Analytical Biochemistry 523 (2017) 50e57
Contents lists available at ScienceDirect
Analytical Biochemistry journal homepage: www.elsevier.com/locate/yabio
Heterogeneous intratumoral distribution of gadolinium nanoparticles within U87 human glioblastoma xenografts unveiled by micro-PIXE imaging phane Roudeau a, b, Baptiste L'Homel a, b, Fre de ric Pouzoulet c, Asuncion Carmona a, b, *, Ste c d a, b Sarah Bonnet-Boissinot , Yolanda Prezado , Richard Ortega a
CNRS, IN2P3, CENBG, UMR 5797, 19 Chemin Du Solarium, 33170 Gradignan, France University of Bordeaux, CENBG, UMR 5797, F-33170 Gradignan, France Institut Curie, Translational Research Department, Experimental Radiotherapy Platform, 15 Rue Georges Cl emenceau, 91898 Orsay, France d CNRS, IN2P3, IMNC, UMR 8165, University of Paris 7 and 11, 5 Rue Georges Clemenceau, 91405 Orsay, France b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 25 November 2016 Received in revised form 9 February 2017 Accepted 15 February 2017 Available online 20 February 2017
Metallic nanoparticles have great potential in cancer radiotherapy as theranostic drugs since, they serve simultaneously as contrast agents for medical imaging and as radio-therapy sensitizers. As with other anticancer drugs, intratumoral diffusion is one of the main limiting factors for therapeutic efficiency. To date, a few reports have investigated the intratumoral distribution of metallic nanoparticles. The aim of this study was to determine the quantitative distribution of gadolinium (Gd) nanoparticles after direct intratumoral injection within U87 human glioblastoma tumors grafted in mice, using micro-PIXE (Particle Induced X-ray Emission) imaging. AGuIX (Activation and Guiding of Irradiation by X-ray) 3 nm particles composed of a polysiloxane network surrounded by gadolinium chelates were used. PIXE results indicate that the direct injection of Gd nanoparticles in tumors results in their heterogeneous diffusion, probably related to variations in tumor density. All tumor regions contain Gd, but with markedly different concentrations, with a more than 250-fold difference. Also Gd can diffuse to the healthy adjacent tissue. This study highlights the usefulness of mapping the distribution of metallic nanoparticles at the intratumoral level, and proposes PIXE as an imaging modality to probe the quantitative distribution of metallic nanoparticles in tumors from experimental animal models with micrometer resolution. © 2017 Elsevier Inc. All rights reserved.
Keywords: Gadolinium Nanoparticles PIXE Tumor
1. Introduction In recent years the use of high-Z (Au, Gd, Pt, Ag, etc.) nanoparticles (NPs) as potential tumor-selective radiosensitizers has been proposed as a breakthrough in cancer radiotherapy [1]. High-Z metallic NPs are capable of accumulating in cancer cells, increasing the radiation absorption coefficient in tumor tissues and leading to fewer adverse effects than conventional radiosensitizers. High-Z NPs can also be used as theranostic agents, combining anticancer therapeutic activity and diagnostic capability as contrast agents in medical imaging modalities [2,3]. High-Z metallic NPs have been
aires de Bordeaux Gradignan, 19 * Corresponding author. Centre d’Etudes Nucle chemin du Solarium, 33175 Gradignan, France. E-mail address:
[email protected] (A. Carmona). http://dx.doi.org/10.1016/j.ab.2017.02.010 0003-2697/© 2017 Elsevier Inc. All rights reserved.
