PINAL
cord injury is devastating for the individual
and costly to society because it requires substantial
long-term healthcare expenditures. Within minutes
of primary injury, a cascade of biochemical events is
ini-tiated, leading to secondary cell death that evolves over a
period of days to weeks. Various molecular pathways
may be involved including hypoxia, ischemia,
intracellu-lar and extracelluintracellu-lar ionic shifts, lipid peroxidation, free
radical production, excitotoxicity, eicosanoid production,
neutral protease activation, prostaglandin production, and
programmed cell death or apoptosis. The result is that,
days or weeks after SCI, some of the neuronal and glial
supporter cells die, even though they survived the initial
injury.
11Currently, high-dose methylprednisolone
admin-istered within 8 hours of injury is the only therapy with
recognized benefit,
3which, unfortunately, is relatively
mi-nor. We have demonstrated that p35-p25-Cdk5 activation,
tau hyperphosphorylation, and apoptosis could be the
rea-sons for neural damage after spinal cord hemisection.
10Any new treatment of SCI that allows for major recovery
of both functional and molecular levels would be a
signif-icant advance.
Insulin-like growth factor–I has been shown to be a
po-tent neurotrophic factor that promotes the growth of
pro-jection neurons, dendritic arborization, and
synaptogene-sis.
4,16Insulin-like growth factor–I acts in an autocrine and
paracrine manner to promote glucose utilization, using
phosphatidylinositol 3–kinase/Akt and the downstream
GSK-3b pathways.
5, 9Its neuroprotective roles may be
co-ordinated by activation of Akt, inhibition of GSK-3b, and
thus inhibition of tau phosphorylation.
5,9The role and
mechanism of IGF-I gene transfer after SCI are, however,
still unknown.
J Neurosurg Spine 6:35–46, 2007
Gene transfer of insulin-like growth factor–I providing
neuroprotection after spinal cord injury in rats
K
UO-S
HENGH
UNG, M.D., P
H.D.,
1S
HIN-H
ANT
SAI, M.D., P
H.D.,
1T
AO-C
HENL
EE, M.D.,
2J
IA-W
EIL
IN, M.D.,
1C
HENG-K
UEIC
HANG, M.D., P
H.D.,
3ANDW
EN-T
AC
HIU, M.D., P
H.D.
11
Department of Neurosurgery, Graduate Institute of Injury Prevention and Control, Taipei Medical
University, Wan Fang Medical Center, Taipei;
2Department of Neurosurgery, Chang Gung University
College of Medicine, Chang Gung Medical Center, Kaohsiung County;
3Department of Neurosurgery,
Mackay Memorial Hospital and Graduate Institute of Injury Prevention and Control, Taipei Medical
University, Taipei, Taiwan
Object. Insulin-like growth factor–I (IGF-I) has been shown to be a potent neurotrophic factor that promotes the growth of projection neurons, dendritic arborization, and synaptogenesis. Its neuroprotective roles may be coordi-nated by activation of Akt, inhibition of glycogen synthase kinase–3b (GSK-3b), and thus inhibition of tau phos-phorylation. The authors investigated the role and mechanism of IGF-I gene transfer after spinal cord injury (SCI).
Methods. Studies were performed in 40 male Sprague–Dawley rats after spinal cord hemisection. The authors conducted hydrodynamics-based gene transfection in which an IGF-I plasmid was rapidly injected into the rat’s tail vein 30 minutes after SCI. The animals were randomly divided into four groups: Group I, sham operated; Group II, SCI treated with pCMV–IGF-I gene; Group III, SCI treated with vehicle pCMV–LacZ gene; and Group IV, SCI only. The results showed that IGF-I gene transfer promoted motor recovery, antiinflammatory responses, and anti-apoptotic effects after SCI. Using techniques of Western blotting and immunohistochemistry, the authors assessed the mechanism of IGF-I gene transfer after SCI in terms of activation of Akt, inhibition of GSK-3b, attenuation of p35, and inhibition of tau phosphorylation. Moreover, they found that IGF-I gene transfer could block caspase-9 cleavage, increase Bcl-2 formation, and thus inhibit apoptosis after SCI.
Conclusions. The intravenous administration of IGF-I after SCI activated Akt, attenuated GSK-3b, inhibited p35 activation, diminished tau hyperphosphorylation, ended microglia and astrocyte activation, inhibited neuron loss, and significantly improved neurological dysfunction. Furthermore, IGF-I attenuated caspase-9 cleavage, increased Bcl2, and thus inhibited apoptosis after SCI.
K
EYW
ORDS• insulin-like growth factor–I • spinal cord injury • gene transfer •
apoptosis
S
Abbreviations used in this paper: BBB = Basso-Beattie-Bres-nahan; Cdk5 = cyclin-dependent kinase 5; ELISA = enzyme-linked immunosorbent assay; GFAP = glial fibrillary acidic protein; GSK-3b = glycogen synthase kinase–GSK-3b; IGF-I = insulin-like growth fac-tor–I; PBS = phosphate-buffered saline; SCI = spinal cord injury; SEM = standard error of the mean; TUNEL = terminal deoxynu-cleotidyl transferase–mediated deoxyuridine triphosphate nick-end labeling.
