以調控缺氧誘發因子(HIF-1α)的觀點探討低能雷射治療慢性發炎及疼痛的分生機制

1

中 國 醫 藥 大 學

1

專題研究計畫成果報告

2 3

計畫名稱：以調控缺氧誘發因子(HIF-1α)的觀點探討

4

低能雷射治療慢性發炎及疼痛的分生機

5

制

6 已經發表於 7

Hsieh YL, Chou LW, Chang PL, Yang CC, Kao MJ, Hong CZ. Low-level laser therapy alleviates 8

neuropathic pain and promotes function recovery in rats with chronic constriction injury-possible 9

involvements in hypoxia-inducible factor 1α (HIF-1α). J Comp Neurol. 2012 Feb 20. doi: 10

10.1002/cne.23072. [Epub ahead of print] 11

計畫編號：CMU99-TC-23

12 13

執行期限：100 年 04 月 01 日至 101 年 03 月 31 日

14

單位名稱：物理治療學系

15

主持人：謝悅齡

16 17

中 華 民 國

101 年 04 月 14 日

18

2

Abstract

19

Background. Nerve inflammation plays an important role in the development and progression 20

of neuropathic pain after chronic constrictive injury (CCI). Recent studies explored 21

hypoxia-inducible factor 1α (HIF-1α) in the process of inflammation. Low-level laser therapy 22

(LLLT) has been suggested to benefit treatment of pain disorders, but few data directly support 23

LLLT for neuropathic pain. Objective. We investigated the effect of LLLT on accumulation of 24

hypoxia-inducible factor-1 alpha (HIF-1α), proinflammatory cytokines tumor necrosis factor-α 25

(TNF-α), and interleukin-1β (IL-1β) for controlling neuropathic pain, as well as on activation 26

of vascular endothelial growth factor (VEGF) and nerve growth factor for promoting 27

functional recovery in rat model of CCI. Methods. CCI was induced by placing four loose 28

ligatures around the sciatic nerve of rats. LLLT (660 nm, 9 J/cm2) at CCI sites was performed 29

after 7 days of CCI. Effects of LLLT in CCI animals were determined by measuring 30

mechanical paw withdrawal threshold (MPWT), sciatic, tibial and peroneal function indexes 31

(SFI, TFI and PFI), and histopathological and immunoassay analyses. Results. Our results 32

demonstrated that LLLT significantly improved MPWT, SFI, TFI and PFI after CCI. LLLT 33

also significantly reduced overexpressions of HIF-1α, TNF-α and IL-1β and increased the 34

amounts of VEGF, NGF and Schwann cells. Conclusions. LLLT can modulate HIF-1α activity 35

and may represent a novel, clinically applicable therapeutic approach for improvement of 36

tissue hypoxia/ischemia and inflammation in nerve entrapment neuropathy as well as for 37

promotion of nerve regeneration, which may lead to sufficient morphologic and functional 38

recovery of the peripheral nerve. 39

40

Key Words: Chronic constrictive injury－Low-level laser therapy－Hypoxia-inducible factor

41

1α－Neuropathic pain－Functional recovery

3

Introduction

43

Neuropathic pain is a common sequela initiated by a primary lesion of the peripheral or 44

central nervous system (Baron, 2000, Zimmermann, 2001). In previous studies, the relationship 45

between proinflammatory cytokines, such as tumor necrosis factor-α (TNF-α), interleukin 1 46

(IL-1) released by inflammatory cells on their activation and the development of hyperalgesia 47

and allodynia in neuropathic pain has been identified (Sommer and Kress, 2004, Sommer and 48

Schäfers, 2004, Li et al., 2011, Liou et al., 2011). These results support the notion that nerve 49

inflammation plays an important contributory role in the development and progression of 50

neuropathic pain. Experimentally, various animal models of peripheral neuropathy have been 51

developed. Chronic constriction injury (CCI) of the sciatic nerve with loose ligatures is the 52

most widely used model for peripheral neuropathy and neuropathic pain (Bennett and Xie, 53

1988, Kingery et al., 1993), simulating the clinical condition of chronic nerve compression as 54

occurs in nerve entrapment neuropathy or spinal root irritation by a lumbar disk herniation 55

