In vivo magnetic resonance imaging of cell tropism, trafficking mechanism, and therapeutic impact of human mesenchymal stem cells in a murine glioma model.
ABSTRACT Stem cells have offered much promise as delivery vehicles for brain tumor therapy, with the development of modalities to track the tumor tropism of stem cells receiving intense focus. Cellular magnetic resonance imaging (MRI) allows serial high-resolution in vivo detection of transplanted stem cells' tropism toward gliomas in the mouse brain once these cells are internally labeled with iron oxide particles, but has been impeded by low labeling efficiencies. In this study, we describe the use of ferucarbotran and protamine (Fer-Pro) complexes for labeling human mesenchymal stem cells (hMSCs) for MRI tracking of glioma tropism in vivo. We found that Fer-Pro was not toxic and was highly efficient for labeling in vitro. Cell labeling with Fer-Pro promoted the migration of hMSCs toward glioma U87MG cells in vitro, which was mediated by stromal-derived factor-1/CXCR4 (SDF-1/CXCR4) signaling. Fer-Pro-labeled hMSCs could migrate specifically toward gliomas in vivo, which was observed with a clinical 1.5-T MRI system. The efficient labeling of Fer-Pro also allowed a tropic mechanism mediated by SDF-1/CXCR4 signaling to be detected by MRI in vivo. Additionally, the potential intrinsic inhibitory effect of hMSCs on glioma progression was estimated simultaneously. This is the first report to have used a clinical MRI modality to simultaneously study the migration, the therapeutic impact on tumors, and above all the trafficking mechanism of bone marrow-derived mesenchymal stem cells from human in a murine glioma xenograft model. The use of Fer-Pro for stem cell labeling may have potential clinical applications in stem cell guided therapy.
- SourceAvailable from: Karen S Aboody[show abstract] [hide abstract]
ABSTRACT: One of the impediments to the treatment of brain tumors (e.g., gliomas) has been the degree to which they expand, infiltrate surrounding tissue, and migrate widely into normal brain, usually rendering them "elusive" to effective resection, irradiation, chemotherapy, or gene therapy. We demonstrate that neural stem cells (NSCs), when implanted into experimental intracranial gliomas in vivo in adult rodents, distribute themselves quickly and extensively throughout the tumor bed and migrate uniquely in juxtaposition to widely expanding and aggressively advancing tumor cells, while continuing to stably express a foreign gene. The NSCs "surround" the invading tumor border while "chasing down" infiltrating tumor cells. When implanted intracranially at distant sites from the tumor (e.g., into normal tissue, into the contralateral hemisphere, or into the cerebral ventricles), the donor cells migrate through normal tissue targeting the tumor cells (including human glioblastomas). When implanted outside the CNS intravascularly, NSCs will target an intracranial tumor. NSCs can deliver a therapeutically relevant molecule-cytosine deaminase-such that quantifiable reduction in tumor burden results. These data suggest the adjunctive use of inherently migratory NSCs as a delivery vehicle for targeting therapeutic genes and vectors to refractory, migratory, invasive brain tumors. More broadly, they suggest that NSC migration can be extensive, even in the adult brain and along nonstereotypical routes, if pathology (as modeled here by tumor) is present.Proceedings of the National Academy of Sciences 12/2000; 97(23):12846-51. · 9.74 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: Glioblastomas, the most frequent and malignant of primary brain tumors, have a very poor prognosis. Gene therapy of glioblastomas is limited by the short survival of viral vectors and by their difficulty in reaching glioblastoma cells infiltrating the brain parenchyma. Neural stem/progenitor cells can be engineered to produce therapeutic molecules and have the potential to overcome these limitations because they may travel along the white matter, like neoplastic cells, and engraft stably into the brain. Retrovirus-mediated transfer of the gene for interleukin-4 is an effective treatment for rat brain glioblastomas. Here, we transferred the gene for interleukin-4 into C57BL6J mouse primary neural progenitor cells and injected those cells into established syngeneic brain glioblastomas. This led to the survival of most tumor-bearing mice. We obtained similar results by implanting immortalized neural progenitor cells derived from Sprague-Dawley rats into C6 glioblastomas. We also documented by magnetic resonance imaging the progressive disappearance of large tumors, and detected 5-bromodeoxyuridine-labeled progenitor cells several weeks after the injection. These findings support a new approach for gene therapy of brain tumors, based on the grafting of neural stem cells producing therapeutic molecules.Nature Medicine 05/2000; 6(4):447-50. · 22.86 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: Malignant gliomas spawn disseminated microsatellites, which are largely refractory to currently employed therapies, resulting in eventual tumor recurrence and death. The use of tumor-tropic neural stem cells (NSCs) as delivery vehicles for therapeutic gene products represents an attractive strategy specifically focused at treating these residual neoplastic foci. We wished to elucidate the biological cues governing NSC tropism for glioma. In this context, we describe that tumor-tropic NSCs comprise largely of astrocytic progenitors expressing chemokine receptor 4 (CXCR4). Blocking of CXCR4 significantly inhibits NSC migration toward the tumor. These findings define specific characteristics associated with the cell populations within transplanted NSCs that demonstrate glioma-tracking behavior.Neoplasia 01/2004; 6(3):287-93. · 5.47 Impact Factor
In vivo magnetic resonance imaging of cell tropsim, trafficking mechanism, and
therapeutic impact of human mesenchymal stem cells in a murine glioma model
Li-Ying Chiena,1, Jong-Kai Hsiaob,c,1, Szu-Chun Hsud, Ming Yaoe, Chen-Wen Lub, Hon-Man Liub,
Yao-Chang Chend, Chung-Shi Yanga, Dong-Ming Huanga,*
aCenter for Nanomedicine Research, National Health Research Institutes, Miaoli 350, Taiwan
bDepartment of Medical Imaging, National Taiwan University Hospital and College of Medicine, National Taiwan University, Taipei 100, Taiwan
cDepartment of Medical Imaging, Buddhist Tzu Chi General Hospital, Taipei branch, Taiwan
dDepartment of Laboratory Medicine, National Taiwan University Hospital and College of Medicine, National Taiwan University, Taipei 100, Taiwan
eDepartment of Internal Medicine, National Taiwan University Hospital and College of Medicine, National Taiwan University, Taipei 100, Taiwan
a r t i c l e i n f o
Received 21 October 2010
Accepted 13 January 2011
Available online 3 February 2011
Mesenchymal stem cell
a b s t r a c t
Stem cells have offered much promise as delivery vehicles for brain tumor therapy, with the development
of modalities to track the tumor tropism of stem cells receiving intense focus. Cellular magnetic resonance
imaging (MRI) allows serial high-resolution in vivo detection of transplanted stem cells’ tropism toward
gliomas in the mouse brain once these cells are internally labeled with iron oxide particles, but has been
impeded by low labeling efficiencies. In this study, we describe the use of ferucarbotran and protamine
(Fer-Pro) complexes for labeling human mesenchymal stem cells (hMSCs) for MRI tracking of glioma
tropism in vivo. We found that Fer-Pro was not toxic and was highly efficient for labeling in vitro. Cell
labeling with Fer-Pro promoted the migration of hMSCs toward glioma U87MG cells in vitro, which was
mediated by stromal-derived factor-1/CXCR4 (SDF-1/CXCR4) signaling. Fer-Pro-labeled hMSCs could
migrate specifically toward gliomas in vivo, which was observed with a clinical 1.5-T MRI system. The
efficient labeling of Fer-Pro also allowed a tropic mechanism mediated by SDF-1/CXCR4 signaling to be
detected by MRI in vivo. Additionally, the potential intrinsic inhibitory effect of hMSCs on glioma
progression was estimated simultaneously. This is the first report to have used a clinical MRI modality to
simultaneously study the migration, the therapeutic impact on tumors, and above all the trafficking
mechanism of bone marrow-derived mesenchymal stem cells from human in a murine glioma xenograft
model. The use of Fer-Pro for stem cell labeling may have potential clinical applications in stem cell guided
? 2011 Elsevier Ltd. All rights reserved.
