Long-Term Effects of Magnetically Targeted Ferumoxide-Labeled Human Neural Stem Cells in Focal Cerebral Ischemia

Cell Culture

A stable, immortal female hNSC line, HB1.F3 (F3), was generated from primary fetal human brain cell cultures by infection with a recombinant replication incompetent retroviral vector pLNX.v-myc (12,13). F3 human NSCs express ATP-binding cassette subfamily G member 2 (ABCG2), nestin, and Musashi1, which are cell type-specific markers for NSCs and were grown in Dulbecco's modified Eagle's medium with high glucose (Sigma, St. Louis, MO, USA) supplemented with 5% fetal bovine serum (Sigma), 5% horse serum (Sigma), 100 U/ml penicillin, and 100 μg/ml streptomycin (Sigma).

Cell Labeling with Ferumoxide

As described in our previous study, F3 human NSCs (5 × 10⁵ cell/ml) were incubated for 60 min at room temperature in culture media containing 25 μg/ml ferumoxide (Feridex IV; Advanced Magnetics, Cambridge, MA, USA) and 2 μg/ml poly-l-lysine (Sigma) prior to transplantation (27).

Animal Model and Transplantation Procedure

All experimental procedures were approved by the Animal Care and Use Committee of Seoul National University Hospital. Male Sprague–Dawley rats weighing 250–300 g (Orient, Seoul, South Korea) were used. Transient focal cerebral ischemia was induced using intraluminal thread occlusion of the left middle cerebral artery (MCAO) as described previously (15).

Three experimental groups were used: ischemia only (IO, n = 10), ischemia with cell injection (IC, n = 12), and ischemia with cell injection and magnet (IM, n = 11). IO rats received only MCAO, IC rats were injected with labeled F3 NSCs after MCAO without placement of magnet, and IM rats were hNSC-administered rats in which a neodymium magnet (5 × 10 × 2 mm, 0.32 T) was attached to the outside of the skull near the damaged left brain with a high-strength adhesive immediately after MCAO modeling. The magnet was removed 28 days after MCAO. Ferumoxide-labeled F3 cells [4 × 10⁶ cells in 500 μl phosphate-buffered saline (PBS; Sigma)] were administered via tail vein injection 24 h post-MCAO.

Measurement of Iron Content

The iron content of rat brains was determined using inductively coupled plasma mass spectrometry [SPS-7000 (Ver. 2.03); Seiko Instruments, Tokyo, Japan].

Briefly, each contralateral and ipsilateral hemisphere was digested for 3 h at 60°C in a mixture (2.4 ml) of 35% hydrochloric acid (1.8 ml; Sigma) and 65% nitric acid (0.6 ml; Sigma). The mixture was diluted by adding sterile water to a volume of 5 ml, followed by filtering. The iron concentration was calculated using a standard curve obtained from ferrous chloride calibration standards containing 0, 1.25, 2.5, and 5 mg/L iron chloride (Sigma) in the above mixture. Iron content was confirmed by ferrozine-based spectrophotometric assays (7) using triplicate aciddigested samples. The average iron content per cell was calculated as the mean value divided by the tissue weight (kg) in each sample.

Histology and Immunohistochemistry

Rats were sacrificed 28 days post-MCAO. Briefly, animals were anesthetized and perfused through the heart with 100 ml cold saline (Sigma), followed by 100 ml 4% paraformaldehyde in 0.1 M phosphate buffer (Sigma). Brains were postfixed in the same fixative for 24 h, followed by immersion in 30% sucrose (Sigma) for 24 h, and sectioning on a cryostat (Leica CM 1900, Wetzlar, Germany) at 30 μm thickness. Five sections at 2-mm intervals were collected for Nissl staining. Briefly, the collected sections are gently mounted on gelatin-coated slides (Marienfeld GmbH & Co., Lauda-Koenigshofen, Germany) with a paintbrush, and then dried at room temperature overnight. The dried slides were dipped in 95% ethanol for 20 min and in chloroform (Merck, Darmstadt, Germany) solution for 10 min. They were dipped in a grades series of ethanol (95%, 100%; Merck) for 2 min each, exchanges of xylene (Sigma) for 5 min each, and then in a graded series of ethanol (100%, 95%, 95%) for 5 min each, after which slides were washed in distilled water (DW) for 5 min and then were dipped in 0.1% thionin solution (Sigma) at room temperature for 7 min. The stained slides were washed in DW for 2 min twice. The slides were dipped in 0.4% formalin acetic solution (Sigma) for 8 s, and then were washed with DW for 2 min. For dehydration, slides were dipped in a graded series of ethanol (95%, 95%) for 2 min, and were differentiated in 100% butanol (Sigma) for 2 min and then in cedarwood oil (Fisher Scientific, Inc.) for 5 min. The sections were cleared in xylene (5 min × 3). Finally, the sections were coverslipped. The area of the lesions was determined by Nissl staining, followed by image analysis (ImageJ; NIH, Bethesda, MD, USA).