We have developed a strategy in which a plasmid
vec-tor containing the genes of interest is injected directly into
a vein.
21,22This approach represents a promising new
strat-egy to deliver and express foreign genes in vivo and has
obvious advantages including the easy preparation of a
large amount of plasmids, continuous expression of
tar-geted protein, and proven safety in vivo. In the present
study, we describe a hydrodynamics-based transfection
procedure utilizing intravenous administration of naked
IGF-I plasmid that results in significant high levels of
ex-ogenous IGF-I protein expression in the spinal cord.
To delineate the nature and mechanism of the IGF-I
gene therapy in SCI, we evaluated the effects of
intrave-nous IGF-I plasmid infusion in rats with spinal cord
hemi-section. The specific aims were to answer the following
questions. 1) Can intravenous administration of naked
IGF-I plasmid transfect the spinal cord after cord
hemi-section? 2) Do alterations in motor function occur after
IGF-I gene transfer? 3) What is the effect of the changes
in neuron survival by IGF-I gene therapy on cord
hemi-section? 4) Can IGF-I gene transfer inhibit glial scarring
and microglia activation after SCI? 5) Does activation of
Akt, inhibition of GSK-3b, attenuation of p35, and
inhibi-tion of tau phosphorylainhibi-tion occur after IGF-I gene
trans-fer? 6) Does cleavage of caspase-9, Bcl2, and apoptosis
change in an IGF-I plasmid–treated group compared with
those treated with control vectors? The overall objective
of this study was to determine the neuroprotective,
anti-inflammatory, and antiapoptotic effects of IGF-I gene
transfer in SCI.
Materials and Methods
Animal CareMale Sprague–Dawley rats (Academia Sinica), weighing 280 to 330 g, were kept two per cage for at least 5 days after their arrival at our laboratory. The rats had access to food and water ad libitum and were housed within a room with a 12:12 hour dark-light cycle. This study was performed in accordance with the guidelines pro-vided by the Experimental Animal Laboratory and approved by the Animal Care and Use Committee at Wan Fang Medical Center. Induction of Spinal Cord Hemisection Injury in Rats
For hemisection, the rats received isoflurane inhalational anes-thesia and were placed in a spinal cord unit of a stereotaxic appara-tus (David Kopf Instruments). Using an adjustable wire knife, we made a lesion on the left side of the rats’ spinal cords (30 rats) or
ensure that a motor deficit of the ipsilateral limb had occurred. The animals exhibiting loss of locomotion in both hindlimbs were ex-cluded from the study. The sham operation consisted of laminecto-my without spinal cord hemisection.
Animal Grouping, Plasmid Injection, and IGF-I Expression The full-length of human IGF-1 cDNA was subcloned into pCMV–MCS vector (Stratagene) to create the human IGF-I expres-sion plasmid (pCMV–IGF-I). The pCMV–LacZ plasmid was used as a vehicle control. The plasmids were purified using the EndoFree Plasmid Giga Kit (Qiagen). Forty male Sprague–Dawley rats were randomly divided into four groups: Group I, treated with sham sur-gery; Group II, underwent SCI and treated with pCMV–IGF-I gene; Group III, underwent SCI and treated with vehicle pCMV–LacZ gene; and Group IV, underwent SCI only. Plasmids of pCMV–IGF-I and pCMV–LacZ were, respectively, injected into pCMV–IGF-IGF-pCMV–IGF-I–treated and vehicle-treated groups. The plasmid DNA was administered to rats using a hydrodynamics-based gene transfer technique that in-volved rapid injection of a large volume of DNA solution through the tail vein 30 minutes after cord hemisection.21,22Briefly, a certain
amount of plasmid DNA (3 mg/g) was diluted in 15 ml of saline and injected via the tail vein into the circulation within 15 seconds.
For analysis of IGF-I peptide expression, we obtained blood sam-ples from cardiac puncture site after induction of isoflurane anesthe-sia 1, 7, and 14 days after the gene transfer procedure. The amount of plasma IGF-I was measured using an ELISA method, with min-imum detectable concentration of 0.15 ng/ml, according to the user’s manual (R&D Systems).
Histological Features, Immunohistochemical Findings, and Cell Count
On the 14th day after hemisection-induced SCI, several rats (five in each group) were deeply anesthetized with isoflurane and re-ceived a left-ventricle perfusion of PBS, followed by cold 4% para-formaldehyde in 0.15 M sodium phosphate buffer, pH 7.4. The spinal cord was removed immediately, postfixed for 8 hours in the same fixative at 4˚C, and cryoprotected for 2 to 3 days in 15 and 30% sucrose. The spinal cord was frozen in powdered dry ice and stored at 2280˚C until needed. Five-micrometer sections of the spinal cord hemisection were cut with a freezing and sliding micro-tome at the center. The sections were prepared for either immuno-staining or apoptosis immuno-staining. For immunohistochemistry, sections were washed in PBS and incubated in 3% normal goat serum with 0.3% Triton X-100 in PBS for 1 hour. The sections were incubated free floating at 4˚C with anti–IGF-I (1:100, Santa Cruz Biotechnol-ogy, Inc.), anti-NeuN (1:500, Chemicon), anti-GFAP (1:500, Dako-Cytomation), anti-CD11b (OX-42;1:100, Chemicon), anti-p35 (C-19; 1:500, Santa Cruz), or anti-phospho-tau (AT8) (1:50, Pierce Biotechnology) antibodies. Immunoreactivity was visualized using the Vectastain Elite ABC Peroxidase method (Vector Laboratories) and diaminobenzidine as the chromagen. Furthermore, apoptosis af-ter hemisection was detected by TUNEL in which we used an apop-tosis detection kit (Oncogene Research Products). Terminal
deoxy-nucleotidyl transferase–mediated deoxyuridine triphosphate nick-end labeling was performed according to the manufacturer’s in-structions. A negative control of TUNEL staining was generated by omission of Klenow enzyme, whereas the negative control sections of other immunohistochemical studies were incubated without pri-mary antibodies. Cell counting was performed on every sixth sec-tion at the center of spinal cord hemisecsec-tion stained with the afore-mentioned antibodies at a magnification of 400. Only cells with clearly visible staining were counted. All data are presented as means 6 SEMs of five consecutive cell quantifications.
