Glioma-Targeted Delivery of a Theranostic Liposome Integrated with Quantum Dots, Superparamagnetic Iron Oxide, and Cilengitide for Dual-Imaging Guiding Cancer Surgery

He-Lin Xu, Jing-Jing Yang, De-Li ZhuGe, Meng-Ting Lin, Qun-Yan Zhu, Bing-Hui Jin, Meng-Qi Tong, Bi-Xin Shen, Jian Xiao,* and Ying-Zheng Zhao*

Abstract

Herein, a theranostic liposome (QSC-Lip) integrated with superparamagnetic iron oxide nanoparticles (SPIONs) and quantum dots (QDs) and cilengitide (CGT) into one platform is constructed to target glioma under magnetic targeting (MT) for guiding surgical resection of glioma. Transmission electron microscopy and X-ray photoelectron spectroscopy confirm the complete coencapsulation of SPIONs and QDs in liposome. Besides, CGT is also effectively encapsulated into the liposome with an encapsulation efficiency of ∼88.9%. QSC-Lip exhibits a diameter of 100 ± 1.24 nm, zeta potential of −17.10 ± 0.11 mV, and good stability in several mediums. Moreover, each cargo shows a biphasic release pattern from QSC-Lip, a rapid initial release within initial 10 h followed by a sustained release. Cellular uptake of QSC-Lip is significantly enhanced by C6 cells under MT. In vivo dual-imaging studies show that QSC-Lip not only produces an obvious negative-contrast enhancement effect on glioma by magnetic resonance imaging but also makes tumor emitting fluorescence under MT. The dual-imaging of QSC-Lip guides the accurate resection of glioma by surgery. Besides, CGT is also specifically distributed to glioma after administration of QSC-Lip under MT, resulting in an effective inhibition of tumors. The integrated liposome may be a potential carrier for theranostics of tumor.patients with glioblastomas. Tumor mass resection is usually guided intraoperatively by anatomically registered images (usually CT and magnetic resonance imaging (MRI)) incorporated into a stereotactically based image-guided surgery platform. Although images from CT and MRI can indicate where tumor, necrosis, or edematous cortex is, it is still difficult to distinguish the invading tumor front for maximal resection of tumor. An additional radiochemotherapy is usually required as an auxiliary treatment of GBM after surgical resection.[4]
Multimodal imaging can integrate structural/functional information from several imaging modalities, thus promising more accurate diagnosis than any single imaging modality.[5] For example, a fibronectintargeting contrast agent using MRI in coregistration with high-resolution fluorescence cryoimaging can provide robust contrast enhancement in the metastatic tumors and enable the detection of micro-

1. Introduction

Glioblastoma (GBM), derived from the glial cells, is one of the most pernicious and vital brain tumors in the central nervous system (CNS).[1] Despite all of the current treatment modalities for malignant gliomas, such as microsurgery, chemotherapy, and radiotherapy, there is no definitive treatment.[2,3] Nonetheless, the maximum extent of surgical resection is associated with a longer recurrence-free period and overall survival of metastases of size <0.5 mm, extending the detection limit of the current clinical imaging modalities.[6] Moreover, it is easy to integrate multiple diagnostic/therapeutic modalities into one platform, which can potentially be applied as personalized theranostic nanomedicine. These theranostic nanomaterials are usually constructed by using a large number of imaging agents, including magnetic resonance imaging agents, fluorescence in vivo imaging agents, photoacoustic imaging agents, etc.[7,8] The multifunctional integrated system combines various properties such as tumor active targeting, imaging, and selective treatment into one system, which supplies more effectively multimodal methods to battle against glioma.[9] Although active targeting is more and more popular, it depends on the specific binding between targeting ligands/aptamers conjugated on nanoparticles and receptors overexpressed on the cancer cells. By contrast, physical targeting (such as extramagnetic field, light, ultrasound, etc.) has been realized for efficient tumor targeting or stimulating drug release because the physical interactions are more controllable in the complicated cancer molecular biology pathways.
Magnetic targeting has emerged as a promising approach to help theranostic nanoparticles to efficiently target tumors. Superparamagnetic nanoparticles have emerged as a versatile agent to magnetic targeting of glioma for magnetic resonance imaging due to their negative contrast capability and good safety.[10–12] In our previous report, SPION coated by polymer shell of PEI, PEG, and tween-80 is capable of delivering its encapsulated doxorubicin into gliomas for theranostics under exogenous magnetic field.[13] Nevertheless, the size of SPIONs was less than 10 nm, which could “leak” from the pores of fenestrated capillaries in normal tissues.[14] Moreover, magnetic resonance (MR) imaging sometimes has low sensitivity in spite of high spatial resolution and is susceptible to interference due to edema zones around tumors or the associated vascular. As an alternative, fluorescence imaging allows for in vivo noninvasive, sensitive, and real-time detection of cells or tissues.[15] Semiconductor nanocrystals, also known as quantum dots (QDs), have attracted tremendous attention as a new optical imaging method for molecular tracing and biomedical diagnostics because of their strong resistance to photobleaching, broad excitation spectra, narrow emission spectra, and longer fluorescent lifetime.[16,17] Due to their in vivo nonspecific distribution and potentially toxic effects caused by the release of heavy metals due to either oxidation of the core or surface defects, the core is encapsulated in a shell of organic material such as polymeric micelles, liposome, and nanoparticles.
Cilengitide (CGT) is a cyclic arginine–glycine–aspartic pentapeptide antagonist that inhibits overexpress of integrin receptors on GBM and tumor-associated endothelial cells.[18] Preclinical and clinical studies have confirmed that CGT can effectively inhibit angiogenesis and GBM growth via actively targeting the glioma cells to destroy their signal transmission for survival and proliferation.[19] However, chemotherapeutic application is restricted by blood–brain barrier (BBB) and its poor targeting efficiency. Recently, a combination of low-intensity ultrasound with microbubbles (MBs), referred to as ultrasound-targeted microbubble destruction (UTMD), has been proved to significantly enhance the transportation of therapeutics across normal BBB without causing any damage on brain.[20,21] Moreover, the microbubbles exposed to glioma will also be oscillated and collapsed to facilitate drug penetrating the deeper tumor zones under the assistance of ultrasound.[22] Our previous study indicated that cilengitide-loaded nanoparticles were effectively delivered to glioma by a transient opening of BBB using UTMD, thereby producing a complete growth inhibition of glioma.[23]
In this study, an integrated liposome, which combines simultaneously multiple-imaging agents (superparamagnetic iron oxide nanoparticles (SPIONs) and QDs) and therapeutic drug, cilengitide into one platform (QSC-Lip), is conceived for glioma-targeted delivery of its loading cargo under exogenous magnetic field. The “all in one” nanocarriers strategy together with UTMD may not only overcome nonspecific distribution of QDs and CGT, but also prevent homogenous leakage of single SPIONs less than 10 nm into normal tissues. The integrated liposome is expected to not only accurately localize tumor for imaging-guided cancer surgery but also used as chemotherapy for pre- and postsurgery. First, we described the preparation and properties of an integrated QSC-Lip. The glioma-targeted delivery capability of the encapsulated QDs/CGT in QSC-Lip was carefully evaluated under external field. The dual imaging of glioma including MRI negative-contrast imaging and fluorescence imaging was also confirmed in vitro and in vivo. Finally, the antitumor efficacy and safety of the QSC-Lip were evaluated in glioma implanted rat models (Figure 1).

2. Materials and Methods

Cilengitide was kindly supplied by Wuxi Kaili Pharmaceutical Company (Jiangsu Province, China). The oil-soluble QDs (CdSe/ZnS core/shell QDs) were purchased from Wuhan Jiayuan Quantum Dot Technology Development Co. Ltd. Iron oxide (II, III) magnetic nanoparticles (oil-soluble SPIONs) were provided by Aladdin Industrial Corporation, Shanghai, China. Dipalmitoyl phosphatidyl choline and polyethylene glycol 2000-distearoylphosphatidylethanolamine (PEG2000-DSPE) and egg yolk lecithin (PC-98T) were obtained from Avanti Polar Lipids (Alabaster, AL). Cholesterol (CHO) was purchased from Biolife Science & Technology Co. Ltd (Shanghai, PR China). Phospholipid-based microbubbles were provided by our own laboratory as described in our previous study.[22]

2.1. Preparation of the Integrated QSC-Lip

In this study, QSC-liposomes were prepared by film hydration followed by sequential extrusion. PC-98T, CHO, PEG2000DSPE and quantum dots, and superparamagnetic iron oxide nanoparticles were dissolved in chloroform at a mass ratio of 50:30:5:2:5. The organic solvent was evaporated by rotating evaporation instrument at 40 °C, and a homogenous dry lipid film was obtained by drying in a vacuum for 24 h. Cilengitide (10 mg) was suspended or dissolved in 10 mL of phosphate buffer solution (PBS), as a hydrated medium. The dry lipid film was hydrated at room temperature and coarse liposome suspension was further processed by probe sonication (400 W, on 5 s, off 5 s, total 10 min). Afterward, the liposome suspension was extruded through polycarbonate membrane with a pore size of 100 µm to control the size of QSC-Lip. Hydrophobic QDs/SPIONs were expected to be encapsulated inside lipid membrane while CGT was incorporated into aqueous inner core of liposome. The blank liposome (BL) was also prepared by the similar procedure above, just without addition of QDs, SPIONs, and cilengitide.

