Pharmacokinetic and toxicological evaluation of multi-functional thiol-6-fluoro-6-deoxy-d-

glucose gold nanoparticles in vivo

This article has been downloaded from IOPscience. Please scroll down to see the full text article.

2012 Nanotechnology 23 375101

(http://iopscience.iop.org/0957-4484/23/37/375101)

Download details:

IP Address: 129.128.216.34

The article was downloaded on 22/11/2012 at 05:34

Please note that terms and conditions apply.

View the table of contents for this issue, or go to the journal homepage for more

Home Search Collections Journals About Contact us My IOPscience

IOP PUBLISHING NANOTECHNOLOGY

Nanotechnology 23 (2012) 375101 (10pp) doi:10.1088/0957-4484/23/37/375101

Pharmacokinetic and toxicologicalevaluation of multi-functionalthiol-6-fluoro-6-deoxy-D-glucose goldnanoparticles in vivo

Wilson Roa1, Yeping Xiong2, Jie Chen2,3,4, Xiaoyan Yang1,2, Kun Song5,Xiaohong Yang1, Beihua Kong5, John Wilson6 and James Z Xing1,7

1 Department of Oncology, University of Alberta, Edmonton, AB, T6G 1Z2, Canada2 Department of Biomedical Engineering, University of Alberta, Edmonton, AB, T6G 2V2, Canada3 National Institute for Nanotechnology, Edmonton, AB, T5H 3B6, Canada4 Department of Electrical and Computer Engineering, University of Alberta, Edmonton, AB, T6G 2V2,Canada5 Department of Gynecological Oncology/Gynecological Cancer Laboratory, Qilu Hospital of ShandongUniversity, Jinan, Shandong, People’s Republic of China6 Department of Oncologic Imaging, Cross Cancer Institute, Edmonton, AB, T6G 2V4, Canada7 Department of Laboratory Medicine and Pathology, University of Alberta, Edmonton, AB, T6G 2S2,Canada

E-mail: jzxing@ualberta.ca

Received 4 January 2012, in final form 26 July 2012Published 24 August 2012Online at stacks.iop.org/Nano/23/375101

AbstractWe synthesized a novel, multi-functional, radiosensitizing agent by covalently linking6-fluoro-6-deoxy-D-glucose (6-FDG) to gold nanoparticles (6-FDG–GNPs) via a thiolfunctional group. We then assessed the bio-distribution and pharmacokinetic properties of6-FDG–GNPs in vivo using a murine model. At 2 h, following intravenous injection of6-FDG–GNPs into the murine model, approximately 30% of the 6-FDG–GNPs weredistributed to three major organs: the liver, the spleen and the kidney. PEGylation of the6-FDG–GNPs was found to significantly improve the bio-distribution of 6-FDG–GNPs byavoiding unintentional uptake into these organs, while simultaneously doubling the cellularuptake of GNPs in implanted breast MCF-7 adenocarcinoma. When combined with radiation,PEG–6-FDG–GNPs were found to increase the apoptosis of the MCF-7 breast adenocarinomacells by radiation both in vitro and in vivo. Pharmacokinetic data indicate that GNPs reach theirmaximal concentrations at a time window of two to four hours post-injection, during whichoptimal radiation efficiency can be achieved. PEG–6-FDG–GNPs are thus novel nanoparticlesthat preferentially accumulate in targeted cancer cells where they act as potent radiosensitizingagents. Future research will aim to substitute the 18F atom into the 6-FDG molecule so that thePEG–6-FDG–GNPs can also function as radiotracers for use in positron emission tomographyscanning to aid cancer diagnosis and image guided radiation therapy planning.

S Online supplementary data available from stacks.iop.org/Nano/23/375101/mmedia

(Some figures may appear in colour only in the online journal)

10957-4484/12/375101+10$33.00 c© 2012 IOP Publishing Ltd Printed in the UK & the USA

Nanotechnology 23 (2012) 375101 W Roa et al

1. Introduction

The use of gold nanoparticles (GNPs) represents a promisingmodality to achieve personalized cancer therapy, includingthe targeted delivery of anti-cancer chemotherapeutics suchas methotrexate [1], tamoxifen [2] and oxaliplatin [3].GNPs have also been successfully applied in radio-frequencyradiation [4], thermal ablation [5, 6] and metal enhancedradiotherapy [7]. GNPs enhance the radiosensitivity of cancercells by preferentially increasing the photoelectric absorptionof low kilovoltage x-rays in tumour cells as compared tonormal soft tissue adjacent to the tumour cells [8–12]. In ourprevious work, GNPs were utilized as a radiation sensitizerto significantly enhance radiation induced cancer cell deathin breast cancer MCF-7 cells [12], prostate cancer DU-140cells [13] and ovary cancer Hela cells [14]. When prostatecancer DU-140 cells were treated with thiol-glucose coatedGNPs (Glu–GNPs) for 6 h followed by x-ray irradiation,we observed a significant enhancement in radiosensitivitycompared to the controls [15].

Utilizing a thiol functional group, glucose molecules canbe covalently linked to the GNP surface to form Glu–GNPs.Since cancer cells are known to metabolize glucose muchfaster than normal cells, Glu–GNPs have been shown todouble the amount of intracellular uptake of Glu–GNP [13,16]. Increased cancer cell uptake of Glu–GNPs comparedto unbound GNPs leads to enhanced radiotherapeuticcytotoxicity in vitro [13, 14] and contrast enhancement oftumours in computerized axial tomography (CT) imaging [17,18]. The mechanisms governing the Glu–GNP enhancementof radiotherapy have been investigated in vitro [15]. Whencombined with 2 Gy of orthovoltage irradiation, Glu–GNPscontribute to a 1.5–2.0 fold enhancement in growth inhibitionof prostate cancer cells compared to those treated withx-rays alone [13]. This enhancement is likely caused by thegeneration of Auger electrons, a significant factor contributingto DNA damage that subsequently leads to cell apoptosis.Other potential mechanisms of the increased radiosensitivityinclude Glu–GNP induced acceleration of the G0/G1 phaseand accumulation of cells in the G2/M phase [15].

