|Home | About | Journals | Submit | Contact Us | Français|
Mesoporous silica-coated hollow manganese oxide (HMnO@mSiO2) nanoparticles were developed as a novel T1 magnetic resonance imaging (MRI) contrast agent. We hypothesized that the mesoporous structure of the nanoparticle shell enables optimal access of water molecules to the magnetic core, and consequently, an effective longitudinal (R1) relaxation enhancement of water protons, which value was measured to be 0.99 (mM−1s−1) at 11.7 T. Adipose-derived mesenchymal stem cells (MSCs) were efficiently labeled using electroporation, with much shorter T1 values as compared to direct incubation without electroporation, which was also evidenced by signal enhancement on T1-weighted MR images in vitro. Intracranial grafting of HMnO@mSiO2-labeled MSCs enabled serial MR monitoring of cell transplants over 14 days. These novel nanoparticles may extend the arsenal of currently available nanoparticle MR contrast agents by providing positive contrast on T1-weighted images at high magnetic field strengths.
In recent years, the use of nanoparticles for biomedical imaging has revolutionized the ability to monitor molecular and cellular events in living organisms. Semiconductor nanoparticles (quantum dots) have been applied as fluorescence probes for cell labeling in optical imaging.1−3 Gold nanoparticles have been investigated in optical imaging and as biomedical sensors because of their unique optical and electrical properties.4,5 Magnetic nanoparticles have also been used as contrast enhancement agents for magnetic resonance imaging (MRI), and as biosensors.6−12 MRI is currently one of the most powerful and widely used imaging modalities that provides high spatial resolution combined with excellent anatomical details.13−15
In the past decade, superparamagnetic iron oxide nanoparticles (SPIO) have become the preferred technique for MRI cell tracking,16−21 and have now entered the clinical arena.22,23 SPIO-labeled cells have much shorter transverse (T2) relaxation time and thus, produce hypointensities (dark regions) on T2/T2*-weighted MR images, which presumably indicate the location of the transplanted cells. Unfortunately, SPIO-labeled cells cannot be distinguished from other hypointense regions, such as hemorrhage and blood clots, which are common in many lesions. Therefore, alternative tracking methods using “positive” contrast agents have been explored, i.e. gadolinium (Gd)-based complexes that can generate hyperintense regions as a result of their predominant effects on the longitudinal (T1) relaxation time of water protons in tissue.24−30 Unfortunately, gadolinium-based contrast agents are now associated with nephrogenic systemic fibrosis (NSF), which makes them less favorable agents,(31) in particular when being retained in cells without rapid clearance from the body. A different approach to enhance MRI contrast is to use manganese as a T1 relaxation agent. Manganese has been mostly used in the form of MnCl2 to study the structure and function of normal or diseased brain.32−34 MnCl2 was used for direct labeling of cells in vitro.(35) Mn has also been chelated and conjugated to proteins to achieve a higher r1 relaxivity.(36) Relaxivity (r1, for longitudinal relaxation) describes the change in the relaxation rates of the water protons in the presence of a given contrast agent. Contrast agents with high r1 values enable lower concentrations of the agent, and consequently, result in higher sensitivity.25,36 Alternatively, the longitudinal relaxation can be improved by increasing the number of Mn atoms per particle that can exchange with the water protons, as in the case of nanoparticles. Indeed, manganese oxide (MnO) nanoparticles have recently been explored as a new T1 MR contrast agent that can delineate fine anatomical features in mouse brain,(37) and to track cells with positive contrast.(38) However, most of the reported manganese oxide nanoparticles provide weak contrast, and the duration of signal is too short for long-term in vivo MR contrast enhancement.37−39
Coating of nanoparticles can significantly improve their stability, biocompatibility, and relaxivity.3,16,21,39 Silica has been recognized as a good candidate for a coating material because it is relatively biocompatible and resistant to biodegradation.40−44 In particular, mesoporous silica material is an excellent candidate due to its stability in aqueous solution and high labeling efficiency.45−53 Furthermore, mesoporous silica allows easy access for water molecules to the magnetic center, which significantly improves the water proton relaxation.54,55 In this study, we report on a novel design of MnO nanoparticles that have a ‘hollow’ MnO core structure and a coating consisting of mesoporous silica (termed HMnO@mSiO2). The high surface area-to-volume ratio and water accessibility through the pores allows for an efficient T1 MR contrast agent. This was demonstrated in vitro and in vivo after labeling of multipotent adipose-derived mesenchymal stem cells (MSCs) with HMnO@mSiO2 nanoparticles.