Glioblastoma multiforme (GBM) is the most common malignant
brain tumor, with a high mortality rate and the worst prognosis
[1e3]. Despite advances in surgical and adjuvant therapies, patients
carrying a primary GBM typically have a mean survival period of
less than a year following diagnosis . The poor outcome of
patients with GBM is related to the resistance to current therapy,
radiation, and chemotherapeutic approaches. Also, it is difficult to
completely resect high-grade gliomas because of the infiltration of
malignant cells into surrounding brain parenchyma [4,5]. More-
over, some new therapeutic modalities have failed to achieve
significant therapeutic effect due to an inability to effectively
deliver therapeutic agents to the invasive tumor cells . Therefore,
new strategies for selective tumor-targeting of therapeutic agents
are needed to substantially improve brain tumor therapy.
Stem cells have offered much promise as delivery vehicles for
brain tumor therapy because of their tropism toward tumor cells.
For instance, the use of neural stem cells (NSCs) as vehicles for gene
therapy in brain tumors has been of great interest [6e9]. In addi-
tion, due to the intrinsic tumor-inhibition effect of NSCs on glio-
mas [7,10], this NSC-based therapeutic approach would overcome
current limitations of conventional clinical therapies. Because of
ethical and technical problems associated with neural stem cells, an
alternative type of is being sought. Mesenchymal stem cells (MSCs)
are increasingly regarded as attractive candidate for delivering
vesicles of therapeutic agents [11,12]. Among the main reasons for
this are that MSCs can be easily isolated and expanded in culture
[13e15], that MSCs can possibly overcome immunological incom-
patibilities because of autologous transplantation , and, above
* Corresponding author. Tel.: þ886 37 246 166x38105; fax: þ886 37 586 447.
E-mail address: firstname.lastname@example.org (D.-M. Huang).
1Both authors contributed equally to this work.
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/biomaterials
0142-9612/$ e see front matter ? 2011 Elsevier Ltd. All rights reserved.
Biomaterials 32 (2011) 3275e3284
all, that MSCs can migrate toward gliomas in vivo [11,12,17]. Also,
there is evidence that MSCs can exert an inhibitory effect on several
tumors [17e19]. Taken together, these facts indicate that MSCs
could replace NSCs as a therapeutic vehicle against gliomas.
Development of any cell-based therapy requires monitoring the
fate and distribution of transplanted stem cells to maximize the
therapeutic benefit. Traditional techniques for examining stem-cell
transplantation in animal models, often performed by postmortem
histological analysis, cannot be appropriate in clinical studies;
hence, it is important to develop noninvasive in vivo stem cell
ideal imaging modality for serial tracking and identifying the fate of
transplanted cells once the cells are internally labeled with suitable
employ superparamagnetic iron oxide (SPIO) nanoparticles as
cellular MRI probes. The efficacy of cellular MRI largely depends on
the internalizing efficiency of SPIO nanoparticles into stem cells;
cellular uptake efficiency. Recently it was reported that ferucarbo-
tran (Resovist), an ionic SPIO nanoparticle with carboxydextran
coating, can more efficiently magnetically label human MSCs
(hMSCs) than ferumoxide (Feridex) without cytotoxicity . Our
study has also demonstrated that ferucarbotran-labeled hMSCs can
1.5-T MRI system . Thus, ferucarbotran seems to be an ideal
labeling probe to investigate the tropism of MSCs toward malignant
gliomas using MRI.
Protamine is a low molecular weight polycationic peptide that
has been approved by the U.S. FDA as an antidote for heparin anti-
coagulation . Arbab et al. showed an enhanced uptake of fer-
umoxides by protamine for cell labeling . Moreover, a recent
study used protamine to accelerate hMSC labeling with ferucarbo-
as MRI contrast probe combined with protamine for efficiently
cellular labeling to study the detailed tropism behavior of hMSCs in
a murine glioma model. Using a clinical 1.5-T MRI system we
observed that hMSCs labeled by ferucarbotran and protamine
complexes could specifically migrate toward gliomas in vivo. The
a tropic mechanism mediated by stromal-derived factor-1/CXCR4
(SDF-1/CXCR4) signaling to be detected also by MRI in vivo. Addi-
tionally, the potential intrinsic inhibitory effect of hMSCs on glioma
progression was estimated simultaneously.
2. Materials and methods
2.1. Cells and culture conditions
Human mesenchymal stem cells (hMSCs) were isolated from bone marrow of
normal donors as described in our previous study  with informed consent
approved according to the procedures of the institutional review board and were
cultured in regular growth medium consisting of low-glucose DMEM (Gibco, Grand
Island, NY) supplemented with 10% fetal bovine serum (FBS) (Gibco), 100 U/ml
penicillin, and 100 mg/ml streptomycin (Gibco). The cells were immunophenotypi-
cally characterized as in our previous study  by flow cytometry for many
markers common to human MSCs  and were able to give rise to adipocytes,
osteocytes, and chondrocytes when incubated with adequate differentiation media
, which shows their multipotentiality (data not shown). hMSCs were used at
passages 6e8 in all experiments.
U87MG cells from American Type Culture collection (Manassas, VA) were
maintained in high-glucose DMEM (Gibco) supplemented with 10% FBS, 100 U/ml
penicillin, and 100 mg/ml streptomycin.
All cultures were kept in an atmosphere of 5% CO2and 95% air at 37?C.
2.2. Formation of ferucarbotran and protamine complexes (Fer-Pro) and magnetic
ready-to-use solution of 28 mg iron/ml (0.5 M). A ferucarbotran stock solution of 1 mg
iron/mlwaspreparedwithphosphate bufferedsaline(PBS).Protamine sulfate (Sigma-
Aldrich, St. Louis, MO) was also prepared with PBS in a stock solution of 1 mg/ml.
Complex formationwasperformedbymixingferucarbotranwitha final concentration
sonication for 10 min in the labeling media (serum-free DMEM), after which the fer-
ucarbotran and protamine complexes (Fer-Pro) were used.
and then cultured for 24 h. After PBS wash, the cells were labeled with or without
ferucarbotran(100mg iron/ml),protamine (100mg/ml), or Fer-Procomplexes (100mg
iron/ml ferucarbotran and 100 mg/ml protamine) in the labeling media for 1 h, after
which cells were washed once with 2% FBS-containing PBS and twice with fresh PBS,
and then processed for further experiments.
2.3. Cell viability assay
The cytotoxic effect of cellular labeling was assessed using 3-[4,5-dimethylth-
iazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT, Sigma-Aldrich) reduction. After
labeling, cells in 96-well plates (6 ? 103cells per well) were either immediately
incubated with fresh serum-free medium containing 0.5 mg/ml of MTT for 1 h at
37?C for acute cytotoxicity assay or allowed to grow in regular growth medium for
24 h, followed by incubation with fresh serum-free medium containing 0.5 mg/ml
MTT for 30 min at 37?C for growth effect assay. After MTT incubation, 400 ml of
dimethyl sulfoxide (DMSO, Sigma-Aldrich) was added to the cells to extract the dark
blue formazan dye generated by the live cells, and the absorbance at 570 nm was
measured using a microplate reader.
