Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

TISSUE-SPECIFIC STEM CELLS

Nitric Oxide Donor Upregulation of Stromal Cell-Derived Factor-1/


Chemokine (CXC Motif) Receptor 4 Enhances Bone Marrow
Stromal Cell Migration into Ischemic Brain After Stroke
XU CUI,a JIELI CHEN,a ALEX ZACHAREK,a YI LI,a CYNTHIA ROBERTS,a ALISSA KAPKE,b
SMITA SAVANT-BHONSALE,c MICHAEL CHOPPa,d
Departments of aNeurology and bBiostatistics and Research Epidemiology, Henry Ford Health Sciences Center,
Detroit, Michigan, USA; cDepartment of Neurobiology, Theradigm, Inc., Baltimore, Maryland, USA;
d
Department of Physics, Oakland University, Rochester, Michigan, USA
Key Words. DETA-NONOate • Stromal cell-derived factor-1/chemokine receptor 4 • Matrix metalloproteinases
Bone marrow stromal cell • Migration • Stroke

ABSTRACT
Stromal cell-derived factor-1 (SDF1) and its chemokine come after stroke compared with BMSC monotherapy
(CXC motif) receptor 4 (CXCR4), along with matrix metal- and MCAo control, and it increased SDF1 expression in
loproteinases (MMPs), regulate bone marrow stromal cell the ischemic brain compared with DETA-NONOate
(BMSC) migration. We tested the hypothesis that a nitric monotherapy and MCAo control. The number of BMSCs
oxide donor, DETA-NONOate, increases endogenous isch- in the ischemic brain was significantly increased after
emic brain SDF1 and BMSC CXCR4 and MMP9 expres- combination BMSC with DETA-NONOate treatment
sion, which promotes BMSC migration into ischemic brain compared with monotherapy with BMSCs. The number
and thereby enhances functional outcome after stroke. of engrafted BMSCs was significantly correlated with
C57BL/6J mice were subjected to middle cerebral artery functional outcome after stroke. DETA-NONOate signif-
occlusion (MCAo), and 24 hours later, the following were icantly increased BMSC CXCR4 and MMP9 expression
intravenously administered (n ⴝ 9 mice per group): (a) and promoted BMSC adhesion and migration to MBECs
phosphate-buffered saline; (b) BMSCs (5 ⴛ 105); (c) 0.4 and astrocytes compared with nontreatment BMSCs. In-
mg/kg DETA-NONOate; (d) combination of CXCR4-inhibi- hibition of CXCR4 or MMPs in BMSCs significantly
tion BMSCs with DETA-NONOate; and (e) combination of decreased DETA-NONOate-induced BMSC adhesion and
BMSCs with DETA-NONOate. To elucidate the mecha- migration. Our data demonstrate that DETA-NONOate
nisms underlying combination-enhanced BMSC migration, enhanced the therapeutic potency of BMSCs, possibly via
transwell cocultures of BMSC with mouse brain endothe- upregulation of SDF1/CXCR4 and MMP pathways, and
lial cells (MBECs) or astrocytes were performed. Combi- increased BMSC engraftment into the ischemic brain.
nation treatment significantly improved functional out- STEM CELLS 2007;25:2777–2785
Disclosure of potential conflicts of interest is found at the end of this article.

receptor [5] chemokine (CXC motif) receptor 4 (CXCR4) reg-


INTRODUCTION ulate the tracking of various types CXCR4-positive cells [6].
BMSCs can be efficiently transduced to express CXCR4, and
The regenerative potential of bone marrow stromal cells transduced BMSCs migrate rapidly toward SDF1␣ [7]. Intrace-
(BMSCs) has been demonstrated in myocardial, limb, and brain rebral injection of recombinant human SDF1␣ stimulates the
ischemia [1]. Systemically administered BMSCs selectively mi- homing of transplanted BMSCs to the site of injection in the brain
grate into the injury site [2] and dose-dependently improve [8]. Nitric oxide (NO) donors significantly enhance the SDF1␣-
neurological functional recovery after stroke [3]. The effect induced cell migration [9]. Intravenous infusion of endothelial
of BMSC transplantation is dependent on the number of en- nitric oxide synthase (eNOS)⫺/⫺ bone marrow (BM) cells into
grafted BMSCs [3]. In addition, the success of a vascular route wild-type mice decreases BM migration into lesion area [10].
for BMSC treatment has been limited by the low migration Thus, NO-mediated SDF1/CXCR4 likely controls the traffick-
efficiency of the transplanted BMSCs into the lesioned area [4]. ing of transplanted BMSCs. In addition, matrix metalloprotein-
Thus, priming the ischemic brain to promote BMSC migration ases (MMPs) are involved in SDF1/CXCR4-induced chemotaxis
into the ischemic brain may augment BMSC regenerative treat- of human hematopoietic progenitor cells across subendothelial
ment of stroke. basement membranes [11]. SDF1␣ gradient increases the chemo-
Chemokines are important factors controlling cellular mi- taxis of bone marrow and blood CD34(⫹) cells, which is blocked
gration. Stromal cell-derived factor-1 (SDF1) and its unique by inhibitors of MMPs [11].