predominantly used since they can generate short-range photoelectrons or Auger electrons, which can enhance the therapeutic dose locally. Despite the attractive potential for NP usage in cancer radiotherapy there have been very few clinical trials using NPs [4,5]. Preclinical studies are needed to determine the most efficient protocols among the multiple combinations that can be proposed. Efficient radio-sensitization with high-Z NPs depends on tumorselective delivery, accumulation, and uniform distribution in the tumor. Radio-sensitization can be modulated and improved by changing the NP type, size, surface charge, shape, vectorization, or the mode of administration [6]. It is of great interest to limit the diffusion of metallic NPs only to the tumor itself. Thus, current research attempts to optimize the size of NPs and/or use chelating ligands of high affinity for tumor cells. The mode in which NPs are delivered to the tumor, intravenously or by direct injection also determines their distribution within the tissue. The intratumoral
A. Carmona et al. / Analytical Biochemistry 523 (2017) 50e57
distribution of NPs is an important parameter to achieve efficient therapeutic and theranostic results. However, to date, few studies have reported metallic NP distribution in tumors [7e11]. The intratumoral distribution of Gd NPs has been reported after systemic injection [7,10], but no results have yet been reported after direct intratumoral delivery. The purpose of the present study was to use PIXE (Particle Induced X-ray Emission) imaging to determine the quantitative distribution of Gd NPs in a xenograft U87 human glioblastoma tumor after intratumoral injection. AGuIX NPs (Activation and Guiding of Irradiation by X-ray) were used as models of high-Z NPs in this study. AGuIX NPs consist of 3 nm polysiloxane Gd NPs and have good potential for as theranostic agents since they are simultaneously magnetic resonance imaging contrast agents and radiotherapy sensitizers [12e14]. Pre-clinical studies on animal models have shown increased tumor cell death and host survival in animals bearing multiple brain melanoma after the combined use of AGuIX and radiotherapy, providing proof-of-concept before human clinical trials [14]. PIXE imaging is based on the detection of X-rays emitted by a sample when irradiated with a MeV ion micro-beam, usually protons. This technique offers two important specific capabilities for trace metal analysis: imaging at submicron resolution and quantification with a detection limit in the mg.g 1 range [15]. PIXE is a multi-elemental technique that allows determining elements with atomic number higher than 11; the technique is more sensitive than an electron microprobe. For studying biological samples, PIXE can be simultaneously carried out with RBS (Rutherford Backscattering Spectrometry), a technique which provides the local sample mass. This multimodal micro-spectrometry enables expressing metal content in terms of mg of element per g of dry mass throughout the scanned area [16]. 2. Material and methods 2.1. Gadolinium nanoparticles The characteristics of AGuIX NPs have been previously reported [12,14]. In brief, AGuIX NPs are composed of a polysiloxane network surrounded by Gd chelates. The chemical composition of the AGuiX NPs is Gd10Si40C200N50O150Hx for a molar mass of 8.5 ± 1 kDa. The hydrodynamic diameter of the AGuIX NPs is 3 ± 0.1 nm; and the AGuIX NPs are characterized by a zeta potential of 9.03 ± 5.5 mV at pH 7.2. 2.2. Animal model Gd NP injection Five-week-old female NMRI nude mice from Janvier Labs were grafted with 4.106 U87 human glioblastoma cells (ATCC HT-B14) diluted in 40 mL PBS (phosphate-buffered saline) to produce a skin tumor on the right hind leg. The implantation was performed under general anesthesia with a mixture of oxygen and isoflourane at a concentration of 3% for induction and 1.5% for maintenance. After the glioblastoma tumors were grown for four weeks, the animals were injected with 50 mL AGuIX NPs at 100 mM directly into the tumor. Mice were housed on a 12 h light/dark cycle, in an environment with controlled temperature and humidity, with food and water available ad libitum. 2.3. Sample preparation for micro-PIXE imaging Fifty minutes after AGuiX NP injection, the xenograft tumors were extracted and cryofixed by rapid immersion into isopentane cooled with liquid nitrogen. This cryo-preparation preserves elemental distributions and tissue structure close to native
51
conditions, avoiding any chemical element redistribution or contamination that could be due to chemical fixatives [17]. Tumors were stored at 80 C and sections of 50 mm thickness centred on the injection point were prepared using a cryo-microtome. Cryosections were deposited onto sample holders designed for microPIXE analysis and freeze-dried according to previously published protocols [17]. Adjacent tissue sections were deposited onto coverslips for haematoxylin-eosin-saffron staining. 