2.4. Prussian blue staining
Magnetic labeling efficiency of hMSCs was also qualitatively examined by
Prussian blue staining as previously described [27,28]. After labeling, cells in 12-well
plate (4.8 ? 104cells per well) were stained with 10% potassium hexacyano-ferrate
(II) trihydrate (Sigma-Aldrich) and 20% hydrochloric acid (HCl) for 30 min. Then the
cells werewashed twicewithdistilled waterand visualized undera light microscope
(CKX41, Olympus, Tokyo, Japan).
Fig. 1. The effect of magnetic labeling on cytotoxicity. After labeling for 1 h, the effects
of 100 mg iron/ml ferucarbotran (Fer), 100 mg/ml protamine (Pro), and the ferucarbo-
tran and protamine complexes (Fer-Pro) on the acute MTT reduction activity (A) and
the cell proliferation (B) of hMSCs, respectively. All data are expressed as
mean ? standard error of five determinations (each in quadruplicate) of three donors.
(*, p < 0.05; ***, p < 0.001).
L.-Y. Chien et al. / Biomaterials 32 (2011) 3275e3284
Magnetic labeling efficiency of hMSCs was quantitatively determined by ICP-MS
for iron content. After labeling, cells in 6-well plates (1.2 ? 105cells per well) were
washed twice by PBS for removing residual culture media, collected by trypsiniza-
tion, and centrifuged in 1-ml eppendorf tubes. The cell pellets were lysed with
0.5 ml of double deionized water. Hereafter, nitric acid (0.1 ml, ultrapure reagent
grade, Mallinckrodt Baker Inc., Phillipsburg, NJ) was admixed to digest the cell lysate
for 10 min, which was then diluted to 5 ml with double deionized water. All of the
digested cell lysates were further analyzed by ICP-MS (Agilent 7500cx, Agilent
Technologies Inc., Tokyo, Japan) for quantifying the concentration of iron.
2.6. In vitro cellular MRI protocol
MRI was performed using a clinical 1.5-T MR system (Signa Excite, GE Health-
care). After magnetic labeling, cells in 6-well plates (1.2 ? 105cells per well) were
collected by trypsinization, centrifuged, and placed in 300-ml eppendorf tubes
(1.2 ?105cells per tubes) positioned in a water tank. Next, the tank was placed in an
8 channel head coil. Gradient echo pulse sequences provided by the vendor were
used (TR/TE ¼ 550/15 ms, Flip angle ¼ 15?, matrix size ¼ 256 ? 192). The slice
thickness was 1.4 mm with a 0.3 mm gap; and the field of view (FOV) was
14 ? 10.5 cm for coronal scanning of the test tubes and 8 min and 2 s for sagittal
scanning at the NEX of 3. The images were then analyzed at the workstation
provided by GE healthcare.
2.7. In vitro migration assay
The migration capacity of hMSCs was analyzed using Costar Transwell chambers
with polycarbonate membrane filters of 6.5 mm diameter and 8 mm pore size
(Corning, NY). After Fer-Pro labeling, cells in 6-well plates (1.2 ? 105cells per well)
were harvested and then added to the top chambers at 2 ? 104cells/100 ml of 15%
FBS-containing DMEM per chamber. The bottom chambers were added with U87MG
cells at 5 ?104cells/600 ml of 10% FBS-containing DMEM per chamber. Then the top
chambers and bottom chambers were separately incubated at 37
?C, 5% CO2
overnight for cell seeding. After seeding, the media in the top chambers and the
bottom chambers were replaced with 0.5% FBS-containing DMEM, and the top
chambers were transferred to the bottom chambers. After that the chamber system
was incubated overnight at 37?C, 5% CO2for 24 h; hMSCs on the top surface of the
filters were wiped off with cotton swabs aftera careful wash with cold PBS. Cells that
had migrated into the lower compartment of the top chambers and attached to the
lower surface of the filter were counted after staining with crystal violet dye. The
bottom chambers without U87MG cells (0.5% FBS-containing DMEM only) were
used as negative control. For neutralization assays, anti-CXCR4 (100 mg/ml) (clone
12G5, R&D Systems, Lille, France) and anti-SDF-1 antibodies (250 mg/ml) (clone
79104, R&D Systems) were added with hMSCs in the top chambers or with U87MG
cells in the bottom chambers, respectively. Five replicates of each sample were
counted. The assay was done six times for each condition. The migration rate was
expressed as the percentage of migrated cell number of anti-CXCR4 antibody-
treated condition or anti-SDF-1 antibody-treated group compared with that of
vehicle control (hMSCs without anti-CXCR4 antibodies in the top chambers and
U87MG cells without anti-SDF-1 antibodies in the bottom chambers).
2.8. Animal study
Seven- to 8-week-old male nude mice were obtained from the Animal Center of
National Taiwan University and maintained in accordance with the guidelines of the
Institutional Animal Care and Use Committee. Stereotaxic implantation of cells was
performed with a NARISHIGE apparatus (51600 Single Manipulator Model, Stoelting
Co., Wood Dale, IL) and a WPI syringe (Hamilton Co., Reno, NV) with a bevel-tipped
26.5-gauge needle. Animals were anesthetized with ketamine/xylazine during the
following cell implantation.
U87MG cells maintained in high-glucose DMEM were labeled with Vybrant
DyeCycle Green stain (Molecular Probes, Eugene, OR) (5 ml/ml medium) for 48 h,
harvested, and suspended in PBS at a concentration of 1 ? 105cells/ml. DyeCycle
Green-labeled U87MG cells were injected into the right frontal lobe of mice. The
injection coordinates were (relative to bregma [AP], the midline [ML], and the dura
[DV]): AP: þ1 mm, ML: 1.5 mm (right), DV: 2 mm. Injections were performed at the
rate of 1 ml/min; and the needle was left in place for 5 min before withdrawal.
Fig. 2. The efficiency of magnetic labeling. (A) Qualitative detection of intracellular iron of unlabeled cells (Control cells), Fer-labeled cells, and Fer-Pro-labeled cells by Prussian blue
staining. (B) Intracellular iron content was quantitatively determined by ICP-MS. Data are expressed as mean ? standard error of three independent experiments. ***Statistically
significant difference (p < 0.001) as compared with unlabeled cells (Control). Fer-Pro demonstrated higher cellular uptake in hMSCs compared to Fer (#, p < 0.05). (C) Repre-
sentative MRI images of centrifuged hMSCs pellets in tubes placed in a water bath.
