDNA complexes withthe same N/P ratios have similar particle size and zeta potentials.The particle size and zeta potential of the Nd@PEI-CyD/pDNAcomplexes with an N/P ratio of 30 are about 150 nm and 28 mV,respectively, which are suitable for cell uptake [36]. Thetransmission electron microscope (TEM) picture of the Nd@PEI-CyD/pDNA complexes with an N/P ratio of 30 in Fig. 4B showsthat the nanoparticles are spherical with a size of about 100 nm. Inaddition to determining the chemical compositions of the 293Tcells incubated with the Nd@PEI-CyD/pDNA complexes, energy-dispersive X-ray spectroscopy (EDS) spectra are also acquired. Asshown in Fig. 5, the existence of Nd in the 293T cells is confirmed.

In vitro transfection is performed using the 293T cells to inves-tigate the transfection ability of Nd@PEI-CyD/pDNA at various N/Pratios. The PEI-CyD without plasma treatment (optimal N/P of 25)and commercial transfection agent Lipofectamine 2000 serve as thecontrols. As shown in Fig S2, the Nd@PEI-CyD/pDNA complexesexhibit the highest transfection at an N/P ratio of 35 in the 293Tcells. The gene expressions of the complexes decreasewhen the N/Pratio is larger than 35 possibly due to the increased cytotoxicity orstrong interaction between polycation and DNA preventing therelease of pDNA [37]. By using the optimal N/P ratio in the 293Tcells, Nd@PEI-CyD shows 27-folds and 15-folds higher luciferaseactivity than PEI-CyD and Lipofectamine 2000, as shown in Fig. 6I.In order to visualize the transfection, the gene expression of theenhanced green fluorescent protein (EGFP) is examined by fluo-rescence microscopy. Fig. 6AeC show, respectively, the images ofthe EGFP-positive 293T cells treated with the PEI-CyD/pEGFPcomplexes, Lipofectamine 2000/pEGFP complexes, and Nd@PEI-CyD/pEGFP complexes with the respective optimal N/P ratios. To

quantify the amounts of transfected cells, the EGFP expression inthe 293T cells is analyzed by flow cytometry (FCM) (Fig. 6DeF). Atthe respective optimal N/P ratios, the percentage of the EGFP-positive 293T cells mediated by Nd@PEI-CyD is up to58.55 � 4.55% which is dramatically higher than 12.04 � 2.51% forPEI-CyD without plasma treatment and 24.25 � 3.61% for Lip-ofectamine 2000.

The transfection efficiency of Nd-PEI-CyD, which is the polymerformed after Nd@PEI-CyD is dissolved and freeze-dried, is aboutthe same as that of PEI-CyD, as shown in Fig. 6D and I. The resultsreveal that freeze drying degrades the NdeO bond (as seen in FTIR)and this affects the transfection efficiency. Ideal non-viral genevectors should have both high transfection efficiency and lowcytotoxicity [7]. Fig. 7 shows the in vitro MTT assay results of thecytotoxicity of the Nd@PEI-CyD/DNA complexes with various N/Pratios in the 293T, 3T3, and MCF-7 cells. The N/P ratio has anobvious impact on the cytotoxicity of Nd@PEI-CyD/DNA complexesin the three types of cells. The cell viability decreases withincreasing N/P ratios. At the optimal N/P ratio of 35 for the in vitrotransfection assay, the cell viability of the Nd@PEI-CyD/pDNA

10/1 15/1 25/1 30/1 35/1 40/1

0

20

40

60

80

100

120

Ce

ll A

via

bility

(%

)

N/P ratio

293T

3T3

MCF-7

Fig. 7. Cell viability in 293T, 3T3, and MCF-7 cells incubated with Nd@PEI-CyD/DNAcomplexes at various N/P ratios (mean � SD, n ¼ 3).

Fig. 8. ATP assay in HEK293 and SGC-7901 cells incubated with PEI-CyD/pDNA,Nd@PEI-CyD/pDNA, respectively (mean � SD, n ¼ 3, *P < 0.01).