Glucose transport rates are estimated in a non-invasivemanner in the physiological and pathological states bykinetic imaging using PET [19]. The common PET traceris the glucose analogue, 2-(18F)fluoro-2-deoxy-D-glucose(18F-2-FDG). In comparison with 18F-2-FDG which is poorlyabsorbed by the intestines and kidneys, 18F-labelled 6-fluoro-6-deoxy-D-glucose (18F-6-FDG) is actively transported by thekidneys and intestines [20]. Therefore, 18F-6-FDG can beutilized as a viable PET tracer for glucose transport in certainspecific organs (e.g., the intestine) in which 18F-2-FDGcannot perform in the same manner [20].

To achieve the functions of imaging contrast and ofimage guided radiotherapy in a unified nanoparticle platform,we proposed to bind 18F-6-FDG, a useful PET tracer andknown radiosensitizer, to the GNP surface, thus forming18F-6-FDG–GNPs. A thiol group on the 6-FDG molecule wasused to covalently bind the 18F-6-FDG onto the surfaces ofGNPs. In this paper, we examine the in vivo toxicological

safety and pharmacokinetic profile of our multi-functionalGNPs. The in vivo radiation enhancement was furtherevaluated in order to demonstrate the promising role thatGlu–GNPs can play in treating cancers.

2. Materials and methods

2.1. General

Reagents and solvents were obtained from Sigma-AldrichChemical Corp. and Fisher Scientific Inc., and were usedwithout further purification. The MTT cytotoxicity assaykit was purchased from Invitrogen (Burlington, Ontario,Canada). The pre-coated silica gel 60 F254 analyticalplates for thin layer chromatography (TLC) were obtainedfrom Merck (Kirkland, Quebec, Canada). The flash columnchromatography was performed on silica gel 60. 1H-NMRspectra were recorded on a Varian 600 MHz spectrometer at298 K in CDCl3 or D2O.

2.2. Synthesis of 6-deoxy-6-fluoro-1-thiol-D-glucose(Thiol-6-FDG)

The synthesis process of thiol-6-FDG is shown in figure 1(A).Compound-1 was 1,2,3,4-tetra-O-acetyl-beta-D-gluco-

pyranose (Sigma-Aldrich).Compound-2 was 6-fluoro-1,2,3,4-tetra-O-acetyl-beta-D-

glucose. Diethylaminosulfur trifluoride (DAST, 6.20 mmol)was added to a solution of compound-1 (1050 mg, 3.02 mmol)in anhydrous CH2Cl2 (3 ml) at −41 ◦C. The solution wasremoved from the cooling bath and then stirred until it waswarmed to room temperature. The solution was cooled to−10 ◦C and quenched by methanol (2 ml), then concentratedunder reduced pressure. The residue was then subjected toflash column chromatography to yield compound-2 as a whitepowder. About 658 mg of final product was obtained with apercentage yield of 62.25%.

Compound-3 was 6-fluoro-2,3,4-tri-O-acetyl-1-S-acetyl-1-thiol-beta-D-glucopyranose. HSAc (20 µl, 0.27 mmol)and compound-2 (425 mg, 1.214 mmol) were added to asuspension of aluminum chloride (AlCl3) (36 mg, 0.27 mmol)in anhydrous dichloromethane (CH2Cl2) (2 ml). The mixturewas refluxed for 20 min using an oil bath at 55 ◦C. Thereaction mixture was then diluted with CH2Cl2 (10 ml)and washed with saturated sodium bicarbonate (NaHCO3)solution and brine. The crude product was subjected to silicagel flash column chromatography (hexanes/EtOAcratio 3:1)to obtain a white powder: compound-3 (34 mg, 65%), Rf 0.37(Hex/EtOAc ratio 1:1). To identify this compound, 1H-NMR(δH) was performed (figure 1(B)): 2.00, 2.01, 2.04 (9H, 3S,OCOCH3), 2.37 (3H, S, SCOCH3), 3.80–3.86 (1H, m, H5),4.35–4.52 (2H, m, H6), 5.08–5.12 (2H, m, H2 and H4),5.26–5.30 (2H, m, H1 and H3). 280 mg of final product wasobtained with a percentage yield of 63%.

Compound-4 was 1,6-dideoxy-6-fluoro-1-thiol-beta-D-glucose. Sodium methoxide (0.37 ml, 0.5M in CH3OH,0.185 mmol, 2 equiv) was added to a solution of 6-fluo-ro-2,3,4-tri-O-acetyl-1-S-acetyl-1-thiol-beta-D-glucopyranose

2

Nanotechnology 23 (2012) 375101 W Roa et al

Figure 1. Thiol-6-FDG synthesis. (A) Schematic of the synthesis: compound-1: 1,2,3,4-tetra-O-acetyl-beta-D-glucopyranose purchasedfrom Sigma; compound-2: 6-fluoro-1,2,3,4-tetra-O-acetyl-beta-D-glucose; compound-3:6-fluoro-2,3,4-tri-O-acetyl-1-S-acetyl-1-thiol-beta-D-glucopyranose; and compound-4 1,6-dideoxy-6-fluoro-1-thiol-beta-D-glucose. (B) The1H-NMR spectrum of 6-fluoro-2,3,4-tri-O-acetyl-1-S-acetyl-1-thiol-beta-D-glucopyranose.

(compound-3) (367 mg, 1.0 equiv) in methanol (CH3OH)(1 ml). The solution was stirred at room temperature for 2 hand then the solvent was removed by a rotary evaporator. Theresidue was purified using a weak cation exchange columnto obtain pure compound-4. 198 mg of final product wasobtained with a yield of 53.95%.