Correspondence: Michael Chopp, Ph.D., Neurology Research, E&R Building, Room 3056, Henry Ford Hospital, 2799 West Grand
Boulevard, Detroit, Michigan 48202, USA. Telephone: 313-916-3936; Fax: 313-916-1318; e-mail: chopp@neuro.hfh.edu Received March
9, 2007; accepted for publication July 12, 2007; first published online in STEM CELLS EXPRESS July 19, 2007. ©AlphaMed Press
1066-5099/2007/$30.00/0 doi: 10.1634/stemcells.2007-0169

STEM CELLS 2007;25:2777–2785 www.StemCells.com


2778 DETA-NONOate Enhances BMSC Migration After Stroke

DETA-NONOate ([Z]-1-[N-(2-aminoethyl)-N-(2-ammonio- paraffin [14]. A standard paraffin block was obtained from the
ethyl) aminio] diazen-1-ium-1,2-diolate), a NO donor, promotes center of the lesion (bregma, ⫺1 mm to ⫹1 mm). A series of
angiogenesis and improves neurological outcome after stroke 6-␮m-thick sections was cut from the block. Every 10th coronal
[12]. Our previous studies have shown that the combination of section for a total five sections was used for immunohistochemical
a subtherapeutic dose of DETA-NONOate and BMSCs acts staining. Antibody against BrdU (1:100; Roche, Indianapolis, http://
www.roche-applied-science.com) and SDF1 (1:250; Santa Cruz
additively to improve the therapeutic outcome after stroke in Biotechnology Inc., Santa Cruz, CA, http://www.scbt.com) single
rats [13]. However, the mechanism of combination treatment- immunostaining was performed to detect the number of engrafted
induced additive functional improvement after stroke has not BMSCs and the expression of SDF1 in the ischemic brain, respec-
been investigated. In this study, we sought to test the mechanism tively. Control experiments consisted of staining brain coronal
by which combination treatment of BMSC with DETA-NONO- tissue sections as outlined above, but the primary antibodies were
ate amplified restorative therapy after stroke in mice. We hy- omitted, as previously described [16].
pothesized that DETA-NONOate treatment promotes BMSC
migration into the ischemic brain by increasing SDF1 within the Double Immunofluorescence Staining
compromised brain. To identify SDF1-reactive cells colocalized with brain endothelial
cells or astrocytes, double immunofluorescence staining was used.
von Willebrand factor (vWF) is a marker for endothelial cells. Glial
MATERIALS AND METHODS fibrillary acidic protein (GFAP) is a marker for astrocytes. Double
immunofluorescence labeling for SDF1 with vWF and SDF1 with
GFAP was performed. Each coronal section was first treated with
All experiments were conducted in accordance with the standards the primary anti-SDF1 antibody with fluorescein isothiocyanate
and procedures of the American Council on Animal Care and the (FITC), followed by anti-vWF (1:400; DAKO, Carpinteria, CA,
Henry Ford Health System Institutional Animal Care and Use
http://www.dako.com) or anti-GFAP (1:1,000; DAKO) with Cy3
Committee.
staining. Control experiments consisted of staining brain coronal
tissue sections as outlined above, but the primary antibodies were
BMSC Culture and Labeling with omitted, as previously described [16]. SDF1 and GFAP double
5-Bromo-2ⴕ-Deoxyuridine immunofluorescent images were acquired using fluorescent micros-
Mouse BMSCs (M-126, P9; Cognate BioServices, Inc., Baltimore, copy (Axiophot2, HB0100 W/2; Carl Zeiss, New York, http://
MD, http://www.cognatetherapeutics.com) were incubated in HyQ www.zeiss.com) with a digital camera (C4742-95; Hamamatsu,
MEM Alpha Modification (HyClone, Logan, UT, http://www. Hamamatsu City, Japan, http://www.hamamatsu.com). SDF1 and
hyclone.com) with 20% fetal bovine serum (Gibco, Grand Island, vWF double immunostaining confocal images were acquired using
NY, http://www.invitrogen.com) and 1% antibiotins (penicillin- an MRC 1024 (argon and krypton; Bio-Rad, Hercules, CA, http://
streptomycin; Gibco) and maintained at 37°C in 5% CO2/95% www.bio-rad.com) laser-scanning confocal imaging system
ambient mixed air. For injection to animals, BMSCs were labeled mounted onto a Carl Zeiss microscope, as previously described
with 5-bromo-2⬘-deoxyuridine (BrdU) (30 ␮g/ml; Sigma-Aldrich, [17].
St. Louis, http://www.sigmaaldrich.com) in vitro for 3 days [3].
Passage 3–5 BMSCs were used. Quantification
For measurement of the numbers of engrafted BMSCs, the total
Middle Cerebral Artery Occlusion Model and numbers of BrdU-labeled BMSCs in both the ipsilateral and con-
Experimental Groups tralateral hemispheres were counted. For quantification, the percent-
Adult male C57BL/6J mice (22–25 g; Charles River Laboratories, age of the SDF1-positive area was measured in the ischemic border
Wilmington, MA, http://www.criver.com) were subjected to tran- area [18].
sient (2.5 hours) monofilament right middle cerebral artery occlu-
sion (MCAo) [14]. Mice were randomly divided into five groups BMSC Culture
(n ⫽ 9 mice per group) 24 hours after MCAo, and the following was To test whether DETA-NONOate regulates BMSC CXCR4 and
injected via the tail vein: (a) 0.2 ml of phosphate-buffered saline MMP9 expression, BMSCs were cultured and treated with (a)
(PBS) for control; (b) 5 ⫻ 105 BMSCs; (c) DETA-NONOate (0.4 nontreatment control or (b) 0.4 ␮M DETA-NONOate. For real-time
mg/kg in 0.2 ml of PBS; Alexis Biochemical, San Diego, http:// polymerase chain reaction (PCR), one set of the experimental group
www.alexis-corp.com); (d) combination of BMSCs (5 ⫻ 105) pre- was terminated at 3 hours after treatment. For immunostaining,
treated with specific CXCR4 inhibitor AMD3100 [15] (20 ␮M for zymography, and flow cytometric analysis (fluorescence-activated
3 hours) with DETA-NONOate (0.4 mg/kg); or (e) combination of cell sorting [FACS]), three additional sets of experimental groups
BMSCs (5 ⫻ 105) with DETA-NONOate (0.4 mg/kg). In an earlier were terminated at 24 hours after treatment.
study, a single dose of 0.4 mg/kg DETA-NONOate did not improve
functional outcome after stroke [13]. In addition, we have shown
that a dose of 5 ⫻ 105 BMSCs does not provide a significant benefit
Real-Time PCR
on functional outcome after stroke (J. Chen, C.L. Zhang, and M. Brain tissue (ischemic boundary zone) from the MCAo control and
Chopp, unpublished data). In this study, we investigated whether a DETA-NONOate monotherapy groups, and cultured BMSCs were
subtherapeutic dose of DETA-NONOate and a subtherapeutic dose collected, and total RNA was isolated with TRIzol (Invitrogen,
of BMSCs induce additive or superadditive effects. Therefore, we Carlsbad, CA, http://www.invitrogen.com) following a standard
selected subtherapeutic doses of 0.4 mg/kg DETA-NONOate (sin- protocol [19]. Quantitative PCR was performed using the SYBR
gle injection) and 5 ⫻ 105 BMSC. Green real-time PCR method on an ABI 7000 PCR instrument (Ap-
plied Biosystems, Foster City, CA, http://www.appliedbiosystems.
Behavioral Tests com) using three-stage program parameters provided by the manu-
facturer. Each sample was tested in triplicate, and relative gene
A modified neurological severity score (mNSS) [14] and foot-fault expression data were analyzed using the 2⫺⌬⌬CT method. The
tests [14] were performed by a blinded investigator before MCAo
following primers for real-time PCR were designed using Primer
and at 1, 7, and 14 days after MCAo, as previously described [14].
Express software (Applied Biosystems): GAPDH (forward, AGA
ACA TCA TCC CTG CAT CC; reverse, CAC ATT GGG GGT
Histological and Immunohistochemistry Assessment AGG AAC AC), CXCR4 (forward, GGC TGT AGA GCG ATG
Mice were sacrificed at 14 days after MCAo. The brains were fixed TTT GC; reverse, GTA GAG GTT GAC AGT GTA), and MMP9
by transcardial perfusion with saline, followed by perfusion and (forward, AAT CTC TTC TAG AGA CTG GGA AGG AG; re-
immersion in 4% paraformaldehyde, before being embedded in verse, AGC TGA TTG ACT AAA GTA GCT GGA).
Cui, Chen, Zacharek et al. 2779