Q. Wang et al. / Biomaterials 35 (2014) 479e488 485

complexes is over 75%, suggesting that the toxicity of the plasma-treated cationic polymers remains low, in fact much lower thanPEI (non-viral gold standard for gene transfection) [38]. The resultsalso indicate that the plasma treatment affects the gene

transfection ability of the PEI-CyD significantly. Nd@PEI-CyD ex-hibits a very high transfection efficiency of about 60% which is closeto that of some viral vectors [37] and at the same time, the in vitrocytotoxicity of Nd@PEI-CyD is quite low.

Fig. 9. (A) Representative CaM, Caveolin, and Metallothionein protein expressions determined by Western blot analysis in HEK-293 cells; (B) Analysis of light intensities of CaM,Caveolin, and Metallothionein protein expressions from Western blot results; (C) Expression of CaM, Caveolin and Metallothionein mRNA determined by RT-PCR in HEK-293 cells;(D) Analysis of light intensities of CaM, Caveolin, and Metallothionein mRNA from RT-PCR results; (E) Representative CaM, Caveolin, and Metallothionein protein expressionsdetermined by Western blot analysis in SGC-7901 cells; (F) Analysis of light intensities of CaM, Caveolin, and Metallothionein protein expressions from Western blot results; (G)Expressions of CaM, Caveolin, and Metallothionein mRNA determined by RT-PCR in SGC-7901 cells; (H) Analysis of light intensities of CaM, Caveolin, and Metallothionein mRNAfrom RT-PCR results (mean � SD, n ¼ 3, *P < 0.01).

Q. Wang et al. / Biomaterials 35 (2014) 479e488486

To investigate the underlying mechanism of the high trans-fection efficiency exhibited by the Nd@PEI-CyD, the ATP assay isperformed to evaluate the energy metabolism of cells and RT-PCRand western blot analysis are conducted to assess several mRNAand proteins. The results of the ATP assay in HEK293 and SGC-7901cells incubated with PEI-CyD/pDNA and Nd@PEI-CyD/pDNA areshown in Fig. 8. Both cells produce more ATP, suggesting thatNd@PEI-CyD can stimulate the energy metabolism of cells. Rare-earth elements have been reported to change the mitochondrialmetabolic activity [16,17]. The synthesis of proteins requires a lot ofATP for energy. Nd@PEI-CyD changes the mitochondrial metabolicactivity of the cells to produce more ATP to enhance the trans-fection efficiency. Calmodulin (CaM) is a ubiquitousmultifunctionalcalcium-binding protein participating in many physiological andbiochemical processes and thus essential in cellular regulation[21,39,40]. Some rare-earth elements can decrease the CaM levelsand regulate the cellular pathways [41]. Fig. 9A and B for HEK293cells and Fig. 9E and F for SGC-7901 show lower levels of CaMmediated by Nd@PEI-CyD. Nd affects the cell activity by downregulating the expression of CaM. PEI-CyD is internalized by thecaveolae pathway as one possible means and the inhibition of thecaveolae pathway can decrease the PEI-CyD mediated gene trans-fection [42]. Caveolins play a vital role in the formation and stabilityof caveolae [43]. There is a remarkable increase in the expression ofcaveolin induced by Nd@PEI-CyD in the HEK293 cells (Fig. 9A andB) and SGC-7901 cells (Fig. 9E and F) according to the western blotanalysis. Moreover, the RT-PCR analysis shown in Fig. 9C and D forthe HEK293 cells and Fig. 9G and H for the SGC-7901 cells confirmthe same results for caveolinmRNA. These results furnish strongevidence that Nd@PEI-CyD enhances cell uptakes by up-regulatingthe expressions of caveolins. Our data show an up-regulation ofcaveolin and down-regulation of CaM, which are the reasons for thehigh transfection.

