Optical and MRI Multifunctional Nanoprobe for Targeting Gliomas

Surgery, UniVersity of Washington School of Medicine, Seattle, Washington 98104, and. Fred Hutchinson Cancer Research Center, Seattle, Washington 9810...
0 downloads 0 Views 786KB Size
Download PDF
NANO LETTERS

Optical and MRI Multifunctional Nanoprobe for Targeting Gliomas

2005 Vol. 5, No. 6 1003-1008

Omid Veiseh,† Conroy Sun,† Jonathan Gunn,† Nathan Kohler,† Patrik Gabikian,§ Donghoon Lee,‡ Narayan Bhattarai,† Richard Ellenbogen,§ Raymond Sze,‡ Andrew Hallahan,| Jim Olson,| and Miqin Zhang*,†,§ Department of Materials Science & Engineering and Department of Radiology, UniVersity of Washington, Seattle, Washington 98195, Department of Neurological Surgery, UniVersity of Washington School of Medicine, Seattle, Washington 98104, and Fred Hutchinson Cancer Research Center, Seattle, Washington 98109 Received February 9, 2005; Revised Manuscript Received March 3, 2005

ABSTRACT A multifunctional nanoprobe capable of targeting glioma cells, detectable by both magnetic resonance imaging and fluorescence microscopy, was developed. The nanoprobe was synthesized by coating iron oxide nanoparticles with covalently bound bifunctional poly(ethylene glycol) (PEG) polymer, which were subsequently functionalized with chlorotoxin and the near-infrared fluorescing molecule Cy5.5. Both MR imaging and fluorescence microscopy showed significant preferential uptake of the nanoparticle conjugates by glioma cells. Such a nanoprobe could potentially be used to image resections of glioma brain tumors in real time and to correlate preoperative diagnostic images with intraoperative pathology at cellular-level resolution.

Gliomas are currently the most common and lethal type of primary brain tumor.1 The effectiveness of tumor resection in neurosurgery is severely limited by the poor visual contrast between neoplastic and normal brain tissue.2,3 Gross tumor resection restricts both patient quality of life and patient survival,4 whereas limited resection leaves residual glioma cells that have the ability to proliferate and migrate rapidly. Various targeting ligands have been labeled with fluorophores to delineate tumor margins to improve surgical outcome.5,6 However, using these probes, preoperative diagnostic images cannot be correlated with intraoperative pathology. This has sparked new research geared toward the development of multimodal probes that can be detected with both magnetic resonance imaging (MRI) and intraoperative optical devices.3 Currently, the major limitation of the multimodal probes is their low specificity and limited internalization by glioma cells. Here we report the fabrication of a multifunctional nanoprobe capable of targeting glioma cells, detectable by both MRI and fluorescence imaging. The probe was fabricated by coating iron oxide nanoparticles (NPs) with covalently bound bifunctional poly(ethylene glycol) (PEG) * Corresponding author. E-mail: [email protected]. Tel: (206) 616 9356. Fax: (206) 543 3100. † Department of Materials Science & Engineering, University of Washington. ‡ Department of Radiology, University of Washington. § Department of Neurological Surgery, University of Washington School of Medicine. | Fred Hutchinson Cancer Research Center. 10.1021/nl0502569 CCC: $30.25 Published on Web 04/30/2005

© 2005 American Chemical Society

polymers that were subsequently functionalized with chlorotoxin (Cltx), a glioma tumor-targeting molecule, and the near-infrared fluorescing (NIRF) molecule Cy5.5. Unlike other targeting ligands that are specific only to certain types of glioma cells (e.g., epidermal growth factor receptor and vascular endothelial growth factor ligands), Cltx is a unique peptide shown to specifically target the vast majority of glioma tumors.3 Cltx is a small 36-amino acid peptide purified from the venom of the giant Israeli scorpion (Leiurus quinquestriatus). This peptide has been shown to bind with high affinity to the membrane-bound matrix metalloproteinase-2 (MMP-2) endopeptidase, which is preferentially upregulated in gliomas, medulloblastomas, and other tumors of the neuroectodermal origin.7-9 The use of small peptides such as Cltx overcomes the limitations of widely used antibodies that are bulky and exhibit limited tissue penetration and cellular uptake when introduced in vivo.10 Conjugating small peptides on NPs can also overcome the limitation of the short half-life in blood and poor tissue retention generally associated with the peptides.11 The use of NIRF molecules minimizes autofluorescence interference from healthy brain tissue and allows the visualization of tissues millimeters in depth because of the efficient penetration of photons in the near-infrared range.12 PEG coatings were used to prevent nanoparticle agglomeration and protein adsorption, thus increasing particle blood circulation time and the efficiency of their internalization by targeted cells when introduced in vivo.