2.3. Synthesis of thiol-6-18FDG

1,6-Dideoxy-6-fluoro-1-thiol-beta-D-glucose (thiol-6-18FDG)was prepared in three steps. First, 1,2,3,4-tetra-O-acetyl-beta-D-glucopyranose was prepared. Reaction of this precursorwith 19F (the non-radioactive counterpart of 18F) resultedin the formation of 6-fluoro-2,3,4-tri-O-acetyl-1-S-acetyl-1-thiol-beta-D-glucopyranose (compound-3) identified by

NMR (figure 1(B)). Subsequent hydrolysis resulted in theproduction of thiol-6-FDG. Finally, when 18F was usedto replace the F in thiol-6-FDG, the thiol-6-18FDG wasproduced.

2.4. Synthesis of 6-FDG–GNPs

The gold nanoparticles were synthesised [13] by the followingthree steps. (i) 2 ml of 25 mM tetrachloroauric acid (HAuCl4)solution was added into 25 ml of deionized water in an icebath with moderate stirring. (ii) 2 ml of 30 mM sodiumborohydride (NaBH4) was then added as a reductant to obtainGNPs without any capping agents. (iii) To functionalizethe GNPs, 4 ml of 25 mM thiol-6-FDG was added to the

3

Nanotechnology 23 (2012) 375101 W Roa et al

previous gold solution to obtain functional 6-FDG–GNPs.Considering that the GNPs in step (ii) were vulnerable toaggregation, we added sodium citrate (TGS) as a cappingagent. The average number of biomolecules (2.5 × 104

thiol-6-FDG) on each gold nanoparticle (20 nm in size) wascalculated by measuring the gold-to-sulfur atom ratio withx-ray photoelectron spectroscopy (XPS) (Kratos Analytical).

2.5. Synthesis of PEG and thiol-6-FDG co-coated goldnanoparticles (PEG–FDG–GNPs)

In order to increase the biological half-life of the nanoparticlesin vivo, a PEG-coated GNP was developed [23]. PEG-coatedGNPs were prepared in two steps. Firstly, GNPs weresynthesized using a well-known citrate reduction method [21].Secondly, the GNPs were covalently linked to PEG by meansof the thiol group of PEG-SH. PEG-SH was chosen as asurface reagent because it is biocompatible and spontaneouslyforms a chemisorbed surface layer on gold surfaces [22, 23].

1.125 ml of 10 mM PEG (Sigma-Aldrich) was added into1.125 ml of 20 mM thiol-6-FDG; the volume of PEG wasadjusted in order to produce the desired molar ratio of FDGto PEG and the mixed solution was added into 30 ml of 20 nmdiameter GNPs in a 50 ml centrifuge tube. The mixture wasvortexed well and incubated at room temperature for 24 h.The GNPs bound with both thiol-6-FDG and PEG were thencentrifuged at 9800 rpm for 20 min. After careful removal ofthe supernatant layer with a pipette, the concentrated GNPcomplex was diluted with deionized water to the desiredconcentration.

2.6. Characterization of the PEG–FDG–GNPs

The hydrodynamic particle size of the PEG–FDG–GNPs wasmeasured by electrophoretic light scattering (ELS) (BeckmanCoulter). The size of the PEG-coated GNPs in their driedstate was measured using a Hitachi HF 3300 transmissionelectronic microscope (TEM). The gold concentration wasmeasured by inductively coupled plasma mass spectrometry(ICP-MS) (PerkinElmer). The surface atom composition ofthe GNPs was assessed by Kratos Axis 165 XPS (KratosAnalytical).

2.7. Cell culture and cytotoxicity assay

The cell culture and culture conditions were as follows.The human breast cancer adenocarcinoma line MCF-7 wasused in all experiments. The MCF-7 cell line was purchasedfrom American Type Culture Collection (Manassas, VA,USA) and maintained in Dulbecco’s modified Eagle medium(DMEM, Invitrogen) containing 10% foetal bovine serum(FBS) (Invitrogen), 100 IU penicillin G and 100 mg mlstreptomycin (Sigma). The cells were incubated at 37 ◦C in ahumidified 5% CO2 atmosphere. The cytotoxicity induced bythe GNPs with and without radiation was assessed by an MTTassay in 96-well plates following the instructions provided bythe manufacturer.

2.8. The tumour-bearing mouse model

The tumour model was constructed as follows. MCF-7 cellswere diluted in a culture medium (containing neither serumnor antibiotics) to a concentration of 5× 106 cells per 0.2 ml.BALB/c female nude mice with intact ovaries at 4–5 weeksof age were used in this experiment. We chose the BALB/cfemale nude mouse model because these mice produce thenecessary hormones for the growth of breast cancer cellsand because they have been demonstrated to work well inconjunction with the human MCF-7 cell line. The mice weremaintained in a specific pathogen-free environment and caredfor in accordance with the institutional animal care guidelines.Each mouse was injected subcutaneously on the right flankwith 5 × 106 cells. The tumour was permitted to grow for10 days at which time the mean tumour volume (V) wasapproximately 0.74 cm3.

2.9. Bio-distribution of 6-FDG–GNPs andPEG–6-FDG–GNPs

On day one of the experiment, the mice were groupedinto separate cages (n = 3). While under light anaesthesia(diethyl ether, Sigma-Aldrich), the animals were subjectedto a tail-vein intravenous injection of 100 µl of GNPfluid (5 mg kg−1). The injections were well tolerated andno adverse effects were observed during the first 24 h.Twenty-four hours post-injection, the mice were anaesthetizedwith light ether and serum was collected in a heparinizedglass tube via retro-orbital bleeding. Following the serumcollection, the mice were sacrificed by cervical dislocation.Their organs including heart, liver, lungs, spleen, kidneysand stomach were collected for further analysis of thebio-distribution of the GNPs. The tissues were washed withnormal saline and dried with blotting paper. The isolatedtissues were weighed and stored at −20 ◦C and the serumwas stored at 4–6 ◦C until the time of analysis. To assessthe bio-distribution of the GNPs in tumour tissue, the tumourtissues were collected at 5, 10, 30, 120 and 240 min after theinjection of the nanoparticles.