SDS-Polyacrylamide Gel Electrophoresis BMSCs with MBECs or GFP-BMSCs with astrocytes was per-
Zymography formed. MBECs or astrocytes (5 ⫻ 104 cells per well) were placed
in the lower chamber. GFP-BMSCs were placed in a six-well
Conditioned media were collected and concentrated using a Centri- chamber. GFP-BMSCs, MBECs, and astrocytes were pretreated
con concentrator (Millipore, Bedford, MA, http://www.millipore. with or without DETA-NONOate (0.4 ␮M) for 24 hours, separately.
com). Protein concentrations were analyzed with a Bio-Rad system. In addition, 24-well Transwell polycarbonate inserts with a pore
Equal amounts of protein (20 ␮g/lane) for each sample were mixed size of 8 ␮m (Corning Life Sciences) coated with 50 ␮g/ml fi-
with 2⫻ sample buffer and loaded on a 10% polyacrylamide gel bronectin (Chemicon) and 0.1% gelatin were used. GFP-BMSCs
incorporated with 0.1% gelatin for electrophoresis. MMP9 zymo- (5 ⫻ 104 cells per well) were placed in the upper chamber and
graphic standards were used as positive controls (Chemicon, Te- cocultured with MBECs or astrocytes for 5 hours. GFP-BMSC
mecula, CA, http://www.chemicon.com). After electrophoresis, gels migration to the lower side of the insert was counted using MCID
were washed in 2.5% Triton X-100 for 1 hour, incubated for 18 software [23, 24].
hours at 37°C in collagenase buffer, and stained for 1 hour with To test whether CXCR4 and MMPs regulate BMSC migration,
0.1% Coomassie Brilliant Blue. Gelatinolytic activity was visual- in separate experiments, MBECs or astrocytes (5 ⫻ 104 cells per
ized as a transparent band against a blue background. Zymography well) were placed in the lower chamber and pretreated with or
was measured for quantification analysis by spot density measure- without DETA-NONOate (0.4 ␮M) for 24 hours. GFP-BMSCs
ment using a digital imaging analysis system (Alpha Innotech, were pretreated with or without AMD3100 (20 ␮M) or GM6001 (10
Mount Prospect, IL, http://www.alphainnotech.com) [20, 21]. ␮M) to block CXCR4 or MMP, respectively, for 3 hours. Then,
GFP-BMSCs were placed in the upper chamber and cocultured with
Immunohistochemistry MBECs or astrocytes for 5 hours. GFP-BMSC migration to the
Antibody against CXCR4 (1:400; Chemicon) and MMP9 (1:200; lower side of the insert was then counted.
Santa Cruz Biotechnology) conjugated with Cy3 (1:200; Jackson
Immunoresearch Laboratories, West Grove, PA, http://www. Statistical Analysis
jacksonimmuno.com) immunostaining was performed. Nuclei were Analysis of variance (ANOVA) was performed to compare the
identified by 4⬘,6-diamidino-2-phenylindole dihydrochloride (1:10,000; functional results, SDF1 immunostaining-positive areas, and BrdU-
Molecular Probes Inc., Eugene, OR, http://probes.invitrogen.com). positive numbers in the brain tissues. A two-independent-sample t
The percentage of CXCR4- and MMP9-positive cells was measured test was used to compare CXCR4- or MMP9-reactive cell numbers
using MCID software (Imaging Research, Saint Catharines, ON, and CXCR4, MMP9, and brain tissue SDF1 mRNA levels. For cell
Canada, http://www.imagingresearch.com). adhesion and migration data, ANOVAs followed by pairwise com-
parisons were performed if the overall treatment effect was signif-
Flow Cytometric Analysis of CXCR4 Expression icant at p ⬍ .05. Running regression models (bivariate correlation)
was used to analyze the correlation of neurological functional
Single-cell suspensions (1 ⫻ 106) were incubated, fixed with 4%
outcome with the presence of BMSC number in the ischemic brain.
paraformaldehyde, and rinsed twice with 1⫻ PBS with CXCR4
All data are presented as mean ⫾ SE; p ⬍ .05 is considered
antibody (1:250; AB1846; Chemicon) conjugated with FITC (1:
significant.
200; Jackson Immunoresearch Laboratories). Immunostaining anal-
ysis was performed on a BD FACSCalibur flow cytometry (BD
Biosciences, San Diego, http://www.bdbiosciences.com). At least
104 cells were analyzed in all specimens. The percentage of RESULTS
CXCR4-positive cells was determined using CellQuest software
(Becton Dickinson Canada Inc., Oakville, ON, Canada, http://www. Combination Treatment with DETA-NONOate and
bd.com).
BMSCs Increases the Numbers of BMSCs in the
BMSC Adhesion Ischemic Brain and Improves Neurological Function
To test whether DETA-NONOate increases BMSC adhesion to
After Stroke
brain endothelial cells and astrocytes, and to investigate the signal- To determine whether DETA-NONOate and BMSC combina-
ing pathway of DETA-NONOate upregulation of BMSC adhesion, tion treatment of stroke improves neurological functional recov-
green fluorescent protein (GFP)-BMSCs were cultured with mouse ery, a battery of functional tests were performed. Figure 1A and
brain endothelial cells (MBECs) (American Type Culture Collec- 1B shows that combination DETA-NONOate with BMSCs sig-
tion, Manassas, VA, http://www.atcc.org) and astrocytes (American nificantly improved the functional outcome (Fig. 1A, mNSS;
Type Culture Collection). MBECs or astrocytes (1 ⫻ 104 cells per Fig. 1B: foot-fault test) compared with MCAo control group at
well) were placed in 96-well chambers and pretreated with or
without DETA-NONOate (0.4 ␮M) for 24 hours, separately. Then, 7 and 14 days (p ⬍ .05), respectively, and compared with the
BMSCs were added to MBECs or astrocytes and treated with or BMSC monotherapy group at 14 days in mNSS test (p ⬍ .05)
without AMD3100 (20 ␮M; AnorMED, Groton, CT, http://www. after stroke. There was no significant functional improvement
anormed.com), a specific antagonist of CXCR4, or GM6001 (10 (mNSS and foot-fault tests) in AMD3100-exposed BMSC com-
␮M; Chemicon), a selective inhibitor of MMPs (MMP1, 2, 3, 8, and bination with DETA-NONOate treatment group compared with
9), for an additional 3 hours in serum-starved Dulbecco’s modified MCAo control.
Eagle’s medium (n ⫽ 6 wells per group) [22]. The numbers of To test whether DETA-NONOate increases BMSC delivery
adherent GFP-BMSCs to MBECs or astrocytes were counted using into the ischemic brain, BMSCs identified by BrdU immuno-
fluorescence microscopy (BH-2; Olympus, Tokyo, http://www. staining were measured. BrdU-positive cell number was counted
olympus-global.com).
in the ipsilateral and contralateral hemispheres, respectively.
Figure 1 shows that BrdU-positive BMSC number significantly
BMSC Migration increased in the combination treatment group (Fig. 1E, 1F)
To test whether SDF1 and DETA-NONOate regulates BMSC compared with the BMSC monotherapy group in both ipsilateral
migration, GFP-BMSCs were placed in the upper chamber of a and contralateral hemispheres (Fig. 1C, 1F, p ⬍ .05). However,
Transwell system (Corning Life Sciences, Acton, MA, http://www.
the number of BrdU-positive BMSCs did not show significantly
corning.com/lifesciences) and treated with or without SDF1␣ (200
ng/ml; Sigma-Aldrich) or DETA-NONOate (0.4 ␮M) for 24 hours. increase in the AMD3100-exposed BMSC combination with
The number of BMSCs migrating to the lower chamber was DETA-NONOate treatment group compared with the BMSC
counted. monotherapy group (Fig. 1D, 1F). Using the running regression
To further test whether DETA-NONOate promotes BMSC mi- models to analyze the correlation of neurological functional
gration toward MBECs or astrocytes, transwell coculture of GFP- outcome with the presence of BMSC number in the ischemic
www.StemCells.com
2780 DETA-NONOate Enhances BMSC Migration After Stroke