2.4. Micro-PIXE and RBS quantitative analyses Micro-PIXE and micro-RBS were performed simultaneously using the high resolution ion beamline at AIFIRA (Applications gion Aquitaine) facility Interdisciplinaires de Faisceaux d’Ions en Re aires de Bordeaux-Gradignan) [18]. CENBG (Center d’Etudes Nucle Protons were accelerated up to 3.0 MeV and focalized at 2 mm diameter, resulting in 600 pA beam intensity. A representative example of the PIXE spectrum is shown in Fig. 1A. PIXE analysis is a multi-elemental method that provides the quantitative content of chemical elements in the analyzed area, expressed in terms of areal mass of the elements (mg.cm 2). RBS allows the areal total mass determination of the sample (g.cm 2), and the accurate measurement of the locally received charge, the number of protons interacting with the sample during the analysis. A representative example of the RBS spectrum is illustrated in Fig. 1B. The RBS spectrum shows the main elemental composition of the biological matrix e carbon and oxygen e with a sharp peak for both elements corresponding to the 2 mm polycarbonate film from the sample holder, and a larger feature at lower energy corresponding to the tumor section. The combination of both PIXE and RBS techniques enables the determination of the elemental content per unit of sample dry mass (mg.g 1) [16]. PIXE and RBS data treatment were performed using GupixWin [19] and Simnra [20] software, respectively. To validate the quantification process using the combined PIXE and RBS methods, several thin calibration standards from MicromatterTM were analyzed. These standards consist of a thin film with an areal density of about 50 mg cm 2 of the certified element deposited on a 6 mm Mylar foil. In addition, a certified reference material, ERM-EC681K, from IRMM (Institute for Reference Materials and Measurements, Geel, Belgium) was analyzed to validate the quantitative analysis [17]. The combined PIXE and RBS standard uncertainty was calculated according to IAEA guidelines for quantifying uncertainty in nuclear analytical measurements [21]. The main contributions to the combined standard uncertainty were due to the determination of the peak areas in PIXE spectra and to the ±5% certified value of the MicromatterTM calibration standards. 2.5. Image reconstruction of element distributions For a 3.0 MeV proton beam, the maximum size of the scanned area achievable with the electrostatic beam deflection plates is a square of 740 740 mm2. Accordingly, in order to analyze entire tumor sections of about 5 mm diameter, a number of sequential square scans were performed. Scans were partially superimposed to facilitate image reconstruction. For each square scan, an X-ray spectrum was obtained. Using SupaVisio software [22] and selecting the peak area of the element of interest, an image of the corresponding element was obtained. Silicon, phosphorus, sulphur, chlorine, potassium, calcium, iron and gadolinium distribution maps were obtained from respective X-ray emission lines, Si Ka (1.74 keV), P Ka (2.01 keV), S Ka (2.30 keV), Cl Ka (2.62 keV), K Ka (3.31 keV), Ca Kb (4.01 keV), Fe Ka (6.40 keV) and Gd La (6.05 keV) (Fig. 1A). Copper and zinc were also detected and quantified in the samples, but their concentrations were too low to provide images
52
A. Carmona et al. / Analytical Biochemistry 523 (2017) 50e57
Fig. 1. (A) Representative PIXE spectrum from a tumor section after injection of Gd NPs. (B) RBS spectrum performed simultaneously with PIXE analysis.
of sufficient quality, so their element maps are not presented here. Each image of element distribution obtained from a single scan was divided by the received charge calculated from RBS data in order to normalize the intensity of all the scans. For each element, chargenormalized images of the selected element from all scans were produced with ImageJ software [23] and juxtaposed to construct the corresponding normalized image of the element for the entire tumor section. Sections of two tumors from two different animals were analyzed. Composite images were constructed by juxtaposing seventy-four 740 740 mm2 square scan analyses of for the first section (Fig. 2), and one hundred and six 500 500 mm2 square
scan analyses of for the second section (Fig. 3). Each square scan analysis was run for 30 min to enable trace element detection, requiring a total of 90 h of effective beamtime to scan both tumor sections.
3. Results and discussion Glioblastoma xenograft tumors injected with Gd NPs were analyzed by micro-PIXE and micro-RBS in order to image and quantify the chemical element distributions in the tumor.
Fig. 2. Si, P, S, Cl, K, Ca, Fe and Gd distribution from the first glioblastoma xenograft tumor. X-ray intensities for each element measured by PIXE (min-max) were normalized by the local deposited charge, measured by RBS. The X-ray intensity levels increase from black/purple to yellow/white, as shown in the color scale. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
A. Carmona et al. / Analytical Biochemistry 523 (2017) 50e57
53
Fig. 3. Si, P, S, Cl, K, Ca, Fe and Gd distributions from the second glioblastoma xenograft tumor section. X-ray emission intensities for each element measured by PIXE (min-max) were normalized by the local deposited charge, measured by RBS. The X-ray intensity levels increase from black/purple to yellow/white, as shown in the color scale. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
