Original Research
Molecular Imaging
April 2007

Superparamagnetic Iron Oxide Labeling and Transplantation of Adipose-Derived Stem Cells in Middle Cerebral Artery Occlusion-Injured Mice

Abstract

OBJECTIVE. Adipose-derived stem cells are an alternative stem cell source for CNS therapies. The goals of the current study were to label adipose-derived stem cells with superparamagnetic iron oxide (SPIO) particles, to use MRI to guide the transplantation of adipose-derived stem cells in middle cerebral artery occlusion (MCAO)-injured mice, and to localize donor adipose-derived stem cells in the injured brain using MRI. We hypothesized that we would successfully label adipose-derived stem cells and image them with MRI.
MATERIALS AND METHODS. Adipose-derived stem cells harvested from mice inbred for green fluorescent protein were labeled with SPIO ferumoxide particles through the use of poly-l-lysine. Adipose-derived stem cell viability, iron staining, and proliferation were measured after SPIO labeling, and the sensitivity of MRI in the detection of SPIO-labeled adipose-derived stem cells was assessed ex vivo. Adult mice (n = 12) were subjected to unilateral MCAO. Two weeks later, in vivo 7-T MRI was performed to guide stereotactic transplantation of SPIO-labeled adipose-derived stem cells into brain tissue adjacent to the infarct. After 24 hours, the mice were sacrificed for high-resolution ex vivo 7-T or 9.4-T MRI and histologic study.
RESULTS. Adipose-derived stem cells were efficiently labeled with SPIO particles without loss of cell viability or proliferation. Using MRI, we guided precise transplantation of adipose-derived stem cells. MR images of mice given injections of SPIO-labeled adipose-derived stem cells had hypointense regions that correlated with the histologic findings in donor cells.
CONCLUSION. MRI proved useful in transplantation of adipose-derived stem cells in vivo. This imaging technique may be useful for studies of CNS stem cell therapies.

Introduction

Stem cell therapy for CNS repair has great clinical potential [1, 2]. Neuronal stem cells and embryonic stem cells are commonly proposed stem cell sources for neuronal therapies, although logistic and ethical concerns limit use of these cells for clinical therapies [3, 4]. Non-neuronal adult tissues, such as mesenchymal stem cells, can differentiate into cells with neuronal and glial characteristics [5-10]. Adult adipose tissue has become recognized as an alternative source of adult stem cells. These cells have been termed adipose-derived stem cells and can differentiate along various pathways [11-16]. Adipose-derived stem cells can differentiate in vitro into cells with several neuronal and glial characteristics, including expression of neuronal and glial proteins [13, 14, 17-19]. Adipose-derived stem cells have been shown to survive after transplantation into rodent brains, but understanding of adipose-derived stem cell function after transplantation remains limited [18].
MRI, which can offer critical information in stem cell studies, requires labeling of donor cells for stem cell localization [20]. Labeling is performed most commonly with superparamagnetic iron oxide (SPIO) nanoparticles. Because of the small crystal size (7-10 nm), SPIO particles exhibit magnetic moments that align in an applied magnetic field, produce microscopic field gradients that dephase the neighboring proton magnetic moments, and reduce T2* relaxation time. The most common method of SPIO labeling is use of a cationic transfection agent, such as poly-l-lysine [20, 21]. These agents are relatively easy to use, but cellular toxicity and effects on differentiation can be troublesome.
MRI may improve the use of SPIO-labeled adipose-derived stem cells in a middle cerebral artery occlusion (MCAO) model. This model is commonly used for study of cerebral ischemia in rodents [22, 23]. Although the variability of infarct size limits use of this model, MRI can be used for precise definition of the area of infarction and for evaluation of the implantation of SPIO-labeled stem cells. Several investigators [21, 24-26] have found it feasible to image SPIO-labeled stem cells in the spine and brain, but the efficacy of this technology with adipose-derived stem cells is unknown. We hypothesized that MRI can be used to define a target adjacent to the infarct for stem cell transplantation and to localize transplanted SPIO-labeled adipose-derived stem cells. The goals of the current study were to establish methods of labeling adipose-derived stem cells with SPIO particles, to use MRI to guide transplantation of adipose-derived stem cells in MCAO-injured mice, and to use MRI to localize donor adipose-derived stem cells in the injured brain.

Materials and Methods

Cell Harvest and Culture

For all experiments, adipose-derived stem cells from the adipose tissue of the inguinal fat pads of four to six 6- to 10-week-old transgenic C57BL/6-Tg(UBC-GFP)30Scha/J mice (Jackson Laboratories) inbred for expression of green fluorescent protein (GFP+) were isolated according to previously published methods [13, 17]. Briefly, we mechanically dissociated the adipose tissue, digested it with collagenase type 1, and centrifuged the suspension to separate the floating adipocytes from the stromal vascular fraction. We cultured the cells in the stromal vascular fraction with Dulbecco minimal essential medium supplemented with 10% fetal calf serum (control medium) in T-75 flasks (Corning). We subcultured adherent adipose-derived stem cells by serial passage and used adipose-derived stem cells after two to five passages. All animal procedures were approved by the Duke University Institutional Animal Care and Use Committee.