2.2. Characterization of the Integrated QSC-Lip

The size and zeta potential of QSC-Lip and BL were determined by using a Zetasizer Nano ZS900 (Malvern Instruments). The morphology of QSC-Lip was observed via transmission electron microscopy (TEM; JEOL-2010, Japan). X-ray photoelectron spectroscopy (XPS) analysis was performed on PHI Quantera SXM equipped with the Ar+ ion and C60+ ion guns. XPS of QSC-Lip before Ar+ ion sputtering was first acquired. Afterward, the XPS spectrum of QSC-Lip after Ar+ ion sputtering was also collected. Accelerating voltage 10 kV, sample current 10 nA, sputtering area 0.5 mm × 2.0 mm, and incidence angle 70° from the direction vertical to the sample surface have been set as the ion sputtering conditions.
The entrapment efficiency of CGT was determined by ultrafiltration method. The unencapsulated CGT was separated from the QSC-Lip suspensions by using an Amicon Ultra-0.5 centrifugal filter units (Amicon Ultracel-10k; EMD Millipore, USA) and centrifuging for 1 h at 8000 × g. The concentration of CGT in the ultrafiltrate was detected by high-performance liquid chromatography (HPLC) equipped with an XTerra RP 18 column (mobile phase: acetonitrile containing 0.1% TFA/ water = 50:50, detect wavenumber at 210 nm). The drug loading percentage (%) and entrapment efficiency (%) were calculated via the following equations, respectively
X-ray photoelectron spectroscopy was used to confirm whether SPIONs had been completely encapsulated into liposomes using imaging photoelectron spectrometer (XPS, Thermo ESCALAB 250). The magnetic property of QSC-Lip was also examined via a Quantum Design MPMS XL-7 superconducting quantum interference device (Quantum Design, USA). The fluorescence stability of QSC-Lip was detected at simulated physiologic conditions. QSC solution and QSCLip were incubated at 37 °C with pH 7.4 PBS containing 10% (v/v) of fetal bovine serum (FBS, Biological Industries) for 24 h. The fluorescence spectra were collected by fluorescence spectrophotometry.

2.3. In Vitro Release of CGT, QDs, and SPIONs from the Integrated QSC-Lip

In vitro release of CGT, QDs, and SPIONs from the integrated QSC-Lip was investigated by dynamic dialysis method depicted in literature.[24] Briefly, 1 mL of QSC-Lip suspension was transferred into a dialysis bag (MWCO 14 000 Da, Spectrumlabs, USA) and the end of dialysis bag was sealed. And then the endsealed dialysis bag was placed into 100 mL of release medium (pH 7.4 PBS) in a thermostatic shaking bath with 120 rpm at 37 °C. At each time internal, 5 mL of released medium was withdrawn and equal volume of fresh medium was added. The concentration of CGT, QDs, and SPIONs in release medium was determined by HPLC and fluorescence spectrophotometry at 605 nm and atomic spectrophotometry, respectively. The cumulative release percentage (%) was calculated by the following formula for each cargo in QSC-Lip where C1, C2, …, Cn are the single cargo concentration in each time point and V is the sample volume.

2.4. In Vitro Imaging-Guided Tracking of QSC-Lip in Glioma Cells

Rat C6 glioma cells were used as glioma model cells to investigate imaging-guided tracking of QSC-Lip in cells. C6 cells were cultured in dulbecco’s modified eagle medium (DMEM) supplemented with 10% FBS (Gibco, USA), 100 IU mL−1 penicillin, and 100 mg mL−1 streptomycin. All cells were incubated at a 37 °C incubator with a 5% CO2 atmosphere. The intrinsic fluorescence of QDs allowed tracing QSC-Lip in the cell to evaluate the cellular uptake of QSC-Lip under the assistance of external magnetic field. C6 cells were grown at a density of 1 × 106 per well on cover slips in 6-well plates. All cells were incubated overnight and then treated with the different formulations including QSC-Solu and QSC-Lip. The concentration of CGT and QDs in QSC solution and QSC liposomes was comparable, 30 µg mL−1 and 50 nmol mL−1, respectively. To evaluate the effect of external magnetic field on cellular uptake, a 0.1 T perpetual magnet was placed under cell-growing plate bottom for 15 min including the QSC-Solu+MT group and the QSC-Lip+MT group. Afterward, the cells were incubated for 4 h without external magnetic field. And then the culture medium was removed, rinsed with fresh PBS for three times, and fixed with 4% paraformaldehyde for 20 min. The cellular nuclei were stained using 4′,6-diamidino-2-phenylindole (DAPI) for analysis by confocal laser microscopy (A1 PLUS, Japan). Meanwhile, cells were collected after trypsin digestion for further quantitative analysis on flow cytometry (FACSCalibur FCM, Becton Dickinson, San Jose, CA).
In order to confirm whether QD-emitting fluorescence represented the amount of the invaginated CGT, the intracellular CGT concentration was also determined. After 4 h of incubation, C6 cells were washed twice with PBS, digested with trypsin, and homogenized by sonication followed by ultrahigh centrifugation. The concentration of CGT in the supernatant was detected by HPLC.

2.5. Cytotoxicity of QSC-Lip against Glioma

In vitro cytotoxicity of QSC-Lip or QSC solution was evaluated using a CCK-8 assay. C6 cells were seeded in a 96-well plate (1 × 104 per well) with 100 µL of culture medium and further incubated for overnight. Then, the culture medium was replaced with different formulations, and the cells were exposed to exogenous magnetic field for 15 min by placing the magnet under the cell dish. After further incubation of 24, 48, and 72 h, the cells were washed twice with PBS and resuspended with fresh medium. 10 µL of CCK-8 was added to the cells and further incubated for 2 h. Absorbance was measured at a wavelength of 450 nm by a microplate reader.
The rat pheochromocytoma (PC12) cells (Shanghai Cell Bank, Chinese Academy of Sciences) were used for the further in vitro biocompatibility study of QSC-Lip and QSC solution. PC12 cells were maintained in DMEM high-glucose medium supplemented with 10% fetal bovine serum and 1% penicillin and streptomycin (complete media) at 37 °C under a humidified atmosphere of 5% CO2. The next operation was the same as above. Cell activity was measured by CCK-8 assay at the same time as the same dose of different formulations.

2.6. Dual Model Imaging and Growth Inhibition of Glioma In Vivo

2.6.1. Rat Glioma Model and Experimental Groups

All animal procedures were performed according to the guidelines of the Institution Animal Care and Use Committee of Wenzhou Medical University. Pathogen-free male Sprague– Dawley rats (250–350 g) were purchased from Shanghai, China. Brain glioma tumor model was established according to the methods previously reported in literature.[21] Before the experiment, all animals were anesthetized by intraperitoneal injection of 10% chloral hydrate and immobilized on a stereotactic frame (KOPF900, USA). Then, 10 µL of cell suspension with 1 × 106 cells was injected into the brain of rats.
From day 7 after tumor implantation, gliomas were confirmed by MR imaging and then started to accept treatments. All animals were randomly divided into four groups (n = 10) as follows: (a) QSC solution (QSC-Solu) + MT; (b) QSC-Lip; (c) QSC-Lip+MT (CGT dose was 2 mg kg−1), and (d) saline control. All formulations were mixed with ultrasound microbubbles (300 µL kg−1, 1108–5108 bubbles mL−1) before injection through the tail vein. After administration of each formulation, UTMD was immediately generated by a linear array transducer (14 MHz, Acuson Sequoia 512C System, Siemens). The linear array transducer was placed on the skull of the hemisphere at the locations of tumor cell implantation as well as the contralateral normal hemisphere. The parameters of the transducer sonication in the experimental group were set as follows: pulse repetition frequency of 1 Hz, sonication time of 60 s, and burst length of 10 ms with an acoustic power of 3 W. After UTMD treatment, the rats in QSC-Solu+MT group and QSC-Lip+MT group were immediately treated with a magnetic field by placing a magnet on the surface of tumor for 1 h. Glioma rats were administered twice weekly for eight injections in 28 d. MR imaging and fluorescence imaging were used to monitor the distribution of CGT in rats’ brain.

