ORIGINAL PAPER

Combination of G-CSF Administration and Human AmnioticFluid Mesenchymal Stem Cell Transplantation PromotesPeripheral Nerve Regeneration

Hung-Chuan Pan Æ Chung-Jung Chen Æ Fu-Chou Cheng Æ Shu-Pen Ho ÆMu-Jung Liu Æ Shiaw-Min Hwang Æ Ming-Hong Chang Æ Yeou-Chih Wang

Accepted: 21 July 2008 / Published online: 9 August 2008

� Springer Science+Business Media, LLC 2008

Abstract Amniotic fluid mesenchymal stem cells (AFS)

harbor the potential to improve peripheral nerve injury by

inherited neurotrophic factor secretion, but present the

drawback of the short-term survival after transplantation.

Granulocyte-colony stimulating factor (G-CSF) has a

diversity of functions, including anti-inflammatory and

anti-apoptotic effects. This study was conducted to evalu-

ate whether G-CSF could augment the neuroprotective

effect of transplanted AFS against peripheral nerve injury.

The potential involvement of anti-inflammation/anti-

apoptosis effect was also investigated. Peripheral nerve

injury was produced in Sprauge-Dawley rats by crushing

left sciatic nerve using a vessel clamp. The AFS were

embedded in fibrin glue and delivered to the injured site.

G-CSF (50 lg/kg) was administrated by intra-peritoneal

injection for 7 consecutive days. Cell apoptosis, inflam-

matory cytokines, motor function, and nerve regeneration

were evaluated 7 or 28 days after injury. Crush injury

induced inflammatory response, disrupted nerve integrity,

and impaired nerve function in sciatic nerve. Crush injury-

provoked inflammation was attenuated in groups receiving

G-CSF but not in AFS only group. In transplanted AFS,

marked apoptosis was detected and this event was reduced

by G-CSF treatment. Increased nerve myelination and

improved motor function were observed in AFS trans-

planted, G-CSF administrated, and AFS/G-CSF combined

treatment groups. Significantly, the combined treatment

showed the most beneficial effect. In conclusion, the con-

comitant treatment of AFS with G-CSF augments

peripheral nerve regeneration which may involve the sup-

pression of apoptotic death in implanted AFS and the

attenuation of inflammatory response.

Keywords Apoptosis � Amniotic fluid mesenchymal

stem cells � G-CSF � Sciatic nerve injury � Inflammatory

cytokines

Introduction

In the past decades there have been significant advance in

the peripheral nerve repair. These have included the

introduction of the microscope, tension free repair by the

H.-C. Pan � M.-J. Liu

Department of Neurosurgery, Taichung Veterans General

Hospital, Taichung, Taiwan

e-mail: hcpan2003@yahoo.com.tw

H.-C. Pan � C.-J. Chen

Institute of Medical Technology, National Chung-Hsing

University, Taichung, Taiwan

H.-C. Pan � S.-P. Ho

Department of Veterinary Medicine, National Chung-Hsing

University, Taichung, Taiwan

F.-C. Cheng

Stem Cell Center, Taichung Veterans General Hospital,

Taichung, Taiwan

S.-M. Hwang

Bioresource Collection and Research Center, Food Industry

Research and Development Institute, Hsinchu, Taiwan

M.-H. Chang

Department of Neurology, Taichung Veterans General Hospital,

Taichung, Taiwan

Y.-C. Wang (&)

Department of Neurosurgery, Chung-Shan Medical University

Hospital, No. 110, Sec. 1, Chien-Kuo N. Road, Taichung 402,

Taiwan, ROC

e-mail: ycwang31@csmu.due.tw

123

Neurochem Res (2009) 34:518–527

DOI 10.1007/s11064-008-9815-5

epineural or perineural suture and accurate nerve apposi-

tion by means of anatomic features, and histochemical or

immunohistochemical method for motor and sensory fiber.

Despite early diagnosis and modern surgical technique,

and no matter how accurate the nerve repair, function

recovery can never reach the pre-injury level. Poor out-

come may result from many factors, both intrinsic and

extrinsic to nervous system, such as the type and level

of injury, the presence of associated injury, the timing

of surgery, change in the spinal cord neuron and end

organ [1].

Several alternative approaches have been proposed to

have a beneficial effect on the peripheral nerve regener-

ation, including application of electric field, trans-

plantation of stem cell, and administration of neurotrophic

factors [2–5]. Recently, cell transplantation has become

the focus of researchers’ attention. The implantation of

embryonic stem cells, neural stem cells, and mesenchymal

stem cells has shown to exert a beneficial effect on

peripheral nerve regeneration. Cell replacement, trophic

factor production, extracellular matrix molecule synthesis,

guidance, remyelination, microenvironmental stabiliza-

tion, and immune modulation have recently been

proposed as beneficial mechanisms after cell implantation

[3, 6, 7].

Granulocyte-colony stimulating factor (G-CSF) is a

member of the hematopoietic growth factor family, which

orchestrates the proliferation, differentiation, and survival

of hematopoietic progenitor cells [8]. However, growing

evidence has suggested that G-CSF also has important non-

hematopoietic functions in other tissues including nervous

tissues. Recently, G-CSF has been shown to exert protec-

tive effects on various tissues and experimental models of

neurological disorders [9–13].Currently, the proposed

mechanisms of G-CSF related neuroprotection are medi-

ated by cell mobilizing, anti-inflammatory, or anti-

apoptotic activity [9–11, 14–17].