Cell Labeling and Viability

We first performed a series of in vitro experiments to evaluate the effect of ferumoxide (Feridex, Berlex Laboratories) and poly-l-lysine hydrobromide (PLL; MW, 388 kd; Sigma) on adipose-derived stem cell viability, iron labeling, and proliferation. Adipose-derived stem cells were labeled with SPIO particles by incubation with a complex of ferumoxide and PLL according to a previously published protocol [27]. Duplicate cultures of adipose-derived stem cells were labeled for 24 hours with various concentrations of ferumoxide (iron, 0, 12.5, or 25 μg/mL), PLL (0, 0.188, 0.375, or 0.75 μg/mL), or both agents. Control adipose-derived stem cells were cultured in control medium without ferumoxide or PLL.
A diluted solution of ferumoxide particles (Feridex; stock elemental iron, 11.2 mg/mL) was added to culture medium without cells and mixed well. PLL was added from a 1.5-mg/mL stock solution for reconstitution of a desired final concentration of elemental iron. The medium was gently mixed for 60 minutes at room temperature to allow complexing of ferumoxide particles to PLL. Labeling was initiated by replacement of the initial control medium with fresh control medium containing ferumoxide-PLL complexes for 24 hours. Cells were then washed twice in phosphate-buffered saline solution (PBS) to remove excess iron-PLL complexes and then were placed in fresh control medium for an additional 24 hours for recovery. After cell recovery, cells were harvested with trypsin digest or were examined in situ for viability and iron staining.
For assessment of cell viability, cultures were washed three times with PBS and stained with trypan blue (Sigma). Viable cells were considered those that did not stain with trypan blue. Two independent observers using light microscopy (10× magnification) enumerated cell viability in 10 separate fields. To assess iron staining, we used Prussian blue (ferric hexacyanoferrate and hydrochloric acid, Sigma) to determine the percentage of cells stained with iron. Cultures were incubated for 30 minutes with 2% potassium ferrocyanide (Perls' Prussian blue reagent) in 3.7% hydrochloric acid, washed again, and counterstained with nuclear fast red. A positive test result was defined as the presence of brown iron particles visible within cell cytoplasm. All experiments were performed on duplicate cultures of adipose-derived stem cells at each concentration of PLL and ferumoxide, and each experiment was repeated three times.
To assess adipose-derived stem cell proliferation, we used the 3-(4,5-dimethyl-thiazol-2-yl)-2,5-diphenyl-tetrazolium bromide (MTT) assay (ATCC) as described [28]. Adipose-derived stem cells were seeded at a density of 10,000 cells/well in a 96-well plate and were allowed to grow to 60% confluence in 5% CO2 at 37°C in control medium. In the three experimental groups, the medium was changed to include ferumoxide, PLL, or ferumoxide-PLL complex. Control cells included control medium alone. On the basis of the results of our initial viability studies (see Results), we used ferumoxide at a concentration of 12.5 μg/mL and PLL at a concentration of 0.375 μg/mL for proliferation studies. After 1-, 3-, 5-, and 7-day culture periods, adipose-derived stem cell proliferation was measured with the MTT assay according to the manufacturer's recommendations. To each culture on the 96-well plate, we added 10 μL of MTT and incubated the cells for 4 hours at 37°C. We then added 100 μL of lysis solution and incubated the sample at room temperature for 2 hours. The absorbance of each sample was recorded at 570 nm in an enzymelinked immunosorbent assay reader and expressed as optic density (OD) after blank subtraction. All experiments were performed in quadruplicate. We compared the proliferation rate (difference between ODs) from days 3-7 in each experimental group with that of the control group. Comparison of CIs showed differences in proliferation rates between each experimental group and the control group.

MRI of SPIO-Labeled Cells In Vitro

To develop techniques for MRI in adipose-derived stem cell transplantation, we performed a series of phantom studies to identify the sensitivity of MRI in the detection of small numbers of adipose-derived stem cells implanted in agar. Cultures of adipose-derived stem cells were incubated with ferumoxide-PLL complex. On the basis of the results of our initial viability studies (see Results), we used ferumoxide at a concentration of 12.5 μg/mL and PLL at a concentration of 0.375 μg/mL. Adipose-derived stem cell cultures were incubated with ferumoxide-PLL complex for 24 hours, washed twice with PBS to ensure removal of unbound ferumoxide-PLL complexes, and recovered for 24 hours in control medium before being released by trypsin digest and enumerated. Cell dilutions were made to suspend various numbers of adipose-derived stem cells (0, 10, 100, 500, 1,000, and 5,000 cells) in 5 μL of control medium. The cell aliquot was imbedded by use of a pipetter in agar (4% weight/weight) in plastic Petri dishes (Corning) for MRI. All experiments were performed in duplicate, and each experiment was repeated three times.
MRI was performed with a 7-T MRI microscopy system consisting of a superconducting 21-cm horizontal-bore magnet (Magnex, Varian) interfaced with a console (Excite, GE Healthcare). The Petri dish sample was placed on top of a 3.0-cm-diameter radiofrequency surface coil. A series of multislice T2-weighted and T2*-weighted images (TR/TE, 250/10-40 in increments of 5 milliseconds; matrix size, 128 × 128; field of view, 4.0 cm; slice thickness, 1 mm) were acquired with standard spin-echo and gradient-echo sequences.

Middle Cerebral Artery Occlusion

To develop techniques for the use of MRI in guiding the transplantation of adipose-derived stem cells in MCAO-injured mice and to localize donor adipose-derived stem cells in the injured brain, we produced MCAO injury in mice as described [29]. Briefly, 8- to 10-week-old male Balb/C mice (Jackson Laboratories) (n = 12) were anesthetized with 1.0-1.5% halothane in 50% O2/N2, intubated, and mechanically ventilated. Rectal temperature was monitored and servo-regulated with surface heating and cooling at 37.0°C throughout the procedure. Inspired halothane concentration was kept between 0.6% and 1.0%. The right common carotid artery was identified through a midline cervical skin incision. The external carotid artery was transected, and the internal carotid artery was dissected for visualization of the origin of the pterygopalatine artery. A 7-0 nylon monofilament suture, tapered at the tip and coated with silicone, was inserted into the external carotid artery stump and advanced approximately 11 mm into the internal carotid artery. The mice were subjected to 60 minutes of MCAO, after which the occlusive filament was removed, the wounds were closed, and the mice were extubated and placed in an O2-enriched environment (Fio2, 50%) for 1 hour. Two of the 12 mice died within several days of MCAO.