Figure 1. Combination DETA-NONOate with BMSC treatment of stroke increases BMSC engraftment into the ischemic brain and improves
functional recovery in mice after stroke. (A): mNSS test. (B): Foot-fault test. (C–E): H&E staining showing immunostaining for BrdU-positive
BMSCs in BMSC-alone treatment ([C], arrow), combination treatment with chemokine (CXC motif) receptor 4-inhibition BMSCs ([D], arrow), and
combination treatment with normal BMSCs ([E], arrow). (F): Quantitative data of BrdU-immunoreactive BMSCs. Scale bar ⫽ 50 ␮m. n ⫽ 9 mice
per group. Abbreviations: BMSC, bone marrow stromal cell; BrdU, 5-bromo-2⬘-deoxyuridine; MCAo, middle cerebral artery occlusion; mNSS,
modified neurological severity score.

brain, the data show that the number of BrdU-positive cells in Double immunofluorescent staining showed that SDF1 pri-
the ischemic brain was significantly (p ⬍ .05) and negatively marily colocalized with GFAP-stained astrocytes (Fig. 2G–
correlated with mNSS (r ⫽ ⫺0.721) and foot-fault (r ⫽ 2I). Confocal images showed that SDF1 partially colocalized
⫺0.713) values. These data suggest that increased BMSCs with vWF-stained endothelial cells in the ischemic boundary
present in the ischemic brain correlate with functional recovery zone (Fig. 2K–2M).
after stroke.