Figure 1. (Left) TEM image of as-synthesized nanoparticles. (Right) X-ray diffraction pattern of nanoparticles.

Iron oxide (Fe3O4) nanoparticles were synthesized via a coprecipitation process of iron chloride and sodium hydroxide. A 1.5 M sodium hydroxide solution was added dropwise to a deoxygenated solution of iron chloride with an Fe(II)/ Fe(III) molar ratio of 0.5 under mechanical stirring and ultrasonication. The precipitation of nanoparticles occurred at a pH of 12. The resulting black precipitate was then isolated with a rare-earth magnet and washed with deionized water until a pH of 10.5 was reached. The morphology and size distribution of nanoparticles were examined by transmission electron microscopy (Philips CM 100 TEM) at an accelerating voltage of 80 kV. TEM sample grids were prepared by dipping 300 mesh silicon-monoxide support films in aqueous suspension of Fe3O4 nanoparticles. The samples were air dried for 24 h prior to analysis. Figure 1A shows a representative image of the prepared nanoparticles. The nanoparticles are well dispersed and uniform in shape and size. Statistical analysis of the TEM micrograph yielded a nanoparticle size of 10.5 ( 1.5 nm. The X-ray diffraction spectrum of the nanoparticles is shown in Figure 1B and was collected on a Phillips PW1820 diffractometer with Cu KR radiation (λ ) 1.541 Å; 40 kV, 20 mA, and 25° < 2θ < 65°). The XRD diffraction pattern of the nanoparticles matches the pattern for the magnetite listed in the ASTM XRD standard card (19-0629), confirming the crystalline structure of the magnetite nanoparticles. The multifunctional nanoprobe, namely, the nanoparticlePEG-Cltx-Cy5.5 (NPC-Cy5.5) conjugate, was synthesized as illustrated in Figure 2. The top panel outlines the overall process that consists of three major steps with each step detailed in the lower panels (labeled as 1, 2, and 3). NPs were first modified with trifluoroethylester terminal PEG silane, which was then converted to an amine-terminated PEG silane (Figure 2-1) following a method reported previously.13 Monofunctional N-hydroxysuccinimide (NHS) esters of Cy5.5 were then utilized to attach the Cy5.5 to the PEG-coated nanoparticles through reaction with the terminal amine (Figure 2-2). Specifically, 1 mg of Cy5.5 NHS ester was dissolved in 100 µL of anhydrous dimethyl formamide (Sigma, St. Louis, MO) and added to amine terminal PEGcoated nanoparticles (5 mg of Fe). The suspension was vortexed and placed on a shaker for 2 h. The remaining terminal amines of the PEG-coated nanoparticles (Figure 2-2) were conjugated with Cltx, as shown in Figure 2-3. This was realized by (1) reacting the amines of PEG-coated nanoparticles with a heterobifunctional linker, succinimidyl iodoacetate (SIA), (2) modifying Cltx with 1004