(Zimmermann, 2001). 56

Hypoxia-inducible factor-1α (HIF-1α) is a transcription factor that is increased in 57

conditions of hypoxia, ischemia and inflammation (Fraisl et al., 2009). HIF-1α is also thought 58

to be essential in maintaining inflammatory processes by promoting the production of 59

proinflammatory cytokines, including TNF-α and IL-1β (Takeda et al., 2009). HIF-1α has been 60

identified as a pivotal transcription factor linking the inflammatory pathways (Dehne and 61

Brune, 2009). Inhibition and/or down-regulation of these molecules may exert anti-hypoxic 62

and anti-inflammatory effects. Therefore, inhibiting HIF-1α accumulation may be a novel 63

therapeutic strategy for neuropathic inflammation. 64

Many experimental and clinical studies have also reported positive effects of low-level 65

laser therapy (LLLT) for promoting the repair processes of peripheral nerve by increasing 66

vascular endothelial growth factor (VEGF) and nerve growth factor (NGF) secretions (Byrnes 67

et al., 2005, Gigo-Benato et al., 2005, Hou et al., 2008, Rochkind, 2009, Rochkind et al., 2009, 68

Gigo-Benato et al., 2010), and by inhibiting the inflammation through reduction of 69

pro-inflammatory cytokines (Albertini et al., 2007). However, to date, there is little evidence 70

directly supporting the anti-allodynia effects of LLLT in neuropathic pain. In this study, 71

therefore, the effects of LLLT on management of neuropathic pain after CCI in sciatic nerve of 72

rat were investigated and possible biological mechanisms through which LLLT may exert its 73

action on functional recovery of peripheral nerve were analyzed. We hypothesized that LLLT 74

can decrease pro-inflammatory cytokines, reduce HIF-1α accumulation, and then promote 75

expressions of VEGF and NGF in the sciatic nerve proximal to the site of CCI on improvement 76

of neuropathic pain and functional recovery. 77

78

MATERIALS AND METHODS

79

80

General Design

81

Neuropathy was induced in all animals by CCI surgery. After surgery, animals (n=40) 82

were divided randomly into four groups (Figure 1) based on the nerve surgery and treatment 83

administration: (1) the CL group (n=10), which consisted of CCI animals that received LLLT; 84

(2) CsL group (n=10), which consisted of CCI animals that received sham-irradiated LLLT; (3) 85

sCL group (n=10), which consisted of sham-operated CCI animals that received LLLT; and (4) 86

sCsL group (n=10), which consisted of sham-operated CCI animals that received 87

sham-irradiated LLLT. Treatments of LLLT or sham-irradiation were given for consecutive 7 88

days. The evaluation instruments were mechanical paw withdrawal threshold (MPWT), sciatic 89

functional index (SFI), tibial functional index (TFI), peroneal functional index (PFI), histology, 90

4 immunohistochemistry and immunoassays. Pain and functional assessments were performed 91

the day before (pre-op, at day 0), immediately after operation (post-op, at day 1), at 7 days (7d 92

post-op, at day 7) after surgery and after the 7-day treatment (post-tr, at day 14). Animals were 93

sacrificed for assessments of histopathology and immunoassays the day after completing the 94

treatments. A flow diagram of the experimental design is presented in Figure 1. 95

96

Animals

97

Experiments were performed on adult male Sprague–Dawley rats (SD, 250 to 300 g, 98

purchased from BioLASCO Co., Ltd, Taiwan). Ambient temperature was maintained at 22 to 99

24 °C and the animals were kept on an artificial 12-h light–dark cycle in the Animal Center of 100

China Medical University. The light period began at 7:00 a.m. with food and water available 101

ad libitum up to the time of testing. Efforts were made to minimize discomfort and reduce the 102

number of animals used. The ethical guidelines of the International Association for Study of 103