Combination Treatment with DETA-NONOate and DETA-NONOate Increases BMSC CXCR4 and
BMSCs Promotes SDF1 Expression in the Ischemic MMP9 Gene and Protein Expression In Vitro
Boundary Zone To investigate the effect of DETA-NONOate on CXCR4 and
MMP9 expression in BMSCs in vitro, CXCR4 and MMP9
The SDF1/CXCR4 axis regulates BMSC recruitment to le-
gene expression and protein expression were measured in
sioned area [25]. To test whether DETA-NONOate enhance-
ment of BMSC recruitment into the ischemic brain occurs via cultured BMSCs. Figure 3 shows that DETA-NONOate treat-
upregulation of the SDF1/CXCR4 pathway, SDF1 mRNA ment (0.4 ␮M for 24 hours) significantly increased the num-
and protein expression were measured using real-time PCR ber of CXCR4 (Fig. 3A, 3B, 3E) and MMP9 (Fig. 3C, 3D,
and tissue immunostaining, respectively. Figure 2 shows that 3H) immunoreactive cells compared with the nontreatment
SDF1 mRNA level significantly increased in the ischemic BMSC control group (p ⬍ .05; n ⫽ 3 mice per group),
boundary zone in DETA-NONOate treatment group com- respectively. Figure 3F and 3I shows that the CXCR4 and
pared with MCAo control group at 14 days after stroke (p ⬍ MMP9 gene expression in the DETA-NONOate treatment
.05; Fig. 2J). Immunostaining showed that SDF1 expression group was significantly increased compared with the non-
significantly increased in BMSC and DETA-NONOate treatment BMSC control group (p ⬍ .05; n ⫽ 3 mice per
monotherapy groups compared with the MCAo control group group). FACS analysis shows (Fig. 3G) that the percentage of
(p ⬍ .05). SDF1 expression in the combination DETA- the CXCR4 protein expression in the DETA-NONOate treat-
NONOate with BMSC treatment group was significantly ment group (34.99%) was significantly increased compared
increased compared with MCAo control and DETA-NONO- with the nontreatment BMSC control group (17.29%). SDS-
ate monotherapy treatment groups, respectively (p ⬍ .05). polyacrylamide gel electrophoresis zymography analysis
However, SDF1 was not significantly increased in the com- shows that the density of the secreted MMP9 was signifi-
bination DETA-NONOate with CXCR4-inhibition BMSC cantly increased in DETA-NONOate treatment BMSCs com-
treatment group compared with MCAo control (Fig. 2A–2F). pared with nontreatment control (Fig. 3J).
Cui, Chen, Zacharek et al. 2781

Figure 2. SDF1 expression in the ischemic brain. (A–E): SDF1 immunohistochemical expression in the ischemic boundary zone at 14 days after
MCAo. (A): MCAo control; (B): BMSC treatment; (C): DETA-NONOate-alone treatment; (D): combination treatment with chemokine (CXC motif)
receptor 4-inhibition BMSCs; (E): combination treatment with normal BMSCs. (F): Quantitative SDF1 data. (G–I): Images of double immunoflu-
orescent staining of SDF1 (G) with GFAP (H). (J): SDF1 gene expression measured by real-time PCR. (K, L): Confocal images of double
immunofluorescent staining of SDF1 ([K1–K4]: 0.2 ␮m/layer; [K]: merged from [K1–K4]) with vWF ([L1–L4]: 0.2 ␮m/layer; [L]: merged from
[L1–L4]). (M): Merged image from (K, L). Scale bars ⫽ 100 ␮m (B), 50 ␮m (I). n ⫽ 9 mice per group. Abbreviations: BMSC, bone marrow stromal
cell; GFAP, glial fibrillary acidic protein; MCAo, middle cerebral artery occlusion; SDF1, stromal cell-derived factor-1; vWF, von Willebrand factor.

The SDF1/CXCR4 Axis Mediates DETA-NONOate- cell adhesion to brain endothelial cell and transendothelial mi-
Enhanced BMSC Adhesion and Migration gration. To investigate whether DETA-NONOate increases
Delivery of BMSCs into ischemic brain is a multistep process. BMSC adhesion to brain endothelial cells or astrocytes in vitro,
The first essential step is the homing cascade, which determines measurements of BMSC adhesion to MBECs and astrocytes
www.StemCells.com
2782 DETA-NONOate Enhances BMSC Migration After Stroke

Figure 3. DETA-NONOate treatment increases BMSC CXCR4 and MMP9 gene and protein expression in vitro. (A, B): CXCR4 immunostaining
in nontreatment Con (A) and DETA-NONOate treatment BMSCs (B). (C, D): MMP9 immunostaining in nontreatment Con (C) and DETA-NONOate
treatment BMSCs (D). (E, H): Quantitative data of CXCR4-positive (E) and MMP9-positive (H) cell expression. (F, I): CXCR4 (F) and MMP9 (I)
gene expression measured by real-time PCR. (G): CXCR4-positive cell expression measured by FACS. (J): MMP9 secretion measured by
zymography (n ⫽ 3 times per group). Scale bar ⫽ 50 ␮m. Abbreviations: Con, control; CXCR4, chemokine (CXC motif) receptor 4; FACS,
fluorescence-activated cell sorting; MMP9, matrix metalloproteinase 9.