Sections were used with Prussian blue staining to detect the presence of transplanted cells in the brain. Prussian blue staining was processed as follows: the sections were washed three times with PBS, incubated for 30 min with 5% potassium ferrocyanide (Sigma) in 5% hydrochloric acid (Sigma), rewashed, and then counterstained with nuclear fast red (Sigma). Prussian blue-stained slices were selected to determine the number of F3 NSCs that migrated into the rat brain. At least five fields of view (objective magnification: 100×) were selected randomly, photographed, and evaluated for Prussian blue staining using the image analyzer program (Qwin Pro, Leica).

We examined the systemic organs to detect any adverse reaction produced by the cells with a magnet. Systemic organs such as liver, spleen, kidney, lung, and heart of IO and IM animals were examined.

To detect transplanted hNSCs in the brain, sections were processed for immunofluorescence microscopy using anti-human nuclear matrix antigen (hNuA, mouse, 1:20; Millipore Biosciences, Temecula, CA, USA). To visualize colocalization of hNuA and cell type-specific markers [neuronal nuclear antigen (NeuN) for neurons or glial fibrillary acidic protein (GFAP) for astrocytes] in the same cell, double staining was employed. Briefly, sections were incubated with the hNuA antibody and cell type-specific antibodies directed against NeuN (1:100, rabbit; Millipore) to identify neurons, or anti-GFAP (1:1,000, rabbit; Millipore) antibody to identify astrocytes, followed by mixture of secondary antibodies [goat anti-mouse IgG Alexa Fluro 568 (1:200; Molecular Probes, Eugene, OR, USA) and goat anti-rabbit IgG Alexa Fluro 488 (1:200; Molecular Probes)]. To investigate if magnetically targeted hNSCs in the IM group enhance angiogenesis, hNSCs that migrated to brain lesions in both the IM and IC groups were stained with von Willebrand factor (vWF, angiogenic marker, rabbit, 1:200; Sigma), followed by secondary antibody goat antirabbit IgG Alexa Fluoro 488 (1:200; Molecular Probes). Cell nuclei were visualized using 4′,6-diamidino-2-phenylindole (DAPI; Sigma) staining. Immunofluorescent-stained sections were viewed using fluorescence microscopy (AH-3, Olympus, Tokyo, Japan).

Behavioral Tests

To determine whether the transplanted hNSCs can improve sensorimotor deficits, behavioral tests were compared (n = 9 for each group). Behavioral testing was performed by experienced testers blinded to the experimental group. The methods for each behavioral test are detailed below. Behavioral tests were performed in a blind fashion using modified neurological severity scores (mNSS). Briefly, the mNSS is a composite of motor, sensory, reflex, and balance tests (2). The scoring system was used with minor modifications, and neurological function was assessed on a scale of 0 to 18 (normal score = 0; maximal deficit score = 18). The limb-placing test was performed weekly, and it was used to assess forelimb and hindlimb responses to tactile and proprioceptive (forefoot, hind leg) stimulation (10,21). Briefly, the test consisted of four limb-placing tasks. The following scores were used to detect impairment of the forelimb and hindlimb: 0 points, the rat performed normally; 1 point, the rat performed with a delay of more than 2 s and/or incompletely; 2 points, the rat did not perform the task. Both sides of the body were tested. The maximum possible score achieved by highly damaged rats was 8.