Immunoblot Analysis
For immunoblotting, spinal cord samples (five in each group) were homogenized in ice-cold modified radioimmunoprecipita-tion buffer (50mM Tris-HCl, pH 7.4, 150 mM NaCl, 1 mM ethyl-enediaminetetraacetic acid, 1 mM phenylmethylsulfonyl fluoride, 1% NP-40, 0.25% Na-deoxycholate, 1 mM Na3VO4, 1 mM NaF,
1 mg/ml each of aprotinin, pepstatin, and leupeptin). Lysates were clarified by centrifugation for 15 minutes at 4˚C. Protein concentra-tions were determined using the BCA protein assay (Pierce Biotechnology, Inc.). Twenty-five-microgram protein extracts were electrophoresed on a 10 or 12% acrylamide sodium dodecyl sulfate-polyacrylamide gel electrophoresis and immunoblotted onto poly-vinylidene fluoride membranes. The membranes were blocked for 1 hour in TBST (10 mM Tris-HCl pH 7.6, 150 mM NaCl, 0.1% Tween-20) containing 5% weight/volume nonfat dry milk and then incubated overnight with the various primary antibodies. The anti-bodies used were anti-phospho-Akt (Ser-473; 1:1000, Cell Signal-ing Technology, Inc.), anti-Akt (1:1000, Cell SignalSignal-ing), anti-phos-pho-GSK-3b (Ser9; 1:1000, Cell Signaling), anti-GSK-3b (1:1000, Cell Signaling), anti–caspase-9 (1:1000, Cell Signaling), Anti-Bcl2 (1:1000, R&D Systems, Inc.), anti-Cdk5 (1:1000, Santa Cruz Bio-technology), and anti–b-actin (1:5000, Sigma-Aldrich Biotech-nology, Inc.). Immunoreactivity was demonstrated with horseradish peroxidase–conjugated goat anti–rabbit or anti–mouse (1:5000; Jackson ImmunoResearch Laboratories, Inc.) and the SuperSignal West Pico Chemiluminescent Substrate (Pierce). The band intensi-ties were quantified by using the Fluor-S Multilmager (Bio-Rad Laboratories, Inc.).
Statistical Analysis
All data are presented as means 6 SEMs (with the means derived from a minimum of three experiments). Statistical analysis was per-formed using a one-way analysis of variance. A probability value less than 0.05 was considered significant.
Results
Expression of IGF-I In Vivo by a Single Administration of
Naked Plasmid
To deliver exogenous IGF-I gene efficiently, we
devel-oped an in vivo gene transfection procedure in which a
large volume of naked plasmid DNA solution is rapidly
injected into the tail vein.
21,22The presence of IGF-I in
se-rum was analyzed with an ELISA kit. As shown in Fig 1,
a single intravenous administration of the pCMV–IGF-I
plasmid resulted in marked IGF-I expression. We used a
certain amount of plasmid DNA (3 mg/g) diluted in 15 ml
of saline and injected the plasmid via the tail vein into the
circulation within 15 seconds. The levels of IGF-I protein
in the circulation could reach as high as 80 ng/ml 1 day
following intravenous injection (Fig. 1). Although the
cir-culating level of IGF-I began to decline thereafter, a
sig-nificant amount of IGF-I protein was still found in the
cir-culation 14 days after the initial injection of pCMV–IGF-I
plasmid compared with that observed in the pCMV–LacZ
group (pCMV–IGF-I compared with pCMV–LacZ: Day 1
p , 0.001, Day 7 p , 0.001, and Day 14 p = 0.001).
Expression of IGF-I in the Spinal Cord 14 Days After
Gene Transfer
We further checked the IGF-I expression in the spinal
cords immunohistochemically 14 days after intravenous
administration of naked IGF-I plasmid (Fig. 2).
Insulin-like growth factor–I was markedly expressed at the spinal
cord hemisection in the pCMV–IGF-I group but not in the
negative control group (Fig. 2B). There was scant staining
in the sham (Fig. 2A), pCMV–LacZ (Fig. 2C), and SCI
(Fig. 2D) groups. In our model, IGF-I gene therapy could
directly transfect the injured spinal cord.