heterobifunctional linker N-succinimidyl-S-acetylthioacetate (SATA) to render sulfhydryl groups, and (3) conjugating iodoacetate-derivatized nanoparticles with sulfhydryl-modified Cltx. Using this reaction scheme, Cltx is attached to the nanoparticle through a stable thioether bond that is not susceptible to breakage under reducing conditions.14 To create iodoacetate-derivatized nanoparticles, 10 mg of succinimidyl iodoacetate (SIA, Molecular Biosciences, Boulder, CO) was dissolved in 1 mL of anhydrous dimethyl sulfoxide (DMSO, Sigma, St. Louis, MO), and the resultant solution was then added to the nanoparticle suspension. The solution was protected from light and placed on a shaker for an additional hour at room temperature. After the reaction was complete, the excess dye and SIA were removed from the suspension through gel filtration chromatography using PD10 desalting columns (Amersham Biosciences, Uppsala, Sweden) equilibrated with 20 mM sodium citrate buffer at pH 8.0. The resulting iodoacetyl-functionalized PEG-coated nanoparticles were readily able to react with free sulfhydryl groups at pH values between 7 and 9 to produce a thioether linkage. To functionalize Cltx with sulfhydryl groups, we prepared a stock solution of Cltx by dissolving 200 µg of Cltx in 200 µL of 50 mM bicarbonate buffer at pH 8.5. A solution of SATA (Molecular Bioscience, Boulder, CO) in anhydrous DMSO was prepared at a concentration of 1 mg/mL.The SATA solution (8 µL) was then added to the stock Cltx solution, and the mixture was allowed to react for 3-4 h at 4 °C. Following the reaction of SATA with Cltx, the excess SATA was removed by dialysis against 50 mM bicarbonate buffer at pH 8.0 using regenerated cellulose membranes (3500 MWCO) in an equilibrium dialyzer (Spectrum Laboratories, Los Angeles, CA). Following dialysis, the thiol groups of SATA-modified Cltx were deprotected by adding 30 µL of 25 mM hydroxylamine and 10 mM EDTA solution. The Cltx solution was then incubated for 1 h at room temperature. The resulting sulfhydryl-modified Cltx was then added to the iodoacetate-modified particles at a molecular ratio of 50 to 1 and shaken overnight in an ice bath. Unreacted Cltx was removed from the suspension through gel filtration chromatography using PD10 desalting columns equilibrated with 20 mM sodium citrate buffer at pH 8.0. The degree of Cy5.5 labeling of NPs was controlled through stoichiometry and reaction conditions and quantified by fluorescence spectroscopy. The emission intensity of a dilute sample of nanoparticle-Cy5.5 (50 µg of Fe/mL) at 689 nm was compared to a linear standard prepared using various concentrations of Cy5.5. The number of NPs was calculated on the assumption that each 10 nm of NP (as determined from TEM analysis, Figure 1A) had a volume of 5.236 × 10-25 m3 and a density of 5.2 kg/m3 based on the determined Fe3O4 crystal structure as identified by X-ray diffraction (Figure 1B). Using this information, we determined the mass of a NP to be 2.728 × 10-18 g, and the reaction yielded 1.22 fluorophores per NP. Similarly, the number of Cltx peptides linked to each NP was quantified using the bicinchoninic acid (BCA) assay (Pierce Biotechnology, Rockford, IL), and NP concentrations were deterNano Lett., Vol. 5, No. 6, 2005

Figure 2. Schematic diagram and reaction scheme for surface modification and preparation of NPC and NPC-Cy5.5 conjugates.

mined by inductively couple plasma atomic emission spectroscopy. From this analysis, the average number of Cltx molecules per NP was determined to be 10.2. Confocal fluorescence microscopy and MR imaging were utilized to monitor the cellular uptake of nanoparticle conjugates. Rat cardiomyocytes (rCM, Cell Applications, San Diego, CA) were grown in rat cardiomyocyte cell culture Nano Lett., Vol. 5, No. 6, 2005

media. 9L glioma cells (American Type Culture Collection, Manassas, VA) were grown in DMEM medium formulation with high glucose supplemented with sodium pyruvate, 1% streptomycin/penicillin, and 10% FBS (Invitrogen, Carlsbad, CA). Trypan blue staining was used to determine cell density and viability, and cell counts were obtained using a hemocytometer. 1005

Figure 3. Confocal fluorescent images of 9L cells cultured with (A) control NP-Cy5.5 and (B) NPC-Cy5.5. (C) MR phantom image of 9L cells cultured with NPCs and control NPs and embedded in agarose (4.7 T, spin-echo pulse sequence, TR 3000 ms, TE 15 ms).