Pain in Animals were followed (Zimmermann, 1983). All experimental procedures were 104

approved by the China Medical University Committee on Animal Care and Use. 105

106

Chronic Constriction Injury of Sciatic Nerve

107

Following the procedure originally proposed by Bennet and Xie (Bennett and Xie, 1988) 108

adapted for mice, CCI of sciatic nerve was used as the model of peripheral nerve injury for 109

evoking neuropathic pain symptoms. Surgery was performed under anesthesia with 4% 110

isoflurane in liquid form for inhalation (AErrane, Baxter Healthcare of Puerto Rico, PR). Using 111

a double-headed operating microscope, the sciatic nerve on one randomly selected side was 112

exposed by skin incision along the femur and separation of biceps femoris and superficial 113

gluteal muscles. At the middle third of the sciatic nerve, four ligatures with 4-0 chromic gut 114

thread (Ethicon, USA) were tied loosely around the nerve with inter-ligation spacing of about 1 115

mm. The wound of muscle layers (with 4/0 reabsorbable suture, Ethicon, USA) and skin (with 116

3/0 non-reabsorbable suture, Ethicon, USA) were then sutured and closed to allow recovery. 117

Sham-operated CCI animals underwent the same procedures. Branches were dissociated and 118

without any lesion for comparison 119

. 120

Low-Level Laser Irradiation

121

Seven days after surgery, a continuous 660-nm Ga-Al-As diode laser (Aculas-Am series, 122

Multi-channel LLLT system; Konftec Corporation, Taipei, Taiwan) was used in this study. 123

After sterilization, the hand-held delivery probe was placed lightly on the skin surface directly 124

above the loose ligation sciatic nerve at 4 spots / per area. The spot size was approximately 0.2 125

cm2. The output power of the laser irradiation was 30 mW per session for 60 sec/ per spot for 7 126

consecutive days. The energy density was 9 J/cm2. The output of the equipment was routinely 127

checked by the Laser Check Power Meter (Coherent, Santa Clara, CA, USA). A similar 128

procedure was applied to the control group with sham-irradiated LLLT with the output power 129

of laser irradiation adjusted to 0. 130

131

Mechanical Allodynia

132

The assessment of mechanical allodynia was performed by a MPWT which was measured 133

by nociceptive thresholds to stimulate von Frey filaments at pre-op, post-op, 7d post-op and 134

post-tr. The test consisted of evoking a hind paw flexion reflex with a handheld force 135

transducer (electronic von Frey anesthesiometer, IITC Inc., CA, USA) adapted with a 0.5 mm2 136

5 polypropylene tip. In a quiet room, the rats were placed in acrylic cages (32 × 22 × 27 cm high) 137

with a wire grid floor for 15-30 min habituation prior to testing. The polypropylene tip was 138

perpendicularly applied to the central area of the hind paw with sufficient force to bend the 139

filaments into an “S” shape for 3-4 sec. The test consisted of poking a hind paw to provoke a 140

flexion reflex followed by a clear flinch response after paw withdrawal. Testing was initiated 141

with the filament corresponding to 20 log of force (g). The filaments were applied with a 142

gradual increase in pressure until a withdrawal reflex response was finally detected from the 143

animal. The response to this filament was defined if a series of weaker or stronger filaments 144

would be tested. The weakest filament able to elicit a response was taken to be the MPWT (g). 145

The intensity of the pressure was recorded and the final value for the response was obtained by 146

averaging five measurements. 147

148

Assessments of Functional Recovery

149

The degree of recovery was monitored by evaluating the rats’ walking patterns in order to 150

obtain SFI, TFI, and PFI according to the method described by Bain et al. (Bain et al., 1989). 151

Before the recording, a few conditioning trials were performed to accustom the animals to the 152

track. All animals underwent preoperative walking-track analysis. Briefly, the plantar surfaces 153

of both hind paws were wetted with red ink in order to obtain clear footprints, and they were 154

allowed to walk along a specially designed alley (84 cm length × 8.5 cm width) lined with 155

scaled paper. Recordings continued until five measurable footprints had been collected. The 156

data used for calculations were taken from the footprint as follows: (1) distance from the heel 157

to the third toe, the print length (PL); (2) distance from the first to fifth toe, the toe spread (TS); 158

and, (3) distance from the second to the fourth toe, the intermediary toe spread (ITS). All three 159

measurements were taken from the experimental (E) and normal (N) sides. Prints were then 160

calculated using the following formulae (Bain et al., 1989): (1) SFI = -38.3 ([EPL−NPL]/NPL) + 161