L.-Y. Chien et al. / Biomaterials 32 (2011) 3275e3284
Fig. 3. Imaging of tumor tropism of hMSCs toward glioma cells in vitro and in vivo. (A) In vitro migration of hMSCs in Transwell assay. Unlabeled hMSCs in the top chambers
migrated toward the bottom chambers with 0.5% FBS-containing media (a), 30% FBS-containing media (b), or U87MG cells in 0.5% FBS-containing media (c). Fer-Pro-labeled
hMSCs in the top chambers migrated toward the bottom chambers with U87MG cells in 0.5% FBS-containing media (d). Comparing (c) with (d), Fer-Pro labeling did not prohibit
but promoted the migration capacity of hMSCs toward U87MG cells (e). Data are expressed as mean ? standard error of five determinations (each in quadruplicate) of three
donors (***, p < 0.001). (B) In vivo MRI of mouse that received Vybrant DyeCycle Green-labeled U87MG cells (green arrow) in the right frontal lobes for 14 days. (C) In vivo MRI of
mouse that received Vybrant DyeCycle Green-labeled U87MG cells alone (green arrow) in the right frontal lobes for 28 days. (D) In vivo MRI of mouse injected with Vybrant
DyeCycle Green-labeled U87MG cells (green arrow) in the right frontal lobes for 28 days and SP-Dil-Fer-Pro colabeled hMSCs (yellow arrow) into the left lateral ventricles for 14
days. Red arrow denotes the migrated SP-Dil-Fer-Pro colabeled hMSCs toward DyeCycle Green-labeled U87MG cells in the right frontal lobes. (E) In vivo MRI of mouse injected
L.-Y. Chien et al. / Biomaterials 32 (2011) 3275e3284
hMSCs in growth media (10% FBS) were labeled with fluorescent dye SP-Dil
(Molecular Probes) (10 mg/ml medium) for 48 h, harvested, and subcultured onto
6-well plates at 1.2 ? 105cells/well for magnetic cellular labeling. At day 14 post-
implantation, SP-Dil-Fer-Pro colabeled hMSCs were resuspended in serum-free
DMEM at a concentration of 3.6 ? 105cells/5 ml and injected into the left lateral
ventricles of nude mice. For in vivo tropism blocking, SP-Dil-Fer-Pro colabeled
hMSCs in serum-free DMEM were mixed with anti-CXCR4 antibodies (200 mg/ml)
before the implantation. The injection coordinates were (relative tobregma [AP], the
midline [ML], and the dura [DV]): AP: 0 mm, ML: 1 mm (left), and DV: 2.0 mm.
Injections were performed at therateof 1ml/min; and the needlewas left inplace for
5 min before withdrawal.
2.9. In vivo MRI protocol
After implantation of hMSCs, the clinical 1.5-T MR system applied in vitro was
also used to observe the tropism of hMSCs in vivo. Under gas anesthesia with 2%
isoflurane, the mouse was placed in a homemade resonance coil with an inner
diameter of 3.7 cm. Fast spin echo pulse sequences provided by the vendor were
used (TR/TE ¼ 4000/101.4 ms, Matrix size ¼ 288 ? 192). The slice thickness was
0.8 mm with a 0.2 mm gap, and the FOV was 5 ? 2.5 cm. Total scan time was 3 min
and 20 s at the NEX of 8. The imageswere then analyzed at the workstationprovided
by GE healthcare.
2.10. Histological analysis of mouse brains
All animals were sacrificed 2 weeks after the implantation of hMSCs with an
overdose of ketamine/xylazine. The brains were harvested, embedded in OCT, and
stored at ?80?C. Brains were cryosectioned at ?20?C by a cryostat (CM1900, Leica,
Heerbrugg, Switzerland), and 20-mm thick sections from representative areas were
directlyexamined with a confocal microscope (LSM510, CarlZeiss, Jena, Germany) to
observe DyeCycle Green-labeled U87MG cells and SP-Dil-labeled hMSCs. The
sections were then stained with hematoxylin and eosin (H&E, Sigma-Aldrich) for
visualization of gliomas, or processed for Prussian blue staining to determine the
localization of Fer-Pro-labeled hMSCs.
2.11. Quantitative MRI for glioma size analysis
We measured glimoa size by performing MRI scanning with T2 weight pulse
sequences described earlier. The tumor growth rate was observed by imaging
identical rat twice at 6e11 day intervals from the beginning of 21e29 days after
tumor implantation. The images were analyzed by two investigators (L.-Y. Chien and
J.-K. Hsiao). The tumor size was measured with plotting the region of interest of each
significant slice. All of the tumor slices were summed up; and the signal intensities
of the slices were recorded for statistical analysis. Up to 8 mice in three group-
sdwith tumor, including SP-Dil-Fer-Pro colabeled hMSCs (N ¼ 3), with tumor-alone
group (N ¼ 3), and with CXCR4 antibody pretreated SP-Dil-Fer-Pro colabeled hMSCs
(N ¼ 2)dwere scanned, with the image data used for further statistical analysis.
2.12. Statistical analysis
Data are presented as the mean ? standard error of mean (SEM) for the indi-
cated numbers of separate experiments. The results were compared using Student’s
t-test. Statistical significance was assigned if the p-value was less than 0.05. For
correlation between observed signal intensity of glioma and its growth rate, linear
regression model was used and F-test was used for accept or reject null hypothesis
(a ¼ 0.05).
3.1. Viability and proliferation potential of magnetic-labeled hMSCs
In our previous study, we showed that ferucarbotran was not
cytotoxic and that it promoted the proliferation of hMSCs ; but
the effect of protamine on the cellular viability and proliferation
potential of hMSCs was not tested. Although some studies have
suggested that protamine has no significant deleterious effect on
hMSC survival [23,29,30], the cytotoxicologic concern of using
polycationic transfection agents for cellular labeling has not been
fully addressed; hence, in this study we used MTT reduction assay
to verify the cytotoxic effect of protamine labeling on hMSCs. As in
our previous study , ferucarbotran (100 mg iron/ml) induced
greater MTT reduction in the acute cytotoxicity (Fig. 1A) and
proliferation (Fig. 1B) assays; however, protamine at a high dose
(100 mg/ml) was shown to significantly decrease the MTT reduction
activity in both assays. Interestingly, the ferucarbotran and prot-
amine complexes (Fer-Pro) did not affect cell viability in the acute
cytotoxicity assay (Fig. 1A) but caused a slight and significant
increase in cell proliferation (Fig.1B). Thus, these results suggested
that ferucarbotran and Fer-Pro were not cytotoxic for labeling
hMSCs and prompted us to investigate the labeling efficiency in
vitro and in vivo.
3.2. Magnetic cellular labeling efficiency
In order totest the utility of ferucarbotran or Fer-Pro for imaging
cell tropism, we used Prussian blue staining and ICP-MS to deter-
mine the cellular iron uptake and employed a clinical 1.5-T MRI
system to examine the MRI-detectable efficiency. By a simple
incubation for 1 h, qualitative immunostaining showed that fer-
ucarbotran-labeled hMSCs obviously displayed Prussian blue
positive; and a stronger staining was observed in Fer-Pro-labeled
cells than cells with ferucarbotran-labeling (Fig. 2A). The iron
amount engulfed by hMSCs was also quantitatively determined by
ICP-MS. The iron contents increased significantly in hMSCs labeled
with ferucarbotran (71.6 ? 8.2 pg/cell), and greatly in Fer-Pro-
labeled cells (157.5 ? 27.5 pg/cell) (Fig. 2B). Under T2-weighted
image mode, the MRI images of ferucarbotran-labeled or Fer-Pro-
labeled cell pellets in test tubes placed in a water bath were easily
and obviously detected as dark spots with hypointensities; more-
over, the MRI signal intensity loss of hMSCs labeled with Fer-Pro
was greater than that of those treated with ferucarbotran (Fig. 2C).
The fact that a higher iron uptake in qualitative Prussian blue
staining and ICP-MS quantification correlated with the finding of
marked contrast effect in MRI suggested the potential of Fer-Pro
labeling for detailed MRI in vivo.
3.3. Tumor tropsim of hMSCs: in vitro and in vivo visualization
glioma cells, an in vitro migration assay using Transwell systemwas
first employed. The migration capacity of hMSCs in the presence of
0.5% FBS-containing medium alone in the bottom chamber was low
(negative control; Fig. 3A, a); it increased when 30% FBS-containing
medium was in the bottom chamber (positive control; Fig. 3A, b).