2.6.2. In Vivo Dual-Modality Imaging and Treatment via QSC-Lip

In vivo glioma imaging was performed on the 7th, 13th, 21th, and 28th days after different administration. Two hours later after administration of different treatments, all MRI images were captured by a 3 T scanner (Trio with Tim, Siemens, Erlangen, Germany) using standard wrist coil (Chenguang Medical Technologies Company, Shanghai, China) with an inner diameter of 13 cm and quantified by turbo-spin-echo T2-weighted imaging according to following parameters: pulse repetition time (TR)/ echo time (TE) = 2300/110 ms; NEX = 2; matrix size = 256 × 256; FOV = 56 mm × 70 mm; slice thickness = 2.4 mm. Image analysis software was used to evaluate the size of the glioma. The volumes (V) of glioma were calculated using an ellipsoid approximation [V = 4/3 × (0.5)3(abc)]. In three orthogonal planes on 2D T2-weighted MRI, the maximum diameters of the tumor measured were represented by a, b, and c.
Meanwhile, at each time internal, all animals were deeply anesthetized with chloral hydrate after MRI images, and the hearts of rats were infused with saline through the cardiac ventricle until colorless infusion fluid was obtained from the atrium. The brains were separated along the transverse suture and then imaged using the Maestro in vivo optical imaging system (CRI, Inc., USA) for fluorescence imaging. The red filter set was applied and image cube files were taken in series while scanning the liquid crystal filter from 500 to 700 nm (with 10 nm step). The fluorescent images were then analyzed.

2.6.3. Surgical Accurate Resection of Glioma under Fluorescent Imaging

Glioma-bearing rats were divided into two groups. One group was received with QSC solution through intravenous injection (i.v.), while the other was administrated with QSC-LIP. Once administration is completed, the glioma-bearing rats in both groups were immediately exposed to UTMD. Meanwhile, a neodymium–iron-bore, disk-shaped magnet (2 cm diameter, 0.3 T, Webcraft GmbH) was placed on the surface of the tumor for 1 h. After 2 h of intravenous injection, a midline incision was made in the skin above the skull of glioma-bearing rats to expose the cranium. A small part of the skull covering the neoplasm was surgically removed using a drill skull. The covering dura was slightly peeled back from the cortical surface to expose GBM. The complete surgical horizon around GBM allowed to be observed under Olympus SZX10 microscope. Fluorescent images of GBM were captured to guild the realtime resection of glioma and surgical horizon was confined to the area with a highly focused red fluorescence. The GBM was surgically excised using microscope forceps and microscissors under the laser until no fluorescence signal can be observed. The bright light and fluorescence images of surgical horizon were taken by using a DP-72 camera before and after surgery, respectively.
To confirm whether the tiny tumor niche on the border of surgical horizon was completely and accurately resected, a biopsy sample with 3 mm thickness was extracted from the remaining surgical border and fixed in 4% paraformaldehyde for hematoxylin and eosin (HE) staining.

2.6.4. Glioma-Targeted Delivery of QSC-Lip under Exogenous Magnetic Field

The capability of simultaneously delivering the encapsulated CGT, QDs, and SPIONs into glioma for QSC-Lip through exogenous magnetic targeting was probed by HPLC detection, fluorescence imaging, and Prussian blue staining, respectively. Glioma rats were administrated with QSC-Lip with or without magnet targeting. Two hours later, rats were anesthetized and perfused with saline as above. The separated brain glioma fixed with optimum cutting temperature compound (OCT Compound) and further cut into 5 µm thick consecutive frozen sections. DAPI was used to stain nuclei for 10 min and observed under confocal laser microscopy for fluorescence distribution. Prussian blue staining was performed following the manufacturer’s instructions to determine the accumulation of Fe3+.
Meantime, CGT deposition in brain glioma was also quantified using HPLC. Briefly, 100 mg of tumor tissues were sliced into small pieces and homogenized in 200 µL of PBS by IKA homogenizer. Afterward, 200 µL of methanol was added to homogenized sample, and CGT was extracted by vortex and centrifugation at 13 000 rpm for HPLC analysis.

2.7. Histological Analysis of Brain Gliomas after Treatment

After the treatment, all animals were sacrificed to harvest the brains and other organs. Brain glioma tissues were fixed in 4% paraformaldehyde, embedded in paraffin, and cut into sections at 5 µm thickness. The tumor histology was viewed under light microscopy (Nikon ECLIPSE Ti-S, Ruikezhongyi Company, Beijing, China) by stained with H&E.

2.8. Apoptosis Assay of Glioma after Treatment

Cell apoptosis in the glioma zone after treatment with QSC-Lip was detected via tunnel staining (Boster Biological Engineering Co., Wuhan, China) following the manufacturer’s instructions. The cell staining and morphology were observed under optical microscope. Alternatively, the immunohistochemical staining of caspase-3 was also performed. The rabbit polyclonal antibody to caspase-3 (1:500, Abcom, Cambridge, UK) was utilized. Brain tumor sections were deparaffinized in a xylene series and then hydrated in distilled water. Samples were blocked with 3% H2O2 and citrate buffer was utilized to retrieve antigen again, followed by treatment with primary antibodies for 2 h at 37 °C and HRP-secondary antibody for 30 min at 37 °C. Finally, samples were stained with diaminobezidin (DAB) and counterstained with hematoxylin. The glioma tissues were examined and recorded under optical microscope. Four random microscopic fields were used to evaluate the number of stained cell.

2.9. In Vivo Toxic Evaluation of QSC-Lip

For the toxic evaluation, the heart, liver, spleen, lung, and kidney of rats treated with different formulations were collected. The excised tissues were treated with 4% paraformaldehyde, paraffin embedded and sectioned, and finally processed with HE staining for histological studies.

2.10. Statistical Analysis

Survival curves between groups were compared using the logrank test. GraphPad software (Prism version 5.0.1) was used to perform statistical analyses. All results are displayed by the mean ± SD for each group. Student’s t-test was carried to confirm the significant difference between two groups. The significant difference between multiple groups was determined using one-way analysis of variance. p < 0.05 was considered statistically significant.