For confocal imaging experiments, 2 × 105 cells were seeded on cover slips 24 h prior to labeling and staining. Cells were cultured with nanoparticle conjugates for 1 h in a 37 °C humidified incubator maintained at 5% CO2. Following labeling, the cover slips were washed twice with cell culture medium and twice with PBS buffer. After washes, cellular membranes were stained with FM 1-43FX (Molecular Probes, Eugene, OR). Cells were incubated in a 1 µM solution of FM 1-43FX for 20 min at room temperature and washed twice with PBS. The cells were then fixed using a 4% paraformaldehyde solution. Following fixation, cellular nuclei were stained with 4′,6-diamidino-2-phenyindole (DAPI) according to the manufacturers’ instructions (Sigma Aldrich, St. Louis, MO). Confocal images were acquired using a DeltaVision SA3.1 wide-field deconvolution microscope (Applied Precision, Inc., Issaquah, WA) equipped with DAPI, TRITC, and Cy5 filters. Image processing was performed using SoftWoRx (Applied Precision, Inc., Issaquah, WA). Using this software, the fluorescence intensity of the images was normalized. For MR phantom imaging, samples were prepared by suspending 106 cells in 50 µL of 1% low-melting agarose (BioRad, Hercules, CA). Cell suspensions were loaded into a prefabricated 12-well agarose sample holder and allowed to solidify at 4 °C. MR imaging was performed using a 4.7-T Varian MR spectrometer (Varian, Inc., Palo Alto, CA) and a Bruker magnet (Bruker Medical Systems, Germany) equipped with a 5-cm volume coil. A spin-echo multisection pulse sequence was selected to acquire MR phantom images. A repetition time (TR) of 3000 ms and variable echo times (TE) of 15-90 ms were used. The spatial resolution parameters were as follows: an acquisition matrix of 256 × 128, field of view of 4 × 4 cm, section thickness of 1 mm, and two averages. Regions of interest (ROIs) of 5.0 mm in diameter were placed in the center of each sample image to obtain signal intensity measurements using NIH ImageJ. T2 values were obtained using the VnmrJ “t2” fit program to generate a T2 map of the acquired images. To evaluate the specificity of NPC-Cy5.5 for glioma cells, 9L cells were cultured with NP-Cy5.5 and NPC-Cy5.5, and their fluorescence confocal images are shown in Figure 1006

3A and B, respectively. In these fluorescence images and the remaining fluorescence images, the cellular membrane, nuclei, and NPC-Cy5.5 are green, blue, and red, respectively. 9L cells cultured with NPC-Cy5.5 (Figure 3B) took up a substantial amount of NPC-Cy5.5 as clearly identified by IR fluorescence signals, whereas those cultured with NPCy5.5 (Figure 3A) took up virtually no NP-Cy5.5. These results confirmed (1) that the nanoparticle-Cltx conjugates exhibit a strong targeting role to glioma tumor cells and (2) that the cellular uptake of the conjugates can be visualized by fluorescence imaging at the cellular level. Figure 3C shows an MR phantom image of 9L cells cultured with NPCs (top) and control NPs (bottom). 9L cells cultured with NPCs show a much greater negative contrast than the cells cultured with the control NPs. The corresponding T2 relaxation time of 9L cells with NPCs and control NPs were 5 ms and 95 ms, respectively. This further confirmed the specific targeting role of the nanoparticle-Cltx conjugates to glioma cells and that the internalization of these conjugates by glioma cells is readily detectable by MRI through both imaging and T2 relaxation. To evaluate the specificity of NPC-Cy5.5 for 9L cells versus noncancerous freshly isolated cells (control cells) that lack MMP-2 expression, NPC-Cy5.5s at a concentration of 100 µg Fe/mL were incubated with 2 × 105 9L and rat cardiomyocyte (rCM) cells, respectively. Parts A and B of Figure 4 show the confocal fluorescence images of rCM and 9L cells, respectively, cultured with NPC-Cy5.5. The images show that the 9L cells have taken up notably higher amounts of NPC-Cy5.5 than the rCM cells. The nanoparticles appeared to be in cytoplasm surrounding the nuclei. Figure 4C shows a MR phantom image of 9L (top) and rCM (bottom) cells incubated with NPCs. The MRI results are consistent with those obtained through confocal imaging. rCM cells were barely detectable from the agarose background, whereas 9L cells showed dramatically preferential uptake of NPC versus rCM cells. The corresponding T2 relaxation values of the 9L cells and rCM cells incubated with NPCs were 15.3 ( 1.7 and 63.8 ( 2.2 ms, respectively. This significant difference in the T2 value and in the MR contrast indicates that Cltx-bound NPs have a much higher Nano Lett., Vol. 5, No. 6, 2005