In Vivo MRI

Two weeks after MCAO injury, the surviving mice (n = 10) were imaged with 7-T MRI. The mice were anesthetized with ketamine (20-60 mg/kg) and xylazine (2-5 mg/kg) administered intraperitoneally and then placed supine with the head above a 1.0-cm-diameter surface coil. After endotracheal intubation, anesthesia was maintained with 2-3% isoflurane. A custom-made ventilator was used to deliver the isoflurane. The mice were mechanically ventilated at a rate of 90 breaths/min at a tidal volume of 0.4 mL. An airway pressure tracing was monitored with a solid-state pressure transducer on the breathing valve, and ECG recording was performed with electrodes taped to the footpads of the mice. Body temperature was recorded with a rectal thermistor that was also used to control heat lamps that maintained core body temperature at 36.5°C.
Multislice T2-weighted spin-echo images (TR, 1.0; matrix size, 128 × 128; field of view, 3.5 cm; thickness of consecutive slices, 1 mm; number of signals averaged, 4) were obtained in the anatomic coronal orientation. We performed all sequences within 40-60 minutes per mouse. Two of these 10 mice did not have a visible infarct and were not subjected to stem cell transplantation.
MRI images were reviewed to assess the infarct and to determine the stereotactic coordinates for transplantation of donor adipose-derived stem cells into either the hippocampus (n = 4) or the striatum (n = 4) immediately adjacent (within 0.5 mm) to the edge of the infarct. Briefly, we identified the coronal image section showing both the hippocampus and the infarct, usually within 0.2 mm rostral or caudal to the bregma. Bregma refers to the junction of the sagittal and coronal sutures at the top of the skull and is a common, easily identifiable landmark on the mouse skull with which to guide stereotactic transplantation. Using the MRI images, we correlated surrounding major structures (lateral ventricles, hippocampus, central aqueduct) with images in a standard mouse brain anatomic atlas [30]. With these coordinates, we defined the stereotaxic coordinates for transplantation of donor adipose-derived stem cells into the area immediately adjacent to the infarct.

Stereotactic Transplantation

Within 24 hours after in vivo MRI, the mice were anesthetized with tribromoethanol (10 mg/kg intraperitoneally) and placed in a standard stereotaxic device for transplantation of adipose-derived stem cells. On the basis of stereotaxic coordinates defined on in vivo MRI, mice received transplants into either the hippocampus (n = 4) or the striatum (n = 4) in the area within 0.5 mm of the mature infarct. A Hamilton syringe was used to inject donor cells (500, 1,000, 2,000, or 5,000 cells/mouse) within a 0.5-μL volume over a 20-minute period to minimize backflow of cells along the injection track. Each cell dose was administered to two mice. Two control MCAO-injured mice received transplants of 1,000 unlabeled adipose-derived stem cells into the striatum.

High-Resolution Ex Vivo MRI

Twenty-four hours after adipose-derived stem cell transplantation, the mice were sacrificed for high-resolution ex vivo MRI and histologic examination for confirmation of placement of donor cells in the area adjacent to the infarct. The mice were anesthetized with xylazine (10 mg/kg) and ketamine (90 mg/kg) administered intraperitoneally. Transcardial perfusion of 20 mL of warm (37°C) PBS was followed by 20 mL of PBS with 10% buffered formalin and 5% gadoteridol (ProHance, Bracco Diagnostics). To avoid physically perturbing the brain, we removed the entire mouse head after at least 24 hours of fixation and placed it inside a 2.0-cm-diameter cylindric holder.
High-resolution ex vivo MRI was performed with either a 7-T (n = 4) or a 9.4-T (400 MHz, vertical bore Oxford superconducting magnet) (n =4) system with a 2.0-cm-diameter birdcage radiofrequency coil. We used both 7-T and 9.4-T magnets to determine whether differences existed in the quality of MR signal and ease of localizing donor cells. An isotropic 100-μm-resolution image (100/7; matrix size, 512 × 256 × 256; number of signals averaged, 2) was obtained with a standard 3D spin-echo sequence. We examined the MR images in the coronal, axial, and sagittal planes to identify the hypointense region corresponding to the donor cell transplant. In some cases, 3D images were volume rendered with VoxelView Ultra 2.5 software (Vital Images) on a Reality Engine2 workstation (Silicon Graphics). Volume-rendered images in all three orientations and reformatted slices were captured on the workstation for annotation with Adobe PhotoShop 4.0 v4.0.1 PowerPC software (Adobe Systems).

Tissue Histology

After ex vivo MRI, brain samples were cryosectioned at 30-μm thickness through the entire coronal rostral-caudal extent of the infarct. Donor adipose-derived stem cells (GFP+) were detected by immunohistochemical analysis with a peroxidase-conjugated anti-GFP marker as described [17]. In addition, brain sections were treated with Prussian blue to detect iron and with Nissl stain. Prussian blue staining (ferric hexacyanoferrate and hydrochloric acid) was performed by incubation of sections for 30 minutes with 2% potassium ferrocyanide (Perls' Prussian blue reagent) in 3.7% hydrochloric acid. The sections were washed again and counterstained with nuclear fast red.
To ensure that the coronal brain sections were in the same coronal plane as the MR images, we correlated MRI and histologic sections with images in a standard mouse brain atlas [30] to determine rostral-caudal distances from the bregma.

Results

We efficiently labeled adipose-derived stem cells with SPIO particles, used MRI to guide the transplantation of adipose-derived stem cells in MCAO-injured mice, and used MRI to localize donor adipose-derived stem cells in the injured brain.

Adipose-Derived Stem Cells Can Be Labeled with SPIO In Vitro with High Efficiency