Five hundred milligrammes of each sample of GNP-containing tumour tissue and murine organ tissue wereminced by scissors and suspended in PBS to a final volumeof 5 ml. 5 ml of 50% nitric acid (HNO3) was added to eachsample in order to lyse the cells, after which the sampleswere boiled until all sample tissues were dissolved. Thelysis solution was then dried and 5 ml of 50% HNO3 wasadded in order to normalize final solution volume to 5 ml.The gold mass in the lysis solution was measured usinginductively coupled plasma mass spectrometry (ICP-MS) andthe concentration of GNPs was calculated as gold mass andexpressed as GNP mass per gramme of tissue [12].

2.10. In vivo acute toxicity

We monitored the mice for acute toxicity at the followingsix different doses in a step-wise fashion: 35, 50, 67.6,84.5, 101.5 and 120 mg Au kg−1. Mice were grouped by

4

Nanotechnology 23 (2012) 375101 W Roa et al

GNP concentration (n = 6) and received tail-vein intravenousinjections of 100 µl of 6-FDG–GNPs at the respectiveconcentration while under ketamine anaesthesia. After 1 h,the mice received a second intravenous injection of 100 µlat the same concentration of 6-FDG–GNPs. Each dose levelwas delivered to a new set of mice in order to determine theacute toxicity for each injected dose level. Animal survivaland morbidity was monitored three times a week using amorbidity score based on institutional guidelines.

2.11. 6-FDG–GNP-based radiotherapy

Irradiation of cells: all irradiation treatments of cell cultureswere carried out using a Pantak Therapax 3 Orthovoltage244 Monitor Units/minute x-ray machine at 200 kVp usinga 0.35 Cu + 1.5 Al filter. The MCF-7 cells were allowedto grow in 96-well plates overnight. After 75% confluencewas reached, the old medium was replaced with 150 ml offreshly prepared medium containing GNPs at the desiredconcentrations. After 2 h of incubation to allow cellular uptakeof the GNPs, the medium was replaced with fresh medium.The cells then received either a mock treatment for control or2 Gy of radiation. After irradiation, the cultures were returnedto incubation at 37 ◦C with 5% CO2 until the time of analysis.The response of the cell culture to the GNPs with and withoutirradiation was assessed by the MTT assay following themanufacturer’s instructions.

2.12. In vivo study

Based on the pharmacokinetic data of 6-FDG–GNPsdetermined in step 2.9, an optimized radiotherapeutic protocolwas established. Twenty tumour-bearing BALB/c nude mice(four to five weeks of age) were used in this part ofthe study. The mice were divided into 4 groups: (1) thecontrol group (n = 5) for which 200 µl saline alone wasintravenously injected; (2) the experimental group receiving6-FDG–GNPs (n = 15); (3) the experimental group treatedwith x-ray radiation alone (n = 15); and (4) the experimentalgroup treated with both 6-FDG–GNPs and x-ray radiation(n = 5). All treatments were given while the mice wereunder anaesthesia with ketamine. The control group wasonly observed for the effect of the anaesthesia and injectionprocedures. Groups (2) and (4) received tail-vein intravenousinjection of either GNPs or 6-FDG–GNPs. The 6-FDG–GNPswere administered at a starting concentration of 100 µg ml−1

for each fixed volume (0.2 ml) of IV solution. To assess theimpact of the GNPs on radiotherapy, the animals in group (4)were irradiated using a Pantak Therapax 3 Orthovoltage 244Monitor Units/minute x-ray machine at 200 kVp using a 0.35Cu + 1.5 Al filter. Mice received either a mock treatment forcontrol or 10 Gy of x-rays. Four weekly fractions of radiationwere administrated to the mice in group 4. The irradiated micewere monitored as follows: continuously for the first 4 h, thenevery 1 h for eight hours, then every 12 h for two days andthen once daily for two weeks. The time-course progressionof tumour values was then plotted.

2.13. Statistics

All the experimental values were determined in triplicate.All measured values including percentage of gold contentwere expressed as means and standard errors (SEs). Theone-way analysis of variance (ANOVA) and Tukey multiplecomparison post-test were used. Differences of less than 0.05(p < 0.05) were considered statistically significant.

3. Results and discussion

We have demonstrated the effectiveness of GNPs as radiationdose enhancing agents for a number of cancers, includingbreast and prostate cancer lines, in vivo and in vitro [8–12].Several mechanisms for GNP enhanced radiotherapy havebeen reported. Cancer cells take up more glucose comparedto adjacent normal tissue due to their elevated metabolicgrowth rate. In our study, thiol-glucose was used tocoat GNPs (Glu–GNPs) resulting in higher concentrationsof GNPs in cancer cells compared to adjacent normaltissues [12]. However, Glu–GNPs have no imaging function.Therefore, we developed thiol-6-FDG–GNPs that can beeventually labelled by the radioisotope 18F for PET imaging(supplementary supporting material available at stacks.iop.org/Nano/23/375101/mmedia). We are the first research groupworldwide to report the pharmacokinetic and toxicologicalevaluation of 6-FDG–GNPs both in vitro and in vivo.