Recent evidence has shown amniotic fluid to be a

novel source of stem cells for therapeutic transplantation.

Amniotic fluid-derived stem cells express characteristics

of both mesenchymal and neural stem cells [18]. In our

previous studies, we demonstrated that transplantation of

amniotic fluid mesenchymal stem cells (AFS) promoted

peripheral nerve regeneration [4, 5]. However, the via-

bility of implanted cells declined dramatically after

transplantation. The short-term survival of implanted stem

cells might diminish cell transplantation-mediated bene-

ficial effects. Therefore, the present study was designed to

evaluate whether the combination of G-CSF and AFS

transplantation could augment the peripheral nerve

regeneration. The potential contribution of the anti-apop-

totic and anti-inflammatory effects of G-CSF was also

conducted.

Experimental Procedure

Animal Model

Sprague Dawley rats weighing from 250 to 300 g were

used in this study; permission was obtained from the Ethics

Committee of Taichung Veterans General Hospital. The

rats were anesthetized with 4% isoforane in induction

followed by a maintenance dose (1–2%). The left sciatic

nerve was exposed under a microscope using the gluteal

muscle splitting method. A vessel clamp (B-3, pressure

1.5 gm/mm2, S&T Marketing LTD, Switzerland) was

applied 10 mm from the internal obturator canal for 20 min

[5]. The animals were categorized into four groups: group I

(n = 27): The crush nerve was wrapped with fibrin glue.

The rats received the intra-peritoneal injection of normal

saline per day for 7 consecutive days; group II (n = 27):

The crush nerve was wrapped with fibrin glue. The rats

were concomitantly injected with G-CSF (Kirin Brewery

Co. Ltd., Japan) (50 lg/kg 9 7 days) intra-peritoneally;

group III (n = 33): AFS was embedded in fibrin glue and

delivered to injured nerve. The rats received the intra-

peritoneal injection of normal saline; group IV (n = 33):

AFS was embedded in fibrin glue and delivered to injured

nerve. The rats were followed by injection of G-CSF

(50 lg/kg 9 7 days) intra-peritoneally. Another group of

animals (n = 22) without crush acted as the control for

some assays (n = 6 for histology, n = 4 for determination

of S-100 expression, n = 6 for determination of neurofil-

ament, and n = 6 for the determination of inflammatory

cells). To avoid the rejection of cell transplantation, the

cyclosporine was used in this study. As known, the mac-

rophage migration and inflammatory cytokine expression

were influenced by the administration of cyclosporine. To

lessen these effects, all animals either as experimental or

control groups, were allowed free for accessing to food and

water supplemented with cyclosporine (Novartis, USA)

(12.5 mg cyclosporine in 125 ml drinking water and kept

daily intake at 50 cc). The administration of cyclosporin

started from 1 day after injury till the day of sacrifice [19].

Preparation and Culture of Human Amniotic

Mesenchymal Stem Cell (AFS)

Amniotic fluid samples (20 ml) were obtained by amnio-

centesis performed between 16 weeks of gestation for fetal

karyotyping. For culturing amniocytes, four primary in situ

cultures were set up in 35 mm tissue culture-grade dishes

using Chang medium (Irvine Scientific, Santa Ana, CA),

Microscopic analysis of Giemsa-stained chromosome

banding was performed, and the rules for metaphase

selection and colony definition were based on the basic

requirements for prenatal cytogenetic diagnosis in

Neurochem Res (2009) 34:518–527 519

123

amniocytes [20]. For culturing AFS, non-adhering amniotic

fluid cells in the supernatant medium were collected on the

fifth day after primary amniocytes culture and maintained

until completion of fetal chromosome analysis. The cells

were then centrifuged and plated in 5 ml of b-minimum

essential medium (b-MEM; Gibco-BRL) supplemented

with 20% fetal bovine serum (FBS; Hyclone, Logan, UT,

USA) and 4 ng/ml basic fibroblast growth factor (bFGF;

R&D system, Minneapolis, MN, USA) in a 25 cm flask and

incubated at 37% with 5% humidified CO2 [4]. This pro-

tocol was approved by the Institutional Review Board

(IRB) of the Veterans General hospital and written

informed consents were obtained from all patients.

Grafting Procedure

AFS were labeled with Hoechst 33342 before grafting. A

volume of 25 ll of AFS with density of 106 cell/ml was

suspended in 25 ll of Fibrin glue (Aventis Behring, Ger-

many) containing the woven Surgicel (Johnson& Johnson,

USA) and transplanted into the injured site immediately

after crush [5].

Analysis of Functional Recovery

Base on our previous report of crush nerve injury model,

the values of SFI at 1 week nearly reached -100 and the

technique was undesirable in investigating SFI due to the

wound pain immediately after the injury [5]. One technical

assistant who was blinded to treatment allocation evaluated

sciatic nerve function weekly after the surgery. The eval-

uation method included sciatic function index (SFI) [4, 5].