Using Prussian blue staining, we found distinctive blue cytoplasmic inclusions clustered in the nuclei of adipose-derived stem cells labeled with ferumoxide-PLL complex. As the concentration of PLL was increased from 0 to 0.375 μg/mL in the presence of ferumoxide, the percentage of iron-labeled cells increased proportionally (Fig. 1). At increasing levels of PLL, there was a progressive loss of cell viability, measured by trypan blue exclusion, with relatively high levels of both iron staining and viability at a PLL concentration of 0.375 μg/mL. Therefore, for all further experiments, we used a PLL concentration of 0.375 μg/mL to maximize both iron uptake and cell viability. Similar experiments performed with variation in the ferumoxide concentration showed that a ferumoxide concentration of 12.5 μg/mL was optimal for maximizing iron uptake while minimizing cell toxicity, and this concentration of ferumoxide was used for all further experiments.
We next used the MTT assay to measure the effect of the 0.375-μg/mL PLL dose, ferumoxide, and ferumoxide-PLL complex on the proliferation of adipose-derived stem cells. We found that proliferation rates from days 3-7 in all experimental and control groups were equivalent (Fig. 2) to proliferation rates of control cells (95% CI of mean, 1.19-1.24): PLL (95% CI, 1.09-1.30), ferumoxide (95% CI, 0.97-1.16), and ferumoxide-PLL (95% CI, 1.15-1.42).
Fig. 1 —Graph shows effect of poly-l-lysine (PLL) concentration on adipose-derived stem cell viability (♦) and percentage iron staining (▪) in presence of ferumoxide (12.5 μg/mL of elemental iron). Error bars represent SE. Increasing concentration of PLL resulted in both increased iron staining and decreased cell viability. These data showed PLL concentration of 0.375 μg/mL as providing acceptable level of iron staining with minimal loss of cell viability. Findings led to choice of this concentration for future experiments.
Fig. 2 —Graph shows effect of ferumoxide-poly-l-lysine (PLL) labeling on proliferation of adipose-derived stem cells over 7 days in culture. Cell proliferation was assessed with 3-(4,5-dimethyl-thiazol-2-yl)-2,5-diphenyl-tetrazolium bromide assay in duplicate cultures of adipose-derived stem cells grown in control media (▪), PLL (▴), ferumoxide (▾), and PLL-ferumoxide complex (•). Results show initial decrease in proliferation of cells grown with PLL-ferumoxide complex, although within 7 days, all experimental groups showed proliferation equivalent to that of control cells.
To test the sensitivity of MRI in the detection of small numbers of adipose-derived stem cells, we labeled adipose-derived stem cells with ferumoxide-PLL complex, immobilized labeled cells in agar, and imaged this plate with 7-T MRI. In these phantom studies, we found that T2-weighted images revealed clear differences in signal intensity between labeled and unlabeled cells and that the number of labeled cells correlated with the decrease in MR signal intensity (Fig. 3A, 3B, 3C, 3D, 3E, 3F). These results showed low MR signal intensity at cell numbers greater than 10, suggesting that MRI can depict as few as 10 labeled adipose-derived stem cell cells with nominal pixels, at least in vitro. Quantitative assessment of the sensitivity of MRI in the detection of adipose-derived stem cells with use of signal strength was not addressed in our study but may be useful for evaluation of detection of adipose-derived stem cells compared with other stem cell populations [31].

Ferumoxide-PLL-Labeled Adipose-Derived Stem Cells Can Be Transplanted Under MRI Guidance into MCAO-Injured Mice

Having established an efficient ferumoxide-PLL labeling protocol for adipose-derived stem cells, we addressed whether MRI can be used to guide the transplantation of adipose-derived stem cells into MCAO-injured mice. Two weeks after transient MCAO, live mice were imaged with MRI. We found variable infarction patterns consistent with those in a previous report on this model [22]. Two of the surviving 10 mice subjected to MCAO injury had no visible infarct on MRI, and the absence of infarct was confirmed at histologic examination. The other eight surviving mice had a large infarct with replacement of ischemic brain by surrounding CSF.
We correlated major structures (e.g., lateral ventricles, hippocampus, central aqueduct) between the MR images and images in a standard mouse brain anatomy atlas [30] (Fig. 4). From these coordinates, we calculated the stereotaxic coordinates for transplantation of donor adipose-derived stem cell into either the striatum or hippocampus next to the infarct. We reanesthetized the mice and injected ferumoxide-PLL-labeled adipose-derived stem cells into the appropriate regions.

High-Resolution Ex Vivo MRI Can Be Used to Confirm Successful Transplantation of Ferumoxide-PLL-Labeled Adipose-Derived Stem Cells