We first synthesized thiol-19fludeoxyglucose coatedGNPs because 19fludeoxyglucose can achieve the samefunction as 18fludeoxyglucose but without radiation. As aresult, we could use non-radioactive 19fludeoxyglucose in thelaboratory to independently test pharmacokinetics, toxicologyand cancer killing efficiency. In our experiments, we usednon-radioactive thiol-19fludeoxyglucose coated GNPs (thiol-6-FDG–GNPs) as a model compound for pharmacologicalstudy and evaluated its biological function in vivo. Finally,we optimized the bio-distribution of thiol-6-FDG–GNPs incancer tissue for enhanced radiotherapy on breast cancer-bearing mice.

3.1. Thiol-6-19FDG and thiol-6-18FDG

18F-labelled 6-FDG has been approved as a valuable tracerthat is actively transported by the kidneys as well as theintestines [20]. 6-FDG’s ability to promote higher uptakeby glucose-intensive tissue, including most cancers, makesit a useful molecule for diagnosing cancer via PET scan.We exploited 6-FDG’s capacity for intercellular uptake andstorage as a means to deliver GNPs into tumour cells bymodifying the structural properties of 6-FDG so that it couldbe bound to the GNP surface. In order to bind 6-18FDGto GNPs, we synthesized a thiol group on the 6-FDGmolecule to form thiol-6-FDG (figures 1(A) and (B)). Weused thiol-6-FDG as a model chemical for initial chemicaland biological studies. After labelling 18F to thiol-6-FDG,the thiol-6-18FDG was used as a tracer for PET with its ownspecific bio-distribution (supplementary supporting materialavailable at stacks.iop.org/Nano/23/375101/mmedia) that is

5

Nanotechnology 23 (2012) 375101 W Roa et al

Figure 2. Synthesis of PEGylated 6-FDG–GNPs: FDG (A) and PEG (B) were coated onto GNPs (C) to form PEGylated 6-FDG–GNPs(D). The surface atom composition of the GNPs was assessed by Kratos Axis 165 x-ray photoelectron spectroscopy (XPS) (KratosAnalytical). (E) S-bind and (F) Au-bind.

different from 2-18FDG, but similar to 6-18FDG reportedpreviously [20].

3.2. Synthesis of 6-FDG–GNPs and PEG–6-FDG–GNPs

In this study, we employed non-radioactive thiol-19fludeoxy-glucose coated GNPs (thiol-6-FDG–GNPs) as a modelcompound to assess the bio-distribution in vivo. Althoughthiol-19Fludeoxyglucose coated GNPs do not carry theradioactive 18F molecule that would enable the GNPsto be localized with a PET scan, thiol-19fludeoxyglucosepossesses the same molecular functional properties as18fludeoxyglucose without the potentially harmful effects ofradiation. As such thiol-19fludeoxyglucose provided us withthe means to conduct our initial pharmacological assessmentswithout the need for RADHAZ protective measures. Thenext step in our research process will be to radiolabel ourmulti-functional molecule with the 18F atom in order to assessits suitability for diagnostic and radiation treatment planningimaging studies.

Figure 2 shows a schematic diagram of the PEGand thiol-6-FDG mixture coated GNPs. Figure 2(A) showstransmission electron microscopy (TEM) images of theGNPs with average size of 20.17 + 1.23 nm (theTEM pictures of other nanoparticles are shown in thesupplementary material available at stacks.iop.org/Nano/23/375101/mmedia). In order to identify the amount of PEGand thiol-6-FDG bound to each GNP (figure 2(B)), the

nanoparticles were characterized using x-ray photoelectronspectroscopy (XPS). The average numbers of FDG and PEGcombined bound to each GNP were 2.5×104 (for 10 nm) and1.15 × 105 (for 20 nm) respectively, which were calculatedby measuring the gold-to-sulfur atom ratio obtained by XPS(figures 2(E) and (F)). In the synthesis process, equal amountsof FDG and PEG were added for binding to the GNPs. Hence,in theory, the PEG to FDG ratio on the GNPs should be 1 to1 since both PEG and FDG had equal chances of thiol–goldbinding.

3.3. Bio-distribution of 6-FDG–GNPs

The 6-FDG-GNP suspension was administered by intravenousinjection into mice. Figure 3(A) lists the bio-distribution ofdifferent sized 6-FDG–GNPs (without PEGylation) in variousorgan tissues (µg g−1 tissue) at 2 h after administration.For nanoparticles used for cancer therapy, the correlationof achieved levels of tumour accumulation with respect tothe size of GNPs is of paramount importance. Our in vitroexperiments demonstrated that the 20 nm 6-FDG–GNPsachieved the highest uptake in tumour tissue (figure 3(C)) andthey were henceforth used in the study. The bio-distributionsof 6-FDG–GNPs (20 nm) at different time intervals afterinjection are tabulated in table 1 and figure 4(B). It isof interest that the data from the in vivo murine modelindicate that 6-FDG–GNPs (20 nm) had also significantlyaccumulated in various organs (figure 3(A)). From the

6

Nanotechnology 23 (2012) 375101 W Roa et al

Figure 3. (A) Bio-distribution of FDG–GNPs of different sizes inmouse organ tissues (µg g−1 tissue). (B) Distribution of 20 nm6-FDG–GNPs in mouse organs (percentage). After 24 h ofadministration of 6-FDG–GNPs, mouse blood and tissues includingblood, liver, spleen, lung, kidney and heart were collected. The goldmass in the lysis solution was measured using ICP-MS. The numberof GNPs was calculated as the gold mass and expressed as GNPmass per gramme of tissue. (C) Bio-distribution of FDG–GNPs ofdifferent sizes in mouse tumour tissues (µg g−1 tissue).