Mesoporous silica-coated MnO nanoparticles were prepared using the following procedure. First, uniformly sized 15 nm MnO nanoparticles were synthesized by the thermal decomposition of Mn-oleate complex.(56) The resulting MnO nanoparticles stabilized with oleic acid were dispersed in chloroform at a concentration of 10.8 mg Mn/mL, as measured by inductively coupled plasma atomic emission spectroscopy (ICP-AES). A typical mesoporous silica coating onto MnO nanoparticles (MnO@mSiO2) was performed using a sol−gel reaction of tetraethyl orthosilicate (TEOS) in an aqueous solution containing cetyltrimethylammonium bromide (CTAB) and MnO nanoparticles stabilized with oleic acid.(49) First, 600 μL of MnO nanoparticles in chloroform was poured into 5 mL of 0.05 M aqueous CTAB solution and the resulting solution was stirred for 1 h. Forming an oil-in-water microemulsion, the mixture was heated up to 60 °C and maintained at that temperature for 20 min under stirring in order to evaporate the chloroform. The transparent resulting solution of MnO/CTAB was added to a mixture of 25 mL of water and 1.8 mL of 2 M NaOH solution, and the mixture was heated. At approximately 70 °C under stirring, 0.3 mL of TEOS and 1.8 mL of ethylacetate were added to the reaction solution in sequence. The reaction was continued for 12 h at approximately 70 °C. The washing steps for MnO@mSiO2 nanoparticles with ethanol were performed to remove unreacted species, and then the nanoparticles were redispersed in 5 mL of ethanol.
For the hollow core structure of mesoporous silica-coated MnO nanoparticles, an etching process was performed: 100 μL of mild hydrochloric acid solution (4%) (pH ≈ 2.4) was added to the dispersion followed by stirring for 1 h at 70 °C. Acid-etched MnO@mSiO2 nanoparticles were then dispersed in 5 mL ethanol. After washing with ethanol two more times, nanoparticles were redispersed in distilled water. After filtration through a 450 nm (pore size) cellulose acetate filter, HMnO@mSiO2 nanoparticles were finally redispersed in phosphate buffered saline (PBS) (10 mM phosphate, 138 mM NaCl, pH = 7.4) solution.
MRI of phantoms was performed using an 11.7 T Bruker Avance system equipped with a 15 mm birdcage RF coil. T1 and T2 relaxation times were measured using a modified multiecho spin echo (MSME) protocol.(38) Typical settings for T1 measurements were as follows: 64 × 64 matrix size; echo time (TE) = 6.4 ms; and a repetition time (TR) series of 0.2, 0.4, 0.8, 1, 1.5, 3, 4, 6, and 10 s; and 2 averages per measurement. For T2 measurements, the parameters were 128 × 128 matrix; 3000 ms repetition time; echo time series of 6.4, 12.8, 19.2, 25.6, 32.0, 38.4, 44.8, 51.2, 57.6, 64.0, 70.4, and 76.8 ms; and 2 averages per measurement. Data processing was performed using Matlab (Mathworks, Natick, MA).
Electroporation was used to label cells with HMnO@mSiO2 nanoparticles. Mouse MSCs were cultured in 80 cm2 flasks overnight to 80−90% confluence. On the next day, cells were suspended using trypsin-ethylene diamine tetraacetic acid (EDTA), washed with PBS, and counted. Cells were resuspended and transferred to sterile 0.4 cm gap electroporation cuvettes (Gene Pulser; Bio-Rad). Each cuvette contained 2 × 106 cells suspended in 580 μL. Nanoparticles dispersed in PBS were added to the cuvette with a final volume of 700 μL. Cuvettes were kept on ice for one min, and cells were electroporated using a BTX electroporation system (ECM830; Harvard Apparatus). The following electroporation conditions were used: pulse strength = 100 V; N pulses = 5; pulse duration = 5 ms; and pulse interval = 100 ms. After 30 s, cells were transferred to ice for 2 min, suspended in culture medium, transferred to six-well plates. At 24 h following electroporation, cells were washed twice with PBS, harvested using trypsin-EDTA, and counted. For gelatin phantoms, 2 × 106 cells suspended in 50 μL of PBS were transferred to 0.5 mL polypropylene tubes and mixed with 50 μL of 10% gelatin in PBS. The final cell concentration was 2 × 104 cells/μL in 5% gelatin.