In the eight mice in which an infarct was visualized on in vivo MRI, we transplanted ferumoxide-PLL-labeled adipose-derived stem cells into either the striatum (n = 4) or the hippocampus (n = 4) immediately adjacent to the infarct. All mice tolerated adipose-derived stem cell infusion. Findings on high-resolution ex vivo MRI performed 24 hours after cell transplantation confirmed the placement of donor cells by depiction of dark (hypointense) regions in the area of transplantation in seven of the eight mice (Fig. 5). In the mice receiving 500 cells, we identified a smaller hypointense region than that in mice receiving 5,000 cells. This finding suggested a dose-dependent effect of donor cells on signal intensity. MRI at 7 T and at 9.4 T yielded equivalent high-resolution images with regard to contrast and clarity among brain structures, infarcts, and donor cells, in accordance with the findings in previous reports [32, 33]. Control mice receiving unlabeled cells did not exhibit loss of MR signal intensity in the region of cell implantation.
Fig. 3A —Phantoms of various numbers of ferumoxide-poly-l-lysine (PLL) labeled adipose-derived stem cells implanted in agar in Petri dish, which was placed on top of 3.0-cm-diameter radiofrequency surface coil. Series of multislice T2-weighted and T2*-weighted 7-T MR images (TR/TE, 250/10-40 in increments of 5 milliseconds; matrix size, 128 × 128; field of view, 4.0 cm; slice thickness, 1 mm) were acquired with standard spin-echo sequences. T2-weighted images reveal clear differences in signal intensity between area without cells (A) and areas with increasing number of superparamagnetic iron oxide (SPIO)-labeled adipose-derived stem cells. Numbers of SPIO-labeled adipose-derived stem cells correlated well with decrease in MR signal intensity. Low signal intensity at all cell numbers greater than 10 cells suggests MRI can depict at least as few as 10 labeled adipose-derived stem cells with nominal pixels. Number of cells, 0.
Fig. 3B —Phantoms of various numbers of ferumoxide-poly-l-lysine (PLL) labeled adipose-derived stem cells implanted in agar in Petri dish, which was placed on top of 3.0-cm-diameter radiofrequency surface coil. Series of multislice T2-weighted and T2*-weighted 7-T MR images (TR/TE, 250/10-40 in increments of 5 milliseconds; matrix size, 128 × 128; field of view, 4.0 cm; slice thickness, 1 mm) were acquired with standard spin-echo sequences. T2-weighted images reveal clear differences in signal intensity between area without cells (A) and areas with increasing number of superparamagnetic iron oxide (SPIO)-labeled adipose-derived stem cells. Numbers of SPIO-labeled adipose-derived stem cells correlated well with decrease in MR signal intensity. Low signal intensity at all cell numbers greater than 10 cells suggests MRI can depict at least as few as 10 labeled adipose-derived stem cells with nominal pixels. Number of cells, 10.
Fig. 3C —Phantoms of various numbers of ferumoxide-poly-l-lysine (PLL) labeled adipose-derived stem cells implanted in agar in Petri dish, which was placed on top of 3.0-cm-diameter radiofrequency surface coil. Series of multislice T2-weighted and T2*-weighted 7-T MR images (TR/TE, 250/10-40 in increments of 5 milliseconds; matrix size, 128 × 128; field of view, 4.0 cm; slice thickness, 1 mm) were acquired with standard spin-echo sequences. T2-weighted images reveal clear differences in signal intensity between area without cells (A) and areas with increasing number of superparamagnetic iron oxide (SPIO)-labeled adipose-derived stem cells. Numbers of SPIO-labeled adipose-derived stem cells correlated well with decrease in MR signal intensity. Low signal intensity at all cell numbers greater than 10 cells suggests MRI can depict at least as few as 10 labeled adipose-derived stem cells with nominal pixels. Number of cells, 100.
Fig. 3D —Phantoms of various numbers of ferumoxide-poly-l-lysine (PLL) labeled adipose-derived stem cells implanted in agar in Petri dish, which was placed on top of 3.0-cm-diameter radiofrequency surface coil. Series of multislice T2-weighted and T2*-weighted 7-T MR images (TR/TE, 250/10-40 in increments of 5 milliseconds; matrix size, 128 × 128; field of view, 4.0 cm; slice thickness, 1 mm) were acquired with standard spin-echo sequences. T2-weighted images reveal clear differences in signal intensity between area without cells (A) and areas with increasing number of superparamagnetic iron oxide (SPIO)-labeled adipose-derived stem cells. Numbers of SPIO-labeled adipose-derived stem cells correlated well with decrease in MR signal intensity. Low signal intensity at all cell numbers greater than 10 cells suggests MRI can depict at least as few as 10 labeled adipose-derived stem cells with nominal pixels. Number of cells, 500.
Fig. 3E —Phantoms of various numbers of ferumoxide-poly-l-lysine (PLL) labeled adipose-derived stem cells implanted in agar in Petri dish, which was placed on top of 3.0-cm-diameter radiofrequency surface coil. Series of multislice T2-weighted and T2*-weighted 7-T MR images (TR/TE, 250/10-40 in increments of 5 milliseconds; matrix size, 128 × 128; field of view, 4.0 cm; slice thickness, 1 mm) were acquired with standard spin-echo sequences. T2-weighted images reveal clear differences in signal intensity between area without cells (A) and areas with increasing number of superparamagnetic iron oxide (SPIO)-labeled adipose-derived stem cells. Numbers of SPIO-labeled adipose-derived stem cells correlated well with decrease in MR signal intensity. Low signal intensity at all cell numbers greater than 10 cells suggests MRI can depict at least as few as 10 labeled adipose-derived stem cells with nominal pixels. Number of cells, 1,000.
Fig. 3F —Phantoms of various numbers of ferumoxide-poly-l-lysine (PLL) labeled adipose-derived stem cells implanted in agar in Petri dish, which was placed on top of 3.0-cm-diameter radiofrequency surface coil. Series of multislice T2-weighted and T2*-weighted 7-T MR images (TR/TE, 250/10-40 in increments of 5 milliseconds; matrix size, 128 × 128; field of view, 4.0 cm; slice thickness, 1 mm) were acquired with standard spin-echo sequences. T2-weighted images reveal clear differences in signal intensity between area without cells (A) and areas with increasing number of superparamagnetic iron oxide (SPIO)-labeled adipose-derived stem cells. Numbers of SPIO-labeled adipose-derived stem cells correlated well with decrease in MR signal intensity. Low signal intensity at all cell numbers greater than 10 cells suggests MRI can depict at least as few as 10 labeled adipose-derived stem cells with nominal pixels. Number of cells, 5,000.

Donor Adipose-Derived Stem Cells Can Be Localized with MRI

To confirm the accuracy of MRI for localization of stem cells, we used various stains for visual correlation of histologic findings and MR signal intensity in the MCAO-injured brain after adipose-derived stem cell transplantation. In all seven mice with a hypointense region at the site of adipose-derived stem cell infusion, MRI findings correlated well with histologic evidence of donor cells (Fig. 6A, 6B, 6C, 6D, 6E). For these experiments, the intrinsic GFP fluorescence of the donor adipose-derived stem cell cells was used to assist with donor cell visualization at immunohistochemical analysis, which showed GFP+ donor cells in the area corresponding to the loss of MR signal intensity. Findings at histologic examination of corresponding brain sections confirmed that MRI findings correlated with the deposition of donor cells: At immunohistochemical analysis sections showed cells containing iron (Prussian blue) and GFP. MRI was sensitive for relatively low numbers of cells. We colocalized MRI signal intensity and histologic findings in mice receiving as few as 500 cells. In the single mouse in which no change in MRI signal intensity was seen after transplantation, no donor cells were seen at histologic examination. This finding, which suggested a technical problem with cell implantation in this mouse, was an indication of the specificity of MRI. These results suggest that MRI can be used, at least immediately after transplantation, to localize small numbers of donor labeled stem cells.
Fig. 4 —Mouse 2 weeks after middle cerebral artery occlusion-induced infarction. In vivo 7-T T2*-weighted MR image of mouse head placed above 1.0-cm-diameter surface coil. Multislice spin-echo images (TR/TE, 1/35; matrix size, 256 × 256; field of view, 2.5 cm; resolution, 98 μm; thickness of interleaved consecutive slices, 1.0 mm; number of signals averaged, 2) were obtained in anatomic coronal orientation. All sequences were performed within 40-60 minutes per mouse. Infarct (arrow) is evident within right cortex. Adipose-derived stem cells were transplanted into ipsilateral hippocampus.
Fig. 5 —Middle cerebral artery occlusion-injured mouse sacrificed 24 hours after transplantation of 1,000 ferumoxide-poly-l-lysine-labeled adipose-derived stem cells. High-resolution ex vivo coronal 9.4-T proton density MR image of mouse brain perfused with 10% buffered formalin/5% gadoteridol shows areas of infarction (solid arrow) and low signal intensity (dashed arrow) corresponding to donor cell implantation.