Several measurements were taken from the footprint by red

ink print: [1] distance from the heel to the third toe, the

print length (PL); [2] distance from the first to fifth toe,

the toe spread (TS); and [3] distance from the second to the

fourth toe, the intermediary toe spread (ITS). All three

measurements were taken from the experimental (E) and

normal (N) sides. The SFI was calculated according to the

equation:

SFI ¼ �38:3 EPL� NPL=NPLð Þþ 109:5 ETS� NTS=NTSð Þþ 13:3 EIT� NIT=NITð Þ � 8:8

The SFI oscillates around 0 for normal nerve function,

whereas SFI around -100 represents total dysfunction.

Electrophysiology Study

Ten left sciatic nerves from individual group were exposed

4 weeks after operation. Electric stimulation was applied to

the proximal side of the injured site; the conduction

latency, and the compound muscle action potential

(CMAP) were recorded with an active electrode needle

10 mm below the tibia tubercle and a reference needle

20 mm from the active electrode. The mean length from

the stimulation to the active recording electrode was

53.6 ± 0.3 mm. The stimulation intensity and filtration

ranges were 5 mA and 20–2,000 Hz, respectively. The

CMAP data and conduction latency were converted to

ratios of injured side divided by the normal side to adjust

for the effect of anesthesia [4, 5].

Quantification of Pro-Inflammatory Cytokines

Five nerve tissues in each group for every single parameter

were removed 7 days after the operation. The regenerating

tissues (10 mm in length) were retrieved and the samples

were stored at -80�C. Subsequently, each tissue sample

was homogenized with Laemmli SDS buffer. The

homogenate was centrifuged for 10 min at 12,000 g at 4�C.

The tissue homogenate, 100 ll in triplicate was applied to

a microtiter plate and allowed to adhere overnight at 4�C.

The microtiter plates were washed with phosphate-buffered

saline (PBS)-Tween-20 and blocked with 1% BSA in PBS

(200 ll) for 1 h. The plates were then treated with

respective primary antibodies and allowed to treat for 6 h

at 37�C. One hundred microliters of the respective poly-

clonal antibodies against TNF-a, IL-1b, Il-6 (R&D system,

Inc) and INF-c (Chemicon, Inc) were applied overnight to

microtiter plates. After further washing in PBS-Tween-20,

the plates were incubated with the respective second anti-

body conjugate to alkaline phosphate 100 ll for 1 h. The

reaction was developed using p-nitrophenyl phosphate,

disodium (3 mM) in carbonate buffer, pH 9.6(100 mM

Na2CO3 and 5 mM MgCl2 (150 ll), and the reaction was

terminated after 30 min using 0.5 N NaOH(50 ll). The

absorbance of colored product was read at 450 nm using a

microplate reader (Bio-Tek instruments). The relative

amount of antigen present was measured from the densi-

tometric reading against a standard curve.

Terminal Dexonucleotidyl Transferase-Mediated

Biotinylated UTP Nick End Labeling (TUNEL) Assay

Serial 8 lm-thick sections of sciatic nerve (7 days after

surgery) were cut on a cryostat and mounted on superfrost/

plus slides (Menzel-Glaser, Braunschweig, Germany).

TUNEL assay (Roche Molecular Biochemicals, Mannheim,

Germany) were carried out as previously described [21].

Apoptotic cells were defined as those cells with TUNEL-

positive nuclei that were condensed and fragmented, as

assay by DAPI (Molecular Probes, Eugene, OR, 1:2,000

dilutions). The number of apoptotic transplanted cells was

expressed as a percentage of the total number of nuclei

counted, with at least 25,000 nuclei for each condition.

520 Neurochem Res (2009) 34:518–527

123

Immunohistochemistry

Serial 8 lm-thick sections of sciatic nerve were cut on a

cryostat and mounted on superfrost/plus slides (Menzel-

Glaser, Braunschweig, Germany) and were subjected to

immunohistochemistry with antibodies against CD68

(Chemicon, 1:200 dilution) (7 days after surgery), S-100

(Neomarkers, 1:400 dilution) (4 weeks after surgery), and

neurofilament(Chemicon, 1:300 dilution) (7 days after

surgery) for the detection of inflammatory cells, schwann

cells, and nerve fibers, respectively. The immunoreactive

signals were observed by goat anti-mouse IgG (FITC)

(Jackson, 1:200 dilution), anti-mouse IgG (Rhodamine)

(Jackson, 1:200 dilution), or 3, 30-diaminobenzidine brown

color. Among longitudinal consecutive resection, five

consecutive resections contiguous to a maximum diameter

were chosen to measure. Of 100 squares with a surface area

of 0.01 mm2 each, 20 were randomly selected in an ocular

gird to count the number of the inflammatory cells. For the

determination of neurofilament and S-100, six nerves in

each group were cut longitudinally into 8 lm-thick sec-

tions, stained with each antibody. The maximum diameter

of the resected nerve tissue with crush mark was chosen to

be examined. Area of activities (0.2 mm2) appeared as

density against the background and were measured by

computer image analysis system (Alpha Innotech Corpo-

ration, IS 1000).

Histological Examination

The sciatic nerve was harvested from the animals after the

electrophysiological testing and the nerve tissue was fixed

on a plastic plate by the stay sutures to keep the nerve

straight [4]. The nerve was embedded, cut longitudinally

into sections 8 lm thick and stained with haematoxylin-

eosin (H&E) for the measurement of vacuole number and

vascular staining. Among longitudinal consecutive resec-

tions, five consecutive resections contiguous to a maximum

diameter were chosen to collect the data for comparison.