3. Results

3.1. Characterization of Liposomes

In this work, the commercially available imaging contrast, QDs for fluorescence imaging, and SPIONs for negative MR imaging, together with therapeutic peptide, cilengitide, were cocapsulated into PEGylated liposomes (QSC-Lip). In brief, QSC-Lip was prepared by premixing egg yolk lecithin, cholesterol, DSPE-mPEG2000, QDs, and SPIONs at a mass ratio of 50:30:5:2:5 followed by hydration in CGT-containing PBS according to standard method for preparation of liposomes. The characterization of QSC-Lip was carried out and results are shown in Figure 2. The prepared QSC-Lip macroscopically exhibited a cloud appearance in the bright field images, while a strong fluorescence was still observed under a laser excitation in dark field. As expected, after 1 h of exposure to a magnetic field, QSC-Lip was attracted and deposited to the side alongside the permanent magnet, meanwhile the cloud suspension become clear. Moreover, the deposited QSC-Lip displayed red fluorescence under a UV lamp, indicating the SPION-guided migration of QDs under exogenous magnetic field (Figure 2A). The particles size, zeta potential, and morphology of QSCLip were also characterized by dyamic laser scattering (DLS) and TEM images, respectively. The particle sizes of QDs and SPIONs were detected to be ≈8 and ≈20 nm by DLS (Figure S1, Supporting Information), respectively, while the particle sizes of QSC-Lip were increased to 100 ± 1.24 nm, larger than that of blank liposome (66.0 ± 1.75 nm), indicating the coencapsulation of QDs and SPIONs in liposome. The negative zeta potential of QSC-Lip (−17.10 ± 0.11 mV) was near to −15.8 mV of blank liposome. The colloidal stability of the QSC-Lip in different dispersion media (pH 7.4 PBS buffer, DMEM) was evaluated by measuring the DLS hydrodynamic size for different time periods. The stability of nanoparticle was associated with surface properties and surface potential. QSC-Lip exhibited not only a hydrophilic PEG surface but also a negative potential of −17.10 ± 0.11 mV, which make them stable for at least 7 d in several dispersion media including PBS buffer or DEMEM (Figure S2, Supporting Information). Morphology of QSC-Lip was roughly spherical vesicle and numbers of fine particles with 7–20 nm were seen in these spherical vesicle, as characterized by TEM in Figure 2B. These fine particles in vesicle represented the encapsulated QDs or SPIONs because their  morphology was very identical to that observed by TEM images of free QDs or free SPIONs (Figure S1, Supporting Information). The poor roundness of QSC-Lip vesicle may be due to its deformation during the sample process for TEM and some of QSC-LIP may even be destroyed, which resulted in appearance of tiny particle of QDs and SPIONs in the TEM image of QSC-LIP (Figure 2B). The drug loading capability (DLC) and the encapsulation efficiency (EE) of CGT were also determined by the ultrafiltration method. DLC was determined to be 1.7 ± 0.2% and EE of CGT was reached to 88.9 ± 3.1%. Any free SPIONs/QDs were not observed in TEM of QSC-Lip (Figure 2B), indicating the complete encapsulation of SPIONs/QDs. And the concentration of SPIONs and QDs in QSC-Lip was 270 ± 12 × 10−6 m and 100 ± 9 × 10−9 m, respectively.
In order to further quantify the encapsulation efficiency of SPIONs and QDs, the depth-dependent XPS spectra of QSCLip were analyzed by using the conventional Ar+ ion sputtering method. The results were displayed in Figure 3C. Before sputtering, the XPS spectrum of QSC-Lip exhibited a group of elements associated with phospholipid including O, C, and P (hydrogen is not accessible for analysis by XPS). Based on the binding energy (BE) values of a strong photoelectron peak, information about the chemical state for each element can be obtained. The BE of P 2p is 133.8 ± 0.1 eV, which was identified as the P5+ state resulting from the PO4 moiety of the lipid headgroup. The peaks for C 1s were analyzed into two components corresponding to the C−C/C−H component at 283.9 ± 0.1 eV, and C−O (H) groups at 285.6 ± 0.1 eV (Figure S3A, Supporting Information). The peaks for O 1s were attributed to the C–O component at 531 ± 0.1 eV for DSPE-PEG2000 (Figure S3B, Supporting Information). The XPS spectra of QSC-Lip after sputtering differed from that before the sputtering. As expected, the peak for P 2p at 133.8 ± 0.1 eV was significantly decreased and the peak corresponding to the Fe element at 700 ± 0.1 eV was observed after sputtering (Figure 3C). Meanwhile, Cd element from QDs at 410 ± 0.1 eV was also detected. Moreover, the C 1s peaks for the CO component at 531 ± 0.1 eV become more obvious because of poly(ethylene glycol) in SPIONs, and the O 1s peaks for the CO component of CGT at 533 ± 0.1 eV appeared (Figure S3C,D, Supporting Information). These results suggested that SPIONs and CGT had been completely encapsulated in liposomes.
The photometric characteristic of QSC-Lip was also examined by fluorescence spectrophotometry. As shown in Figure S4B (Supporting Information), the fluorescence emission spectra of QSC-LIP were similar to that of QSC solution, indicating its potential application as fluorescence diagnostic agent. Some studies have shown that serum protein led to fluorescence quenching of QD solution.[25] In order to confirm the fluorescence stability of QSC-LIP in vivo, the fluorescence spectra of QSC-LIP were further characterized after incubation with PBS containing 10% of FBS at 37 °C. Indeed, the fluorescence of QSC solution was significantly decreased after incubation with pH 7.4 PBS containing FBS. However, the fluorescence of QSC-LIP was not significantly compromised after incubation, suggesting its good stability in the physiologyrelevant medium.
The superparamagnetic character of QSC-Lip was demonstrated by the magnetization–magnetic field strength curves (M–H curves) and results are exhibited in Figure 3B. The magnetic hysteresis curves indicate that the QSC-Lip with saturation magnetizations (Ms) of 34.1 emu g−1 maintained the superparamagnetic property of SPIONs. Although the saturation magnetization of QSC-Lip was lower than that of free SPIONs with saturation magnetizations (Ms) of 57.3 emu g−1, the zero coercivity of the QSC-Lip suggested that the encapsulating process did not influence the superparamagnetic character of the SPIONs. This result suggested the encapsulated QDs or CGT may be transferred together with the encapsulated SPIONs to reach the tumor site under the guidance of an external magnetic field. To evaluate the performance of QSC-Lip in T2weighted MR imaging, their spin–spin relaxation times (T2) and spin–lattice relaxation times (T1) were measured using a 3.0 T MRI system. With the increase of the Fe concentration, the MR signal intensity decreased, indicating a potential application of QSC-Lip as an effective negative T2 contrast agent. The T2 relaxation rate (1/T2) was linearly fit as a function of Fe concentration, and the r2 relativity of the QSC-Lip was estimated to be 172.6 mm−1 s−1, significantly higher than that of free SPIONs with r2 relativity of 88.56 mm−1 s−1 reported in our previous study.[13] Alternatively, the r1 relativity of the QSC-Lip has also been detected and results are shown in Figure S5 (Supporting Information). The T1 relaxation rate (1/T1) (1.56 mm−1 s−1) of QSC-Lip in response to Fe concentration was significantly lower than T2 relaxation rate (1/T2). The calculated r2/r1 ratio was highly reached to 110.6, suggesting the potential ability of QSC-Lip as T2-negative contrast agents. Similar results that encapsulation of multiple SPIONs in nanocarrier increased the sensitivity of imaging contrast have been reported in several literatures.[26,27] To determine whether the T2 relaxometry of the QSC-Lip was compromised after incubation with cells in vitro, the MRI images of C6 cells were acquired after incubating them with different concentrations of QSC-Lip for 24 h. As shown in Figure S3 (Supporting Information), the T2 relaxation rate was still sensitive to low Fe concentrations, and the r2 relativity was estimated to be 162.60 mm−1 s−1, indicating that the detection threshold of QSC-Lip was not significantly compromised after incubation.

3.2. In Vitro Release of the Encapsulated Cargos from QSC-Lip

The release profiles of the encapsulated cargos including CGT, QDs, and SPIONs from QSC-Lip were examined in pH 7.4 PBS (10 mm) containing 0.5% (v/v) of Tween-80. As shown in Figure 3D, each of the encapsulated cargos in QSCLip showed a relatively slower release profile than their solution forms (CGT solution, QD solution, and SPION solution). Moreover, although cumulative release percentage for each cargo from QSC-Lip was different at each time point, each cargo showed a similar release rhythm, that is, a biphasic release pattern consisting of a relatively rapid initial release within initial 10 h followed by 3 d of a sustained release. The slow rhythm release profile was due to encapsulation of liposome, which hindered release of its contents. The difference of cumulative release percentage between each cargo may be due to their different molecular size and interactions between them. For example, CGT was a water-soluble small molecule with fast diffusing ability in liposome, exhibiting a highest cumulative release percentage at each time point. In contrast, for QDs or SPIONs, they were intrinsically supermolecular particle with 10–20 nm of size and possessed a slow diffusing rate in medium, which resulted in a slower release profile. Besides, self-interaction between SPION particles in single liposome made them difficult release from liposome. The cumulative release of SPIONs from QSC-Lip was about 6% within 10 h, far below that of QDs and CGT, indicating that the majority of SPIONs still remained in the liposomes. The retention of SPIONs in liposome was very beneficial to guild intact QSC-Lip to tumor tissues under external magnetic field.