U87MG cells in the bottom chamber (with 0.5% FBS-containing
medium) significantly stimulated the migration of non-labeled
hMSCs (Fig. 3A, c) and Fer-Pro-labeled hMSCs (Fig. 3A, d). By
counting crystal violet-stained cells, we found that Fer-Pro labeling
did not prohibit but promoted the migration capacity of hMSCs
toward U87MG cells (Fig. 3A, e).
Because the migration capacities of hMSCs and Fer-Pro-labeled
hMSCs in vitro were established in the study, we next investigated
whether intraventricularly-transplanted Fer-Pro-labeled hMSCs are
capable of migrating toward human gliomas and then being
imaged by MRI in vivo. For postmortem histological examination,
U87MG cells and hMSCswerestainedbeforeinjectionwith Vybrant
DyeCycle Green and SP-Dil, respectively, as described in Section 2.
Fourteen days after intracranial inoculation of Vybrant DyeCycle
with SP-Dil-Fer-Pro colabeled hMSCs (yellow arrow) alone into the left lateral ventricles for 14 days. (FeH) Immunohistological analyses of mice from (CeE), respectively. (F): SP-
Dil (a), DyeCycle Green (b), merge image (c), and Prussian blue staining (d) of right brain samples. (G): SP-Dil (a), DyeCycle Green (b), merge image (c), and Prussian blue staining
(d) of right brain samples. (H): SP-Dil (a) and Prussian blue staining (c) of left brain samples; SP-Dil (b) and Prussian blue staining (d) of right brain samples. L: left site; R: right
site. Scale bar 50 mm.
L.-Y. Chien et al. / Biomaterials 32 (2011) 3275e3284
Green-labeled U87MG cells in the right frontal lobes of nude mice
when xenografts were established and an intrinsic bright positive
signal (a white spot) of glioma cells could be observed under
T2-weighted images in vivo (Fig. 3B green arrow), SP-Dil-Fer-Pro
colabeled hMSCs were injected into the left lateral ventricles. The
stem cell images of MRI in animals were acquired 14 days after
injection of SP-Dil-Fer-Pro colabeled hMSCs, animals were sacri-
ficed, and the brains were removed for being multi-observed.
Twenty-eight days after intracranial inoculation of Vybrant Dye-
Cycle Green-labeled U87MG cells, the area of intrinsic bright
positive signal of glioma cells increased (green arrows in Fig. 3C
vs. 3B) in the right frontal lobe of the control animal that only
received Vybrant DyeCycle Green-labeled U87MG cells but without
SP-Dil-Fer-Pro colabeled hMSCs; however, in the animal that
received both Vybrant DyeCycle Green-labeled U87MG cells and
SP-Dil-Fer-Pro colabeled hMSCs, the area of intrinsic bright positive
signal of glioma in the right frontal lobe was obviously less than
that of control (green arrows in Fig. 3D vs. C). Interestingly,
although few of intraventricularly-transplanted SP-Dil-Fer-Pro
colabeled hMSCs remained at the original injection site (left
ventricle) (Fig. 3D yellow arrow), most of the cells appeared as
hypointense regions in the right ventricle (Fig. 3D red arrow). The
animal injected with SP-Dil-Fer-Pro colabeled hMSCs but without
Vybrant DyeCycle Green-labeled U87MG cells showed a main area
of hypointensities on the left ventricle (Fig. 3E yellow arrow). These
results suggest the migration capacity of SP-Dil-Fer-Pro colabeled
hMSCs to the glioma cells and the therapeutic impact of the
migrated SP-Dil-Fer-Pro colabeled hMSCs on gliomas.
Correlating the MRI findings with histological and immunos-
taining evidence provide more evidence of the migratory and
therapeutic capacities of hMSCs. From the animal of Fig. 3C, no
SP-Dil-Fer-Pro colabeled hMSCs could be seen because of no
Prussian blue staining, respectively), and Vybrant DyeCycle Green-
labeled U87MG cells were clearly observed in the right frontal lobe
Fig. 3D), SP-Dil-Fer-Pro colabeled hMSCs migrated and were
detected around Vybrant DyeCycle Green-labeled U87MG cells in
the right frontal lobe (Fig. 3G; panel a for SP-Dil, panel b for Vybrant
DyeCycle Green, panel c for the merge image of panels a and b, and
panel d for Prussian blue staining). In addition, there was no
receiving the 1st stereotaxic injection of Vybrant DyeCycle Green-
labeled U87MG cells (Fig. 3E and H). Colocalization of SP-Dil and
brain samples; panel a for SP-Dil and panel c for Prussian blue
staining respectively) but not in the right side of brain samples
(Fig. 3H, right brain samples; panel b for SP-Dil, and panel d for
Prussian blue staining), indicating the migration of SP-Dil-Fer-Pro
colabeled hMSCs is due to a specific attraction of gliomas.
3.4. Mechanism study of glioma tropism of hMSCs in vitro
and in vivo
To further explore the possibility of using the labeling of fer-
ucarbotran combined with protamine (Fer-Pro) for identifying the
tropismmechanisminvivo, usinginvitro migrationassay wetested
is widely believed to be involved in the tropism of hMSCs toward
tumors. First, we demonstrated that hMSCs in the study did express
CXCR4 and that the labeling of Fer-Pro did not diminish its expres-
sion (Fig. 4A). As shown in Fig. 3A, Fer-Pro labeling did not prohibit
hMSCs were used as control in the mechanism experiments. By the
incubation of Fer-Pro-labeled hMSCs with an anti-CXCR4-blocking
antibody in the top chamber or with the addition of anti-SDF-1
Pro-labeled hMSCs toward U87MG cells in the bottom chamber
significantly decreased (Fig. 4B). The inhibitory effect on Fer-Pro-
labeled hMSCs migration was not observed with isotype-matched
hMSCs toward U87MG gliomas in the study. In the in vivo experi-
ments, similar to those in Fig. 3D and G, the migration of intraven-
tricularly-transplanted SP-Dil-Fer-Pro colabeled hMSCs to the
gliomas and the therapeutic impact of the migrated SP-Dil-Fer-Pro
colabeled hMSCs on gliomas were observed in Fig. 4C and E.
However, asshownin Fig.4D,thetreatmentof anti-CXCR4-blocking
antibodies not only inhibited the in vivo migration of SP-Dil-Fer-Pro
efficacy. In vivo MRI showed that anti-CXCR4-blocking antibodies-
treated SP-Dil-Fer-Pro colabeled hMSCs failed to migrate but were
still imaged in the original injection site (Fig. 4D yellow arrow) and
that a greater area of intrinsic bright positive signal represented the
expansion of gliomas (Fig. 4D green arrow). Histological examina-
tions alsoshowedthatSP-Dil-Fer-Pro colabeledhMSCsstayedleft in
the animal brain (Fig. 4F, left brain samples; panel a for SP-Dil and
panel c for Prussian blue staining respectively) and that Vybrant
(Fig. 4F, right brain samples; panel b for Vybrant DyeCycle Green).