Effects of IGF-I Gene Transfer on Motor Function and
Neuron Survival After Hemisection
To evaluate the presence and extent of neurological
impairment after SCI, we used the BBB locomotor scales.
The BBB scores of the affected hindlimbs significantly
improved in the pCMV–IGF-I group compared with those
in the pCMV–LacZ (vehicle) and SCI groups 7 and 14
days after SCI (on Days 7 and 14: pCMV–IGF-I
pared with pCMV–LacZ p , 0.001; pCMV–IGF-I
com-pared with SCI p , 0.001) (Fig. 3). Immunohistochemical
assessment of NeuN, a neuron-specific marker, disclosed
severe neuron loss in the vehicle and SCI groups 14 days
after hemisection (sham compared with SCI p , 0.001;
sham compared with pCMV–LacZ p , 0.001) (Fig. 4).
Significant preservation of neurons was noted in the
pCMV–IGF-I group (pCMV–IGF-I compared with SCI
p , 0.001; pCMV–IGF-I compared with pCMV–LacZ
p , 0.001) (Fig. 4). Evaluation of these data suggested
that IGF-I gene therapy can prevent neurological deficits
and neuronal loss after spinal cord hemisection and thus
provided the reason why the BBB motor scores improved.
FIG. 1. Graph showing the distribution of IGF-I levels following a single injection of naked pCMV–IGF-I or pCMV–LacZ plasmid. At different time points as indicated, the plasma was collected, and the IGF-I levels were determined by a specific ELISA for IGF-I protein. Significantly higher levels of IGF-I were noted in IGF-I groups than in the LacZ Group 1, 7, and 14 days after administra-tion (pCMV–IGF-I compared with pCMV–LacZ [p , 0.001 on Days 1 and 7; p = 0.001 on Day 14]). Data are presented as means
Effects of IGF-I Gene Therapy on Astrocyte and Microglia
Activation After SCI
Astrocyte and microglia activation are the major
inflammatory responses after SCI.
10The expression of
GFAP and CD11b, a marker of activated astrocytes and
microglia, was investigated. Strong immunostaining of
GFAP and CD11b was demonstrated in the spinal cord
lesion, but these activations were greatly attenuated in the
pCMV–IGF-I group (Fig. 5). These results showed that
IGF-I gene transfer attentuates activation of astrocytes and
microglia after SCI.
Insulin-Like Growth Factor–I Gene Transfer Activates Akt,
Inhibits GSK-3
b, and Attenuates p35 Activation and Tau
Phosphorylation
To investigate the mechanism of IGF-I gene therapy
after SCI, we first used immunohistochemistry and
FIG. 2. Representative photomicrographs of IGF-I–stainedspinal cord sections obtained in rats. A: There was scant staining in sham, LacZ, and SCI groups. B1: Marked staining for IGF-I was detected in the injured cord 14 days following gene transfer in the IGF-I group. B2: There was no staining for IGF-I in negative controls (omission of the first antibody). C: The LacZ group. D: The SCI group. Original magnification 3 400.
FIG. 3. Bar graph showing that IGF-I gene transfer, compared with vehicle, improves BBB scores of the affected hindlimbs 7 and 14 days after SCI. A normal limb would receive a score of 21 on the BBB scale. Scores significantly improved in pCMV–IGF-I group compared with the pCMV–LacZ (vehicle) and SCI groups 7 and 14 days after SCI (pCMV–IGF-I compared with pCMV–LacZ
#p , 0.001; pCMV–IGF-I compared with SCI $p , 0.001 on Days
7 and 14 [10 animals in each group]).
FIG. 4. Representative photomicrographs (A–C and E–H) of NeuN-stained sections of the spinal cords acquired in sham controls (A and E), hemisection with IGF-I gene transfer (B and F), hemisection with LacZ vehicle treatment (C and G), and hemisection (SCI) without treatment (H). Note that there are normal neurons in the sham group (E) and rel-ative preservation of NeuN-stained neurons in the IGF-I group (F) compared with severe neuron loss in the vehicle (G) and SCI (H) groups. Original magnifications 3 12 (A–C) and 3 400 (E–H). D: Bar graph plotting results of the histo-pathological findings. Vertical bars indicate the mean (6 SEM) number of neurons per tissue section for the sham con-trols (sham), hemisection with IGF-I gene transfer (IGF-I), hemisection with LacZ vehicle treatment (LacZ), and hemi-section without treatment (SCI) (#p , 0.001, sham compared with SCI; p , 0.001, sham compared with LacZ; $p , 0.001
Western blotting to test the phosphorylated Akt at Ser473.