Metallothioneins (MTs) which have a low molecular weight arecysteine-rich metal-binding proteins existing widely in thebiosphere. MTs have a wide range of biological functions in thestorage, transport, and metabolism of metallic elements [44,45],and REEs can induce the expression of MTs [46]. Expression of MTinduced by Nd@PEI-CyD is assessed by the western blot analysis(Fig. 9). Obvious increases in MTs are observed from both theHEK293 cells (Fig. 9A and B) and SGC-7901 cells (Fig. 9E and F).Additionally, the increased MT mRNA expression is verified by theRT-PCR analysis illustrated in Fig. 9C and D for HEK293 cells andFig. 9G and H for SGC-7901 cells. The results indicate that Nd@PEI-CyD up-regulates the expression of MTs which participate in themetabolism of Nd.

4. Conclusion

A new type of gene vector, Nd@PEI-CyD which combines therare-earth element Nd and non-viral polymeric gene vector, isobtained by plasma treatment and efficient delivery of DNA intocells is demonstrated. Nd@PEI-CyD exhibits a transfection effi-ciency of 58.6% which is dramatically higher than that of PEI-CyDwithout the plasma treatment and commercial transfection agentLipofectamine 2000. The luciferase expression level mediated byNd@PEI-CyD is 15 times and 27 times higher than that inducedby Lipofectamine 2000 and PEI-CyD. Because of the special bio-logical characteristics of Nd, Nd@PEI-CyD enhances the energymetabolism of cells to produce more ATP, strengthens cell up-takes by increasing the caveolin expression, and regulates thecellular pathways by adjusting the CaM. In addition, metal-lothioneins are involved in the discharge of Nd. Our study pro-vides a new strategy for highly effective metal-incorporatednon-viral gene vectors.

Acknowledgments

This work was jointly supported by the National High TechnologyDevelopment Program of China (863 Program 2007AA03Z355,2009AA02Z416), National Natural Science Foundation of China(grant # 30970711 and 21074111), Major Scientific and TechnologicalInnovation Project of Hangzhou (20122511A43), Hong KongResearch Grants Council (RGC) General Research Funds (GRF) Nos.112510 and 112212, and City University of Hong Kong AppliedResearch Grants Nos. 9667066 and 9667069.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.biomaterials.2013.09.069.

References

[1] Ferber D. Safer and virus-free? Science 2001;294:1638e42.[2] Pack DW, Hoffman AS, Pun S, Stayton PS. Design and development of poly-

mers for gene delivery. Nat Rev Drug Discov 2005;4:581e93.[3] Kay MA, Glorioso JC, Naldini L. Viral vectors for gene therapy: the art of turning

infectious agents into vehicles of therapeutics. Nat Med 2001;7:33e40.[4] Thomas CE, Ehrhardt A, Kay MA. Progress and problems with the use of viral

vectors for gene therapy. Nat Rev Genet. 2003;4:346e58.[5] Pannier AK, Shea LD. Controlled release systems for DNA delivery. Mol Ther

2004;10:19e26.[6] Merdan T, Kopec

ek J, Kissel T. Prospects for cationic polymers in geneand oligonucleotide therapy against cancer. Adv Drug Deliver Rev. 2002;54:715e58.

[7] Kunath K, von Harpe A, Fischer D, Petersen 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 Releas2003;89:113e25.

[8] Tang GP, Guo HY, Alexis F, Wang X, Zeng S, Lim TM, et al. Low molecularweight polyethylenimines linked by beta-cyclodextrin for gene transfer intothe nervous system. J Gene Med 2006;8:736e44.

[9] Hu Y, Tang G, Liu J, Cheng W, Yue Y, Li J, et al. FGF receptor-mediated genedelivery using ligands coupled to PEI-b-CyD. J Biomed Biotechnol 2012;2012.

[10] Jiang Q-Y, Lai L-H, Shen J, Wang Q-Q, Xu F-J, Tang G- P. Gene delivery to tumorcells by cationic polymeric nanovectors coupled to folic acid and the cell-penetrating peptide octaarginine. Biomaterials 2011;32:7253e62.