pharmacodynamic perspective, 6-FDG–GNPs accumulatedin the liver and the spleen as soon as 5 min after the6-FDG–GNP injection, at 4.35 and 1.62 µg g−1 respectively.At 2 h post-injection, the concentrations of nanoparticlesin the liver and the spleen reached peak levels of 7.9 and1.56 µg g−1, respectively (table 1 and figure 3(B)). Theconcentration of GNPs in the kidneys was 2.16 µg g−1 at2 h post-injection, and gold concentrations were also detectedin mice urine as soon as 10 min post-injection (table 1).Therefore, although the 6-FDG–GNPs (20 nm) achieveda relatively high concentration in tumour tissue, selectednormal organs accumulated a very high level of GNPs as

Figure 4. PEGylation improved bio-distribution of 6-FDG–GNPsin vivo: (A) distribution of PEG–GNPs, 6-FDG–GNPs andPEG–FDG–GNPs in mouse organs; (B) in tumour tissue; and (C)the dynamic pattern of cell uptakes of PEGylatedthiol-6-FDG–GNPs by tumour tissue.

well. In order to shift the bio-distribution of the 20 nm6-FDG–GNPs preferentially to tumour cells over normaltissue, we PEGylated the 6-FDG–GNPs.

3.4. Bio-distribution of PEGylated 6-FDG–GNPs in normaltissue

PEG is known to reduce non-specific uptake of variousmolecules by the liver and the spleen. The coating ofour 6-FDG–GNPs with PEG (PEG–FDG–GNPs) resulted insuperior bio-distribution in various organs when comparedto uncoated 6-FDG–GNPs. For instance, PEG significantlyreduced non-specific uptake by the liver from 5.41 µg g−1

for 6-FDG–GNPs to 1.61 µg g−1 for PEG–FDG–GNPs.Decreased accumulation of gold in other organs, such as

7

Nanotechnology 23 (2012) 375101 W Roa et al

Table 1. Bio-distribution of FDG–GNPs (20 nm) in mouse organ tissues (µg g−1 tissue) at various time periods after injection.

Organs and tissues 5 min 10 min 2 h

Blood 0.732 ± 0.061 0.451 ± 0.075 0.112 ± 0.021Liver 4.35 ± 0.223 6.06 ± 0.234 7.977 ± 0.341Spleen 1.62 ± 0.085 1.662 ± 0.094 1.561 ± 0.232Lung 0.16 ± 0.013 1.074 ± 0.121 0.975 ± 0.123Kidney 0.265 ± 0.082 0.079 ± 0.006 2.166 ± 0.082Heart 0.405 ± 0.031 0.207 ± 0.145 0.708 ± 0.021Cancer 0.055 ± 0.031 0.105 ± 0.145 0.207 ± 0.031Connective tissue 0.045 ± 0.033 0.037 ± 0.122 0.058 ± 0.024Muscle 0.042 ± 0.003 0.0197 ± 0.003 0.062 ± 0.022Urine NDa 0.15 ± 0.0122 ND

a ND: not detected.

the spleen, the lung, the kidneys and muscle, was alsoobserved (figure 4(A)). The significantly lower accumulationsof PEG–FDG–GNPs in various organs may reduce the acutetoxicity to these tissues; however, the organ toxicity inducedby PEG–FDG–GNPs needs to be further investigated.

It has been reported that nanoparticles administered intothe blood stream, without surface modification, are eliminatedfrom the circulation within seconds to minutes through thereticulo-endothelial system (RES) [25]. Similar results wereobserved in our experiments. Five minutes after the injectionof thiol-6-FDG–GNPs into mice, GNPs accumulated in theliver and the spleen at concentrations of 4.35 µg g−1 and1.62 µg g−1, respectively (table 1). Two hours after injection,thiol-6-FDG–GNPs in the liver reached a peak concentrationof 7.9 µg g−1. This particularly high thiol-6-FDG-GNPconcentration in the liver is attributable to the hepatic roleof glucose-storage and reticulo-endothelial clearance. WithFDG–GNPs, although the amount of gold accumulation in thetumour is twice that of adjacent normal tissue, it is relativelylow at only one-fiftieth of that in the liver. To improve thebio-distribution of thiol-6-FDG–GNPs in vivo, we coatedGNPs with both thiol-6-FDG and PEG (figure 2) becausePEG helps GNPs to avoid the reticulo-endothelial system toimprove GNP bio-distribution in vivo [24, 26]. In our studies,PEGylation of GNPs greatly improved their bio-distributionin vivo while decreasing accumulation in the liver and thespleen by 70.3%, and 52.1%, respectively (figure 4(A)). Inthe meantime, PEG–FDG–GNPs resulted in increased GNPconcentration in cancer tissue by 108.8% when comparedto 6-FDG–GNPs without PEGylation (figure 4(B)). WithPEG–FDG–GNPs, the gold accumulation found in the tumouris about one-eighth of that in the liver. Cancer tissueswould uptake 2.23 times more 6-FDG–GNPs and 6.09times more PEG–FDG–GNPs than normal connective tissues,respectively (figure 4(B)).

PEGylation of the GNPs has, in a sense, improvedthe targeting of the GNPs by avoiding disproportionatelyhigh uptake by the reticulo-endothelial system. Although notclinically apparent at the dose levels employed in our study,this critical shift of the GNPs away from the liver and kidneysmay be an important means in reducing radiation inducedkidney and liver toxicity in the next step of our research whenthe dose levels of the GNPs may be potentially higher and wewill be employing 18F-labelled PEG–FDG–GNPs.

3.5. Bio-distribution of PEG–FDG–GNPs in cancer tissue

The ability for 18F-labelled 6-FDG to promote higher uptakeby glucose-intensive tissue (including most cancers) is usefulfor diagnosing cancer by PET scan [20]. We used this abilityof 6-FDG to deliver more nanoparticles into cancer tissue ascompared to adjacent normal tissue, enabling cancer to bedistinguished from normal tissue.