3.3. In Vitro Imaging-Guided Tracking of QSC-Lip in Glioma Cells

C6 cells were incubated with QSC-Lip followed by UTMD in the presence or absence of a magnetic field. Fluorescence images of QDs in C6 cells after incubation for 4 h were tracked by confocal laser scanning microscopy (CLSM) and quantitatively assayed by flow cytometer analysis. The results are shown in Figure 4. Compared with QSC solution, the higher QD fluorescence in cytoplasm was observed for C6 cells treated with QSCLip regardless of whether an external MF existed, indicating that the enhanced cellular uptake of QDs encapsulated in QSCLip through endocytosis. Moreover, fluorescence of QDs in cytoplasm of C6 cells treated with QSC-Lip was more obvious in the presence of an external MF, indicating that the external magnetic field could significantly enhance the cellular uptake of QSC-Lip. However, there was no significant difference in fluorescence of QDs in cytoplasm of C6 cells treated with QSC-Solu, whether an external MF existed or not. This result suggested that free SPION nanoparticles in QSC-Solu were not able to guild the free QDs or CGT toward cytoplasm of C6 cells even if the external MF was imposed. Similar results were observed by flow cytometry analysis, that is, the QD amount internalized by C6 cells decreased in the order of QSC-Lip in the presence of an MF, QSC-Lip and QSC solution with or without MF. It was easily inferred that the SPIONs in liposome effectively guided the encapsulated QDs entering cytoplasm under an external magnetic field. The amount of internalized QDs for QSC-Lip in the presence of an MF was approximately twofold higher than that of QSC-Lip alone (Figure S5, Supporting Information).
To confirm whether the fluorescence of QDs in cells reflects the cellular uptake of encapsulated CGT in liposome, the amount of CGT in cytoplasm of C6 cells treated with various formulations was also detected. The results are exhibited in Figure 4C. The trends in the amount of CGT uptake by cells were consistent with fluorescence of QDs. The CGT amount in C6 cells treated with QSC-Lip was significantly higher than that of QSC solution. Likewise, the uptake amount of CGT by cells treated QSC-Lip in the presence of MT was significantly higher than that without MT, approximately twofold higher. The good correlation between QD fluorescence and intracellular CGT amount suggested that the integrated magnetic QSC-Lip could effectively carry QDs and CGT into cells in the presence of an external magnetic field.

3.4. Cytotoxicity of QSC-Lip against Glioma

The QDs or SPIONs may have potential cellular toxicity due to their heavy-metal content. The blank liposome coencapsulating QDs and SPIONs (QS-Lip) was prepared to evaluate its toxicity in vitro. The cytotoxicity of QS-Lip on normal neuron cell line, PC12 cells, and glioma cell line, C6 cells, was assayed by CCK-8 kits. Meanwhile, the cytotoxicity of the simple mixture of free QDs and SPIONs (QS-Solu) was also evaluated as control group. As shown in Figure 5, there is no significant cytotoxicity of QS-Solu against the two cell lines at lower QS concentration. But its toxicity was slightly increased when the total QS concentration reached to 80 µg mL−1. The cell viability of PC12 cells was decreased to 85% after incubation of 72 h (Figure S6, Supporting Information). By contrast, there was no significant cytotoxicity for the blank QS-Lip against normal neuron cell line and glioma cell line at all tested concentration. The cell viability rate was higher than 90% at tested concentration, indicating encapsulation of liposome decreased the potential toxicity of QS-Solu. The low cytotoxicity for blank QS-Lip is beneficial to in vivo application.
The growth inhibition of the encapsulated CGT in QSC-Lip on glioma cells was also detected by the CCK-8 kit in the presence or absence of an external MF. As shown in Figure 5, dose- and time-dependent cytotoxicity was observed for all groups. Moreover, as expected, the cytotoxicity of QSC-Lip with or without MT was stronger than that of CGT-Solu or QSC-Solu regardless of incubation time. The calculated half inhibition concentration (IC50) of CGT-Solu, QSC-Solu+MT, QSCLip, and QSC-Lip+MT after 24 h of treatment was 79 ± 0.82, 78 ± 1.02, 55 ± 0.98, and 35 ± 0.89 µg mL−1, respectively. As the incubation time increased to 48 h, the corresponding IC50 values decreased to 60.6 ± 1.15, 62.4 ± 0.22, 23.8 ± 1.24, and 15.4 ± 2.11 µg mL−1, respectively. These may due to the different uptake mechanisms of liposome-encapsulated CGT and free CGT. As mentioned above, QSC-Lip was mainly internalized into cells by endocytosis, which made more CGT available in target sites. Interestingly, QSC-Lip in the presence of external magnetic fields displayed a slightly stronger cytotoxicity than those outside magnetic fields. Under the bottom side of the cell culture plates, imposing a magnetic field endowed with orientation and attachment of QSC-Lip alongside the cell layer enhances the availability of encapsulated CGT to the tumor cells.

3.5. In Vivo Evaluation of QSC-Lip in Glioma-Bearing Rats

3.5.1. In Vivo Dual-Modal Imaging and Glioma Treatment via QSC-Lip

The rats bearing glioma were constructed by implanting C6 cells inside corpus striatum of brain. After 7 d of growth, tumor formation was confirmed by MRI imaging using Gd-DTPABMA as positive MRI contrast agent. As shown in Figure 6, the corpus striatum zone of brain shows hyperintense signals in the T1-weighted imaging compared with neighboring normal tissues, indicating the existence of glioma. To confirm whether QSC-Lip can enhance the contrast imaging to accurately localize glioma under assistance of external magnetic targeting, MR T2-weighted images of rat brains in the axial directions were determined after 2 h postinjection of magnetic formulations including QSC-LiP+MT, QSC-Lip, and QSC-Solu+MT. The MR images of the brains are shown in Figure 6. Compared with saline, the tumor zone was obviously darkening postinjection of all magnetic formulations, indicating the enhanced negative contrast of SPIONs regardless of their free or encapsulated forms. Moreover, the negative contrast enhancement of QSC-Lip was more significant than QSC solution, indicating that encapsulation of liposome promoted the localization of SPIONs to glioma. This phenomenon may be associated with the fact that simultaneous encapsulation of numbers of SPIONs particles in single liposome facilitated more SPIONs transporting into glioma during the transient opening of blood–brain barriers through UTMD. Expectedly, exposing an external magnetic field could further promote the targeted delivery of QSC-Lip to the glioma zone and produced a clearer negative imaging of tumor than that of QSC-Lip without MT, which was helpful for accurate localization of glioma to guild the corresponding surgery.
The accurate MRI tracking of glioma by combination of QSCLip and external magnetic field was further confirmed by fluorescence imaging of the excised brain-bearing glioma ex vivo. After 2 h of administration with QSC-Lip, the rats were sacrificed, and the brains were separated and then imaged using in vivo optical imaging system. The fluorescence images are exhibited in Figure 7. There was almost not any fluorescence visualized in the saline group. The QSC-Lip group without an external magnetic field exhibited a dispersed fluorescence distribution along bulk brain despite strong fluorescence inside the tumor zone. By contrast, a strong fluorescence was mainly concentrated in the tumor site in the QSC-Lip group followed by an external magnetic field (QSC-Lip+MT), with a decreased distribution in neighboring brain tissue near tumor site. Moreover, the morphology and size of tumor indicated by fluorescence imaging in the QSC-Lip+MT group were similar to that reflected by MRI. The accurate dual-modal imaging of QSC-Lip combined MT was required for the accurate diagnosis and in situ imaging-guided cancer surgery.
To confirm whether the encapsulated CGT in magnetic liposomes together with the encapsulated SPIONs can be targeted to tumor site in the presence of external magnetic fields, the real time of tumor was monitored by MRI and tumor size was also calculated by MRI software. The volume of glioma at different times is exhibited in Figure 7. The tumors grew fast in rats treated with saline or QSC solution with MT. The volume of tumor was calculated via ellipsoid approximations on MR images basis. The tumors of rats in saline or QSC solution with MT grew fast from 12.32 ± 7.12 and 13.13 ± 6.22 mm3 on day 7 to 620.26 ± 65.76 and 480.33± 62.58 mm3, respectively, on day 28 (day 7 vs day 28, all p < 0.001) in Figure 7B. Nevertheless, the tumor displayed slow growth in rats administrated with QSC-Lip, especially QSC-Lip with MT. The tumor in rats administrated with QSC-Lip or QSC-Lip with MT displayed a growth of tumor volume from 13.28 ± 6.15 and 12.38 ± 5.76 mm3 on day 7 to 189.28 ± 6.15 and 10.38 ± 5.76 mm3, respectively, on day 28. The tumors growth of rats treated with QSC-Lip together with MT was completely inhibited. The treatment efficacy of QSC-Lip with MT was significantly enhanced compared with that of QSC solution with the MT group or QSC-Lip group.
Furthermore, as shown in Figure 7C, the survival of rats administrated with QSC-Lip together with an external magnetic field (59.6 ± 2.3 d) was greatly prolonged as compared with that of the control group (20.2 ± 1.8 d, p < 0.001), QSC-solution group (25.6 ± 2.3 d, p < 0.01), and QSC-Lip group (38.1 ± 2.4 d, p < 0.01). In this study, survival is expressed as median days of survival, and a few rats were still alive in the control group on day 28.