3.5. Evaluation of therapeutic impact of migrated hMSCs
Asabove shown inFigs.3and4,thetherapeuticimpactofhMSCs
on gliomas could be observed in vivo by MRI due to the intrinsic
bright positive signal of gliomas that originate from edematous
change, we analyzed the area of tumor to quantify the antitumor
effect of hMSCs. The tumor growth rate of SP-Dil-Fer-Pro colabeled
hMSCs group shows shrinkage of the glioma after the migration of
SP-Dil-Fer-Pro colabeled hMSCs into the tumor part. The mean
tumor growth size is ?0.8 mm3whereas the tumor-alone group is
6.67 mm3(p < 0.03) (Fig. 5A). The measured signal intensity of
glioma at first MRI observation was 3.37 absolute units (AU) in the
SP-Dil-Fer-Pro colabeled hMSCs group and 8.17 AU in the tumor-
alone group; the results reached statistical significance. We also
found the tumor growth and glimoa signal intensity of the CXCR4-
pretreated SP-Dil-Fer-Pro colabeled hMSCs group was between that
of the SP-Dil-Fer-Pro colabeled hMSCs and tumor-alone groups
(Fig. 5B). We further analyzed the relationship between measured
signal intensity and tumor size. Interestingly, the signal intensity
correlated well with glioma tumor growth under the linear regres-
significant under the F-test (Fig. 5C). Histological examination by
H&E staining also qualitatively demonstrated that migrated hMSCs
inhibited the expansion of gliomas in vivo. As typically shown in
Fig. 5D, enormous glioma masses (denoted by black dotted line) not
only occupied the right hemisphere and extended the growth into
the left hemisphere in the mouse that only received Vybrant Dye-
Cycle Green-labeled U87MG cells. In the animal that received both
Vybrant DyeCycle Green-labeled U87MG cells and SP-Dil-Fer-Pro
colabeled hMSCs (Fig. 5E), the gliomas retained in the right hemi-
sphere and were obviously smaller than those in Fig. 5B. However,
when SP-Dil-Fer-Pro colabeled hMSCs were pretreated with anti-
occupation of the right hemisphere shifted the midline toward the
left hemisphere (Fig. 5F). These results confirmed the antitumor
effect of hMSCs and the involvement of SDF-1/CXCR4 in the migra-
tion and antitumor mechanisms and showed that antitumor effect
could be determined by a clinical MRI system.
L.-Y. Chien et al. / Biomaterials 32 (2011) 3275e3284
The use of tumor-tropic stem cells for targeted delivery of
tumoricidal therapeutic agents to tumor sites is generally believed
to be able to improve therapeutic efficacy in glioma patients. MRI
represents an attractive approach for monitoring the trafficking of
transplanted stem cells in vivo in order to achieve effective use of
stem cell-based cancer therapy; however, the cells of interest have
to be internally labeled with suitable MRI contrast agents, such as
SPIO nanoparticles, prior to transplantation. Therefore, it is highly
required to develop efficient and biosafe labeling strategies, man-
ifested as the requirements of short-term incubation and low
Fig. 4. Imagingof themechanismoftumortropismofhMSCs.(A)TheimmunofluorescentdetectionofCXCR4expressioninhMSCs.(B)InvitromigrationofFer-Pro-labeledhMSCswas
blocked byanti-CXCR4-blocking antibodies and anti-SDF-1 neutralization antibodies. Data are expressed as mean ?standard errorof seven independent experiments. ***Statistically
significant difference (p< 0.001) as comparedwith untreated cells (Control).(C)Invivo MRI of mouse injectedwithVybrantDyeCycle Green-labeled U87MG cells(green arrow) in the
right frontal lobes for 28 days and SP-Dil-Fer-Pro colabeled hMSCs (yellow arrow) into the left lateral ventricles for 14 days. Red arrow denotes the migrated SP-Dil-Fer-Pro colabeled
right frontal lobes for 28 days and anti-CXCR4-blocking antibodies-pretreated SP-Dil-Fer-Pro colabeled hMSCs (yellow arrow) into the left lateral ventricles for 14 days. No tropism of
SP-Dil-Fer-Procolabeled hMSCs was observed.(E) and (F):immunohistological analyses of mice from (C)and (D), respectively. (E): SP-Dil (a), DyeCycle Green(b), merge image (c), and
bar 50 mm.
L.-Y. Chien et al. / Biomaterials 32 (2011) 3275e3284
concentration of labeling agents, the employment of low number of
labeled cells and a clinical-used MRI, and lower toxicity. Whereas
protamine has been employed to improve the labeling efficiencies
of SPIO nanoparticles in many cell types [23,24,31,32], in this study
we have demonstrated that the combination of ferucarbotran and
protamine would be a clinically applicable strategy for achieving
good MRI performance when studying cellular behaviors in stem
cell-mediated glioma therapy. Although several MRI studies show
promise for the monitoring of cellular trafficking in animal models
of regenerative medicine as well as stem cell-mediated tumor
therapy [33e35], to our knowledge this is the first report in which
a clinical MRI modality was used to simultaneously study the
migration, the therapeutic impact on tumors, and above all the
trafficking mechanism of bone marrow-derived mesenchymal stem
cells from human in a murine glioma xenograft model.
The primary concern associated with using MRI for cellular
tracking is the uptake/labeling efficiency of the MRI-visible particles
selected for use. Up to now, of the two main classes of MRI-visible
agents for labeling cellsdi.e., SPIO nanoparticles and Gd-based che-
probe. SPIO nanoparticles act as negative contrast agents, producing
strong hypointensities (dark spots) on T2- or T2*-weighted images.
Gd-based contrast agents, however, produce bright positive signal
intensity in T1-weighted MRI images. Due to the intrinsic bright
Fig. 5. Imaging and evaluation of therapeutic impact of hMSCs on gliomas. The tumor growth observed during two MR imaging observation in SP-Dil-Fer-Pro colabeled hMSCs
(hMSCs), anti-CXCR4-blocking antibodies-pretreated SP-Dil-Fer-Pro colabeled hMSCs (CXCR4), and tumor-alone (Tumor) groups. The U87MG glioma growth shows statistical
significance between hMSCs treated group and U87MG tumor-alone group (p < 0.03). The anti-CXCR4-blocking antibodies-pretreated SP-Dil-Fer-Pro colabeled hMSCs group exhibit
intermediate tumor growth rate between the other groups (A). The signal intensity of U87MG glioma at first MRI observation among SP-Dil-Fer-Pro colabeled hMSCs group (hMSCs),
anti-CXCR4-blocking antibodies-pretreated SP-Dil-Fer-Pro colabeled hMSCs (CXCR4), and tumor-alone (Tumor) groups. Statistical significance could be found in the hMSCs treated
group and U87MG cell alone group (p < 0.005). The findings suggest more stem cell tropism toward glioma in the SP-Dil-Fer-Pro colabeled hMSCs group (B). Relationship between
observed U87MG glioma signal intensity at first MRI session and the U87MG cells growth rate was visualized by MRI. Strong relationship (Relation coefficient ¼ 0.83, p < 0.05) was
observed between these two parameters (C). Macroscopic view (H&E staining) of brain specimens of mice that were injected with Vybrant DyeCycle Green-labeled U87MG cells
alone in the right frontal lobes for 28 days (D), Vybrant DyeCycle Green-labeled U87MG cells in the right frontal lobes for 28 days and SP-Dil-Fer-Pro colabeled hMSCs into the left
lateral ventricles for 14 days (E), or Vybrant DyeCycle Green-labeled U87MG cells in the right frontal lobes for 28 days and anti-CXCR4-blocking antibodies-pretreated SP-Dil-Fer-Pro
colabeled hMSCs into the left lateral ventricles for 14 days (F). Gliomas are denoted by black dotted lines.