Discussion

The major findings of this study were as follows. First, adipose-derived stem cells can be efficiently labeled with SPIO particles by use of ferumoxide-PLL complex with minimal loss of cell viability and proliferation. Second, MRI can be used to guide transplantation of adipose-derived stem cells into MCAO-injured mice. Third, MRI can be used to localize donor ferumoxide-PLL-labeled adipose-derived stem cells after transplantation into the injured brain. Histologic examination of brain sections showed the presence of donor labeled cells in the area of decreased MR signal intensity, confirming a correlation between MRI findings and conventional histologic evidence. These findings have important ramifications for the eventual transplantation of stem cells into the human brain and the role of MRI in facilitating the process and tracking stem cell migration.
Stem cell studies in animal stroke models are limited by variability in stroke patterns that results in difficulty placing donor stem cells into the ischemic border zone immediately adjacent to the injury. A variety of techniques have been described for localization of ischemic tissue in rodents, including laser Doppler flowmetry [22] and measurement of global and regional cerebral blood flow [34]. These techniques, however, do not easily lend themselves to guidance of stem cell transplantation. We based our work on the leading studies of Jiang and collaborators, who defined parameters for the use of MRI in tracking neuronal progenitor cells in animal stroke models [35]. Our study appears to be the first report of MRI guidance of transplantation of donor adipose-derived stem cells within a brain injury model.
The fate of transplanted adipose-derived stem cells and their effect on the functional outcome of cerebral ischemia were not addressed in our study and remain important issues. The efficient labeling of adipose-derived stem cells with ferumoxide concurs with results of stem cell studies showing the usefulness of SPIO preparations for labeling of mesenchymal stem cells [36], embryonic stem cells [21], and neural progenitor cells [24]. We recognize that our short-term study did not specifically address the long-term differentiation of labeled adipose-derived stem cells in vivo. However, other stem cell studies have shown both conservation and limited differentiation after SPIO labeling [24, 36]. In most studies of embryonic stem cell and neuronal stem cell transplantation into the brain, donor cells generally have differentiated along glial lineages [37, 38]. In contrast, transplantation of stem cells into neurogenic regions, such as the subventricular zone, may increase differentiation toward neurons [37]. Results [18] have shown that donor adiposederived stem cells transplanted into either the normal or injured brain may lead to both neuronal and glial differentiation. These topics require further study. Before clinical use of adipose-derived stem cells, characterization of the effect of SPIO labeling on long-term adipose-derived stem cell differentiation and function is needed.
We recognize that before in vivo MRI is used to track the activity of stem cells over time after transplantation, it is essential to define the resolution and validity of an MRI protocol. To design our initial imaging protocol, we chose ex vivo rather than in vivo MRI for posttransplantation localization of stem cells. The resolution of our ex vivo studies was 0.043 × 0.043 × 0.043 mm (7.9 × 10-5 mm3), nearly 1,000 times greater spatial resolution than in our in vivo studies and providing a much more sensitive method for detection. Therefore, the use of ex vivo MR histologic examination allows much higher confidence in localization because of correlation with conventional histologic findings and detection of far fewer cells than would be possible in vivo.
Results of studies by Heyn and collaborators [39] have suggested that single SPIO-labeled stem cells can be detected in vivo. Several authors [20, 25, 26] have performed longitudinal studies to track labeled stem cells in vivo. The spatial resolution of our in vivo studies at 0.273 × 0.273 × 1.0 mm (7.4 × 10-1 mm3) is on the same scale as that in the previous in vivo studies. Our current work validates the use of MRI for adipose-derived stem cell localization with sensitive ex vivo imaging and provides the foundation for subsequent longitudinal in vivo studies. Those studies will be conducted with a larger number of cells and longer in vivo scan times to achieve the required sensitivity.
Labeling adipose-derived stem cells with SPIO particles enhances cell-to-background contrast and renders the cells visible on MRI. The intensity and area of the low MRI signal intensity from the SPIO particles depend on the concentration of iron oxide particles per cell and the image acquisition parameters. In the current study, our phantom experiments showed that MRI can depict as few as 10 cells labeled with a relatively low concentration of SPIO particles (12.5 μg/mL). In contrast, a higher iron content for cell labeling, for example, up to 1,200 μg/mL, has been used in other stem cell MRI studies [21]. Frank et al. [40] reported that iron in a concentration of 25 μg/mL was well tolerated by various stem cells and allowed MRI detection. However, we agree with Magnitsky et al. [26] that given the wide variation of labeling techniques, imaging hardware (radiofrequency coils), and acquisition parameters (TE, imaging matrix, and acquisition time) used by various investigators, it is difficult to compare the quantitative results of stem cell labeling among laboratories.
Fig. 6A —Middle cerebral artery occlusion-injured mouse sacrificed 24 hours after receiving transplant of 1,000 superparamagnetic iron oxide (SPIO)-labeled green fluorescent protein (GFP)-positive adipose-derived stem cells into hippocampus. Coronal MR image shows hypointense region in area of transplantation. Annular hyperintense ring is thought to represent artifact from SPIO intensity given absence of corresponding abnormality on histologic sections in B-E.
Fig. 6B —Middle cerebral artery occlusion-injured mouse sacrificed 24 hours after receiving transplant of 1,000 superparamagnetic iron oxide (SPIO)-labeled green fluorescent protein (GFP)-positive adipose-derived stem cells into hippocampus. Low-magnification Nissl stain of brain section similar to that in A shows donor cell deposit (box).
Fig. 6C —Middle cerebral artery occlusion-injured mouse sacrificed 24 hours after receiving transplant of 1,000 superparamagnetic iron oxide (SPIO)-labeled green fluorescent protein (GFP)-positive adipose-derived stem cells into hippocampus. High-magnification Prussian blue stain of section adjacent to B shows cell deposit.
Fig. 6D —Middle cerebral artery occlusion-injured mouse sacrificed 24 hours after receiving transplant of 1,000 superparamagnetic iron oxide (SPIO)-labeled green fluorescent protein (GFP)-positive adipose-derived stem cells into hippocampus. Immunohistochemical results with use of anti-GFP antibody and 3,3′-diaminobenzidine conjugate of adjacent section show brown staining of donor cells.
Fig. 6E —Middle cerebral artery occlusion-injured mouse sacrificed 24 hours after receiving transplant of 1,000 superparamagnetic iron oxide (SPIO)-labeled green fluorescent protein (GFP)-positive adipose-derived stem cells into hippocampus. High-magnification Nissl stain shows iron-laden donor cells within CA3 region of hippocampus.
With our tissue fixation and imaging protocol, both 7-T and 9.4-T MRI yielded high-resolution ex vivo images of donor stem cells. Each image element (pixel) in an MR image represents the signal intensity from a volume (voxel) of tissue. As described earlier, 0.043 × 0.043 × 0.043 mm voxels are encoding a volume of 7.9 × 10-5 mm3. For context, such a voxel is almost equal in size to a single Purkinje neuron in the mouse cerebellum [41]. We consider our method of tissue fixation important for rendering high-quality images, as reported previously [33]. This method of fixation exploits differences in tissue magnetization, which both uniformly reduces the T1 intensity of all tissues, resulting in a 10-fold increase in signal intensity, and the detection of T2 differences in tissues with short TRs. Furthermore, the MRI data set we used yields 3D images (matrix size, 256 × 256 × 512) that can be viewed and sectioned in any plane (coronal, horizontal, and sagittal) without loss of spatial resolution.
Although histochemical examination showed the presence of iron-positive cells in tissue sections that corresponded to hypointense regions on MR images, we recognize that several endogenous sources of iron can result in similar appearances. The areas of iron also costained for GFP. We do not believe, however, that the iron originated in hemorrhage and iron degradation product release from iron-containing proteins, such as hemoglobin, ferritin, and hemosiderin, from endogenous cell uptake. Long-term studies of donor cells in vivo, including phenotype analysis, would be helpful for assessment of the role of phagocytic processes in iron sequestration that may affect MRI.
We conclude that adipose-derived stem cells can be safely and efficiently labeled with ferumoxide particles and that MRI can be used to guide transplantation of donor stem cells in MCAO-injured mice and localization of ferumoxide-PLL-labeled adipose-derived stem cells in MCAO-injured mice. As shown in this immediately posttransplantation model, these techniques can assist the study of stem cell implantation. Use of these labeling techniques may facilitate long-term longitudinal study of donor stem cells after transplantation.