Of 100 squares with a surface area of 0.01 mm2 each, 20

were randomly selected in an ocular grid and used to count

the vacuole number and vascular staining.

Statistical Analysis

Data were expressed as the meant ± SE (standard error).

The statistical significance of differences between groups

was determined by one-way analysis of variance

(ANOVA) followed by Dunnett’s test. In SFI study, the

results were analyzed by repeated-measurement of ANOV

followed by multiple comparison method of Bonferroni.

P value less than 0.05 was considered significant.

Results

Motor Function and Electrophysiology Improvement

by the Concomitant Administration of AFS and G-CSF

Increased nerve regeneration was accompanied by the

improvement of sciatic nerve function index, increased

compound muscle action potential, and reduced nerve

conduction latency [5]. The amplitude of muscle com-

pound action potential reflected the number of axon

reinnervating the muscle and was related to the amount of

actylocholine release [22]. The nerve conduction latency

was reciprocal to motor function improvement [23]. The

SFI in different time points and treatment groups was

shown in Table 1. Treatment with either AFS or G-CSF

treatment exerted significant improvement in SFI as

compared to non-treatment (P = 0.007 and P = 0.02,

respectively). Improvement of SFI was also demonstrated

in nerve crush injury treated by AFS + G-CSF as com-

pared with those treated either with AFS or G-CSF alone

(P = 0.03 and P = 0.013, respectively). But there was no

significant difference between AFS and G-CSF groups

(P = 0.98) (Fig. 1). The electrophysiological study

Table 1 The values of SFI in different time points and treatment

groups

Groups Time

1 Week 2 Weeks 3 Weeks 4 Weeks

Crush -96.5 ± 2.1 -75.9 ± 7.4 -62.5 ± 5.6 -44.9 ± 3.6

G-CSF -83.0 ± 5.3 -69.7 ± 4.8 -39.4 ± 6.1 -28.8 ± 3.8

AFS -66.4 ± 7.5 -57.1 ± 6.2 -38.4 ± 5.3 -23.0 ± 3.8

AFS +

G-CSF

-59.2 ± 9.8 -42.7 ± 8.3 -23.0 ± 5.6 -7.6 ± 2.9

Crush, G-CSF, AFS, AFS + G-CSF: see text; data presented

mean ± standard errors

Fig. 1 Neurobehavioral evaluation. A representative illustration

of SFI in four treatment groups is depicted. ** P \ 0.01 and

*** P \ 0.001 versus crush control, n = 10

Neurochem Res (2009) 34:518–527 521

123

showed the similar trends. The average percentage of

CMAP in four different groups were 24 ± 3% (crush),

56 ± 5% (G-CSF), 50 ± 3% (AFS), and 70 ± 7%

(AFS + G-CSF), respectively. There was significant dif-

ference between crush and G-CSF (P \ 0.001), crush and

AFS (P \ 0.001), crush and G-CSF + AFS (P \ 0.001),

G-CSF and AFS + G-CSF (P \ 0.001), and AFS and

G-CSF + AFS (P \ 0.001), respectively. There existed no

significant difference between G-CSF and AFS (P = 0.37).

The ratio of conduction latency in four different groups

were 2.7 ± 0.1 (crush), 1.75 ± 0.11 (G-CSF), 1.9 ± 0.04

(AFS), and 1.31 ± 0.07 (AFS + G-CSF), respectively.

There was significant difference between crush and G-CSF

(P \ 0.001), crush and AFS (P \ 0.001), crush and

G-CSF + AFS (P \ 0.001), G-CSF and AFS + G-CSF

(P \ 0.001), AFS and G-CSF + AFS (P \ 0.001),

respectively. No significant difference existed between

G-CSF and AFS (P = 0.19). The impaired CMAP

(Fig. 2a) and conduction latency (Fig. 2b) were restored in

all three treated groups. Among the three treated groups,

AFS + G-CSF showed the best improvement. The findings

revealed that the nerve regeneration could be promoted by

the concomitant treatment of AFS and G-CSF.

Early and Late Nerve Regeneration by the Concomitant

Treatment of AFS + G-CSF

Axonal degeneration took place dramatically from 3 to

7 days after the nerve crush injury. Evidence indicated that

an increased expression of neurofilament reflected the early

regenerative potential [24]. Treatment with either AFS

(969.7 ± 50.8 relative density/mm2) or G-CSF (522.5 ±

73.3 relative density/mm2) enhanced significant expression

of neurofilament as compared to non-treatment (204.5 ±

16.6 relative density/mms2) (P \ 0.001 and P = 0.002,

respectively), but treatment with AFS + G-CSF (1258.5 ±

28.6 relative density/mm2) produced higher expression

than either AFS (P = 0.002) or G-CSF (P \ 0.001) alone

(Fig. 3). Increased myelination and vascular organization

and decreased vacuoles are positively correlated to the

integrity of nerve tissues and may reflect the strength of

nerve regeneration at later phase [5]. The parameters of late

nerve regeneration such as vacuole number, vascular

staining, and myelination as evidenced by the expression of

S-100 in this study in line with these findings (Fig. 4).

Based on the early expression of neurofilament and late

regeneration marker, treatment with either AFS or G-CSF

alone promoted greater nerve regeneration than those

without treatment; however, the combined treatment

aroused remarkable regeneration than either of the single

treatments.