L.-Y. Chien et al. / Biomaterials 32 (2011) 3275e3284
positive signal of gliomas in T2-weighted images in vivo, SPIO
nanoparticles are more suitable than Gd-based nanoparticles in dis-
tinguishing labeled stem cells intargetedsites withhigh MRI signals.
Although the internalization mechanisms were not examined in this
study, an important role of endocytosis has been strongly sugges-
ted in the uptake/labeling process of ferucarbotran and Fer-Pro
complexes [20,21,24]. High cellular labeling can be achieved by
combining several strategies,including using simple incubationwith
SPIO nanoparticles alone for a longer incubation time, combination
with a transfection agent but a shorter incubation period, surface
modifications of SPIO nanoparticles with other materials, HIV tat
peptide, or monoclonal antibody [36e39], and electroporation .
perform Fer-Pro complexes but a shorter incubation time to label
hMSCs than previously reported by Golovko et al. , resulting in
a huge cellular loading of iron (157.5 ? 27.5 pg/cell in Fig. 2B). The
in ourcase offers an instant but still biosafe labeling for hMSCs in the
glioma model (as discussed below); moreover, this method may
With regard to the toxicity of labeling strategies of the combi-
nation of protamine and SPIO particles, protamine is usually
believed to be biosafe for enhancing SPIO nanoparticle labeling.
Studies with protamine, however, have focused mainlyon the effect
of complexes of SPIO nanoparticles and protamine on cellular
attributes; thus, the potential of polycationic transfection agents to
harm stem cell behavior remains uncertain [41,42]. This prompted
us to examine the cytotoxic effects of all the labeling agents, espe-
cially protamine itself, in hMSCs. The labeling of hMSCs with prot-
amine alone at a high concentration (100 mg/ml) in this study
resulted in the cell number decreases, suggesting a cytotoxic
potential of protamine for stem cell labeling. However, the pro-
moting activity of ferucarbotran on cell growth alleviated the threat
of protamine to cellular viability, suggesting that protamine may be
an uptake-enhancing adjuvant with ferucarbotran for stem cell
labeling. In addition to protamine, SPIO nanoparticles (such as,
in this study, ferucarbotran) play an important role in toxicity
concerns, although they are generally believed to be inert and
biocompatible. Previously we have shown that ferucarbotran can
promote cell growth of hMSCs , as verified in the present study,
leaving the question of whether this growth promotion is good or
bad for stem cells themselves. Recently we demonstrated that fer-
ucarbotran could activate the migration of hMSCs in the osteogenic
medium and subsequently induce the abolishment of cellular
osteogenesis , suggesting the need for caution about using
ferucarbotran to label stem cells for osteogenic MRI tracking.
Furthermore, the applicability of protamine in combination with
ferucarbotran for stem cell labeling would be greatly dependent on
the mission of stem cells in different circumstances. In the glioma
models, the migration capacity toward gliomas and the expression
activity of therapeutic transgene but not multi-potent differentia-
tions are of vital importance for stem cells to be used as a tool for
targeted therapeutic transgene delivery and then tumor therapy. In
the present study the glioma tropism of hMSCs is demonstrated by
in vitro Transwell assay and by in vivo MRI; and the labeling of Fer-
Pro complexes has been shown to activate the in vitro migration of
of hMSCs in vivo for that unlabeled hMSCs cannot be observed by
MRI, the caution of migration activation for using stem cells in
osteogenesis or other regenerative medicines would advantage Fer-
Pro complexes to label/image hMSCs in stem cell-mediated tumor
therapy. More comprehensive analyses of the effects of Fer-Pro
complexes on the in vivo tropism and the expression of engineered
genes are warranted.
NSCs as well as MSCs from various species (e.g., mouse, rat, and
human) have been demonstrated to be able to migrate toward
gliomas in vitro and in vivo [6e12,17,33e35]. Although several
growth factors/chemokines maybe possible candidates responsible
of CXCR4 and SDF-1 is known to be involved in the tropism of
hMSCs. Although the tropism behaviors in above studies can be
determined by histochemical staining, fluorescent images, or
noninvasive and dynamic MRI, no report has shown the utility of
we not only demonstrate the tropism of hMSCs toward gliomas in
vitro and in vivo by histochemical staining, fluorescent images, and
a clinical MRIsystem butalso indicatethatthe SDF-1/CXCR4 system
of hMSCs in vivo. Additionally, it should be noted that MSCs in the
study were from human subjects, suggesting an autologous trans-
plantation of these cells in patients. Furthermore, the observed
SDF-1/CXCR4 mechanism may provide insights into strategies for
enhancing the glioma tropism of hMSCs for better therapeutic
The strong relationship between signal intensity of U87MG
glioma was detected under MRI and tumor growth. The signal
intensity of glioma is determined by the tumor itself and the
migrated hMSCs. The finding reinforces that hMSCs implantation is
feasible as a treatment method and evaluation technique for
understanding stem cell behavior in vivo. As our data show, the
tropism of hMSCs could be visualized clearly under 1.5 Tclinical MR
system after SPIO labeling.
Labeling of hMSCs by ferucarbotran and protamine complexes
(Fer-Pro) is not toxic to cells but can promote the migration of
hMSCs toward glioma cells in vitro. Fer-Pro-labeled hMSCs can
specifically migrate toward gliomas in vivo, which was observed
with a clinical 1.5-T MRI system. The efficient labeling of Fer-Pro
allows a tropic mechanism mediated by SDF-1/CXCR4 signaling to
be also detected by MRI in vivo. Additionally, the potential intrinsic
inhibitory effect of hMSCs on glioma progression can be estimated
simultaneously. This MRI technique described in the study has
clinical application potential for stem cell guided therapy.
This work was supported by grants from the National Health
Research Institutes (NHRI) (NM-098-PP-02, NM-098-PP-09, NM-
099-PP02 and NM-098-PP-12) and the National Science Council
(96-2628-B-400-001-MY3 and 97-2314-B-303-018-MY2), both of
are difficult to interpret in black and white. The full color images
 Walker AE, Robins M, Weinfeld FD. Epidemiology of brain tumors: the
national survey of intracraninal neoplasms. Neurology 1985;35:219e26.
 Louis DN, Holland EC, Cairncross JG. Glioma classification: a molecular reap-
praisal. Am J Pathol 2001;159:779e86.
 Ohgaki H, Dessen P, Jourde B, Horstmann S, Nishikawa T, Di Patre PL, et al.
Genetic pathways to glioblastoma: a population-based study. Cancer Res
L.-Y. Chien et al. / Biomaterials 32 (2011) 3275e3284
 Grossman SA, Batara JF. Current management of glioblastoma multiforme.
Semin Oncol 2004;31:635e44.
 Ehtesham M, Stevenson CB, Thompson RC. Stem cell therapies for malignant
glioma. Neurosurg Focus 2005;19:E5.
 Aboody KS, Brown A, Rainov NG, Bower KA, Liu S, Yang W, et al. Neural stem
cells display extensive tropism for pathology in adult brain: evidence from
intracranial gliomas. Proc Natl Acad Sci USA 2000;97:12846e51.
 Benedetti S, Pirola B, Pollo B, Magrassi L, Bruzzone MG, Rigamonti D, et al.
Gene therapy of experimental brain tumors using neural progenitor cells. Nat
 Ehtesham M, Yuan X, Kabos P, Chung NH, Liu G, Akasaki Y, et al. Glioma tropic
neural stem cells consist of astrocytic precursors and their migratory capacity
is mediated by CXCR4. Neoplasia 2004;6:287e93.