3.1. Gd and element distributions in glioblastoma xenograft tumors Figs. 2 and 3 show Si, P, S, Cl, K, Ca, Fe and Gd distributions in two sections of glioblastoma xenograft tumors from different animals. In both cases, the distributions of Gd and Si at the tumor level were found to be heterogenous. Both elements follow the same distribution pattern since AGuIX NPs are composed of a polysiloxane network surrounded by Gd chelates. Gd and Si are found at high concentrations near the injection point and decrease rapidly in the rest of the tumor. This heterogeneous intratumoral distribution may have important implications for the theranostic efficiency of metallic NPs, as discussed below. Iron is another element that displays heterogenous distribution in the tumor sections (Figs. 2 and 3), with a spot-shaped pattern in the center of the tumor (Fig. 2), and areas of concentration on the periphery of the tumor. A possible explanation for this distribution pattern could be the high Fe content of blood vessels. The distribution of Fe is highly heterogeneous, as could be predicted from the large heterogeneity and chaotic arrangement of the vascular system in human xenotransplanted tumors [24]. The Ca distribution was also heterogeneous in the tumor sections. Calcium is generally globally homogeneous, except for some regions with higher Ca content, often also rich in Gd. However, the distributions of Ca and Gd are not correlated in terms of intensity or exact co-localization, suggesting that Ca could be redistributed to some tumor areas in response to NP accumulation, possibly as the signature of local toxicity. The other elements, P, S, Cl, and K, are found quite uniformly distributed in the tumor sections; these elements correspond to the main inorganic components of biological tissues. However, their imaging shows some intratumoral structures. The distribution of K highlights intratumoral nodules of about 1 mm diameter delimiting cellular clusters of variable densities (Figs. 2 and 3). Gadolinium NPs accumulate around these intratumoral subdomains (Figs. 2 and 3). These intratumoral subdomains are also visible on the RBS images (Figs. 4B and 5B), which represent the organic mass distribution showing millimeter-sized nodules of slightly higher mass density co-localized with the structures highlighted by K maps
(Figs. 4C and 5C). The overlay images of K and Gd intratumoral distributions clearly illustrate the heterogeneous accumulation of Gd NPs from the injection point area to the rest of the tumor corresponding to the regions of lower K and mass densities. The comparison of the RBS and PIXE images with the histological sections (Figs. 4A and 5A) show that the higher signals for most of the endogenous elements (Figs. 2 and 3) are found in the skin layer surrounding the tumor. This can be explained by the higher matrix density of the skin. On the other hand, non-tumoral tissue was analyzed, as shown in Fig. 5, revealing the presence of Gd in the adjacent muscle tissue (Fig. 3). The distribution of Gd AGuiX NPs has been previously reported in non-tumoral tissues, in kidneys from experimental animals, to study the mechanisms of NP clearance after systemic injection [13,25,26]. In these studies, Gd, Na, Si and Fe distributions in kidney sections were determined by laser-induced breakdown spectroscopy (LIBS), a multi-elemental imaging method with 10 mm spatial resolution and a mg.g 1 detection limit. Gd distribution was found heterogeneous in the kidney cortex and was mainly located in the periphery of the kidney, where the initial stages of blood filtration occur [13,25,26]. A few studies have reported the metallic NP distribution in tumors. In 2010, laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) using a 25 mm diameter laser beam of liposomal Gd NPs injected intravenously to a mouse tumor model indicated heterogenous distribution for Gd, with higher Gd content in the more vascularized and peripheral regions [7]. In another study, imaging of Au NPs using micro-computed X-ray tomography was performed on brain tumors after intravenous injection of NPs in mice bearing orthotopic or subcateneous tumors [8]. With subcateneous tumors, the Au NPs were predominantly located on the periphery; in contrast, with orthotopic brain tumors, they were distributed throughout the tumors. Autoradiography of radiolabeled 89Zr NP tumor sections after intravenous administration in orthotopic mouse models of breast cancer showed heterogeneous distribution of the NPs with 89Zr deposition in vascularized regions [9]. The tumoral distribution of Gd NPs after intravenous injection was recently investigated with synchrotron X-ray fluorescence
54
A. Carmona et al. / Analytical Biochemistry 523 (2017) 50e57
Fig. 4. (A) Histological section of the first tumor; (B) RBS image of the matrix distribution and (C) overlay PIXE images of K (in red), Fe (in blue) and Gd (in green) distributions. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 5. (A) Histological section of the second tumor; (B) RBS image of the matrix distribution and (C) overlay PIXE images of K (in red), Fe (in blue) and Gd (in green) distributions. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
imaging, showing that Gd reaches tumors implanted in the brains of rats quite rapidly, helping to determine the best time lapse for radiotherapy after injection [10]. These four studies were performed after intravenous injection of the NPs. They all reported heterogeneous distribution of the metallic NPs, with some correlation with vascularization. In the present study, the distribution of Gd NPs was also heterogeneous. However, since the mode of administration was direct injection into the tumor, we did not observe a correlation with tumor vascularization.