Upregulation of phosphorylated Akt was shown in the
pCMV–IGF-I group 14 days after hemisection by both
immunohistochemistry (Fig. 6A–E) and Western blot
analysis (Fig. 6F). Additionally, an increasing ratio of
inhibitory phosphorylation of GSK-3b at Ser9/GSK-3b
FIG. 5. A–E: Representative photomicrographs of CD11b-stained sections of the spinal cords obtained in the shamcontrol (A), IGF-I (B), LacZ (C), SCI (D), and negative control (omission of primary antibody) of LacZ (E) groups 14 days after hemisection. Strong staining of CD11b-positive cells was demonstrated in the spinal cord lesion (C and D), whereas IGF-I gene administration markedly attenuated this upregulation (B). Original magnification 3 400. F: Bar graph plotting results of histopathological findings. Vertical bars indicate the mean (6 SEM) number of CD11b-stained cells per tissue section for the sham control (sham), hemisection with IGF-I gene transfer (IGF-I), hemisection with LacZ vehicle treatment (LacZ), and hemisection without treatment (SCI) groups (#p , 0.001, sham compared with SCI; p ,
0.001, sham compared with LacZ; $p , 0.001, IGF-I compared with SCI; p , 0.001, IGF-I compared with LacZ [five
FIG. 6. A–D: Representative immunostained sections of phosphorylated Akt in sham control (A), IGF-I group (B), LacZ group (C), SCI group (D) and Western blotting (F)—all obtained 14 days after hemisection. Upregulation of phos-phorylated Akt was noted in the IGF-I group compared with other groups. E: Graph. Vertical bars indicate the mean (6 SEM) number of phosphorylated Akt-stained cells per tissue section for the sham controls (sham), hemisection with IGF-I gene transfer (IGF-I), hemisection with LacZ vehicle treatment (LacZ), and hemisection without treatment (SCI) (#p = 0.018, sham compared with SCI; $p , 0.001, IGF-I compared with SCI; p , 0.001, IGF-I compared with LacZ;
p , 0.001, IGF-I compared with sham [five animals in each group]). F: Western blot analysis confirming the up-regulation of phosphorylated Akt/Akt in the IGF-I group compared with the LacZ group (##p , 0.01, in arbitrary units,
was noted in the pCMV–IGF-I group compared with the
pCMV–LacZ group (Fig. 7A). In a recent study we have
revealed p35-p25-Cdk5 activation and tau
hyperphos-phorylation in the pathogenesis of SCI.
10In the present
study, we further demonstrated that IGF-I gene therapy
could also significantly attenuate p35 activation (Fig. 8)
and tau phosphorylation (Fig. 9) without significantly
interfering in Cdk5 protein levels (Fig. 7D). Therefore, we
have confirmed that the mechanisms of IGF-I gene
trans-fer in SCI include activation of phosphorylated Akt,
inhi-bition of GSK-3b, inactivation of p35, and attenuation of
phosphorylated tau.
Insulin-Like Growth Factor–I Gene Transfer Inhibits
Cleavage of Caspase-9, Increases Bcl2 Expression, and
Attenuates Apoptosis
To investigate signaling within the apoptotic pathway,
we tested the cleavage of caspase-9 and Bcl2 expression
by immunoblotting. Insulin-like growth factor–I gene
therapy was able to markedly reduce the amount of
cas-pase-9 cleavage (Fig. 7B) and increase Bcl2 expression 14
days after gene transfer therapy (Fig. 7C). Furthermore,
apoptosis 14 days after SCI was detected by TUNEL.
Terminal deoxynucleotidyl transferase–mediated
deoxy-uridine triphosphate nick-end labeling–positive cells were
FIG. 7. Representative panels and graphs showing inhibitory phosphorylation of GSK-3b at Ser9 (A), caspase-9 (B),Bcl2 (C), and Cdk5 (D) expressions in spinal cord by immunoblotting 14 days after hemisection. Upregulation of inhibitory phosphorylation of GSK-3b and Bcl2 expression was noted in the IGF-I group compared with the LacZ group (phospho–GSK-3b; IGF-I compared with LacZ [#p , 0.05]) (Bcl2; IGF-1 compared with LacZ [p , 0.01]; sham vs.
LacZ [* p , 0.05]). Downregulation of cleaved caspase-9 expression (P37 and P39) was noted 14 days after IGF-I gene transfer (IGF-I compared with LacZ [##p , 0.01]; sham compared with LacZ [**p , 0.01]). However, there was no
sig-nificant difference in Cdk5 expression. Arbitrary unit was defined as phosphor–GSK-3b/GSK-3b (A), P37/caspase-9 (B), Bcl2/b-actin (C), Cdk5/b-actin (D). There were five animals in each group.
easily demonstrated in the pCMV–LacZ and SCI groups
(Fig. 10C and D), whereas apoptosis was hardly detected
in the pCMV–IGF-I group (Fig. 10B). Insulin-like growth
factor–I gene transfer significantly attenuated apoptosis
14 days after SCI (Fig. 10F) (pCMV–IGF-I compared
with pCMV–LacZ p , 0.001; pCMV–IGF-I compared
with SCI p , 0.001). In our study we confirmed the
anti-apoptotic effect of IGF-I gene therapy with the inhibition
of caspase-9 cleavage and the upregulation of Bcl2.
Discussion
In the present study we have shown that a single
intra-venous IGF-I gene injection after spinal cord hemisection
provides neuroprotective, antiinflammatory, and
antiapop-totic effects in rats, and we hope that the findings of this
study will open a new window for the treatment of SCI.
We developed an in vivo gene transfection procedure in
which a large volume of naked plasmid DNA solution is
rapidly injected into the rat’s tail vein.