Reduction of Apoptosis by the Concomitant Treatment

of AFS + G-CSF

Neurobehavioral and histological examination and other

related studies [4, 5] show that transplantation of AFS can

alleviate neurological deficits in a concentration-dependent

manner. Therefore, it is reasonable that the preservation of

viable implanted AFS is a strategy for improving periph-

eral nerve regeneration. The Hoechst 33342-positive

implanted AFS were found in the retrieved nerve tissues

7 days after grafting. Apoptotic AFS (7.8 ± 0.7%) were

detected by the TUNEL-positive nuclei. The apoptosis of

implanted AFS (2.3 ± 0.34%) was attenuated by G-CSF

treatment (P \ 0.001) (Fig. 5). The findings indicate that

one of beneficial effects of G-CSF is to strengthen the

viability of implanted AFS so as to prevent apoptosis.

Attenuation of inflammation by the concomitant

treatment of AFS + G-CSF

Over-activated inflammatory response is a detrimental

stress on the nerve tissues and is a potential cytotoxic factor

in the survival of implanted cells. Immunohistochemical

results showed an accumulation of inflammatory cells in

the injured nerve tissues (27.5 ± 1.1/0.05 mm2) (Fig. 6a).

The accumulation of inflammatory cells was not changed

in AFS group (28.8 ± 1.25/0.05 mm2) (P = 0.2) (Fig. 6b),

but was remarkably alleviated in G-CSF (13 ± 0.96/

0.05 mm2) (P \ 0.001) (Fig. 6c) and G-CSF + AFS

(10.5 ± 0.76/mm2) (P \ 0.001) (Fig. 6d) groups. On the

Fig. 2 Electrophysiological evaluation. Electrophysiological exami-

nation, including CMAP (a) and conduction latency (b), was

conducted 4 weeks after injury in four treatment groups. P values

in G-CSF, AFS, AFS + G-CSF were determined relative to the crush

group. * P \ 0.05, ** P \ 0.01, and *** P \ 0.001, n = 10

522 Neurochem Res (2009) 34:518–527

123

Fig. 3 Determination of

neurofilament. The nerve tissues

were retrieved 7 days after

injury and were subjected to

immunohistochemistry with

antibody against neurofilament

in four treatment groups, (a)

Normal (b) Crush (c) G-CSF (d)

AFS (e) AFS + G-CSF. The

relative density of neurofilament

was depicted in (f). P values in

G-CSF, AFS, and AFS +

G-CSF were determined relative

to crush group. ** P \ 0.01 and

*** P \ 0.001; n = 6. Bar

length = 50 lm

Fig. 4 Histological evaluation.

The nerve tissues were retrieved

4 weeks after injury and were

subjected to H&E stain (a–e)

and immunohistochemistry with

antibody against S-100 (f–j) in

four treatment groups, (a, f)Normal (b, g) Crush, (c, h)

G-CSF, (d, i) AFS, (e, j)AFS + G-CSF. The results of

quantitative analysis were

shown in (k) vacuole counts, (l)vascular stain, and (m) S-100. Pvalues in G-CSF, AFS, and

AFS + G-CSF were determined

relative to the crush group.

* P \ 0.05, ** P \ 0.01, and

*** P \ 0.001; n = 6 for H&E;

n = 4 for S-100. Bar

length = 50 lm (a–e) and

100 lm (f–j)

Neurochem Res (2009) 34:518–527 523

123

other hand, crush injury triggered the production of

inflammatory cytokines including IL-1b, IL-6, TNF-a, and

IFN-c (Fig. 7). The elevated production of IL-1b, TNF-aand IFN-c was attenuated in G-CSF and G-CSF + AFS

groups. Inducible inflammatory cytokines were not abro-

gated by AFS transplantation alone. The findings indicate

that the G-CSF but not the AFS possesses immunosup-

pressive effect.

Discussion

Utility of AFS delivered to the injured nerve is regarded as

one of treatment strategy in peripheral nerve injury. Either

of immunomodulation or neurotrophic factors secretion

was postulated to exert its effect on regeneration [4, 5].

However, the short-term survival of implanted cells

restricted the clinical application and reduced its efficacy

[25, 26]. G-CSF has been shown to harbor anti-apoptotic/

anti-inflammatory effects [17, 27–30]. In this study, we

found that the administration of G-CSF decreased inflam-

matory cell infiltration and attenuated the elevated

production of inflammatory cytokines including TNF-a,

IL-1b, and IFN-c as well as exerted the anti-apoptotic

effect on implanted AFS. Treatment with either AFS or

G-CSF alone increased better nerve regeneration than that

of control. Moreover, the combination of G-CSF and AFS

significantly augmented peripheral nerve repair. It has been

postulated that the combined effect is due to the decreased

production of cytotoxic inflammatory mediators and

increased survival of implanted AFS modulated by G-CSF.

The continuous survival and successful integration of

implanted cells are regulated by multiple factors. There are

several possible reasons for the short-term survival of the

implanted cells such as detrimental effect of inflammatory

cytokines, inadequate niches, abnormal apoptosis, and

other un-identified mechanisms [25, 26, 31]. Our results

showed that, despite functional improvement, apoptotic

Fig. 5 Determination of

apoptosis. The nerve tissues

were retrieved 7 days after

injury and were subjected to

apoptotic assay by TUNEL in

AFS (a) and AFS + G-CSF (c)

groups. The implanted AFS

within the corresponding areas

was demonstrated by the

positivity of Hoechst 33342 in

AFS (b) and AFS + G-CSF (d)

groups. Quantitative analysis of

TUNEL test was depicted in (e).