were performed. Figure 4A and 4B shows that DETA-NONOate SDF1/CXCR4 axis, along with MMP9, regulates BMSC migra-
significantly increased BMSC adhesion to MBECs and astro- tion.
cytes compared with nontreatment control (p ⬍ .05; n ⫽ 6 wells To further test whether DETA-NONOate promotes BMSC
per group). Blocking CXCR4 with AMD3100 (20 ␮M) and migration to brain endothelial cells and astrocytes, coculture of
MMPs with GM6001 (10 ␮M) in BMSCs significantly attenu- BMSCs with MBECs or coculture of BMSCs with astrocytes
ated DETA-NONOate-induced BMSC adhesion to MBECs or was performed. MBECs or astrocytes were placed in the lower
astrocytes, respectively, compared with the DETA-NONOate chamber of the well and pretreated with or without DETA-
treatment group (p ⬍ .05; n ⫽ 6 wells per group). NONOate. BMSCs were placed in the upper chamber. BMSC
To investigate the effect of DETA-NONOate on BMSC migration to the lower chamber was measured. Figure 4D shows
migration and the mechanism of DETA-NONOate regulation of that DETA-NONOate significantly increased BMSC migration
BMSC migration, a transwell culture system was used. DETA- toward MBECs compared with nontreatment MBECs alone
NONOate or SDF1␣ was added to the lower chamber. BMSCs (p ⬍ .05; n ⫽ 4 wells per group). Blocking CXCR4 and MMP
were placed in the upper chamber of the insert. The migration of in BMSCs significantly decreased DETA-NONOate-induced
BMSCs into the lower chamber of the insert was measured. BMSC migration toward MBECs (p ⬍ .05; n ⫽ 4 wells per
Figure 4C shows that DETA-NONOate and SDF1 significantly group). In addition, DETA-NONOate significantly increased
increased BMSC migration numbers compared with nontreat- BMSC migration toward astrocytes compared with nontreat-
ment control BMSCs (p ⬍ .05; n ⫽ 4 wells per group). Blocking ment astrocytes alone (p ⬍ .05; n ⫽ 4 wells per group).
CXCR4 (ADM3100) or MMP (GM6001) in BMSCs signifi- Blocking CXCR4 or MMP in BMSCs significantly attenuated
cantly attenuated DETA-NONOate-induced BMSC migration DETA-NONOate-induced BMSC migration toward astrocytes
(p ⬍ .05; n ⫽ 4 mice per group). These data indicated that the (p ⬍ .05; n ⫽ 4 wells per group; Fig. 4E). These data suggest
Cui, Chen, Zacharek et al. 2783

Figure 4. Stromal cell-derived factor-1/chemokine (CXC motif) receptor 4 axis and matrix metalloproteinase mediate DETA-NONOate-induced
BMSC adhesion and migration in vitro. (A, B): BMSC adhesion assay. Coculture BMSCs with MBECs and astrocytes. BMSC adhesion to MBECs
(A) and astrocytes (B) (n ⫽ 6 wells per group). (C–E): BMSC migration assay. Culture BMSCs alone or coculture BMSCs with MBECs and
astrocytes were used. Shown are BMSC migration alone (C), BMSC migration to MBECs (D), and BMSC migration to astrocytes (E) (n ⫽ 4 wells
per group). Abbreviations: BMSC, bone marrow stromal cell; MBEC, mouse brain endothelial cell; SDF1, stromal cell-derived factor-1.

that stimulation of MBECs or astrocytes with DETA-NONOate Chemokines are important factors controlling cellular mi-
enhances BMSC migration. The SDF1/CXCR4 axis and MMP gration. SDF1 has CXCR4 as its only receptor [5]. SDF1 is a
regulate DETA-NONOate-induced BMSC migration. master regulator of tracking of various types of CXCR4-positive
cells [6]. Overexpression CXCR4 increases BMSC migration
toward SDF1 but does not render a survival advantage to MSCs
under serum-deprived conditions [7]. Inhibition of CXCR4 ex-
DISCUSSION pression in BMSCs decreases BMSC migration. Thus, the in-
teraction of SDF1␣/CXCR4 likely controls the trafficking of
In this study, we demonstrated that DETA-NONOate treatment transplanted BMSCs. Our data show that DETA-NONOate
of stroke enhances BMSC migration into the ischemic brain, treatment of stroke increases endogenous brain astrocyte and
which correlates with functional outcome after stroke in mice. endothelial cell SDF1 expression in the ischemic border. Com-
DETA-NONOate increases SDF1 expression in the ischemic bination treatment significantly increases SDF1 expression
brain. In addition, DETA-NONOate upregulates CXCR4 and compared with DETA-NONOate monotherapy and increases
MMP9 expression in cultured BMSCs. DETA-NONOate pro- the number of engrafted BMSCs compared with BMSC mono-
motes BMSC adhesion to MBECs and increases BMSC migra- therapy group; however, inhibition of CXCR4 in BMSCs does
tion. The SDF1/CXCR4 axis, along with MMPs, mediates not significantly increase SDF1 expression and the number of
DETA-NONOate-increased BMSC migration. engrafted BMSCs. In addition, DETA-NONOate also increases
NO is a reactive molecule with numerous physiological and BMSC CXCR4 gene and protein expression in vitro. Inhibition
pathophysiological roles affecting the nervous and cardiovascu- of CXCR4 expression in BMSCs significantly decreases BMSC
lar systems. NO produced by eNOS has a crucial role in the migration and attenuates DETA-NONOate-induced BMSC mi-
regulation of systemic blood pressure, vascular tone, vascular gration. Therefore, DETA-NONOate augmentation of the
SDF1/CXCR4 axis may facilitate BMSC migration into the
remodeling, and angiogenesis [26]. DETA-NONOate in combi-
ischemic brain.
nation with BMSCs promotes functional outcome after stroke in
Circulating stem cells release factors (e.g., MMPs) that
rats [13], which is consistent with our data that treatment of enable them to cross the endothelial barrier and home to the
stroke in adult mice with DETA-NONOate in combination with organ [6]. The CXCR4/SDF1␣ system mediates active MMP9
BMSCs significantly improves functional outcome compared and MMP2 secretion from Hca-F and hepatocarcinoma cell
with control MCAo. In this study, we demonstrate for the first lines; this mediation facilitates lymphogenous metastasis [27].
time that DETA-NONOate in combination with BMSC treat- SDF1 recruits osteoclast precursors partly mediated through
ment of stroke enhances BMSC migration into the ischemic increases in MMP9 [28]. SDF1/CXCR4, along with MMPs,
brain. In addition, the increased BMSC migration into the isch- recruits BMSCs to damaged tissues [29]. DETA-NONOate in-
emic brain correlated with functional outcome after stroke. creases BMSC CXCR4 and MMP9 expression. Inhibition of
Therefore, increased BMSC delivery to the ischemic brain im- CXCR4 and MMPs in BMSCs significantly decreases DETA-
proves functional outcome after stroke. NONOate-induced BMSC adhesion and migration. Therefore,
www.StemCells.com
2784 DETA-NONOate Enhances BMSC Migration After Stroke