A solution of 2 × 106 MSCs suspended in 580 μL PBS was mixed with 120 μL HMnO@mSiO2 (or PBS as a control) and electroporated or incubated. MRI scans of in vitro phantoms were obtained using a Bruker 9.4 T MRI scanner. T1 relaxation was measured using an MSME pulse sequence (for T1; TE = 9 ms, and TR = 0.2, 0.3, 0.5, 0.8, 1, 1.5, 2, 4, 6, and 10 s). Data processing was performed using Matlab (Mathworks, Natick, MA).
Animal experiments were performed in accordance with a protocol approved by Animal Care and Use Committee of the Johns Hopkins University School of Medicine. C15/BL6 male mice (weighing 20 g) were anesthetized with ketamine/xylazine (100/15 mg per kg), and positioned in a stereotaxic device (Stoelting, Wood Dale, IL, U.S.A.). A small skin incision was made in the midline to expose the skull. Using a motorized nanoinjector (Stoelting, Inc.) and 10 μL Hamilton syringe (Hamilton, Reno, NV, U.S.A.) with an attached 33G needle, 1.0 × 105 labeled or unlabeled MSCs were injected into the putamen of mice (n = 2), according to the following coordinates from bregma: anteroposterior [AP] = 1.5 mm; mediolateral [ML] = 1.34 mm; and dorsoventral [DV] = −3.5 mm. Cells were injected slowly over 4 min, and the needle was left in place for 1 min before being withdrawn. The incision was closed with surgical glue, and postoperative analgesia was provided with Ketofen (2 mg/kg) for 72 h.
Serial in vivo MRI scans were performed on a Bruker 9.4 T horizontal bore magnet, equipped with a 30-mm Sawtooth resonator (Bruker), and using an MSME pulse sequence. For T1 measurements, the following parameters were used: TE = 10.5 ms; TR = 1.5 s; FOV = 2.1 × 2.1 cm; and matrix size = 256 × 96.
Scheme 1 describes the synthesis of HMnO@mSiO2 nanoparticles and labeling of MSCs using electroporation. Manganese oxide nanoparticles with a mean diameter of 15 nm, stabilized with oleic acid, were synthesized by the thermal decomposition of manganese oleate complex.(56) Then, mesoporous silica shells were coated with a silica sol−gel reaction of tetraethyl orthosilicate (TEOS) in aqueous solution containing cetyltrimethylammonium bromide (CTAB) and MnO nanoparticles under basic conditions.(49) Hydrophobic MnO nanoparticles capped with oleic acid were transferred to aqueous solution using CTAB, and the subsequent formation of a mesoporous silica shell was achieved. CTAB molecules were used not only as the stabilizing agent to transfer hydrophobic nanoparticles to the aqueous phase, but also as the organic structure-directing template to establish mesopores on the silica shell. The creation of the hollow interior of the MnO cores, along with template removal, was achieved by acid etching and refluxing in ethanol solutions (pH ≈ 2.4). TEM revealed discrete and uniformly sized HMnO@mSiO2 nanoparticles. The average diameter of the particles was about 65 nm and the MnO core size was 15 nm. The nanoparticles were all well-dispersed and separated from one another. The mesoporous silica shell and hollow MnO core structures can be clearly seen on the TEM (Figure (Figure1a)1a) and HRTEM (Figure (Figure1b)1b) images. We also confirmed the mesoporous structures of HMnO@mSiO2 nanoparticles using N2 adsorption/desorption isotherms. The average pore diameter calculated using the Barrett−Joiner−Halenda (BJH) method was 3.3 nm. The Brunauer−Emmett−Teller (BET) surface area and pore volume of the HMnO@mSiO2 nanoparticles were measured to be 181.3 m2 g−1 and 0.38 cm3 g−1 (Figure S1, Supporting Information), respectively. These BET results indicate that HMnO@mSiO2 nanoparticles are highly porous and have large surface area, which enables rapid access for water molecules to the manganese core through the nanopores of the particles. In addition, the hollow interior of the MnO cores also enables more Mn ions to be exposed to water molecules at the inner surface.(57) HMnO@mSiO2 nanoparticles that have a large surface area at the manganese center resulting from these novel structures can be expected to have high T1 relaxivities.