21,22One major
con-FIG. 8. A–D: Representative photomicrographs of p35(C-19)–stained sections of the spinal cords obtained in the sham (A), IGF-I (B), LacZ (C), and SCI (D) groups 14 days after SCI. Specific staining of p35 was upregulated in the LacZ and SCI groups (C and D), whereas IGF-I gene transfer prevented this activation (B). Original magnification 3 400. E: Bar graph plotting results of histopathological findings. Vertical bars indicate the mean (6 SEM) number of p35-stained cells per tissue section for the sham control (sham), hemisection with IGF-I gene transfer (IGF-I), hemisection with LacZ vehi-cle treatment (LacZ), and hemisection without any treatment (SCI) groups (sham compared with SCI [p , 0.001]; sham compared with LacZ [p , 0.001]; IGF-I compared with SCI [$p , 0.001]; IGF-I compared with LacZ [p , 0.001]; fivecern regarding this gene transfer technique, however, is
that it requires a fast injection of a large volume of
solu-tion, which may alter the physiological conditions of the
liver and heart. Indeed, after the injection, we noticed
ele-vated levels of some liver enzymes, which eventually
re-turned to baseline 3 to 10 days later.
15FIG. 9. A–E: Representative photomicrographs of phosphorylated tau (AT8)–stained sections of the spinal cords
acquired in the sham controls (A), IGF-I group (B), LacZ group (C), SCI group (D), and negative control (omission of primary antibody) of LacZ group (E) 14 days after hemisection. Specific staining of AT8 was upregulated in the LacZ (C) and SCI (D) groups, whereas IGF-I gene transfer prevented this activation (B). Original magnification 3 400. F: Bar graph plotting results of histopathological findings. Vertical bars indicate the mean (6 SEM) number of AT8-stained cells per tissue section for the sham controls (sham), hemisection with LacZ vehicle treatment (LacZ), and hemisection without any treatment (SCI) (sham compared with SCI [#p , 0.001]; sham compared with LacZ [p , 0.001]; IGF-I
We confirmed the presence of exogenous gene
expres-sion in both serum (Fig. 1) and the injured spinal cord
(Fig. 2B1). Whetstone et al.
23reported that SCI produces
a biphasic opening of the blood–spinal cord barrier. The
first peak of abnormal leakage occurs within the first
sev-eral hours after injury, whereas a second peak is evident
between 3 and 7 days postinjury. We took advantage of
this kind of time window to transfect IGF-I into the
in-FIG. 10. Representative photomicrographs of TUNEL-stained sections of the spinal cords acquired in sham control(A), hemisection with IGF-I gene transfer (B), hemisection with LacZ vehicle treatment (C), hemisection without any treatment (D), and negative control (without Klenow enzyme) of hemisection with LacZ vehicle treatment (E) groups. Original magnification 3 400. F: Bar graph plotting results of histopathological findings. Vertical bars indicate the mean (6 SEM) number of TUNEL-stained cells per tissue section for the sham controls (sham), hemisection with IGF-I gene transfer (IGF-IGF-IGF-I), hemisection with LacZ vehicle treatment (LacZ), and hemisection without any treatment (SCIGF-I) groups (sham compared with SCI [#p , 0.001]; sham compared with LacZ [p , 0.001]; IGF-I compared with SCI [$p , 0.001]; IGF-I compared with LacZ [p , 0.001]; five animals in each group).
jured cord (Fig. 2B1). Additionally, significant elevation
of serum IGF-I levels was shown for as many as 14 days
after IGF-I gene transfer (Fig. 1). Within 14 days of
inject-ing the IGF-I gene solution, there were relatively high
lev-els of circulating IGF-I protein in the IGF-I–treated rats
(Fig. 1). Furthermore, Pan and Kastin
17showed that
sys-temically administered IGF-I enters the central nervous
system by a saturable transport system at the blood–brain
barrier. This is one of the pharmacokinetic reasons why a
single injection of the IGF-I plasmid can protect the spinal
cord from SCI. The other direct immunohistochemical
evidence of high-level IGF-I protein expression in injured
spinal cord was present for as many as 14 days after gene
transfer (Fig. 2B1).
In terms of motor function recovery, given that the
pri-mary injuries caused by the intentional SCI were equal
in all three groups, we found no significant differences
among pCMV–IGF-I, pCMV–LacZ, and SCI-alone
groups 1 day after SCI (Fig. 3). The primary injury
imme-diately causes cell death or necrosis at the injury site and
then initiates a secondary injury process leading to an
ex-tension of the lesion into rostral and caudal areas of the
spinal cord. Insulin-like growth factor–I gene transfer
sig-nificantly improved the BBB Locomotor Scores 7 and 14
days after SCI (Fig. 3), partly because neuronal survival is
preserved after hemisection (Fig. 4B and F). Moreover,
within 14 days of the gene transfer therapy, there were
rel-atively high levels of circulating I protein in the
IGF-I–treated rats, and IGF-I is an anabolic growth factor for
skeletal muscle that can stimulate myoblast proliferation
and myofiber hypertrophy.
18Another reason for near-total
recovery after hemisection would be the adaptive
plastic-ity of the motor pathways in the spinal cord hemisection
model.