*** P \ 0.001, n = 6. Bar

length = 50 lm. The vertical

axis presented the percentage of

positive TUNEL assay

524 Neurochem Res (2009) 34:518–527

123

damage was detected in implanted AFS (Fig. 5). We found

that G-CSF exerted an anti-apoptotic effect against injured

cells, especially the implanted AFS (Fig. 5). The activation

of the signal transducer and activator of transcription

(STAT), extracellular signal-regulated kinase (ERK), and

Akt has been implicated in anti-apoptotic effect of G-CSF

[11, 28–30]. The direct anti-apoptotic effect of G-CSF is in

turn mediated by the G-CSF receptor, which is expressed in

neuron, astrocyte, microglia, and other immune cells [15].

In our study, G-CSF receptor was also expressed in AFS

(data not shown). Therefore, an anti-apoptotic action

against implanted cells is one of the potential mechanisms

of G-CSF in relation to augmented neuroprotection.

Nerve injury initiates inflammatory response and indu-

ces expression of pro-inflammatory cytokines expression

such as TNF-a, IL-1b, and IFN-c [32, 33]. Inflammatory

cells and inflammatory mediators not only cause tissue

damage and second wave injury but also play a role in the

regenerative process. In consideration of cell transplanta-

tion, an alternative role of inflammatory cytokines is to be

an important determinant for the survival and fate of

implanted cells. In our study, a significant accumulation of

inflammatory cells was detected in the injured sites after

nerve crush injury (Fig. 6). The injured nerve tissues pro-

duced elevated levels of pro-inflammatory cytokines,

including TNF-a, IL-1b, IL-6, and IFN-c (Fig. 7). The

over-activated inflammatory response in the injured nerve

tissues was associated well with the deficit of nerve func-

tion (Figs. 1 and 2), pathophysiological change (Figs. 3

and 4), and apoptosis of implanted AFS (Fig. 5). Studies

have shown that G-CSF possesses immunomodulatory

effect [15, 16]. In this study, the administration of G-CSF

attenuated crush injury-induced inflammatory cell accu-

mulation (Fig. 6) and the production of pro-inflammatory

cytokine production such as TNF-a, IL-1b, and IFN-c(Fig. 7). This immunosuppressive effect of G-CSF was

Fig. 6 Determination of

inflammatory cells. The nerve

tissues were retrieved 7 days

after injury and were subjected

to immunohistochemistry with

antibody against CD68 in four

treatment groups, (a) Normal

(b) Crush (c) G-CSF (d) AFS

(E) AFS + G-CSF.

Quantitative analysis of

inflammatory cells was depicted

in (f). *** P \ 0.001 versus

crush group, n = 6. Bar

length = 50 lm

Fig. 7 Determination of pro-inflammatory cytokines. The left sciatic

nerve of rats was injured by crushing. The injured nerves (10 mm) in

different treatment groups were retrieved 7 days after injury and

subjected to ELISA for the determination of TNF-a, IL-1b, IL-6, and

IFN-c. *** P \ 0.001, n = 5. P value in crush group was determined

relative to normal group and P values in G-CSF, AFS, and

G-CSF + AFS were determined relative to the crush group

Neurochem Res (2009) 34:518–527 525

123

paralleled to histological and functional improvement

(Figs. 1–4) and decreased apoptosis in implanted AFS

(Fig. 5). Cells potentially responsible for the secretion of

pro-inflammatory cytokines include intrinsic cells such as

Schwann cells, fibroblast, or resident macrophage and

extrinsic cells such as neutrophil or migrating macrophage

[24, 34]. The immunosuppressive effect could be accom-

plished by down-regulated chemoattraction, activation, or

gene induction. In addition, the prevention of early massive

destruction could alleviate the initiation and progression of

inflammation. IL-6, a multipotent cytokine possessing pro-

inflammatory and anti-inflammatory effects has been

shown to be involved in cell proliferation, survival, dif-

ferentiation, and death [35]. In this study, the elevation of

IL-6 production after crush injury was not attenuated by

G-CSF (Fig. 7).Therefore, the characteristics of cytokine

production after crush injury and the immunomodulatory

effect of G-CSF require further investigation.

Previously, we reported that transplantation of AFS

improved functional deficits caused by crush injury in rats

involving neurotrophin secretion [4, 5]. In this study, AFS

alone had little effect against crush injury-induced

inflammatory cell accumulation (Fig. 6), and pro-inflam-

matory cytokine production (Fig. 7). The concomitant

treatment of G-CSF and AFS augmented functional

improvement (Figs. 1–4). However, the immunosuppres-

sive effect was not escalated by the combination of G-CSF

and AFS as compared to G-CSF alone (Figs. 6 and 7). It

has been shown that mesenchymal stem cells harbor an

immuomodulatory effect, mainly by regulating T cells

[36]. In this study, we did not detect paramount counts of

lymphatic cells over the injured area (data not shown).

Thus, the absence of lymphatic cells within injured nerve

tissues might partly explain the little effect of AFS against

inflammation.