[11] LeGeros RZ. Calcium phosphate-based osteoinductive materials. Chem Rev.2008;108:4742e53.

[12] Peeling P, Dawson B, Goodman C, Landers G, Trinder D. Athletic induced irondeficiency: new insights into the role of inflammation, cytokines and hor-mones. Eur J Appl Physiol 2008;103:381e91.

[13] Vallee BL, Auld DS. Zinc coordination, function, and structure of zinc enzymesand other proteins. Biochemistry 1990;29:5647e59.

[14] Raven J, Evans MW, Korb R. The role of trace metals in photosynthetic electrontransport in O2-evolving organisms. Photosynth Res 1999;60:111e50.

[15] Dominiczak AF, Bohr DF. Cell membrane abnormalities and the regulation ofintracellular calcium concentration in hypertension. Clin Sci Lond 1990;79:415e23.

[16] Dai J, Zhang Y-Z, Liu Y, Li Q- G. Microcalorimetric study of the effect of Ce(III)on metabolic activity of mitochondria isolated from indice rice 9311. ChemBiodivers 2008;5:1321e6.

[17] Peng L, Yi L, Jia H, Fengjiao D, Daiwen P, Songsheng Q. Microcalorimetricinvestigation of the effect of La3þ on mitochondria isolated from avianchicken liver tissue cells. J Therm Anal Calorim 2003;73:843e9.

[18] Smith J, Smith L. Initiation of DNA synthesis in quiescent Swiss 3T3 and 3T6cells by lanthanum. Biosci Rep 1984;4:777e82.

[19] Zeng F, Tian H, Wang Z, An Y, Gao F, Zhang L, et al. Effect of rare earth elementeuropium on amaranthin synthesis in Amarathus caudatus seedlings. BiolTrace Elem Res. 2003;93:271e82.

[20] Sun Y, Liu DL, Yu ZQ, Zhang Q, Bai J, Sun DY. An apoplastic mechanism forshort-term effects of rare earth elements at lower concentrations. Plant CellEnviron 2003;26:887e96.

[21] Cheung WY. Calmodulin plays a pivotal role in cellular regulation. Science1980;207:19e27.

[22] Tyler G. Rare earth elements in soil and plant systems - a review. Plant Soil2004;267:191e206.

[23] He M, Ranz D, Rambeck W. Study on the performance enhancing effect of rareearth elements in growing and fattening pigs. J Anim Physiol Anim Nutr2001;85:263e70.

[24] Trachsel H, Erni B, Schreier MH, Staehelin T. Initiation of mammalian proteinsynthesis: II. The assembly of the initiation complex with purified initiationfactors. J Mol Biol 1977;116:755e67.

Q. Wang et al. / Biomaterials 35 (2014) 479e488 487

[25] Zeng J, Heuchel R, Schaffner W, Kägi JHR. Thionein (apometallothionein) canmodulate DNA binding and transcription activation by zinc finger containingfactor Spl. FEBS Lett 1991;279:310e2.

[26] Swirnoff AH, Milbrandt J. DNA-binding specificity of NGFI-A and related zincfinger transcription factors. Mol Cell Biol 1995;15:2275e87.

[27] Zhao Y, Wong SM, Wong HM, Wu S, Hu T, Yeung KW, et al. Effects of carbonand nitrogen plasma immersion ion implantation on in vitro and in vivobiocompatibility of titanium alloy. ACS Appl Mater Interfaces 2013;5:1510e6.

[28] Wu S, Liu X, Yeung A, Yeung KW, Kao RY, Wu G, et al. Plasma-modified bio-materials for self-antimicrobial applications. ACS Appl Mater Interfaces2011;3:2851e60.

[29] Inagaki N, Tasaka S, Narushima K, Teranishi K. Surface modification of pol-y(tetrafluoroethylene) with pulsed hydrogen plasma. J Appl Polym Sci2002;83:340e8.