concentration. In Gd-rich areas the mass ratio of Gd over Si ranges from about 1.1 to 1.8 corresponding to a molar ratio for Gd/Si ranging from 0.20 to 0.33. This is close to the expected ratio in
3.2. Quantitative analysis of Gd heterogeneous distribution The combination of PIXE and RBS offers the possibility of calculating element concentrations expressed in mg.g 1 of dry tissue [16]. We quantified elemental content in different areas of each tumor section (Figs. 6 and 7): areas within the tumor far from the injection point, areas close to the injection point but showing a high degree of Gd heterogeneity, and areas close to the injection point showing the highest Gd content. Quantitative data for each of these selected regions are presented in Figs. 8 and 9. Gadolinium quantitative distribution in the tumor is highly heterogeneous; 57% of Gd is accumulated in 13% of the tumor surface. Gadolinium minimum concentrations in the two analyzed samples are 37 and 61 mg g 1, respectively; maximal concentrations are 10700 and 10200 mg g 1, respectively. Gd NPs are present throughout the tumor sections and no Gd-free areas could be identified. Some other element concentrations are affected by the Gd NP distribution. Silicon increases proportionally to Gd
Fig. 6. Images of K and Gd distribution and selected regions for quantitative analysis in the first tumor section (see Fig. 8 for quantitative multi-element concentrations). Regions 1 and 2 correspond to tumor areas far from the injection point, whereas regions 3 and 4 are close to the injection point. Region 3 is centred at the border between low and high Gd concentrations, as illustrated in the zoomed area to determine element concentrations in sub-regions 3a and 3b.
A. Carmona et al. / Analytical Biochemistry 523 (2017) 50e57
Fig. 7. Images of K and Gd distribution and selected regions for quantitative analysis in the second tumor section (see Fig. 9). Regions 5 and 6 correspond to tumor areas far from the injection point, whereas regions 7 and 8 are close to the injection point. Region 8 is centred at the edge between low and high Gd concentrations, as illustrated in the zoomed area to determine element concentrations in sub-regions 8a and 8b. Region 9 is within the muscle tissue.
AGuiX NPs, which contain 3.8 atoms of Si per atom of Gd [12]. In Gd-poor regions the Gd/Si ratio is much lower, indicating that the tumor tissue contains endogenous trace levels of Si. P, S and K concentrations which decrease when Gd concentrations increase (Figs. 8 and 9). This trend was also suggested by the K PIXE images showing a depletion of K in areas of Gd maximal distribution, as clearly illustrated in the zoomed areas of both tumor sections (Figs. 6 and 7). P, S and K are mainly intracellular elements, suggesting that in Gd-rich areas the cellular density is slightly lower, about 10e20% for the K, P or S concentrations in Gd-poor vs Gd-rich regions. Potassium images enable visualizing subdomains in the
55
tumor, arranged in cellular nodules. Loss of these elements could be attributed to either the toxic effects of Gd NPs when they accumulate at very high concentrations producing cell loss, and/or to the preferential diffusion of Gd NPs within regions of lower cellular density. Calcium concentrations increase in some, but not all, Gdrich areas, as also evidenced by Ca imaging (Figs. 2 and 3). Zn and Cu are detected with the same concentrations in all analyzed regions. These two elements are important intracellular elements and are also involved in the extracellular matrix, which might explain their homogeneous content in the tumor sections independent of the cellular density. Finally, Gd NPs were detected outside the tumor, within the muscle tissue, at higher concentrations than in some tumor regions far from the injection point. Overall, these quantitative data suggest that the direct injection of Gd NPs in U87 human glioblastoma xenografts results in the heterogeneous diffusion of Gd NPs within the tumor mass. All the regions of the tumor contain Gd NPs but with markedly different concentrations, and the diffusion of Gd NPs is not restricted to the tumor since it can diffuse into the normal adjacent tissue. After intravenous injection Gd NPs are also detected outside the tumors in the contralateral region of rats implanted with brain tumors, as evidenced by synchrotron X-ray fluorescence imaging [10]. This study revealed that the Gd content is much lower in healthy tissue than in the tumor. Both after intratumoral or systemic delivery, the undesired Gd distribution outside the tumor could exert secondary toxicity to healthy tissue. By quantifying the amount of metallic NPs outside the tumor, micro-analytical methods can be very helpful for evaluating the benefits and risks related to the different modalities of administration. 4. Conclusions Considerable research efforts have recently been dedicated to the development of metallic NPs to be used as chemotherapeutic or theranostic agents. The inability to deliver anticancer agents throughout the entire tumor mass strongly limits the chemotherapeutic efficacy in most cancers, including primary brain tumors. The use of metallic NPs with improved tumor penetration has been proposed to overcome this issue [6,27,28]. The delivery mode of
Fig. 8. Element concentration, expressed in mg.g 1 of dry mass and combined standard uncertainty; in the selected regions of the first tumor section (as illustrated in Fig. 6), quantification was performed combining PIXE and RBS.
56
A. Carmona et al. / Analytical Biochemistry 523 (2017) 50e57
Fig. 9. Element concentration, expressed in mg.g 1 of dry mass and combined standard uncertainty; in the selected regions of the second tumor section (as illustrated in Fig. 7) quantification was performed combining PIXE and RBS.