Statistical Analyses

Statistical analysis was conducted by using GraphPad Prism version 6 for Windows (GraphPad Software, San Diego, CA, USA). Difference between two groups in terms of continuous variables was analyzed by using the Student's t test, and for multiple comparisons, one-way ANOVA was used with the Bonferroni correction. A value of p < 0.05 was considered statistically significant.

Long-Term External Magnet-Enhanced Migration of Injected Ferumoxide-Labeled hNSCs in Focal Cerebral Ischemic Rat Brains

Twenty-eight days after injection of hNSCs in focal cerebral ischemic rats, Prussian blue staining of brain slices showed a higher number of Prussian blue-positive F3 NSCs in IM brains (MCAO + F3 NSCs + magnet) compared to IC brains (MCAO + F3 NSCs) (Fig. 1a). In addition, the area of Prussian blue staining was significantly larger in brain slices from IM rats compared to IC rats (2.1 ± 0.3 mm² vs. 0.8 ± 0.6 mm², p < 0.05) (Fig. 1b). OPEN IN VIEWER

In the IO (MCAO alone) group, there was no significant difference between the iron content of the contralateral (14.4 ± 3.5 mg/kg) and ipsilateral (15.5 ± 3.6 mg/kg) hemispheres of the brain. However, the iron content of the ipsilateral hemisphere was higher than the contralateral hemisphere iron content in both IC (23.3 ± 1.8 vs. 14.7 ± 1.6 mg/kg, p < 0.05) and IM (30.1 ± 6.1 vs. 14.8 ± 0.4 mg/kg, p < 0.05) brain slices. Comparison of iron concentration in the ipsilateral hemisphere between the IC and IM groups showed that there was a significantly higher iron concentration in the IM group (p < 0.05) (Fig. 1c).

No Prussian blue-positive NSCs were found in other organs including liver, spleen, kidney, lung, and heart in IO and IM animals.

Magnetically Targeted hNSCs Induce Long-Term Infarct Volume and Functional Recovery

Control rats exhibited significant functional recovery over the course of the 28 days following MCAO, as assessed by several measurements, including mNSS and limb-placing tests.

mNSS results showed that functional deficits in the IM group were significantly reduced compared with the control IO group 21 days after MCAO (IO vs. IM on day 21: 6.1 vs. 5.2, p < 0.05). Twenty-eight days post-MCAO, the IM group had significantly reduced mNSS results compared to the IO and IC groups (IO vs. IM: 5.7 vs. 4.7, p < 0.05; IC vs. IM; 5.5 vs. 4.7, p < 0.05) (Fig. 2a). OPEN IN VIEWER

In limb-placing tests, the result in the IM group was slightly reduced compared to those in IC and IO groups on days 14 to 21 following MCAO. However, no significant difference was found among the groups on days 1 to 21 after MCAO (Fig. 2b).

In order to examine the effects of magnetically targeted hNSCs on MCAO lesions, rats were sacrificed 28 days post-MCAO, and infarct volumes were analyzed in brain sections processed for Nissl staining. There was a minor difference in infarct volume between IO and IC brains (221.9 ± 34.48 vs. 188.2 ± 22.27 mm³). However, the average lesion size in IM brains (152.4 ± 19.7 mm³) was significantly smaller than lesions observed in either IO or IC brains (p = 0.001 and 0.04, respectively) (Fig. 2c).

Magnetically Targeted hNSCs Enhance Angiogenesis

The migrated hNSCs were positive for hNuA staining, a specific marker for human cellular nuclei. Differentiated hNSCs were also positive for NeuN, a neuronal-specific marker, and GFAP, an astrocyte-specific marker. Twenty-eight days after MCAO, no difference was noted in the differentiation pattern into neurons or astrocytes from magnetically targeted or lesion-targeted hNSCs (Fig. 3a). Thus, we conclude that the presence of the magnetic field did not affect long-term differentiation of hNSCs into neurons or astrocytes. However, the number of cells positive for vWF, an endothelial cell-specific protein, in the IM group was significantly increased compared to the IC group 28 days after MCAO (Fig. 3b). This observation indicates that the IM group significantly promotes angiogenesis. OPEN IN VIEWER