Figure 4. Confocal fluorescent images of cells incubated with NPC-Cy5.5. (A) Rat cardiomyocytes (rCM). (B) 9L glioma. (C) MR phantom image of 9L and rCM cells cultured with NPC and embedded in agarose (4.7 T, spin-echo pulse sequence, TR 3000 ms, TE 30 ms).

Figure 5. Confocal fluorescent images of 9L cells cultured with NPC-Cy5.5. (A) Top section of cells. (B) Middle section of cells. (C) Bottom section of cells.

specificity for 9L cells than noncancerous cells and that glioma cells can be differentiated from normal cells by MRI with a high degree of detectability when Cltx-coated nanoparticles are used as a contrast-enhancement agent. To ascertain that NPC-Cy5.5 was indeed internalized by cells rather than bound to cellular membranes, 9L cells cultured with NPC-Cy5.5 were examined sectionally by confocal fluorescence microscopy at different depths. Figure 5 shows images from three sectional depths of 9L cells. Illustrated from left to right are the corresponding top (Figure 5A), middle (Figure 5B), and bottom (Figure 5C) sections of the cells, with the strongest Cy5.5 fluorescence observed in the middle section (i.e., within the cells and with decreasing fluorescence intensity toward the top and bottom sections). These confocal images confirmed that the nanoparticles were indeed internalized by the cells and that the NPs accumulated uniformly within the cytoplasm. The internalized nanoparticle probes as magnetic contrast agents would allow prolonged imaging during tumor resection and serve as therapeutic drug carries for tumor treatment. One of the major challenges in targeting brain tumors is the blood-brain barrier (BBB) formed by intercellular tight junctions of the endothelial cells of the brain capillaries, which limits the access of drugs or targeting therapeutics to the brain tissue.15,16 It has been demonstrated that the BBB may be overcome by particles with a size dimension smaller Nano Lett., Vol. 5, No. 6, 2005

than ∼50 nm or by lipid-mediated transport or receptormediated and PEG-assisted processes.17-20 Thus, the PEGlinked nanoprobe developed in this study, with a size of ∼15 nm, may promise the penetration of the nanoprobe across the BBB. Compared to dextran-coated nanoparticles currently used in MRI intraoperative examination during glioma resection, which label macrophages situated at the tumor boundary rather than the tumor cells themselves,3 the Cltxbounded nanoprobe directly targets tumor cells and thus can potentially “follow” the cell migration to delineate the tumor boundary in real time. This is particularly beneficial for monitoring high-grade glioma cells that are highly invasive and quickly infiltrate the surrounding healthy tissue.21 In conclusion, iron oxide NPs have been conjugated with Cltx and Cy5.5 to create a multifunctional nanoprobe that specifically targets glioma cells and that is detectable both magnetically and optically. MRI and confocal fluorescence analysis showed strong preferential uptake of Cltx-bound NPs by glioma cells over control NPs. A significantly higher degree of internalization of NPC-Cy5.5 conjugates by glioma cells versus control cells was observed, indicating the preferential targeting abilities of NPC-Cy5.5 for gliomas. The high stability and prolonged retention (at least 24 h) of NPC-Cy5.5 within targeted cells, as demonstrated by confocal imaging, are particularly advantageous in intraoperative imaging applications as compared to conventional 1007