 Lesniak MS. Targeted therapy for malignant: neural stem cells. Expert Rev
 Staflin K, Honeth G, Kalliomäki S, Kjellman C, Edvardsen K, Lindvall M. Neural
progenitor cell lines inhibit rat tumor growth in vivo. Cancer Res 2004;64:
 Nakamizo A, Marini F, Amano T, Khan A, Studeny M, Gumin J, et al. Human
bone marrow-derived mesenchymal stem cells in the treatment of gliomas.
Cancer Res 2005;65:3307e18.
 Sonabend AM, Ulasov IV, Tyler MA, Rivera AA, Mathis JM, Lesniak MS.
Mesenchymal stem cells effectively deliver an oncolytic adenovirus to intra-
cranial glioma. Stem Cells 2008;26:831e41.
 Friedenstein AJ, Chailakhyan RK, Gerasimov UV. Bone marrow osteogenic
stem cells: in vitro cultivation and transplantation in diffusion chambers. Cell
Tissue Kinet 1987;20:263e72.
 Bianco P, Riminucci M, Gronthos S, Robey PG. Bone marrow stromal stem
cells: nature, biology, and potential applications. Stem Cells 2001;19:180e92.
 Tocci A, Forte L. Mesenchymal stem cells: use and perspectives. Hematol J
 Vogel G. Embryonic stem cells. Stem cells not so stealthy after all. Science
 Nakamura K, Ito Y, Kawano Y, Kurozumi K, Kobune M, Tsuda H, et al. Anti-
tumor effect of genetically engineered mesenchymal stem cells in a rat glioma
model. Gene Ther 2004;11:1155e64.
 Khakoo AY, Pati S, Anderson SA, Reid W, Elshal MF, Rovira II, et al. Human
mesenchymal stem cells exert potent antitumorigenic effects in a model of
Kaposi’s sarcoma. J Exp Med 2006;230:1235e47.
 Qiao L, Xu Z, Zhao T, Zhao Z, Shi M, Zhao RC, et al. Suppression of tumori-
genesis by human mesenchymal stem cells in a hepatoma model. Cell Res
 Mailänder V, Lorenz MR, Holzapfel V, Musyanovych A, Fuchs K, Wiesneth M,
et al. Carboxylated superparamagnetic iron oxide particles label cells intra-
cellularly without transfection agents. Mol Imaging Biol 2008;10:138e46.
 Huang DM, Hsiao JK, Chen YC, Chien LY, Yao M, Chen YK, et al. The promotion
of human mesenchymal stem cell proliferation by superparamagnetic iron
oxide nanoparticles. Biomaterials 2009;30:3645e51.
 Bull BS, Huse WM, Brauer FS, Korpman RA. Heparin therapy during extra-
corporeal circulation. II. The use of a dose-response curve to individualize
heparin and protamine dosage. J Thorac Cardiovasc Surg 1975;69:685e9.
 Arbab AS, Yocum GT, Kalish H, Jordan EK, Anderson SA, Khakoo AY, et al.
Efficient magnetic cell labeling with protamine sulfate complexed to ferum-
oxides for cellular MRI. Blood 2004;104:1217e23.
 Golovko DM, Henning T, Bauer JS, Settles M, Frenzel T, Mayerhofer A, et al.
Accelerated stem cell labeling with ferucarbotran and protamine. Eur Radiol
 Huang DM, Hung Y, Ko BS, Hsu SC, Chen WH, Chien CL, et al. Highly efficient
cellular labeling of mesoporous nanoparticles in human mesenchymal stem
cells: implication for stem cell tracking. FASEB J 2005;19:2014e6.
 Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R, Mosca JD, et al.
Multilineage potential of adult human mesenchymal stem cells. Science 1999;
 Metz S, Bonaterra G, Rudelius M, Settles M, Rummeny EJ, Daldrup-Link HE.
Capacity of human monocytes to phagocytose approved iron oxide MR
contrast agents in vitro. Eur Radiol 2004;14:1851e8.
 Hsiao JK, Tai MF, Chu HH, Chen ST, Li H, Lai DM, et al. Magnetic nanoparticle
labeling of mesenchymal stem cells without transfection agent: cellular
behavior and capability of detection with clinical 1.5 T magnetic resonance at
the single cell level. Magn Reson Med 2007;59:717e24.
 Bulte JWM, Kraitchman DL, Mackay AM, Pittenger MF. Chondrogenic differ-
entiation of mesenchymal stem cells is inhibited after magnetic labeling of
with ferumoxides. Blood 2004;104:3410e2.
 Arbab AS, Yocum GT, Rad AM, Khakoo AY, Fellowes V, Read EJ, et al. Labeling
of cells with ferumoxides-protamine sulfate complexes does not inhibit
function or differentiation capacity of hematopoietic or mesenchymal stem
cells. NMR Biomed 2005;18:553e9.
 Mani V, Adler E, Briley-Saebo KC, Bystrup A, Fuster V, Keller G, et al. Serial in
vivo positive contrast MRI of iron oxide-labeled embryonic stem cell-derived
cardiac precursor cells in a mouse model myocardial infraction. Magn Reson
 Sadek H, Latif S, Collins R, Garry MG, Garry DJ. Use of ferumoxides for stem
cells labeling. Regen Med 2008;3:807e16.
 Zhang Z, Jiang Q, Jiang F, Ding G, Zhang R, Wang L, et al. In vivo magnetic
resonance imaging tracks adult neural progenitor cell targeting of brain tumor.
 Thu MS, Najbauer J, Kendall SE, Harutyunyan I, Sangalang N, Gutova M, et al.
Iron labeling and pre-clinical MRI visualization of therapeutic human neural
stem cells in a murine glioma model. PLoS ONE 2009;4:e7218.
 Josephson L, Tung CH, Moore A, Weisseleder R. High-efficiency intracellular
magnetic labeling with novel superparamagnetic-Tat peptide conjugates.
Bioconjug Chem 1999;10:186e91.
 Lewin M, Carlesso N, Tung CH, Tang XW, Cory D, Scadden DT, et al. Tat
peptide-derivatized magnetic nanoparticles allow in vivo tracking and
recovery of progenitor cells. Nat Biotechnol 2000;18:410e4.
 GuptaAK, GuptaM. Cytotoxicity suppressionand cellular uptake enhancement
of surface modified magnetic nanoparticles. Biomaterials 2005;26:1565e73.
 Wilhelm C, Billotey C, Roger J, Pons JA, Bacri JC, Gazeau F. Intracellular uptake
of anionic superparamagnetic nanoparticles as a function of their surface
coating. Biomaterials 2003;24:1001e11.
 Walczak P, Kedziorek DA, Gilad AA, Lin S, Bulte JWM. Instant MR labeling of
stem cells using magnetoelectroporation. Magn Reson Med 2005;54:769e74.
to polycation induced cytotoxicity. Adv Drug Deliv Rev 2006;58:1523e31.
 Symonds P, Murray JC, Hunter AC, Debska G, Szewczyk A, Moghimi SM. Low
and high molecular weight poly(L-lysine)s/poly(L-lysine)-DNA complexes
initiate mitochondrial-mediated apoptosis differently. FEBS Lett 2005;579:
 Chen YC, Hsiao JK, Liu HM, Lai IY, Yao M, Hsu SC, et al. The inhibitory effect of
superparamagnetic iron oxide nanoparticle (Ferucarbotran) on osteogenic
differentiation and its signaling mechanism in human mesenchymal stem
cells. Toxicol Appl Pharmacol 2010;245:272e9.
L.-Y. Chien et al. / Biomaterials 32 (2011) 3275e3284