3.5.2. Surgical Accurate Resection of Glioma under Fluorescent Imaging

Glioma-bearing rats were divided into two groups. One group was received with QSC solution through intravenous injection, while the other was administrated with QSC-LIP. The bright light and fluorescence images of surgical horizon were taken before and after surgery. The results are shown in Figure 8A. An obvious glioma plaque could be observed in brain in bright images of both groups before surgical resection but its boundary with healthy brain parenchyma was very fuzzy. Alternatively, in fluorescent images, the tiny fluorescence dots were scattered in bulk brain tissue and any fluorescence was not focused inside glioma for the QSC-solution group. By contrast, a strong fluorescence was concentrated in neoplasm area of rats in the QSC-Lip group. Thus, the glioma from the QSC-solution group was empirically resected under bright light while the fluorescence-guided resection was practiced for glioma-bearing rats in the QSCLip group. The glioma plaque was surgically excised until no fluorescence signal can be observed in the tumor zone. After resection, there was an obvious cavity in brain and no fluorescence was visualized in the cavity for the QSC-Lip group.
To confirm whether the tiny tumor niche on the border of surgical horizon was completely and accurately resected, a biopsy sample with 3 mm thickness was evaluated by HE staining. The results are shown in Figure 8B. Numbers of tiny tumor residue were remained at the periphery of the cavity in the QSC-solution group. On the contrary, there were not any tumor cells in sight for the QSC-Lip group, indicating that the real-time fluorescence imaging allowed guiding the accurate and complete surgical resection of intracranial GBM.

3.5.3. Glioma-Targeted Delivery of QSC-Lip under Exogenous Magnetic Field

To further confirm the correlation of dual-modal imaging and effective inhibition of tumor growth with the targeted distribution of QSC-Lip under an external magnetic field in glioma, frozen sections of OCT-embedded brains were either directly observed by fluorescent microscopy or stained with Prussian blue. The results are shown in Figure 9A. No fluorescence dots were detected in glioma tissues of rats treated with saline, while red fluorescence dots could be observed in QSC-Lip regardless of an external magnetic field existed or not. Moreover, it was also found that more fluorescence QDs in the form of QSC-Lip were migrated to tumor tissues in the presence of an external magnetic field compared with QSC-Lip without MT, further verifying the targeting of QSC-Lip under an external MT. Similarly, the identical results were also obtained for staining of Prussian blue. Prussian blue staining was usually utilized to label the Fe3+ in glioma tissues. More Fe3+-labeled cells were visualized within the glioma under an external MT. In contrast, fewer SPION spots were observed in the brains of rats treated with QSC-Lip alone (Figure 9A), and no blue-colored spots were observed in the brains of rats treated with saline solution. These results indicated that QSC-Lip could truly enhance the negative contrast effect and could be used in noninvasive MRI monitoring techniques.
The correlation between growth inhibitions of glioma with the enhanced distribution of CGT was also investigated by detecting CGT amount in glioma. As shown in Figure 9B, the deposited amount of CGT in tumor was only 10.21 ± 3.45 µg g−1 tissues for QSC-solution groups even in the presence of MT, while the deposited CGT amount increased to 40.15 ± 6.82 µg g−1 and 60.83 ± 5.81 µg g−1 tissues for alone QSC-Lip and QSC-Lip+MT, respectively. These results suggested that the distribution of CGT inside glioma was obviously enhanced by QSC-Lip together with MT. CGT was nonspecifically distributed in vivo and its enhanced distribution in brain by UTMD was highly dependent on the retention time of CGT. It was reported that in vivo half-lives of CGT in mouse was very short, only 0.34 h, and it was cleared through renal excretion.[28] Thus, the delivery of CGT solution to brain was very limited even if the transient opening of BBB by using UTMD. In contrast, PEGlyated liposome exhibited a prolonged circulating time in blood, making more the encapsulated CGT transporting to brain under the assistance of UTMD. Meanwhile, external magnetic targeting was an important factor to enhance its distribution inside glioma. The specific distribution of QSC-Lip in glioma under an external MF was profitable to reduce the toxicity of CGT against normal tissue.

3.6. HE and Immunohistochemistry Staining of Glioma

The histological morphology and cellular apoptosis of glioma after different treatments were tested by HE staining and immunohistochemistry staining, respectively. As shown in Figure 10A, the HE-staining sections from the control and QSC-solution group displayed a significantly dense nuclear polymorphism, whereas group treated with QSC-Lip, especially QSC-Lip+MT, appeared more hypocellular in the necrotic zone, indicating stronger antitumor effects. The level of cellular apoptosis was highly correlated with the upregulation of cleaved caspase-3. The caspase-3 expression was also determined by immunohistochemistry staining. Apparently, compared with the tumor treated with QSC solution, there were more brown dots in tumor treated with QSC-Lip regardless of the presence of an external magnetic field, indicating most of the cells were going through the apoptosis. Moreover, the caspase-3 positive cells in rats administrated with QSC-Lip+MT were more than that of the QSC-Lip group. Alternatively, tunnel immunofluorescence staining was also utilized to detect the cellular apoptosis of glioma after various treatments. As shown in Figure 10B, similar results were observed. The majority of tumor cells in rats treated with QSC-Lip, whether or not the presence of a magnetic field, suffered from apoptosis. However, a few observable apoptosis cells were found in rats administrated with saline or QSC solution. These results suggested that the imposition of an external magnetic field after injection of QSC-Lip further improved the distribution of encapsulated CGT inside the tumors, thereby resulting in a greater number of apoptotic tumor cells.

3.7. The Toxicity of QSC-Lip against Normal Organs

To evaluate the toxicity of QSC-Lip against normal organs, pathological HE-stained sections from the heart, liver, spleen, lung, and kidney were assayed (Figure 11). Regardless of treatment with QSC solution or QSC-Lip at an effective dose, there was not any noticeable organ damage or inflammatory lesions in the heart, liver, spleen, lung, or kidney of rats, indicating the nonexistence of significant toxicity in the treated rats. Moreover, after treatment with QDs-Lip, the rats did not show any change in their drinking and eating behavior. These suggested that QSC-Lip exhibited a good safety in vivo application.