NPs strongly affects their distribution in the tumor as well as in healthy tissues. When NPs are applied intravenously, diffusion in the tumor is limited by non-uniform delivery via blood vessels [7e9], by the rapid clearance of the NPs [13] and by the systemic toxicity towards normal tissues due to the broad distribution of NPs in the entire body [2]. On the other hand, direct intratumoral injection fails to reach metastatic lesions, and is not adapted to the treatment of tumors deeply located in the body. Reported metallic NP distributions in tumors after systemic injection have shown the accumulation of NPs in the periphery of the tumor, indicating some limitations in intratumoral diffusion [10]. Our data obtained after direct intratumoral injection show that Gd NP diffusion is also highly heterogeneous in the tumor, but with higher localization close to the injection point. Therefore, both routes of NP administration, systemic or intratumoral, fail to reach the tumor. The targeting of cancer cells is hindered by limited interstitial diffusion of NPs into the tumor independent of the injection route. These results must be considered for several reasons. First, efforts should be undertaken to improve NP diffusion into the tumors by modifying, for example, the type of ligands, the size of the NPs, or by proposing new modalities such as using ultrasound to reduce intratumoral pressure during NP delivery [29]. Second, heterogeneous intratumoral distribution should be also considered in simulations of dose-enhancement for radiotherapy treatments. Metallic NPs show a great potential for tumor enhanced radiotherapy in animal models [10,11]. A better understanding of NP intratumoral distribution could help to improve simulation radiation parameters. Micro-PIXE and other chemical element imaging methods, such as LIBS [13,25,26] mass spectrometry-based techniques [30,31] or synchrotron radiation X-ray fluorescence microscopy [10,32] are good analytical techniques for imaging NP distribution in tumors and should be more widely used to determine the level of homogeneity of NP distribution in tumors. The micro-PIXE images presented in this study show chemical element distributions from multi-millimeter-sized samples acquired with a micrometric resolution. The multi-element and multi-scale imaging capabilities of micro-PIXE are assets for explaining the fine interactions between NP distribution and tumor heterogeneity, such as illustrated by this
study revealing intratumoral cellular nodules of variable densities and Gd NP limited diffusion inside these tumor sub-domains. Imaging NP distribution in tumors using quantitative micro-analytical techniques, such as micro-PIXE, could help evaluate quantitatively the efficiency of NP diffusion throughout the entire tumor to validate new therapeutic vectors and delivery approaches. Acknowledgments This project was funded in part by IN2P3. The authors are grateful to AIFIRA facility staff for technical support. We acknowledge Dr. Erika Mitchell, Mitchell Cameron, and Dr. Seth Frisbie for reading the manuscript and providing useful comments. References [1] J.F. Hainfeld, D.N. Slatkin, H.M. Smilowitz, The use of gold nanoparticles to enhance radiotherapy in mice, Phys. Med. Biol. 49 (2004) 309e315, http:// dx.doi.org/10.1088/0031-9155/49/18/N03. [2] W.T. Phillips, A. Bao, A.J. Brenner, B.A. Goins, Image-guided interventional therapy for cancer with radiotherapeutic nanoparticles, Adv. Drug Deliv. Rev. 76 (2014) 39e59, http://dx.doi.org/10.1016/j.addr.2014.07.001. [3] H. Sharma, P.K. Mishra, S. Talegaonkar, B. Vaidya, Metal nanoparticles: a theranostic nanotool against cancer, Drug Discov. Today 20 (2015) 1143e1151, http://dx.doi.org/10.1016/j.drudis.2015.05.009. , A. Vallard, J.B. Guy, C. Rodriguez-Lafrasse, E. Deutsch, [4] C. Rancoule, N. Magne