Acknowledgments

We thank Boma Fubara for excellent technical assistance.

Footnotes

Imaging studies were performed at the Duke Center for In Vivo Microscopy, an NCRR/NCI National Resource (P41 RR005959/R24 CA092656).
Address correspondence to H. E. Rice.

References

1.
Gage FH. Mammalian neural stem cells. Science 2000; 287:1433-1438
2.
Rakic P. Adult neurogenesis in mammals: an identity crisis. J Neurosci 2002; 22:614-618
3.
Brustle O, Jones KN, Learish RD, et al. Embryonic stem cell-derived glial precursors: a source of myelinating transplants. Science 1999; 285:754-756
4.
Bain G, Kitchens D, Yao M, Huettner JE, Gottlieb DI. Embryonic stem cells express neuronal properties in vitro. Dev Neurobiol 1995; 168:342-357
5.
Kopen GC, Prockop DJ, Phinney DG. Marrow stromal cells migrate throughout forebrain and cerebellum, and they differentiate into astrocytes after injection into neonatal mouse brains. Proc Natl Acad Sci U S A 1999; 96:10711-10716
6.
Mezey E, Chandross KJ, Harta G, Maki RA, McKercher SR. Turning blood into brain: cells bearing neuronal antigens generated in vivo from bone marrow. Science 2000; 290:1779-1782
7.
Sanchez-Ramos J, Song S, Cardozo-Pelaez F, et al. Adult bone marrow stromal cells differentiate into neural cells in vitro. Exp Neurol 2000; 164:247-256
8.
Woodbury D, Schwarz EJ, Prockop DJ, Black IB. Adult rat and human bone marrow stromal cells differentiate into neurons. J Neurosci Res 2000; 61:364-370
9.
Eglitis MA, Mezey E. Hematopoietic cells differentiate into both microglia and macroglia in the brains of adult mice. Proc Natl Acad Sci U S A 1997; 94:4080-4085
10.
Munoz-Elias G, Marcus AJ, Coyne TM, Woodbury D, Black IB. Adult bone marrow stromal cells in the embryonic brain: engraftment, migration, differentiation, and long-term survival. J Neurosci 2004; 24:4585-4595
11.
Erickson G, Gimble J, Franklin D, Rice H, Awad H, Guilak F. Chondrogenic potential of adipose tissuederived stromal cells in vitro and in vivo. Biochem Biophys Res Commun 2002; 290:763-769
12.
Halvorsen YD, Franklin D, Bond AL, et al. Extracellular matrix mineralization and osteoblast gene expression by human adipose tissue-derived stromal cells. Tissue Eng 2001; 7:729-741
13.
Safford KM, Hicok KC, Safford SD, et al. Neurogenic differentiation of murine and human adipose-derived stromal cells. Biochem Biophys Res Comm 2002; 294:371-379
14.
Zuk P, Zhu M, Ashjian P, et al. Human adipose tissue is a source of multipotent stem cells. Mol Biol Cell 2002; 13:4279-4295
15.
Cowan CM, Shi Y, Aalami OO, et al. Adipose-derived adult stromal cells heal critical-size mouse calvarial defects. Nat Biotech 2004; 22:560-567
16.
De Ugarte DA, Morizono K, Elbarbary S, et al. Comparison of multi-lineage cells from human adipose tissue and bone marrow. Cells Tissues Organs 2003; 174:101-119
17.
Safford KM, Safford SD, Gimble JM, Shetty AK, Rice HE. Characterization of neuronal/glial differentiation of murine adipose-derived adult stromal cells. Exp Neurol 2004; 187:319-328
18.
Kang SK, Lee DH, Bae YC, Kim HK, Baik SY, Jung JS. Improvement of neurological deficits by intracerebral transplantation of human adipose tissue-derived stromal cells after cerebral ischemia in rats. Exp Neurol 2003; 183:355-366
19.
Fujimura J, Ogawa R, Mizuno H, Fukunaga Y, Suzuki H. Neural differentiation of adipose-derived stem cells isolated from GFP transgenic mice. Biochem Biophys Res Comm 2005; 333:116-121
20.
Arbab AS, Bashaw LA, Miller BR, et al. Characterization of biophysical and metabolic properties of cells labeled with superparamagnetic iron oxide nanoparticles and transfection agent for cellular MR imaging. Radiology 2003; 229:838-846
21.
Hoehn M, Kustermann E, Blunk J, et al. Monitoring of implanted stem cell migration in vivo: a highly resolved in vivo magnetic resonance imaging investigation of experimental stroke in rat. Proc Natl Acad Sci U S A 2002; 99:16267-16272
22.
Harada H, Wang Y, Mishima Y, Uehara N, Makaya T, Kano T. A novel method of detecting rCBF with laser-Doppler flowmetry without cranial window through the skull for a MCAO rat model. Brain Res Brain Res Protoc 2005; 14:165-170
23.
Connolly ES, Winfree CJ, Stern DM, Solomon RA, Pinsky DJ. Procedural and strain-related variables significantly affect outcome in a murine model of focal cerebral ischemia. Neurosurgery 1996; 38:523-531
24.
Bulte JW, Douglas T, Witwer B, et al. Magnetodendrimers allow endosomal magnetic labeling and in vivo tracking of stem cells. Nat Biotechnol 2001; 19:1141-1147
25.