The MRI properties of the HMnO@mSiO2 nanoparticles in water were characterized using an 11.7 T MR scanner. The paramagnetic nanoparticles shortened the T1 of water protons significantly (Figure (Figure1c).1c). The molar relaxivity (effectiveness as contrast agent) was obtained by measuring the relaxation rate with increasing concentrations of nanoparticles (Figure (Figure1c),1c), and was calculated to be 0.99 mM−1 s−1 (Figure (Figure1d).1d). This relaxivity is significantly higher as compared to that measured for MnO nanoparticles encapsulated with PEG-phospholipids (MnO@PEG-phospholipid),37,38 dense silica-coated MnO nanoparticles (MnO@dSiO2)(39) (Figure S2, Supporting Information), and nonetched mesoporous silica-coated MnO nanoparticles (MnO@mSiO2). Summarized in Table Table1,1, the r1 values of HMnO@mSiO2 nanoparticles were 12.4, 9.0, and 1.5 times higher than those of the MnO@dSiO2, MnO@PEG-phospholipid, and MnO@mSiO2 nanoparticles, respectively. These findings suggest that the majority of the improvement over other nanoparticles is derived from the coating, consistent with our hypothesis that the mesoporous shell allows better water exchange with the MnO core. Water diffusion across mesoporous silica has been studied previously using pulsed field gradient (PFG) diffusion NMR.(58) It was shown that water diffuses anisotropically inside mesoporous silica in comparison to isotropic diffusion in the surrounding environment, with the fastest diffusion component occurring along the channels in the mesoporous silica. The existence of nanopores allows water diffusion to the manganese core leading to efficient relaxation of water in the vicinity of the nanoparticles. In fact, using PFG NMR and magic angle spinning NMR, similar diffusion properties were shown for larger molecules such as pentane, cyclohexane, n-dodecane, and amino acids such as alanine.(59) An additional improvement over MnO@mSiO2 (1.5 fold, see above) may be a result of the acidic etching, which increases the surface area-to-volume ratio, and further increases the exchange of the water with Mn2+ ions. This improvement is in agreement with previously reported relaxivity for hollow manganese oxide nanoparticles.(57) Recent reports on other manganese-based nanoparticles describe wide range of relaxivity (r1) values.39,57,60,61 It is difficult to compare the relaxivities of those nanoparticles since the relaxivity was measured at different field strengths (Table (Table1).1). When measured at lower field (1.5 T), the HMnO@mSiO2 nanoparticles gave higher r1 but similar r2 values (r1 = 1.72 mM−1 s−1, r2 = 11.30 mM−1 s−1) in comparison to those measured at 11.7 T. As expected, the r2/r1 ratio at lower field is lower. The closer this ratio is to 1, the better the nanoparticle will function as “positive” contrast agent. For example, “Magnevist” (a commercial gadolinium complex) has an optimal r2/r1 ratio of 1 (r1 = 4.6 mM−1 s−1, r2 = 4.5 mM−1 s−1) at 1.5 T,(62) in comparison to 6.56 for HMnO@mSiO2 at the same field strength. When compared to manganese compounds33,36 such as MnCl2 and chelated manganese, the HMnO@mSiO2 nanoparticles have lower relaxivity. Nevertheless, the nanoparticles were designed specifically for cell labeling and tracking which is not possible with these compounds, even if the relaxivity had to be compromised.
The X-ray diffraction (XRD) pattern (Figure S3, Supporting Information) revealed polycrystalline properties of the HMnO@mSiO2 nanoparticles, consisting mainly of tetragonal Mn3O4 phase and small fraction of cubic MnO phase. The existence of an amorphous silica shell also contributes to the broad peaks of the patterns. In particular, it was found that the crystallinity of the nanoparticle was decreased after the acid etching. The dissolution of the cubic MnO phase from the core part was also examined. After acid etching, HMnO@mSiO2 nanoparticles were mainly composed of the tetragonal Mn3O4 phase.