C. Chargari, Nanoparticles in radiation oncology: from bench-side to bedside, Cancer Lett. 375 (2016) 256e262, http://dx.doi.org/10.1016/ j.canlet.2016.03.011. [5] H.S. Choi, J.V. Frangioni, Nanoparticles for biomedical imaging: fundamentals of clinical translation, Mol. Imaging 9 (2010) 291e310, http://dx.doi.org/ 10.2310/7290.2010.00031. [6] M.O. Durymanov, A.A. Rosenkranz, A.S. Sobolev, Current approaches for improving intratumoral accumulation and distribution of nanomedicines, Theranostics 5 (2015) 1007e1020, http://dx.doi.org/10.7150/thno.11742. [7] N. Kamaly, J.A. Pugh, T.L. Kalber, J. Bunch, A.D. Miller, C.W. McLeod, J.D. Bell, Imaging of gadolinium spatial distribution in tumor tissue by laser ablation inductively coupled plasma mass spectrometry, Mol. Imaging Biol. 12 (2010) 361e366, http://dx.doi.org/10.1007/s11307-009-0282-4. [8] J.F. Hainfeld, H.M. Smilowitz, M.J.O. Connor, F. Avraham, D.N. Slatkin, Gold nanoparticle imaging and radiotherapy of brain tumors in mice, Nanomedicine 8 (2013) 1601e1609, http://dx.doi.org/10.2217/nnm.12.165.Gold. rez-Medina, J. Tang, D. Abdel-Atti, B. Hogstad, M. Merad, E.A. Fisher, [9] C. Pe Z.A. Fayad, J.S. Lewis, W.J.M. Mulder, T. Reiner, PET imaging of tumorassociated macrophages with 89Zr-labeled high-density lipoprotein nanoparticles, J. Nucl. Med. 56 (2015) 1272e1277, http://dx.doi.org/10.2967/
A. Carmona et al. / Analytical Biochemistry 523 (2017) 50e57 jnumed.115.158956. , V. Bentivegna, L. Sancey, E. Bra €uer-Krisch, [10] S. Dufort, G. Le Duc, M. Salome H. Requardt, F. Lux, J.L. Coll, P. Perriat, S. Roux, O. Tillement, The high radiosensitizing efficiency of a trace of gadolinium-based nanoparticles in tumors, Sci. Rep. 6 (2016) 29678, http://dx.doi.org/10.1038/srep29678. [11] A. Detappe, S. Kunjachan, L. Sancey, V. Motto-Ros, D. Biancur, P. Drane, R. Guieze, G.M. Makrigiorgos, O. Tillement, R. Langer, R. Berbeco, Advanced multimodal nanoparticles delay tumor progression with clinical radiation therapy, J. Control. Release 238 (2016) 103e113, http://dx.doi.org/10.1016/ j.jconrel.2016.07.021. [12] G. Le Duc, S. Roux, A. Paruta-Tuarez, S. Dufort, E. Brauer, A. Marais, C. Truillet, L. Sancey, P. Perriat, F. Lux, O. Tillement, Advantages of gadolinium based ultrasmall nanoparticles vs molecular gadolinium chelates for radiotherapy guided by MRI for glioma treatment, Cancer Nanotechnol. 5 (2014) 4, http:// dx.doi.org/10.1186/s12645-014-0004-8. [13] L. Sancey, S. Kotb, C. Truillet, F. Appaix, A. Marais, E. Thomas, B. Van Der Sanden, J.P. Klein, B. Laurent, M. Cottier, R. Antoine, P. Dugourd, G. Panczer, F. Lux, P. Perriat, V. Motto-Ros, O. Tillement, Long-term in Vivo clearance of gadolinium-based AGuIX nanoparticles and their biocompatibility after systemic injection, ACS Nano 9 (2015) 2477e2488, http://dx.doi.org/10.1021/ acsnano.5b00552. [14] S. Kotb, A. Detappe, F. Lux, F. Appaix, E.L. Barbier, V.L. Tran, M. Plissonneau, H. Gehan, F. Lefranc, C. Rodriguez-Lafrasse, C. Verry, R. Berbeco, O. Tillement, L. Sancey, Gadolinium-based nanoparticles and radiation therapy for multiple brain melanoma metastases: proof of concept before phase I trial, Theranostics 6 (2016) 418e427, http://dx.doi.org/10.7150/thno.14018. [15] R. Ortega, A. Carmona, I. Llorens, P.L. Solari, X-ray absorption spectroscopy of biological samples. A tutorial, J. Anal. Atomic Spectrom. 27 (2012) 2054e2065, http://dx.doi.org/10.1039/c2ja30224a. s, R. Ortega, Quantitative micro-analysis of metal ions in [16] A. Carmona, G. Deve subcellular compartments of cultured dopaminergic cells by combination of three ion beam techniques, Anal. Bioanal. Chem. 390 (2008) 1585e1594, http://dx.doi.org/10.1007/s00216-008-1866-6. [17] L. Perrin, A. Carmona, S. Roudeau, R. Ortega, Evaluation of sample preparation methods for single cell quantitative elemental imaging using proton or synchrotron radiation focused beams, J. Anal. Atomic Spectrom. 