4. Discussions

Molecular imaging is a noninvasive technique for visualizing the development of glioma and imaging-guided surgery.[29,30] MRI is effective for anatomic, functional, and molecular imaging due to its high spatial resolution and imaging capability of deep tissue, but it has relatively low sensitivity to detect molecular targets expressed on the cancer cell surface.[31] Paramagnetic substances, including gadolinium (Gd) chelates and superparamagnetic iron oxide nanoparticles, are used as contrast agents to alter the T1 or T2 relaxation times around the molecular targets for signal enhancement in molecular MRI. The T1-contrast imaging is typically confined to imaging of edema zones surrounding tumors or tumor-relevant blood vessels.[32,33] In comparison, the negative T2 effect is less dependent on contrast agent compartmentalization. T2 contrast agents increase the rate of transverse magnetization decays, resulting in dark signals.[34] Many researchers utilized the SPIONs as T2 imaging contrast enhancers to monitor growth of glioma.[35] But it was difficult for T2 imaging to achieve the useful information for guiding the precise resection of glioma in surgery.
As an alternative, optical imaging is a versatile imaging modality using fluorophores to visualize the expression of molecular targets and to delineate tumor margins in situ. The near-infrared fluorescent (NIRF) dyes can provide deeper tissue penetration and reduced background fluorescence. NIRF-based optical imaging provides real time in vivo visualization of the tumor with minimized background signal and allows for better discrimination between tumor and normal tissue. This unique property makes optical imaging a promising tool for intraoperative image-guided surgery.[36,37] The quantum dots, possess many advantages over traditional organic NIRF fluorophores, including high sensitivity, fluorescence stability, and easy access of availability.[38] However, potential heavy metalrelated toxicity from constituent leakage such as Cd3+ is still a problem.[39]
Cilengitide is a cyclic arginine-glycine-aspartic acid pentapeptide antagonist of integrins, the corresponding receptors of which were overexpressed on glioma and tumor-invaded endothelial cells.[40] However, plasma clearance of CGT in mice or rats was very rapid, with a mean t1/2 of ≈0.34 h, which resulted in its lower distribution inside brain.[28]
Liposomes are gaining wide attention as an efficient carrier for delivery of various molecules to the CNS. In this study, a magnetic liposome was constructed to simultaneously encapsulate QDs, SPIONs, and CGT for glioma-targeted delivery. Coencapsulation of three composites in liposome was carefully confirmed by TEM and XPS. QSC-Lip exhibited an appropriate particle size of ≈100 nm, zeta potentials of −17 mV, and good magnetic characteristics as SPIONs. Huang et al. previously synthesized multifunctional magnetic nanoparticles for diagnosis and therapy of gastric cancer, in which spin–spin relaxation rate was only 118 mm−1 s−1.[41] In this study, the r2 relaxation of the QSC-Lip was determined to be 172.6 mm−1 s−1, higher than that of nanoparticles in the literature.[42] Moreover, T2-weighted MR imaging of C6 cells showed that QSC-Lip possessed a strong ability of negative contrast. The designed QSC-Lip provided a higher relaxation rate, with a dose of iron 100 times lower than the previously reported dextrancoated SPIONs.[43] More importantly, in vitro release showed that a biphasic release pattern for each cargo in QSC-Lip was observed, that is, a relatively rapid initial release followed by a sustained release. Among the three encapsulated cargos, SPIONs exhibited the lowest release rhythm from QSC-Lip, with the majority of SPIONs still remained in the liposomes, which was very beneficial to guild intact QSC-Lip to tumor tissues under an external magnetic field. The interaction between serum proteins and QSC-Lip may destroy the microscopic structure of the liposomes, resulting in premature leakage and rapid clearance of its encapsulated cargo from body.[44,45] The superior magnetic and fluorescent stability of QSC-Lip in physiology-relevant medium was also proved in this study.
Cellular toxicity of QS-Lip was performed by CCK-8 kit assay and results indicated QS-Lip exhibited no significantly cytotoxicity, displaying that QS-Lip was a safe imaging agent. By contrast, QSC-Lip together with external magnet exhibited stronger anticancer activity against C6 cells compared with that of QSC solution (Figure 5). Besides, in vitro cellular uptake also showed that more fluorescence inside C6 cells treated with QSC-Lip in the presence of external magnet was observed than that of QSC-solution-treated cells (Figure 4). The results were inconsistent with that of cellular toxicity, suggesting that an external magnet can directly guide QSC-Lip to C6 cells and increase the concentration of QSC-Lip.
In this study, dual-modal imaging capability of QSC-Lip was also evaluated by MRI and florescence imaging. As exhibited in Figure 6, MRI signal of rats treated with QSC-Lip+MT was obviously decreased within the glioma lesion compared with the surrounding normal tissue, suggesting that the liposomes were capable of carrying SPIONs to rapidly reach the tumor lesion under the guidance of external magnetic fields. SPIONs can pronouncedly decrease spin-echo T2-weighted pulse time to produce negative contrast. The weakening of signal intensity for MR imaging indicated the existence of brain neoplasm and the accumulation of SPOIN in that zone. Alternatively, the tumor margin delineated by fluorescence imaging of the excised brain-bearing glioma ex vivo correlated well with that defined by MRI in rats treated with QSC-Lip+MT, that is, the morphology and size of tumor indicated by fluorescence imaging was similar to that reflected by MRI. The accurate fluorescence imaging was important for imaging-guided braintumor resection in situ surgery.[46]
Whether the accurate imaging against glioma was associated with the enhanced accumulation and retention of QSC-Lip inside tumor tissue after systemic administration was further explored by Prussian staining and fluorescent detecting of QDs. As shown in Figure 9, more blue spots and stronger fluorescence appeared in the glioma zone of rats treated with QSC-Lip+MT, indicating the high relation between capability of imaging for QSC-Lip and their enhanced distribution in glioma. Moreover, this conclusion was further verified by the implementation of fluorescenceguiding resection of glioma. The tiny tumor residue was still remained at the periphery of the cavity after bright-light-guided surgery in the QSC-solution group, while there were not any tumor cells in sight after fluorescence imaging-guided surgery in the QSC-Lip group, suggesting that QSC-Lip may be served as an effective navigating probe to guide the accurate and complete surgical resection of intracranial GBM.
Although previous studies showed that CGT presented an obvious anticancer effect on glioma in preclinical[47] and clinical study,[48,49] a few studies focused on the distribution of CGT in the glioma site. In this study, the CGT distribution in glioma after treatment with QSC-Lip was detected by the HPLC assay. The distribution of CGT in glioma after QSC-Lip under an external MT was significantly higher than that of QSC-Lip without MT. The antitumor efficacy of QSC-Lip was also evaluated by real-time MR/NIR fluorescence imaging. In our study, a lower dose of QSC-Lip (equivalent CGT dose of 2 mg kg−1) was effective to inhibit growth of glioma than the reported dose of free CGT solution (4 mg kg−1) in this publication.[49] Moreover, it was proved that the therapeutic mechanism of QSC-Lip was highly correlated with the enhanced distribution of CGT. Finally, in vivo toxicity study indicated there was no pathological damage of QSC-Lip against major organs of rats. All these results indicated that QSC-Lip might be a potential theranostic platform for imaging-guild surgery and postoperational therapy.

5. Conclusions

In this study, a theranostic liposome (QSC-Lip) was prepared based on QDs, SPIONs, and CGT for MRI/NIR imaging and treatment of glioma. First, the physiochemical properties of QSC-Lip including the particles size, zeta potential, magnetic saturation intensity, and entrapment efficiency were carefully characterized. The results showed that both SPIONs and QDs were completely coencapsulated in liposome. Besides, CGT was also effectively encapsulated into the therapeutic liposome with an encapsulation efficiency of ≈88.9%. QSC-Lip exhibited a diameter of 100 ± 1.24 nm, zeta potential of −17.10 ± 0.11 mV, and good stability in several mediums. Second, in vitro cellular toxicity and cellular uptake of QSC-Lip were further evaluated by using the C6 cell line. Cellular uptake of QSC-Lip was Cilengitide significantly enhanced by C6 cells under imposing MT, which resulted in its stronger cytotoxicity against C6 cells than free CGT. Moreover, QSC-Lip can effectively cross blood–brain barrier and mainly target to tumor site in assistance with an external magnet and UTMD. The specific distribution of QSCLip in glioma not only produced an obvious negative-contrast enhancement effect on glioma by MRI but also made tumor zone emitting strong fluorescence under MT. Besides, the dualimaging ability of QSC-Lip to guide the accurate resection of glioma was confirmed by surgery. Finally, in vivo inhibition of tumors was also confirmed after intravenous injection of QSCLip under MT. Overall, the integrated liposome as a theranostic carrier may be very useful for the accurate surgery resection.