This increased accumulation of GNPs in tumourtissue may have resulted from improved uptake due toFDG- and PEG-ligands. Compared to the cell uptakeof GNPs by non-cancerous connective and muscle tissue(0.0433 µg g−1), the intra-tumour gold concentrations werealmost doubled after the injection of either 6-FDG–GNPs(from 0.0433 to 0.0952 µg g−1) or PEG–GNPs (from0.0433 to 0.1125 µg g−1) (figure 4(B)). PEGylation ofthe FDG–GNPs further improved the tumour uptake of theGNPs to 0.198 µg g−1, which is almost five times morethan citrate GNPs (i.e., 2.02 fold for PEG–GNPs, 2.33 foldfor 6-FDG–GNPs and 6.09 fold for PEG–FDG–GNPs whencompared to the adjacent normal connective and muscletissue).

3.6. The kinetics of PEG–FDG–GNP uptake by canceroustissue

The uptake of 20 nm PEG–FDG–GNPs by the MCF-7 breastadenocarcinoma in cancer-bearing mice was measured. 5 minpost-injection, PEG–FDG–GNPs were initially observed incancerous tissue with a concentration of 0.043 mg Au kg−1

of cancer tissue. As time elapsed, the concentration of GNPssteadily increased until a peak concentration of 0.1988 mgAu kg−1 was observed 2 h after injection. Subsequently,the gold concentration in cancer tissue remained appreciablyhigh, at 0.127 mg Au kg−1 4 h after GNP injection(figure 4(C)). The dynamic change of GNP bio-distributionindicates that the GNPs reach their maximal concentrationsat a time window of two to four hours post-injection(figure 4(C)). In order to achieve the optimal radiationefficiency, it will be logical to administer the radiation withinthis time window.

The timescale of the uptake and maintenance of peaklevels of GNPs in tumour cells observed in the murine modelexperiments (2–4 h post-injection) is of clinical significance.

8

Nanotechnology 23 (2012) 375101 W Roa et al

Figure 5. In vitro cytotoxicity: (A) cell survival rate induced by 6-FDG, thiol-6-FDG and thiol-6-FDG–GNPs; (B) GNP enhancedradiotherapy.

Two to four hours of time represents a window of clinicalfeasibility that would easily allow modern day radiationoncology clinics enough time to assess, set up and deliver adaily prescribed dose of radiation following an IV injection ofGNPs.

3.7. In vitro cytotoxicity and in vivo acute toxicity

In vitro tests were performed in order to study GNP inducedcytotoxicity. The cancer cells were treated with chemicalsfor 2 h; only small cytotoxicities were observed and the cellcultures maintained cell survival rates of 96.4, 96.23 and93.28%, for 6-FDGs, thiol-6-FDGs and thiol-6–FDG–GNPsrespectively (figure 5(A)). Furthermore, the toxicity ofthiol-6-FDG–GNPs was also evaluated in vivo. In our study,six doses, 35, 50, 67.6, 84.5, 101.5 and 120 mg Au kg−1,were chosen. The highest concentration of 6-FDG–GNPs,120 mg Au kg−1, and all other lower dose levels didnot induce apparent acute toxicity in the mice over aone-week observation period. The dose corresponding toan acute adverse response was never reached within theobserved period, and thus the maximum tolerable dose of6-FDG–GNPs was not reached.

No significant difference in cytotoxicity was observedbetween the control, 6-FDG and thiol-6-FDG (p < 0.05) (seefigure 5). Therefore, neither 6-FDG nor thiol-6-FDG inducedsignificant cytotoxicity in MCF-7 cells and thiol-6-FDG maybe used as a means to deliver GNPs to cancer tissue.

3.8. PEG–FDG–GNP enhanced radiotherapy sensitivityin vitro and in vivo

3.8.1. In vitro study. MCF-7 cells with and withoutGNP pre-treatment were irradiated with 200 kVp x-rays ata dose of 10 Gy. We assessed the resulting cytotoxicityusing the clonogenic assay five days after the irradiation.Responses of the MCF-7 cell line with and without x-raytreatment are shown in figure 5(B). Following irradiation,MCF-7 cancer cells maintained a cell viability of 59.32% fivedays after irradiation (figure 5(B)). The addition of 15 nMPEG–FDG–GNPs (here the concentration refers to the goldconcentration) induced enhanced radiation cell killing with aresultant cell survival rate of 32.34%.

Figure 6. 6-FDG–GNP enhanced radiotherapy sensitivity in vivo. Asignificantly superior tumour control is found with radiotherapy plusPEG–FDG–GNPs over radiotherapy alone treatment (P < 0.05).

3.8.2. In vivo study. The results displayed in figure 6 show asignificant reduction in the tumour size when x-ray radiationtreatment was combined with PEG–FDG–GNP injections ascompared to treatment with x-rays alone (P < 0.05).

The in vitro and in vivo experiments comparingradiotherapy with and without GNPs have demonstratedthat PEG–FDG–GNP molecules, when administered prior toradiotherapy, significantly enhance the cell-killing effects ofx-ray radiation. In vitro, the relative reduction in survivingMCF-7 tumour cells in the cultures treated with bothPEG–FDG–GNPs and radiation therapy was approximately50% when compared to radiation alone. In vivo, the MCF-7tumours in the murine model were also significantly smallerwhen treated with both PEG–FDG–GNPs and radiation asopposed to radiation alone. As such, the PEG–FDG–GNPs,despite having PEG and FDG molecules appended to theGNPs, maintain the much desired radiosensitizing propertiesthat GNPs have been previously found to have [8–12]. Futureresearch will be geared to investigating whether this sameradiosensitizing effect is preserved across different cancer celltypes as a means to determine the extent of the applicabilityof the PEG–FDG–GNPs.