30 (2015), http:// dx.doi.org/10.1039/c5ja00303b. [18] S. Sorieul, P. Alfaurt, L. Daudin, L. Serani, P. Moretto, Aifira: an ion beam facility for multidisciplinary research, Nucl. Instrum. Methods Phys. Res. Sect. B Beam Interact. Mater. Atoms 332 (2014) 68e73, http://dx.doi.org/10.1016/ j.nimb.2014.02.032. [19] J.L. Campbell, N.I. Boyd, N. Grassi, P. Bonnick, J.A. Maxwell, The Guelph PIXE software package IV, Nucl. Instrum. Methods Phys. Res. Sect. B Beam Interact. Mater. Atoms 268 (2010) 3356e3363, http://dx.doi.org/10.1016/ j.nimb.2010.07.012. [20] M. Mayer, SIMNRA User's Guide, Max-Planck-Institut Für Plasmaphysik,
57
Garching, Germany, 1997. [21] International Atomic Energy Agency, Quantifying uncertainty in nuclear analytical measurements, IAEA-TECDOC-1401, IAEA, Vienna, 2004. ISBN: 920-108404-8. [22] Barbotteau, SupaVisio, (n.d.). http://supavisio.software.informer.com/. [23] Rasband, ImageJ, U.S.National Institutes of Health, Bethesda, Maryland USA, 2016. http://imagej.nih.gov/ij/. [24] F. Steinberg, M.A. Konerding, C. Streffer, The vascular architecture of human xenotransplanted tumors: histological, morphometrical, and ultrastructural studies, J. Cancer Res. Clin. Oncol. 116 (1990) 517e524, http://dx.doi.org/ 10.1007/BF01613005. [25] L. Sancey, V. Motto-Ros, B. Busser, S. Kotb, J.M. Benoit, A. Piednoir, F. Lux, O. Tillement, G. Panczer, J. Yu, Laser spectrometry for multi-elemental imaging of biological tissues, Sci. Rep. 4 (2014) 6065, http://dx.doi.org/10.1038/ srep06065. [26] Y. Gimenez, B. Busser, F. Trichard, A. Kulesza, J.M. Laurent, V. Zaun, F. Lux, J.M. Benoit, G. Panczer, P. Dugourd, O. Tillement, F. Pelascini, L. Sancey, V. Motto-Ros, 3D imaging of nanoparticle distribution in biological tissue by laser-induced breakdown spectroscopy, Sci. Rep. 6 (2016) 29936, http:// dx.doi.org/10.1038/srep29936. [27] V.P. Chauhan, T. Stylianopoulos, J.D. Martin, Z. Popovi c, O. Chen, W.S. Kamoun, M.G. Bawendi, D. Fukumura, R.K. Jain, Normalization of tumour blood vessels improves the delivery of nanomedicines in a size-dependent manner, Nat. Nanotechnol. 7 (2012) 383e388, http://dx.doi.org/10.1038/nnano.2012.45. [28] E. Nance, C. Zhang, T.Y. Shih, Q. Xu, B.S. Schuster, J. Hanes, Brain-penetrating nanoparticles improve paclitaxel efficacy in malignant glioma following local administration, ACS Nano 8 (2014) 10655e10664, http://dx.doi.org/10.1021/ nn504210g. [29] K.D. Watson, C.Y. Lai, S. Qin, D.E. Kruse, Y.C. Lin, J.W. Seo, R.D. Cardiff, L.M. Mahakian, J. Beegle, E.S. Ingham, F.R. Curry, R.K. Reed, K.W. Ferrara, Ultrasound increases nanoparticle delivery by reducing intratumoral pressure and increasing transport in epithelial and epithelial-mesenchymal transition tumors, Cancer Res. 72 (2012) 1485e1493, http://dx.doi.org/10.1158/00085472.CAN-11-3232. [30] A. Tata, J. Zheng, H.J. Ginsberg, D.A. Jaffray, D.R. Ifa, A. Zarrine-Afsar, Contrast agent mass spectrometry imaging reveals tumor heterogeneity, Anal. Chem. 87 (2015) 7683e7689, http://dx.doi.org/10.1021/acs.analchem.5b01992. [31] S. Theiner, E. Schreiber-Brynzak, M.A. Jakupec, M. Galanski, G. Koellensperger, B.K. Keppler, LA-ICP-MS imaging in multicellular tumor spheroids e a novel tool in preclinical development of metal-based anticancer drugs Metallomics Accepted Manuscript, Metallomics 8 (2016) 398e402, http://dx.doi.org/ 10.1039/C5MT00276A. [32] T. Liu, I. Kempson, M. de Jonge, D.L. Howard, B. Thierry, Quantitative synchrotron X-ray fluorescence study of the penetration of transferrinconjugated gold nanoparticles inside model tumour tissues, Nanoscale 6 (2014) 9774e9782, http://dx.doi.org/10.1039/c4nr02100b.