Bulte JW, Duncan ID, Frank JA. In vivo magnetic resonance tracking of magnetically labeled cells after transplantation. J Cereb Blood Flow Metab 2002; 22:899-907
26.
Magnitsky S, Watson DJ, Walton RM, et al. In vivo and ex vivo MRI detection of localized and disseminated neural stem cell grafts in the mouse brain. Neuroimage 2005; 26:744-754
27.
Arbab AS, Jordan EK, Wilson LB, Yocum GT, Lewis BK, Frank JA. In vivo trafficking and targeted delivery of magnetically labeled stem cells. Hum Gene Ther 2004; 15:351-360
28.
Mosmann T. Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J Immunol Methods 1983; 65:55-63
29.
Homi HM, Mixco JM, Sheng H, Grocott HP, Pearlstein RD, Warner DS. Severe hypotension is not essential for isoflurane neuroprotection against forebrain ischemia in mice. Anesthesiology 2003; 99:1145-1151
30.
Paxinos G, Franklin KJ. The mouse brain in stereotaxic coordinates. San Diego, CA: Academic Press,2001
31.
Chu WJ, Simor T, Elgavish GA. In vivo characterization of Gd(BME-DTTA), a myocardial MRI contrast agent: tissue distribution of its MRI intensity enhancement, and its effect on heart function. NMR Biomed 1997; 10:87-92
32.
Cyr M, Caron MG, Johnson GA, Laakso A. Magnetic resonance imaging at microscopic resolution reveals subtle morphological changes in a mouse model of dopaminergic hyperfunction. Neuroimage 2005; 26:83-90
33.
Johnson GA, Cofer GP, Fubara B, Gewalt SL, Hedlund LW, Maronpot RR. Magnetic resonance histology for morphologic phenotyping. J Magn Reson Imaging 2002; 16:423-429
34.
Demchenko IT, Boso AE, Natoli MJ, et al. Measurement of cerebral blood flow in rats and mice by hydrogen clearance during hyperbaric oxygen exposure. Undersea Hyperb Med 1998; 25:147-152
35.
Jiang Q, Zhang ZG, Ding GL, et al. Investigation of neural progenitor cell induced angiogenesis after embolic stroke in rat using MRI. Neuroimage 2005; 28:698-707
36.
Bos C, Delmas Y, Desmoulière A, et al. In vivo MR imaging of intravascularly injected magnetically labeled mesenchymal stem cells in rat kidney and liver. Radiology 2004; 233:781-789
37.
Fricker RA, Carpenter MK, Winkler C, Greco C, Gates MA, Bjorklund A. Site-specific migration and neuronal differentiation of human neural progenitor cells after transplantation in the adult rat brain. J Neurosci 1999; 19:5990-6005
38.
Winkler C, Fricker RA, Gates MA, et al. Incorporation and glial differentiation of mouse EGF-responsive neural progenitor cells after transplantation into the embryonic rat brain. Mol Cell Neurosci 1998; 11:99-116
39.
Heyn C, Ronald JA, Mackenzie LT, et al. In vivo magnetic resonance imaging of single cells in mouse brain with optical validation. Magn Reson Med 2006; 55:23-29
40.
Frank JA, Miller BR, Arbab AS, et al. Clinically applicable labeling of mammalian and stem cells by combining superparamagnetic iron oxides and transfection agents. Radiology 2003; 228:480-487
41.
Sotelo C. Anatomical, physiological and biochemical studies of the cerebellum from mutant mice. Part II. Morphological study of cerebellar cortical neurons and circuits in the weaver mouse. Brain Res 1975; 94:19-44

Information & Authors

Information

Published In

American Journal of Roentgenology
Pages: 1101 - 1108
PubMed: 17377054

History

Submitted: May 16, 2006
Accepted: September 12, 2006

Keywords

  1. brain
  2. ischemia
  3. MRI
  4. stem cells
  5. transplantation

Authors

Affiliations

Henry E. Rice
Department of Surgery, Duke University Medical Center, Box 3815, Durham, NC 27710.
Edward W. Hsu
Department of Biomedical Engineering, Duke University Medical Center, Durham, NC.
Huaxin Sheng
Department of Anesthesiology, Duke University Medical Center, Durham, NC.
Debra A. Evenson
Department of Surgery, Duke University Medical Center, Box 3815, Durham, NC 27710.
Alex J. Freemerman
Department of Surgery, Duke University Medical Center, Box 3815, Durham, NC 27710.
Kristine M. Safford
Department of Surgery, Duke University Medical Center, Box 3815, Durham, NC 27710.
James M. Provenzale
Department of Radiology, Duke University Medical Center, Durham, NC.
David S. Warner
Department of Anesthesiology, Duke University Medical Center, Durham, NC.
G. Allan Johnson
Department of Radiology, Duke University Medical Center, Durham, NC.

Metrics & Citations

Metrics

Citations

Export Citations

To download the citation to this article, select your reference manager software.

Articles citing this article

View Options

View options

PDF

View PDF

PDF Download

Download PDF

Media

Figures

Other

Tables

Share

Share

Copy the content Link

Share on social media