4. Conclusion

The PEG–FDG–GNP represents a novel multi-functionalnanoparticle with the ability to gain entry into and residein breast adenocarinoma cells with greater affinity thancitrate–GNPs alone. Furthermore, PEGylation simultaneously

9

Nanotechnology 23 (2012) 375101 W Roa et al

decreases hepatic and renal uptake of GNPs while increasingthe relative amount of GNPs taken up by tumour cells.PEG–FDG–GNPs have also been shown in vitro and in vivoto act as a potent radiosensitizing agent that increases thelethality of radiation therapy towards breast cancer cells.In future investigations, we will use 18F to replace 19F forradiolabelling of PEG–FDG–GNPs in order to allow thePEG–FDG–GNPs to be used to aid in the diagnosis and imageguided radiation treatment planning of tumours.

Acknowledgments

This work was carried out with funding support fromthe Canadian Breast Cancer Foundation, Canadian CancerSociety Research Institute and National Research Coun-cil/National Institute for Nanotechnology, Canada. We thankJulian Kim and Ray Yang for their help in revising theEnglish.

References

[1] Chen Y H et al 2007 Methotrexate conjugated to goldnanoparticles inhibits tumour growth in a syngeneic lungtumour model Mol. Pharm. 4 713–22

[2] Dreaden E C et al 2009 Tamoxifen-poly(ethylene glycol)-thiolgold nanoparticle conjugates: enhanced potency andselective delivery for breast cancer treatment Bioconjug.Chem. 20 2247–53

[3] Brown S D et al 2010 Gold nanoparticles for the improvedanticancer drug delivery of the active component ofoxaliplatin J. Am. Chem. Soc. 132 4678–84

[4] Curley S A et al 2008 Noninvasive radiofrequencyfield-induced hyperthermic cytotoxicity in human cancercells using cetuximab-targeted gold nanoparticles J. Exp.Ther. Oncol. 7 313–26

[5] Terentyuk G S et al 2009 Laser-induced tissue hyperthermiamediated by gold nanoparticles: toward cancerphototherapy J. Biomed. Opt. 14 021016

[6] Cherukuri P and Curley S A 2010 Use of nanoparticles fortargeted, noninvasive thermal destruction of malignant cellsMethods Mol. Biol. 624 359–73

[7] Hainfeld J F, Dilmanian F A, Slatkin D N andSmilowitz H M 2008 Radiotherapy enhancement with goldnanoparticles J. Pharm. Pharmacol. 60 977–85

[8] Castillo M H et al 1988 Effects of radiotherapy on mandibularreconstruction plates Am. J. Surg. 156 61–3

[9] Niroomand-rad A, Razavi A, Thobejane S andHarter K W 1996 Radiation dose perturbation attissue-titanium dental interfaces in head and neck cancerpatients Int. J. Radiat. Oncol. Biol. Phys. 34 475–80

[10] Li X A, Chu J C, Chen W and Zusag T 1999 Doseenhancement by a thin foil of high-Z material: a MonteCarlo study Med. Phys. 26 1245–51

[11] Regulla D F, Hieber L B and Seidenbusch M 1998 Physicaland biological interface dose effects in tissue due tox-ray-induced release of secondary radiation from metallicgold surfaces Radiat. Res. 150 92–100

[12] Kong T et al 2008 Enhancement of radiation cytotoxicity inbreast-cancer cells by localized attachment of goldnanoparticles Small 4 1537–43

[13] Zhang X et al 2008 Enhanced radiation sensitivity in prostatecancer by gold-nanoparticles Clin. Invest. Med. 31 E160-7

[14] Geng F et al 2011 Thiol-glucose bound gold nanoparticlesenhance radio-cytotoxic targeting of ovarian cancerNanotechnology 22 285101

[15] Roa W et al 2009 Gold-nanoparticles sensitize radiotherapy ofprostate cancer cells by regulation of cell cycleNanotechnology 20 375101

[16] Connor E E et al 2005 Gold nanoparticles are taken up byhuman cells but do not cause acute cytotoxicity Small1 325–7

[17] Chen J et al 2010 Gold nanocages as photothermal transducersfor cancer treatment Small 7 811–7

[18] Aydogan B et al 2010 Deoxyglucose-labeled goldnanoparticles as x-ray computed tomography contrastagents for cancer imaging Mol. Imaging Biol. 12 463–7

[19] Bustamante E and Pedersen P L 1977 High aerobic glycolysisof rat hepatoma cells in culture: role of mitochondrialhexokinase Proc. Natl Acad. Sci. USA 74 3735

[20] Landau B R et al 2007 6-Fluoro-6-deoxy-D-glucose as a tracerof glucose transpor Am. J. Physiol. Endocrinol. Metab.293 E237–45

[21] Perrault S D and Chan W C W 2009 Synthesis and surfacemodification of highly monodispersed, spherical goldnanoparticles of 50–200 nm J. Am. Chem. Soc. 131 17042

[22] Kim D, Park S, Lee J H, Jeong Y Y and Jon S 2007Antibiofouling polymer-coated gold nanoparticles as acontrast agent for in vivo x-ray computed tomographyimaging J. Am. Chem. Soc. 129 7661–5

[23] Wang C H et al 2008 Optimizing the size and surfaceproperties of polyethylene glycol (PEG)–gold nanoparticlesby intense x-ray irradiation J. Phys. D: Appl. Phys.41 195301

[24] Oyewumi M O, Yokel R A, Jay M, Coakley T andMumper R J 2004 Comparison of cell uptake,biodistribution and tumour retention of folate-coated andPEG-coated Gadolinium nanoparticles in tumour-bearingmice J. Control. Release 95 613–26

[25] Owens D E III and Peppas N A 2006 Opsonization,biodistribution, and pharmacokinetics of polymericnanoparticles Int. J. Pharm. 307 93–102

[26] Bergen J M et al 2006 Gold nanoparticles as a versatileplatform for optimizing physicochemical parameters fortargeted drug delivery Macromol. Biosci. 6 506–16

10