7The beneficial effects of IGF-I treatment were not
sole-ly restricted to neurons. We postulate that the size of a
sec-ondary neurodegeneration after SCI depends on the
mag-nitude of the inflammatory events. In the immediate
vicinity of the injury site, reactive astrocytes interweave to
form a barrier, creating a glial scar, which can be an
im-pediment to regenerating axons. Increased GFAP
expres-sion is a hallmark of reactive astrocytes, and this
cyto-skeletal protein contributes to a barrier effect of the glial
scar for axonal extension. This response is fortified by the
migration of microglia and macrophages to the damaged
area. Astrocytes and microglia deserve special attention
because of their roles in promoting the glial scar formation
during and after the inflammatory process in spinal cord
lesions.
6,10Recently, Kaspar and colleagues
12also obtained
similar results in a mouse model of amyotrophic lateral
sclerosis, suggesting a delayed activation of astrocytes in
the IGF-I–treated animals. In our study we also
demon-strated marked activation of astrocytes and microglia
af-ter SCI, with results shown by immunostaining of GFAP
and CD11b, and IGF-I gene transfer effectively abolished
these inflammatory responses (Fig. 5).
Regarding the downstream IGF-I expression, one
mol-ecule that may serve as a convergent point of different
sur-vival-promoting signaling pathways is Akt, a 60-kDa
ser-ine/threonine kinase that can be activated by IGF-I. The
mechanism of IGF-I has been shown to increase the
phos-phorylated state of Akt, a protein kinase that is involved in
blocking proapoptotic pathways through
receptor-mediat-ed phosphatidylinositol 3–kinase signaling.
20We found
that IGF-I–treated animals had 100% higher levels of
phosphorylated Akt than LacZ-treated controls (Fig. 6F).
Phosphorylated Akt has been shown to prevent cleavage
of caspase-9, thereby inhibiting apoptosis. Signaling
with-in the apoptotic pathway, with-includwith-ing the cleavage of
cas-pase-3 and caspase-9, could be a target for SCI
interven-tion. Insulin-like growth factor–I significantly reduced the
amount of caspase-9 cleavage. At 14 days after SCI,
IGF-I treatment had decreased the cleaved 37- and 39-kD
sub-units by more than 70% compared with LacZ treatment in
the control group, indicating that IGF-I can block caspase
activation involved in the apoptotic pathway (Fig. 7B). In
addition, IGF-I–treated animals had 166% higher levels of
Bcl2 than LacZ-treated controls (Fig. 7C). Furthermore,
TUNEL staining was less evident in IGF-I–treated
ani-mals than in the LacZ group (8.4 6 1.14 cells compared
with 65.4 6 14.84 cells, respectively [five rats in each
group], p , 0.001) (Fig. 10). These findings provide
mo-lecular evidence of the antiapoptotic effect of IGF-I.
Another target of IGF-I signaling via Akt is GSK-3b.
Insulin-like growth factor–I has been shown to stimulate
the inhibitory serine phosphorylation of this pivotal
en-zyme in cultured neurons.
9,13Insulin-like growth
factor–I-induced inhibitory phosphorylation of GSK-3b at Ser9
19relieves GSK-3b’s inhibition of glycogen synthase and
the translation initiation factor eIF2B, thus promoting
gly-cogen and protein synthesis. We found that IGF-I–treated
rats had 30% higher levels of inhibitory GSK-3b at Ser9
than LacZ-treated controls (Fig. 7A). Tau, a
microtubule-associated protein involved in neurofilament stabilization,
is also a GSK-3b substrate. The authors of some studies
have shown that IGF-I inhibits GSK-3b in neural cells
and results in inhibition of tau hyperphosphorylation.
8,22When it is hyperphosphorylated, tau is prone to form
in-tracelllular neurofibrillary tangles that contribute to
neu-ronal degeneration. We have found that tau
phosphoryla-tion increased in the SCI and LacZ groups compared with
the sham control group (Fig. 9), and IGF-I gene transfer
could significantly inhibit tau hyperphosphorylation (Fig.
9B and F).
Recently we documented p35-p25-Cdk5 activation
and tau hyperphosphorylation in rats that had undergone
SCI.
10Hyperphosphorylated tau is a major component of
neurofibrillary tangles, one of the hallmarks of Alzheimer
disease. It has been shown that Cdk5 is a kinase that
phos-phorylates the tau protein and that its endogenous
activa-tors, p35 and p25, regulate its activity.
14We did not find
significant change in the Cdk5 protein level after IGF-I
gene transfer (Fig. 7D). On immunohistochemical
evalua-tion, however, p35 was less stained in IGF-I–treated rats
than LacZ-treated rats (63.8 6 14.11 cells compared with
293.0 6 33.20 cells, respectively [five rats in each group],
p , 0.001) (Fig. 8E). It has been shown that p35 regulates
not only the overall kinase activity of Cdk5 but also the
sequential phosphorylation of Ser202 and Thr205 in tau.
8Most of the tau studies have been focused on
neurodegen-erative Alzheimer disease, and our findings confirm the
roles of p35 activation and tau hyperphosphorylation in
SCI. Moreover, IGF-I gene transfer could inhibit this
pathological cascade.
This study is the first step in confirming the potential
ef-fects of IGF-I gene transfer in SCI. Hemisection of the
more, IGF-I attenuated caspase-9 cleavage, increased
Bcl2, and thus inhibited apoptosis 14 days after SCI.