References

[1] T. Schneider, C. Mawrin, C. Scherlach, M. Skalej, R. Firsching, Dtsch. Aerztebl. Int. 2010, 107, 799, quiz 808.
[2] C. Alifieris, D. T. Trafalis, Pharmacol. Ther. 2015, 152, 63.
[3] D. LM, New Engl. J. Med. 2001, 344, 114.
[4] M. Roger, A. Clavreul, M. C. Venier-Julienne, C. Passirani, C. Montero-Menei, P. Menei, Biomaterials 2011, 32, 2106.
[5] W. Shang, C. Zeng, Y. Du, H. Hui, X. Liang, C. Chi, K. Wang, Z. Wang, J. Tian, Adv. Mater. 2017, 29.
[6] Z. Zhou, M. Qutaish, Z. Han, R. M. Schur, Y. Liu, D. L. Wilson, Z. R. Lu, Nat. Commun. 2015, 6, 7984.
[7] X. Gao, M. Zhai, W. Guan, J. Liu, Z. Liu, A. Damirin, ACS Appl. Mater. Interfaces 2017, 9, 3455.
[8] F. Ye, A. Barrefelt, H. Asem, M. Abedi-Valugerdi, I. El-Serafi, M. Saghafian, K. Abu-Salah, S. Alrokayan, M. Muhammed, M. Hassan, Biomaterials 2014, 35, 3885.
[9] Z. Cao, R. Tong, A. Mishra, W. Xu, G. C. Wong, J. Cheng, Y. Lu, Angew. Chem., Int. Ed. Engl. 2009, 48, 6494.
[10] C. Sanson, O. Diou, J. Thevenot, E. Ibarboure, A. Soum, A. Brulet, S. Miraux, E. Thiaudiere, S. Tan, A. Brisson, V. Dupuis, O. Sandre, S. Lecommandoux, ACS Nano 2011, 5, 1122.
[11] M. Muthiah, I. K. Park, C. S. Cho, Biotechnol. Adv. 2013, 31, 1224.
[12] Y. Sun, Z. L. Chen, X. X. Yang, P. Huang, X. P. Zhou, X. X. Du, Nanotechnology 2009, 20, 135102.
[13] H. L. Xu, K. L. Mao, Y. P. Huang, J. J. Yang, J. Xu, P. P. Chen, Z. L. Fan, S. Zou, Z. Z. Gao, J. Y. Yin, J. Xiao, C. T. Lu, B. L. Zhang, Y. Z. Zhao, Nanoscale 2016, 8, 14222.
[14] K. Hayashi, M. Nakamura, W. Sakamoto, T. Yogo, H. Miki, S. Ozaki, M. Abe, T. Matsumoto, K. Ishimura, Theranostics 2013, 3, 366. [15] T. M. Jovin, Nat. Biotechnol. 2003, 21, 32.
[16] X. Gao, L. Yang, J. A. Petros, F. F. Marshall, J. W. Simons, S. Nie, Curr. Opin. Biotechnol. 2005, 16, 63.
[17] Z. Li, P. Huang, J. Lin, R. He, B. Liu, X. Zhang, S. Yang, P. Xi, X. Zhang, Q. Ren, D. Cui, J. Nanosci. Nanotechnol. 2010, 10, 4859.
[18] T. Mikkelsen, C. Brodie, S. Finniss, M. E. Berens, J. L. Rennert, K. Nelson, N. Lemke, S. L. Brown, D. Hahn, B. Neuteboom, S. L. Goodman, Int. J. Cancer 2009, 124, 2719.
[19] C. Mas-Moruno, F. Rechenmacher, H. Kessler, Anti-Cancer Agents Med. Chem. 2010, 10, 753.
[20] H. Chen, J. H. Hwang, J. Ther. Ultrasound 2013, 1, 10.
[21] Y. Y. Liao, Z. Y. Chen, Y. X. Wang, Y. Lin, F. Yang, Q. L. Zhou, Biomed. Res. Int. 2014, 2014, 872984.
[22] C. Y. Ting, C. H. Fan, H. L. Liu, C. Y. Huang, H. Y. Hsieh, T. C. Yen, K. C. Wei, C. K. Yeh, Biomaterials 2012, 33, 704.
[23] Y. Z. Zhao, Q. Lin, H. L. Wong, X. T. Shen, W. Yang, H. L. Xu, K. L. Mao, F. R. Tian, J. J. Yang, J. Xu, J. Xiao, C. T. Lu, J. Controlled Release 2016, 224, 112.
[24] S. S. Desale, S. M. Raja, J. O. Kim, B. Mohapatra, K. S. Soni, H. Luan, S. H. Williams, T. A. Bielecki, D. Feng, M. Storck, V. Band, S. M. Cohen, H. Band, T. K. Bronich, J. Controlled Release 2015, 208, 59.
[25] C. C. Fleischer, U. Kumar, C. K. Payne, Biomater. Sci. 2013, 1, 975.
[26] Z. Gao, X. Liu, Y. Wang, G. Deng, F. Zhou, Q. Wang, L. Zhang, J. Lu, Dalton Trans. 2016, 45, 19519.
[27] S. Fang, J. Lin, C. Li, P. Huang, W. Hou, C. Zhang, J. Liu, S. Huang, Y. Luo, W. Fan, D. Cui, Y. Xu, Z. Li, Small 2017, 13.
[28] H. Dolgos, A. Freisleben, E. Wimmer, H. Scheible, F. Kratzer, T. Yamagata, D. Gallemann, M. Fluck, Pharmacol. Res. Perspect. 2016, 4, e00217.
[29] E. Neill, T. Luks, M. Dayal, J. J. Phillips, A. Perry, L. E. Jalbert, S. Cha, A. Molinaro, S. M. Chang, S. J. Nelson, J. Neurooncol. 2017, 132, 171.
[30] S. Cheng, R. Mi, Y. Xu, G. Jin, J. Zhang, Y. Zhou, Z. Chen, F. Liu, J. Cancer Res. Clin. Oncol. 2017, 143, 941.
[31] C. Y. Cheng, K. L. Ou, W. T. Huang, J. K. Chen, J. Y. Chang, C. H. Yang, ACS Appl. Mater. Interfaces 2013, 5, 4389.
[32] M. A. Eckert, P. Q. Vu, K. Zhang, D. Kang, M. M. Ali, C. Xu, W. Zhao, Theranostics 2013, 3, 583.
[33] H. Liu, J. Zhang, X. Chen, X. S. Du, J. L. Zhang, G. Liu, W. G. Zhang, Nanoscale 2016, 8, 7808.
[34] Y. X. Wang, S. Xuan, M. Port, J. M. Idee, Curr. Pharm. Des. 2013, 19, 6575.
[35] C. Sun, J. S. Lee, M. Zhang, Adv. Drug. Delivery Rev. 2008, 60, 1252.
[36] H. Liu, Y. Tan, L. Xie, L. Yang, J. Zhao, J. Bai, P. Huang, W. Zhan, Q. Wan, C. Zou, Y. Han, Z. Wang, J. Colloid Interface Sci. 2016, 478, 217.
[37] B. Xie, M. A. Stammes, P. B. van Driel, L. J. Cruz, V. T. Knol-Blankevoort, M. A. Lowik, L. Mezzanotte, I. Que, A. Chan, J. P. van den Wijngaard, M. Siebes, S. Gottschalk, D. Razansky, V. Ntziachristos, S. Keereweer, R. W. Horobin, M. Hoehn, E. L. Kaijzel, E. R. van Beek, T. J. Snoeks, C. W. Lowik, Oncotarget 2015, 6, 39036.
[38] M. A. Bruckman, K. Jiang, E. J. Simpson, L. N. Randolph, L. G. Luyt, X. Yu, N. F. Steinmetz, Nano Lett. 2014, 14, 1551.
[39] R. Hardman, Environ. Health Perspect. 2006, 114, 165.
[40] T. Mikkelsen, C. Brodie, S. Finniss, M. E. Berens, J. L. Rennert, K. Nelson, N. Lemke, S. L. Brown, D. Hahn, B. Neuteboom, S. L. Goodman, Int. J. Cancer 2009, 124, 2719.
[41] P. Huang, Z. Li, J. Lin, D. Yang, G. Gao, C. Xu, L. Bao, C. Zhang, K. Wang, H. Song, H. Hu, D. Cui, Biomaterials 2011, 32, 3447.
[42] H. L. Liu, P. H. Hsu, P. C. Chu, Y. Y. Wai, J. C. Chen, C. R. Shen, T. C. Yen, J. J. Wang, J. Magn. Reson. Imaging 2009, 29, 31.
[43] Z. Medarova, N. V. Evgenov, G. Dai, S. Bonner-Weir, A. Moore, Nat. Protoc. 2006, 1, 429.
[44] G. Caracciolo, O. C. Farokhzad, M. Mahmoudi, Trends Biotechnol. 2017, 35, 257.
[45] W. T. Al-Jamal, K. T. Al-Jamal, B. Tian, A. Cakebread, J. M. Halket, K. Kostarelos, Mol. Pharm. 2009, 6, 520.
[46] X. Gao, Q. Yue, Z. Liu, M. Ke, X. Zhou, S. Li, J. Zhang, R. Zhang, L. Chen, Y. Mao, C. Li, Adv. Mater. 2017.
[47] R. Elmghirbi, T. N. Nagaraja, S. L. Brown, S. Panda, M. P. Aryal, K. A. Keenan, H. Bagher-Ebadian, G. Cabral, J. R. Ewing, Radiat. Res. 2017, 187, 79.
[48] E. R. Gerstner, X. Ye, D. G. Duda, M. A. Levine, T. Mikkelsen, T. J. Kaley, J. J. Olson, B. L. Nabors, M. S. Ahluwalia, P. Y. Wen, R. K. Jain, T. T. Batchelor, S. Grossman, Neuro-Oncology 2015, 17, 1386.
[49] L. B. Nabors, K. L. Fink, T. Mikkelsen, D. Grujicic, R. Tarnawski, D. H. Nam, M. Mazurkiewicz, M. Salacz, L. Ashby, V. Zagonel, R. Depenni, J. R. Perry, C. Hicking, M. Picard, M. E. Hegi, B. Lhermitte, D. A. Reardon, Neuro-Oncology 2015, 17, 708.