DETA-NONOate induced BMSC adhesion and migration may lating progenitor endothelial cells by exogenous cell and phar-
be mediated by the SDF1/CXCR4 and MMP pathways. We note macological monotherapies may also enhance functional recov-
that the MMP inhibitor GM6001, used in this study, was not ery from stroke [2, 3, 13, 37– 41]. Whether these endogenous
specific for MMP9. Therefore, we cannot exclude the possibility progenitor cells are enhanced by combination treatment is un-
that MMPs other than MMP9 contribute to BMSC adhesion and known.
migration. In addition, MMPs may have both beneficial and
detrimental effects on developing and adult brain and spinal
cord. Detrimental effects include dysfunction of the blood-brain SUMMARY
barrier, demyelination, neuroinflammation, and neurotoxicity
[30, 31]. Beneficial effects include roles in the development and DETA-NONOate treatment of stroke enhances BMSC migra-
neurogenesis of the central nervous system, growth and regen- tion into the ischemic brain by increasing SDF1/CXCR4 and
eration of axons, myelogenesis, angiogenesis, and termination MMP. Priming of the ischemic brain with DETA-NONOate
of neuroinflammation [21, 32, 33]. The double-edged effects of enhances the efficacy of BMSC therapy.
MMP9 depend on the temporal and spatial distribution of
MMP9 within the cells of the neurovascular unit [31]. MMP9
contributes to damage early after a stroke [34]. In addition, ACKNOWLEDGMENTS
MMP9 takes on a reparative role days after stroke [32].
On the basis of our data, the following additional studies are
warranted. Testing of a ceiling effect to identify the limit of We thank Qinge Lu for technical assistance. This work was
therapeutic efficacy of using higher doses of BMSCs and supported by National Institute of Neurological Disorders and
DETA-NONOate would be informative and potentially clini- Stroke (NINDS) Grant RO1-NS047682, American Heart Asso-
cally relevant. In addition, our studies in no way exclude the ciation Grant 0750048Z, and NINDS Grant PO1-NS23393.
possibility that other chemokines and growth factors (e.g., he-
patocyte growth factor, platelet-derived growth factor, and an- DISCLOSURE OF POTENTIAL CONFLICTS
giogenic factors [e.g., vascular endothelial growth factor
(VEGF)/VEGF receptor 2 and Angiopoietin 1/Tie2]) also con- OF INTEREST
tribute to BMSC migration into the ischemic brain and are likely
increased by DETA-NONOate [35, 36]. Upregulation of circu- The authors indicate no potential conflicts of interest.

14 Chen J, Zhang C, Jiang H et al. Atorvastatin induction of VEGF and


REFERENCES BDNF promotes brain plasticity after stroke in mice. J Cereb Blood Flow
Metab 2005;25:281–290.
15 Watanabe M, Matsuyama W, Shirahama Y et al. Dual effect of
1 Dezawa M, Hoshino M, Nabeshima Y et al. Marrow stromal cells: AMD3100, a CXCR4 antagonist, on bleomycin-induced lung inflamma-
Implications in health and disease in the nervous system. Curr Mol Med tion. J Immunol 2007;178:5888 –5898.
2005;5:723–732. 16 Li Y, Jiang N, Powers C et al. Neuronal damage and plasticity identified
2 Li Y, Chen J, Chen XG et al. Human marrow stromal cell therapy for by microtubule-associated protein 2, growth-associated protein 43, and
stroke in rat: Neurotrophins and functional recovery. Neurology 2002;
cyclin D1 immunoreactivity after focal cerebral ischemia in rats. Stroke
59:514 –523.
1998;29:1972–1980; discussion 1980 –1981.
3 Chen J, Li Y, Wang L et al. Therapeutic benefit of intravenous admin-
istration of bone marrow stromal cells after cerebral ischemia in rats. 17 Chen J, Li Y, Katakowski M et al. Intravenous bone marrow stromal cell
Stroke 2001;32:1005–1011. therapy reduces apoptosis and promotes endogenous cell proliferation
4 Muller-Ehmsen J, Krausgrill B, Burst V et al. Effective engraftment but after stroke in female rat. J Neurosci Res 2003;73:778 –786.
poor mid-term persistence of mononuclear and mesenchymal bone mar- 18 Shen LH, Li Y, Chen J et al. Therapeutic benefit of bone marrow stromal
row cells in acute and chronic rat myocardial infarction. J Mol Cell cells administered 1 month after stroke. J Cereb Blood Flow Metab
Cardiol 2006;41:876 – 884. 2007;27:6 –13.
5 Bagri A, Gurney T, He X et al. The chemokine SDF1 regulates migration 19 Livak KJ, Schmittgen TD. Analysis of relative gene expression data
of dentate granule cells. Development 2002;129:4249 – 4260. using real-time quantitative PCR and the 2(-delta delta C(T)) method.
6 Kucia M, Reca R, Miekus K et al. Trafficking of normal stem cells and Methods 2001;25:402– 408.
metastasis of cancer stem cells involve similar mechanisms: Pivotal role 20 Romanic AM, White RF, Arleth AJ et al. Matrix metalloproteinase
of the SDF-1-CXCR4 axis. STEM CELLS 2005;23:879 – 894. expression increases after cerebral focal ischemia in rats: Inhibition of
7 Bhakta S, Hong P, Koc O. The surface adhesion molecule CXCR4 matrix metalloproteinase-9 reduces infarct size. Stroke 1998;29:
stimulates mesenchymal stem cell migration to stromal cell-derived 1020 –1030.
factor-1 in vitro but does not decrease apoptosis under serum deprivation. 21 Wang L, Zhang ZG, Zhang RL et al. Matrix metalloproteinase 2 (MMP2)
Cardiovasc Revasc Med 2006;7:19 –24. and MMP9 secreted by erythropoietin-activated endothelial cells pro-
8 Ji JF, He BP, Dheen ST et al. Interactions of chemokines and chemokine mote neural progenitor cell migration. J Neurosci 2006;26:5996 – 6003.
receptors mediate the migration of mesenchymal stem cells to the im- 22 Fernandez-Patron C, Zouki C, Whittal R et al. Matrix metalloproteinases
paired site in the brain after hypoglossal nerve injury. STEM CELLS regulate neutrophil-endothelial cell adhesion through generation of en-
2004;22:415– 427. dothelin-1 1 . FASEB J 2001;15:2230 –2240.
32
9 Cherla RP, Ganju RK. Stromal cell-derived factor 1 alpha-induced che- 23 Wang L, Li Y, Chen X et al. MCP-1, MIP-1, IL-8 and ischemic cerebral
motaxis in T cells is mediated by nitric oxide signaling pathways. tissue enhance human bone marrow stromal cell migration in interface
J Immunol 2001;166:3067–3074. culture. Hematology 2002;7:113–117.
10 Aicher A, Heeschen C, Mildner-Rihm C et al. Essential role of endo-
24 Annabi B, Thibeault S, Lee YT et al. Matrix metalloproteinase regulation
thelial nitric oxide synthase for mobilization of stem and progenitor cells.
of sphingosine-1-phosphate-induced angiogenic properties of bone mar-
Nat Med 2003;9:1370 –1376.
11 Janowska-Wieczorek A, Marquez LA, Dobrowsky A et al. Differen- row stromal cells. Exp Hematol 2003;31:640 – 649.
tial MMP and TIMP production by human marrow and peripheral 25 Zhou B, Han ZC, Poon MC et al. Mesenchymal stem/stromal cells
blood CD34(⫹) cells in response to chemokines. Exp Hematol 2000; (MSC) transfected with stromal derived factor 1 (SDF-1) for therapeutic
28:1274 –1285. neovascularization: enhancement of cell recruitment and entrapment.
12 Zhang R, Wang L, Zhang L et al. Nitric oxide enhances angiogenesis via Med Hypotheses 2007;68:1268 –1271.
the synthesis of vascular endothelial growth factor and cGMP after stroke 26 Chen J, Zacharek A, Zhang C et al. Endothelial nitric oxide synthase
in the rat. Circ Res 2003;92:308 –313. regulates brain-derived neurotrophic factor expression and neurogenesis
13 Chen J, Li Y, Zhang R et al. Combination therapy of stroke in rats with after stroke in mice. J Neurosci 2005;25:2366 –2375.
a nitric oxide donor and human bone marrow stromal cells enhances 27 Chu H, Zhou H, Liu Y et al. Functional expression of CXC chemokine
angiogenesis and neurogenesis. Brain Res 2004;1005:21–28. recepter-4 mediates the secretion of matrix metalloproteinases from
Cui, Chen, Zacharek et al. 2785