These findings suggest that IGF-I gene transfer is a
prom-ising neuroprotective approach after SCI. Further studies
involving alternative doses, routes, and vector paradigms
are necessary to determine the potential clinical
applica-tion of IGF-I in SCI.
Disclaimer
The authors have no financial investment in any of the products mentioned in this manuscript.
Acknowledgments
We thank Ms. Angel Sun and Yilu Lee for the sample collection and technical support.
References
1. Anderson DK, Beattie M, Blesch A, Bresnahan J, Bunge M, Dietrich D, et al: Recommended guidelines for studies of hu-man subjects with spinal cord injury. Spinal Cord 43:453–458, 2005
2. Basso DM, Beattie MS, Bresnahan JC: A sensitive and reliable locomotor rating scale for open field testing in rats. J Neuro-trauma 12:1–21, 1995
3. Bracken MB: Methylprednisolone and acute spinal cord injury: an update of the randomized evidence. Spine 26 (24 Suppl): S47–S54, 2001
4. Cheng CM, Mervis RF, Niu SL, Salem N Jr, Witters LA, Tseng V, et al: Insulin-like growth factor–I is essential for normal den-dritic growth. J Neurosci Res 73:1–9, 2003
5. Cheng CM, Reinhardt RR, Lee WH, Joncas G, Patel SC, Bon-dy CA: Insulin-like growth factor–I regulates developing brain glucose metabolism. Proc Natl Acad Sci U S A 97:10236– 10241, 2000
6. Davies SJ, Field PM, Raisman G: Regeneration of cut adult axons fails even in the presence of continuous aligned glial pathways. Exp Neurol 142:203–216, 1996
7. Fujiki M, Kobayashi H, Inoue R, Ishii K: Immediate plasticity in the motor pathways after spinal cord hemisection: implica-tions for transcranial magnetic motor-evoked potentials. Exp Neurol 187:468–477, 2004
8. Hashiguchi M, Saito T, Hisanaga S, Hashiguchi T: Truncation of CDK5 activator p35 induces intensive phosphorylation of Ser202/ Thr205 of human tau. J Biol Chem 277:44525–44530, 2002 9. Hong M, Lee VM: Insulin and insulin-like growth factor-1
reg-ulate tau phosphorylation in cultured human neurons. J Biol Chem 272:19547–19553, 1997
15. Miao CH, Thompson AR, Loeb K, Ye X: Long-term and ther-apeutic-level hepatic gene expression of human factor IX after naked plasmid transfer in vivo. Mol Ther 3:947–957, 2001 16. Niblock MM, Brunso-Bechtold JK, Riddle DR: Insulin-like
growth factor I stimulates dendritic growth in primary somato-sensory cortex. J Neurosci 20:4165–4176, 2000
17. Pan W, Kastin A: Interactions of IGF-1 with the blood-brain barrier in vivo and in situ. Neuroendocrinology 72:171–178, 2000
18. Shansky J, Creswick B, Lee P, Wang X, Vandenburgh H: Paracrine release of insulin-like growth factor–I from a bioen-gineered tissue stimulates skeletal muscle growth in vitro. Tis-sue Eng 12:1833–1841, 2006
19. Summers SA, Kao AW, Kohn AD, Backus GS, Roth RA, Pes-sin JE, et al: The role of glycogen synthase kinase 3b in insu-lin-stimulated glucose metabolism. J Biol Chem 274:17934– 17940, 1999
20. Vincent AM, Feldman EL: Control of cell survival by IGF sig-naling pathways. Growth Horm IGF Res 12:193–197, 2002 21. Wang CH, Jawan B, Lee TH, Hung KS, Chou WY, Lu CN, et
al: Single injection of naked plasmid encoding b-melanocyte-stimulating hormone protects against thioacetamide-induced acute liver failure in mice. Biochem Biophys Res Commun 322:153–161, 2004
22. Wang CH, Liang CL, Huang LT, Liu JK, Hung PH, Sun A, et al: Single intravenous injection of naked plasmid DNA encod-ing erythropoietin provides neuroprotection in hypoxia-isch-emia rats. Biochem Biophys Res Commun 314:1064–1071, 2004
23. Whetstone WD, Hsu JY, Eisenberg M, Werb Z, Noble-Haeusslein LJ: Blood-spinal cord barrier after spinal cord in-jury: relation to revascularization and wound healing. J Neuro-sci Res 74:227–239, 2003
Manuscript received January 5, 2006. Accepted in final form September 15, 2006.
This work was supported by the National Science Council (Grant Nos. NSC 95-2314-B-182A-110, NSC 94-2314-B-182-032, and NSC 93-2314-B-182A-163), Department of Health (Grant No. DOH-TD-B-111-002), a Topnotch Stroke Research Center grant, and the Ministry of Education and Kaohsiung Chang Gung Mem-orial Hospital (Grant No. CMRPG 83058).
Address reprint requests to: Wen-Ta Chiu, M.D., Ph.D., Depart-ment of Neurosurgery, Taipei Medical University, Wan Fang Med-ical Center, 111 Section 3, Hsing-Long Road, Taipei 116, Taiwan. email: [email protected].