HMnO@mSiO2 nanoparticles are stable in water and PBS solution for over a month. The hydrodynamic diameter of HMnO@mSiO2 nanoparticles in PBS solution was measured to be 86 nm using Dynamic Light Scattering (Figure S4, Supporting Information). This demonstrates that nanoparticles are stable in PBS and do not form aggregates, which is desirable for its use as intracellular labeling agent. This good colloidal stability of the HMnO@mSiO2 nanoparticles results from the negative charge of the silica surface. As shown by zeta potential (ζ) measurements (Figure S5, Supporting Information), the surface potential of the HMnO@mSiO2 nanoparticles was −30.43 (mV) in PBS (10 mM phosphate, 138 mM NaCl, pH = 7.4) solution.
In order to assess the feasibility of HMnO@mSiO2 nanoparticles for cell labeling, we initially evaluated the cellular uptake in vitro. In this study, we have chosen to use multipotent mesenchymal stem cells (MSCs), which possess the ability to migrate to sites of tissue injury, have many possible therapeutic applications, including the replacement of damaged tissue and the treatment of cancer.63−65 Successful clinical translation for the application of MSCs requires an imaging modality that noninvasively permits their visualization and localization in a living organism. Mouse adipose-derived MSCs were labeled using either electroporation or a simple incubation approach. Electroporation uses an electrical pulse to induce a change in the electrochemical permeability of the cell membrane and enables efficient intracytoplasmic labeling of cells.(66) This approach is commonly used for transferring DNA and chemotherapeutic drugs into cells, and was recently applied to cell labeling with superparamagnetic iron oxide nanoparticles (SPIO)(67) and MnO nanoparticles.(38) We hypothesized that electroporation could also be efficiently used for cell labeling with HMnO@mSiO2 nanoparticles. Since the cell membrane is negatively charged,(68) the negative charge of the nanoparticle would result in repulsion of the nanoparticles by the cell membrane, and thus, uptake would be hampered. However, with electroporation, negatively charged HMnO@mSiO2 nanoparticles can be effectively introduced into cells without using other transfection agents, such as cationic, positively charged molecules.(69)
Figure Figure2a2a shows a T1-weighted MR image of tubes containing MSCs electroporated with different concentrations of HMnO@mSiO2 nanoparticles suspended in 5% gelatin. The higher the nanoparticle concentration used for labeling, the brighter the tube appeared on the image. Figure Figure2b2b demonstrates that electroporation with nanoparticles shortens T1 much more efficiently than direct incubation of the cells with the nanoparticles. Figure Figure2c2c shows the linear correlation between the relaxation rate (R1 = 1/T1) and the concentration of nanoparticles that was used to label the cells, both by electroporation (R2 = 0.994) and by incubation (R2 = 0.989). At higher nanoparticle concentration, the electroporation is much more efficient than incubation. For these samples, the manganese ion content was measured using ICP measurement. The Mn concentration was 9.0 μg Mn/mL (= 0.35 pg Mn/cell) and 1.8 μg Mn/mL (= 0.09 pg Mn/cell) for phantom of MSCs labeled with electroporation and incubation, respectively. These findings indicate that 26.23% and 5.24% of the added nanoparticles were taken up by MSCs using electroporation and incubation, respectively. Next, both cellular labeling using electroporation and simple incubation were investigated using nanoparticles conjugated with rhodamine B isothiocyanate (RITC) (Figure (Figure3).3). This fluorescent organic dye could be easily conjugated to the silica surface using simple silane conjugation chemistry, as previously reported.42,49,70,71 For the same nanoparticle concentration, MSCs labeled with electroporation (Figure (Figure3a)3a) showed a higher uptake of nanoparticles than MSCs labeled with incubation (Figure (Figure3b).3b). There was also a clear concentration-dependent cellular uptake of the nanoparticles. As was previously shown, compounds that were introduced into the cells via incubation are more likely to accumulate in endosomes, while with electroporation the compounds are freely distributed within the cytoplasm. This may also explain in part the lower relaxivity observed for cells incubated with the particles in comparison to electroporation, since with incubation the particles are in confined region (endosome) and have restricted exchange with water in contrast to electroporated nanoparticles.(72)