Conclusion

The combination of G-CSF and AFS potentiated peripheral

nerve regeneration. Immunosuppressive effect was one of

G-CSF related neuroprotective mechanisms. Administra-

tion of G-CSF exerted an anti-apoptotic effect on the

injured cells including the implanted AFS in sciatic nerve

injury. Therefore, the additional effect of G-CSF on AFS

against nerve injury can be attributed to anti-inflammatory/

anti-apoptotic effects directly protecting nerve tissues from

injury or indirectly augmenting the action of AFS.

Acknowledgements This study was supported by grants from

TCVGH-964906D and NSC 96–2314-B-075A-001, Taiwan. This

statistical analysis is supported by the biostatistics task force of

Taichung Veterans General Hospital, Taiwan, ROC.

References

1. Frostick SP (1995) The physiological and metabolic conse-

quences of muscle denervation. Int Angiol 14:278–287

2. Pollock M (1995) Nerve regeneration. Curr Opin Neurol 8:354–

358. doi:10.1097/00019052-199510000-00005

3. Murakami T, Fujimoto Y, Yasunaga Y, Ishida O, Tanaka N, Ikuta

Y et al (2003) Transplanted neuronal progenitor cells in a

peripheral nerve gap promote nerve repair. Brain Res 974:17–24.

doi:10.1016/S0006-8993(03)02539-3

4. Pan HC, Yang DY, Chiu YT, Lai SZ, Wang YC, Chang MH et al

(2006) Enhanced regeneration in injured sciatic nerve by human

amniotic mesenchymal stem cell. J Clin Neurosci 13:570–575.

doi:10.1016/j.jocn.2005.06.007

5. Pan HC, Cheng FC, Chen CJ, Lai SZ, Lee CW, Yang DY et al

(2007) Post-injury regeneration in rat sciatic nerve facilitated by

neurotrophic factors secreted by amniotic fluid mesenchymal

stem cells. J Clin Neurosci 14:1089–1098. doi:10.1016/j.jocn.

2006.08.008

6. Shen ZL, Lassner F, Becker M, Walter GF, Bader A, Berger A

(1999) Viability of cultured nerve grafts: an assessment of pro-

liferation of Schwann cells and fibroblasts. Microsurgery 19:

356–363. doi :10.1002/(SICI)1098-2752(1999)19:8\356::AID-

MICR2[3.0.CO;2-N

7. Dezawa M, Takahashi I, Esaki M, Takano M, Sawada H (2001)

Sciatic nerve regeneration in rats induced by transplantation of in

vitro differentiated bone-marrow stromal cells. Eur J Neurosci

14:1771–1776. doi:10.1046/j.0953-816x.2001.01814.x

8. Lyman GHaS M (2007) Granulocyte colony-stimulating factors:

finding the right indication. Curr Opin Oncol 19:299–307. doi:

10.1097/CCO.0b013e3281a3c0ba

9. Nishio Y, Koda M, Kamada T, Someya Y, Kadota R, Mannoji C

et al (2007) Granulocyte colony-stimulating factor attenuates

neuronal death and promotes functional recovery after spinal cord

injury in mice. J Neuropathol Exp Neurol 66:724–731. doi:

10.1097/nen.0b013e3181257176

10. Koda M, Nishio Y, Kamada T, Someya Y, Okawa A, Mori C et al

(2007) Granulocyte colony-stimulating factor (G-CSF) mobilizes

bone marrow-derived cells into injured spinal cord and promotes

functional recovery after compression-induced spinal cord injury

in mice. Brain Res 1149:223–231. doi:10.1016/j.brainres.2007.

02.058

11. Solaroglu I, Tsubokawa T, Cahill J, Zhang JH (2006) Anti-

apoptotic effect of granulocyte-colony stimulating factor after

focal cerebral ischemia in the rat. Neuroscience 143:965–974.

doi:10.1016/j.neuroscience.2006.09.014

12. Meuer K, Pitzer C, Teismann P, Kruger C, Goricke B, Laage R

et al (2006) Granulocyte-colony stimulating factor is neuropro-

tective in a model of Parkinson’s disease. J Neurochem 97:675–

686. doi:10.1111/j.1471-4159.2006.03727.x

13. Schabitz WR, Kollmar R, Schwaninger M, Juettler E, Bardutzky

J, Scholzke MN et al (2003) Neuroprotective effect of granulo-

cyte colony-stimulating factor after focal cerebral ischemia.

Stroke 34:745–751. doi:10.1161/01.STR.0000057814.70180.17

14. Pan HC, Cheng FC, Lai SZ, Yang DY, Wang YC, Lee MS (2008)

Enhanced regeneration in spinal cord injury by concomitant

treatment with granulocyte colony-stimulating factor and neuro-

nal stem cell. J Clin Neurosci 15:656–664. doi:10.1016/j.jocn.

2007.03.020

15. Hartung T (1998) Anti-inflammatory effects of granulocyte col-

ony-stimulating factor. Curr Opin Hematol 5:221–225. doi:

10.1097/00062752-199805000-00013

16. Xiao BG, Lu CZ, Link H (2007) Cell biology and clinical

promise of G-CSF: immunomodulation and neuroprotection.

J Cell Mol Med 11:1272–1290

526 Neurochem Res (2009) 34:518–527

123

17. Schneider A, Kruger C, Steigleder T, Weber D, Pitzer C, Laage R

et al (2005) The hematopoietic factor G-CSF is a neuronal ligand

that counteracts programmed cell death and drives neurogenesis.