optical fluorophores conjugated to targeting ligands. The cellular-level resolution as demonstrated here may promise accurate delineation of otherwise poorly defined glioma interfaces resulting from their highly invasive morphology. The application of the NPC-Cy5.5 nanoprobe for preoperative and postoperative diagnostic imaging with MRI and real-time intraoperative visualization of tumor margins with optical devices is a novel approach intended to improve the effectiveness of diagnostic and therapeutic modalities available for brain tumor patients. Acknowledgment. Financial support from NIH/NCI (N01-C037122) and Children’s Hospital and Regional Medical Center Brain Tumor Research Endowment is gratefully acknowledged. Also acknowledged are an NSF IGERT fellowship for C.S. and N.K. and a UW/PNNL Joint Institute Fellowship for J.G. Laboratory assistance from Kim Du is also acknowledged. References (1) Wrensch, M.; Minn, Y.; Chew, T.; Bondy, M.; Berger, M. S. NeuroOncology 2002, 4, 278-299. (2) Giese, A.; Bjerkvig, R.; Berens, M. E.; Westphal, M. J. Clin. Oncol. 2003, 21, 1624-1636. (3) Kircher, M. F.; Mahmood, U.; King, R. S.; Weissleder, R.; Josephson, L. Cancer Res. 2003, 63, 8122-8125. (4) Whittle, I. R. Curr. Opin. Neurol. 2002, 15, 663-669. (5) Stummer, W.; Stocker, S.; Wagner, S.; Stepp, H.; Fritsch, C.; Goetz, C.; Goetz, A. E.; Kiefmann, R.; Reulen, H. J. Neurosurgery 1998, 42, 518-525.

1008

(6) Ichioka, T.; Miyatake, S. I.; Asai, N.; Kajimoto, Y.; Nakagawa, T.; Hayashi, H.; Kuroiwa, T. J. Neuro-Oncology 2004, 67, 47-52. (7) Kachra, Z.; Beaulieu, E.; Delbecchi, L.; Mousseau, N.; Berthelet, F.; Moumdjian, R.; Del Maestro, R.; Beliveau, R. Clin. Exp. Metastasis 1999, 17, 555-566. (8) Deshane, J.; Garner, C. C.; Sontheimer, H. J. Biol. Chem. 2003, 278, 4135-4144. (9) Lyons, S. A.; O’Neal, J.; Sontheimer, H. Glia 2002, 39, 162-173. (10) Aina, O. H.; Sroka, T. C.; Chen, M. L.; Lam, K. S. Biopolymers 2002, 66, 184-199. (11) Yamamoto, Y.; Tsutsumi, Y.; Mayumi, T. Curr. Drug Targets 2002, 3, 123-130. (12) Weissleder, R.; Ntziachristos, V. Nat. Med. 2003, 9, 123-128. (13) Kohler, N.; Fryxell, G. E.; Zhang, M. Q. J. Am. Chem. Soc. 2004, 126, 7206-7211. (14) Koch, A. M.; Reynolds, F.; Kircher, M. F.; Merkle, H. P.; Weissleder, R.; Josephson, L. Bioconjugate Chem. 2003, 14, 1115-1121. (15) Yang, C. S.; Chang, C. H.; Tsai, P. J.; Chen, W. Y.; Tseng, F. G.; Lo, L. W. Anal. Chem.y 2004, 76, 4465-4471. (16) Kreuter, J. AdV. Drug DeliVery ReV. 2001, 47, 65-81. (17) Fenart, L.; Casanova, A.; Dehouck, B.; Duhem, C.; Slupek, S.; Cecchelli, R.; Betbeder, D. J. Pharmacol. Exp. Ther. 1999, 291, 1017-1022. (18) Brigger, I.; Morizet, J.; Aubert, G.; Chacun, H.; Terrier-Lacombe, M. J.; Couvreur, P.; Vassal, G. J. Pharmacol. Exp. Ther. 2002, 303, 928-936. (19) Calvo, P.; Gouritin, B.; Villarroya, H.; Eclancher, F.; Giannavola, C.; Klein, C.; Andreux, J. P.; Couvreur, P. Eur. J. Neurosci. 2002, 15, 1317-1326. (20) Masciangioli, T.; Zhang, W. X. EnViron. Sci. Technol. 2003, 37, 102A-108A. (21) Giese, A.; Bjerkvig, R.; Berens, M. E.; Westphal, M. J. Clin. Oncol. 2003, 21, 1624-1636.

NL0502569

Nano Lett., Vol. 5, No. 6, 2005