mouse hepatocarcinoma cell lines with different lymphatic metastasis 35 Zacharek A, Chen J, Zhang C et al. Nitric oxide regulates Angiopoietin1/
ability. Int J Biochem Cell Biol 2007;39:197–205. Tie2 expression after stroke. Neurosci Lett 2006;404:28 –32.
28 Yu X, Huang Y, Collin-Osdoby P et al. Stromal cell-derived factor-1 36 Zacharek A, Chen J, Cui X et al. Angiopoietin1/Tie2 and VEGF/Flk1
(SDF-1) recruits osteoclast precursors by inducing chemotaxis, matrix induced by MSC treatment amplifies angiogenesis and vascular stabili-
metalloproteinase-9 (MMP-9) activity, and collagen transmigration. zation after stroke. J Cereb Blood Flow Metab 2007 [Epub ahead of
J Bone Miner Res 2003;18:1404 –1418. print].
29 Son BR, Marquez-Curtis LA, Kucia M et al. Migration of bone marrow 37 Zhang ZG, Zhang L, Jiang Q et al. VEGF enhances angiogenesis and
and cord blood mesenchymal stem cells in vitro is regulated by stromal- promotes blood-brain barrier leakage in the ischemic brain. J Clin Invest
derived factor-1-CXCR4 and hepatocyte growth factor-c-met axes and 2000;106:829 – 838.
involves matrix metalloproteinases. STEM CELLS 2006;24:1254 –1264. 38 Zhang ZG, Zhang L, Tsang W et al. Correlation of VEGF and angio-
30 Asahi M, Wang X, Mori T et al. Effects of matrix metalloproteinase-9 poietin expression with disruption of blood-brain barrier and angiogen-
gene knock-out on the proteolysis of blood-brain barrier and white matter esis after focal cerebral ischemia. J Cereb Blood Flow Metab 2002;22:
components after cerebral ischemia. J Neurosci 2001;21:7724 –7732. 379 –392.
31 Zlokovic BV. Remodeling after stroke. Nat Med 2006;12:390 –391. 39 Chen J, Zhang ZG, Li Y et al. Intravenous administration of human bone
32 Zhao BQ, Wang S, Kim HY et al. Role of matrix metalloproteinases in marrow stromal cells induces angiogenesis in the ischemic boundary
delayed cortical responses after stroke. Nat Med 2006;12:441– 445. zone after stroke in rats. Circ Res 2003;92:692– 699.
33 Sun CY, Hu Y, Wang HF et al. Brain-derived neurotrophic factor 40 Zhang Z, Chopp M. Vascular endothelial growth factor and angiopoietins
inducing angiogenesis through modulation of matrix-degrading pro- in focal cerebral ischemia. Trends Cardiovasc Med 2002;12:62– 66.
teases. Chin Med J 2006;119:589 –595. 41 Chen J, Zacharek A, Li A et al. Vascular endothelial growth factor
34 Kim YS, Lee KY, Koh SH et al. The role of matrix metalloproteinase 9 mediates atorvastatin-induced mammalian achaete-scute homologue-1
in early neurological worsening of acute lacunar infarction. Eur Neurol gene expression and neuronal differentiation after stroke in retired
2006;55:11–15. breeder rats. Neuroscience 2006;141:737–744.

www.StemCells.com

You might also like