J Clin Invest 115:2083–2098. doi:10.1172/JCI23559

18. Tsai MS, Hwang SM, Tsai YL, Cheng FC, Lee JL, Chang YJ

(2006) Clonal amniotic fluid-derived stem cells express charac-

teristics of both mesenchymal and neural stem cells. Biol Reprod

74:545–551. doi:10.1095/biolreprod.105.046029

19. Taskinen HS, Olsson T, Bucht A, Khademi M, Svelander L,

Roytta M (2000) Peripheral nerve injury induces endoneurial

expression of IFN-gamma, IL-10 and TNF-alpha mRNA. J

Neuroimmunol 102:17–25. doi:10.1016/S0165-5728(99)00154-X

20. Moertel CA, Stupca PJ, Dewald GW (1992) Pseudomosacism,

true mosaicism, and maternal cell contamination in amniotic fluid

processed with in situ culture and robotic harvesting. Prenat Di-

agn 12:671–683. doi:10.1002/pd.1970120808

21. Syroid DE, Maycox PJ, Soilu-Hanninen M, Petratos S, Bucci T,

Burrola P et al (2000) Induction of postnatal schwann cell death

by the low-affinity neurotrophin receptor in vitro and after

axotomy. J Neurosci 20:5741–5747

22. Aminoff MJ (1999) Electrodiagnosis in clinical neurology, vol 4,

4th edn. Churchill Livingstone, New York, pp 257–263

23. Rosen JM, Pham HN, Hentz VR (1989) Fascicular tubulization: a

comparison of nerve experimental nerve repair technique in the

cat. Ann Plast Surg 22:467–478. doi:10.1097/00000637-1989060

00-00002

24. Omura K, Ohbayashi M, Sano M, Omura T, Hasegawa T, Nagano

A (2004) The recovery of blood-nerve barrier in crush nerve

injury—a quantitative analysis utilizing immunohistochemistry.

Brain Res 1001:13–21. doi:10.1016/j.brainres.2003.10.067

25. Chiavegato A, Bollini S, Pozzobon M, Callegari A, Gasparotto L,

Taiani J et al (2007) Human amniotic fluid-derived stem cells are

rejected after transplantation in the myocardium of normal,

ischemic, immuno-suppressed or immuno-deficient rat. J Mol

Cell Cardiol 42:746–759. doi:10.1016/j.yjmcc.2006.12.008

26. Moore KAaL IR (2006) Stem cells and their niches. Science

311:1880–1885. doi:10.1126/science.1110542

27. Hartung T, Docke WD, Gantner F, Krieger G, Sauer A, Stevens P

et al (1995) Effect of granulocyte colony-stimulating factor

treatment on ex vivo blood cytokine response in human volun-

teers. Blood 85:2482–2489

28. Dong FaL AC (2000) Activation of Akt kinase by granulocyte

colony-stimulating factor (G-CSF): evidence for the role of a

tyrosine kinase activity distinct from the Janus kinases. Blood

95:1656–1662

29. Chakraborty A, White SM, Schaefer TS, Ball ED, Dyer KF,

Tweardy DJ (1996) Granulocyte colony-stimulating factor acti-

vation of Stat3 alpha and Stat3 beta in immature normal and

leukemic human myeloid cells. Blood 88:2442–2449

30. Watson FL, Heerssen HM, Bhattacharyya A, Klesse L, Lin MZ,

Segal RA (2001) Neurotrophins use the Erk5 pathway to mediate

a retrograde survival response. Nat Neurosci 4:981–988. doi:

10.1038/nn720

31. Noronha A, Toscas A, Jensen MA (1992) Contrasting effects of

alpha, beta, and gamma interferons on nonspecific suppressor

function in multiple sclerosis. Ann Neurol 31:103–106. doi:

10.1002/ana.410310119

32. Taskinen HS, Olsson T, Bucht A, Khademi M, Svelander L,

Roytta M (2000) Peripheral nerve injury induces endoneurial

expression of IFN-gamma, IL-10 and TNF-alpha mRNA. J

Neuroimmunol 102:17–25. doi:10.1016/S0165-5728(99)00154-X

33. Shamash S, Reichert F, Rotshenker S (2002) The cytokine net-

work of Wallerian degeneration: tumor necrosis factor-alpha,

interleukin-1alpha, and interleukin-1beta. J Neurosci 22:3052–

3060

34. Al-Shatti T, Barr AE, Safadi FF, Amin M, Barbe MF (2005)

Increase in inflammatory cytokines in median nerves in a rat

model of repetitive motion injury. J Neuroimmunol 167:13–22.

doi:10.1016/j.jneuroim.2005.06.013

35. Hodge DR, Hurt EM, Farrar WL (2005) The role of IL-6 and

STAT3 in inflammation and cancer. Eur J Cancer 41:2502–2512.

doi:10.1016/j.ejca.2005.08.016

36. Krampera M, Pasini A, Pizzolo G, Cosmi L, Romagnani S,

Annunziato F (2006) Regenerative and immunomodulatory

potential of mesenchymal stem cells. Curr Opin Pharmacol

6:435–441. doi:10.1016/j.coph.2006.02.008

Neurochem Res (2009) 34:518–527 527

123