To evaluate the possible cytotoxic effects of the HMnO@mSiO2 nanoparticles on MSCs, cell viability was assessed with the MTS (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium) assay. As shown in Figure S6, high cell viability (more than 75%) was still attained at 24 h after electroporation with HMnO@mSiO2 nanoparticles at a concentration of 68.6 μg Mn/mL. The effect of electroporation with HMnO@mSiO2 nanoparticles on the differentiation of adipose-derived MSCs was also studied. It was found that MSCs retained the ability to differentiate into adipocytes; however, differentiation into an osteogenic phenotype could not be induced (Figure S7, Supporting Information). These findings, taken together, indicate that HMnO@mSiO2 nanoparticles by electroporation have a minimal impact on cell viability, but its effect on cell differentiation should be carefully evaluated for each specific application. In this respect, it is important to note that while SPIOs were reported to inhibit chondrogenesis,(73) SPIO-cell tracking clinical studies have nevertheless been initiated.(22)
In order to study the feasibility of HMnO@mSiO2 nanoparticles for long-term in vivo cell tracking, the fate of transplanted, labeled MSCs was serially monitored with MRI. In mice transplanted with unlabeled MSCs, no hyperintensity could be detected (red arrow, Figure Figure4a).4a). However, in mice transplanted with HMnO@mSiO2-labeled MSCs, a hyperintense region was observed at the transplantation site (green arrows, Figure Figure4b).4b). Serial imaging over 14 days demonstrated that the nanoparticles were stable enough to produce sustained contrast in vivo and, therefore, are suitable for monitoring noninvasively the fate of transplanted cells.
The current HMnO@mSiO2 nanoparticles can be further applied to MRI contrast agents for detecting tumors when conjugated with antibodies(37) or RGD peptides(60) that can be targeted directly to cancer cells or to the tumor vasculature, respectively. Furthermore, stem cells labeled with the HMnO@mSiO2 nanoparticles can assist in detecting tumors because stem cells tend to migrate into tumor sites.74,75 And most importantly, such stem cells can be engineered to perform an antitumor activity.
In summary, we have synthesized and characterized mesoporous silica-coated hollow manganese oxide nanoparticles. These HMnO@mSiO2 nanoparticles showed a significantly higher r1 relaxivity over other existing manganese oxide nanoparticle-based contrast agents. The porous coating, which enables water exchange across the shell, combined with the large surface area-to-volume ratio resulting from the novel structure increases water accessibility to the manganese core and consequently provides enhanced T1 contrast. These nanoparticles showed high cellular uptake by adipose-derived MSCs, using electroporation, and were detected with MRI both in vitro and intracranially in vivo over a prolonged time period. In light of these findings, HMnO@mSiO2 nanoparticles have a great potential for MRI cell tracking using positive contrast.
This work was supported by NIH Grants 1R21EB008769, 5R21NS065284, and 2RO1 NS045062. T.H. acknowledges financial support by the Korean Ministry of Education, Science and Technology through National Creative Research Initiative (R16-2002-003-01001-0), Strategic Research (2010-0029138), and World Class University (R31-10013) Programs of National Research Foundation (NRF) of Korea. This work was also supported, in part, by the NIH Grant 5K08NS055851 and an HHMI grant for A.Q.H., K.Y., and H.Z. as well as the Doris Duke grant for E.M. We thank Segun M. Bernard and Derrick Pyomin Jeon for technical assistance.
Detailed experimental procedures, N2 adsorption/desorption isotherm of HMnO@mSiO2, TEM images of PEG-phospholipid capped MnO nanoparticles and dense silica-coated MnO nanoparticles, XRD pattern, DLS data and zeta potential data of HMnO@mSiO2 nanoparticles, cell viability data at 24 h after electroporation or incubation with HMnO@mSiO2 nanoparticles, and differentiation images of MSCs labeled with HMnO@mSiO2 nanoparticles. This material is available free of charge via the Internet at http://pubs.acs.org.