[30] Amaral IF, Cordeiro AL, Sampaio P, Barbosa MA. Attachment, spreading andshort-term proliferation of human osteoblastic cells cultured on chitosan filmswith different degrees of acetylation. J Biomater Sci Polym Ed. 2007;18:469e85.

[31] López GP, Castner DG, Ratner BD. XPS O 1s binding energies for polymerscontaining hydroxyl, ether, ketone and ester groups. Surf Interface Anal1991;17:267e72.

[32] Li Y, Chen LY, Zhang Z, Wu YG, Qiao YI, Xu WX. A study on oxide film growthon Nd surface using XPS. Acta Physica Sinica 2001;50:79e82.

[33] Toli�nski T, Kowalczyk A, Che1kowska G. XPS studies of the hybridization ef-fects in RNi4B (R ¼ Ce, Pr, Nd) compounds. Phys Lett A 2003;308:75e9.

[34] Vijaya Kumar B, Velchuri R, Prasad G, Sreedhar B, Ravikumar K, Vithal M.Preparation, characterization, photoactivity and XPS studies of Ln2ZrTiO7(Ln ¼ Sm and Nd). Ceram Int 2010;36:1347e55.

[35] Fuggle J, Campagna M, Zolnierek Z, Lässer R, Platau A. Observation of a rela-tionship between core-level line shapes in photoelectron spectroscopy andthe localization of screening orbitals. Phys Rev Lett. 1980;45:1597e600.

[36] Shen J, Zhao DJ, Li W, Hu QL, Wang QW, Xu FJ, et al. A polyethylenimine-mimetic biodegradable polycation gene vector and the effect of aminecomposition in transfection efficiency. Biomaterials 2013;34:4520e31.

[37] De Smedt SC, Remaut K, Lucas B, Braeckmans K, Sanders NN, Demeester J.Studying biophysical barriers to DNA delivery by advanced light microscopy.Adv Drug Deliver Rev. 2005;57:191e210.

[38] Liu M, Li ZH, Xu FJ, Lai LH, Wang QQ, Tang GP, et al. An oligopeptide ligand-mediated therapeutic gene nanocomplex for liver cancer-targeted therapy.Biomaterials 2012;33:2240e50.

[39] Meyer T, Hanson PI, Stryer L, Schulman H. Calmodulin trapping by calcium-calmodulin-dependent protein kinase. Science 1992;256:1199e202.

[40] Braun AP, Schulman H. The multifunctional calcium/calmodulin-dependentprotein kinase: from form to function. Ann Rev Physiol 1995;57:417e45.

[41] Ji Y, Xiao B, Wang Z, Cui M, Lu Y. The suppression effect of light rare earthelements on proliferation of two cancer cell lines. Biomed Environ Sci BES2000;13:287.

[42] Hu Q, Wang J, Shen J, Liu M, Jin X, Tang G, et al. Intracellular pathways andnuclear localization signal peptide-mediated gene transfection by cationicpolymeric nanovectors. Biomaterials 2012;33:1135e45.

[43] Fra AM, Williamson E, Simons K, Parton RG. De novo formation of caveolae inlymphocytes by expression of VIP21-caveolin. Proc Nat Acad Sci 1995;92:8655e9.

[44] Kaegi JH, Schaeffer A. Biochemistry of metallothionein. Biochemistry 1988;27:8509e15.

[45] Andrews GK. Regulation of metallothionein gene expression by oxidativestress and metal ions. Biochem Pharmacol 2000;59:95.

[46] Kawagoe M, Hirasawa F, Cun Wang S, Liu Y, Ueno Y, Sugiyama T. Orallyadministrated rare earth element cerium induces metallothionein synthesisand increases glutathione in the mouse liver. Life Sci 2005;77:922e37.

Q. Wang et al. / Biomaterials 35 (2014) 479e488488