109.5 ([E_TS−N_TS]/N_{TS) + 13.3 ([}E_IT−N_IT]/N_{IT) - 8.8; (2) TFI = -37.2 ([}E_PL−N_PL]/N_{PL) + 104.4}

162

([ETS−NTS]/NTS) + 45.6 ([EIT−NIT]/NIT) - 8.8; (3) PFI = 174.9 ([EPL−NPL]/NPL) + 80.3 163

([ETS−NTS]/NTS) - 13.4. Values of these tests equal to -100 indicated total impairment of the 164

sciatic, posterior tibial and peroneal nerves, whereas SFI, TFI and PFI oscillating around 0 165

were considered to reflect normal function (Bain et al., 1989). 166

167

Sciatic Nerve Obtainment and Tissue Preparations

168

After completing the treatments at day 14, rats were sacrificed after being deeply 169

anaesthetized with saturated KCl (300 g/ml, i.p.), then sciatic nerve segment was harvested, 170

which included the four ligatures as well as 1 cm of sciatic nerve proximal to the site of CCI. 171

The biopsied nerve specimens were divided into two portions for histopathology and 172

immunoassays. For histopathological assessments, nerve specimens randomly selected from 5 173

animals of each group were fixed in 10% neutral formalin, and embedded in paraffin for 12 h 174

at room temperature. All of the biopsied nerve specimens obtained from each animal for 175

immunoassays were immediately frozen in liquid nitrogen and stored at −80℃ for later 176

homogenization and subsequent assay of cytokine and protein expression. The homogenization 177

buffer was freshly prepared by adding protease inhibitor (P8340 cocktail Sigma, NY, USA) to 178

T-PER™ Tissue Protein Extraction Reagent (Pierce Chemical Co., USA) and centrifuged for 179

40 min. The supernatant was extracted and stored at −80 °C. 180

181

Histopathological, Immunohistochemical and Immunofluorescent Stainings

182

The specimens were submitted to diafanization with xylene, then dehydrated by graded 183

ethanol, embedded in paraffin and cut in 4-μm-thick sections longitudinally using a microtome. 184

6 Ten consecutive longitudinal resections contiguous to a maximum diameter were chosen for 185

data collection and subsequent comparisons. Histopathologic changes were evaluated on 186

sections stained with hematoxylin and eosin (H&E, Muto Pure Chemicals Co., Ltd., Tokyo, Japan) to 187

determine infiltration of inflamed cells in nerves. Slides were examined by a light microscope 188

and photographed using the Automatic Photomicrographic System PM10SP (Olympus, PA, 189

USA). The area of inflamed cell and nerve nuclei was measured in a 200× magnification field 190

by an ImageScope program (Aperio, Vista, CA, USA). 191

For immunohistochemical staining, the slides of sciatic nerve sections were first incubated 192

overnight at 4℃ with the monoclonal mouse antibodies, including anti-HIF-1α (1:200, 193

Thermo, CA, USA), anti-monocytes/macrophages (ED1, 1:200, Millipore, CA, USA) primary 194

antibodies, with the polyclonal rabbit antibodies, including anti-Schwann cells (S100, 1:400, 195

DakoCytomation, Denmark) and anti-VEGF (1:200, Abbiotec, CA, USA) primary antibodies, 196

as well as with rabbit monoclonal anti-NGF-β (1:2500, Millipore, CA. USA) primary antibody. 197

After washing three times in PBS, the nerve sections were then incubated with biotinylated 198

goat anti-mouse and goat anti-rabbit IgG secondary antibody (Jackson ImmunoResearch 199

Laboratories, Inc., West Grove, PA, USA) for 1 hour at room temperature. Following washing 200

with phosphate buffer three times, sections were incubated with a streptavidin-horseradish 201

peroxidase conjugate (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA). 202

Finally, sections were visualized as brown precipitates yields using 3,3′-diaminobenzidine 203

(DAB, 0.2 mg/ml, Pierce, Rockford, IL, USA) as a substrate and then counterstained with 204

hematoxylin. Negative control sections received the same treatment without the addition of 205

primary antibody. Slides were examined at a minimum of five sections in the more 206

representative fields using a light microscope and then photographed. The area sizes of positive 207

nuclear and cytoplasmic staining cells for HIF-1α, ED1, S100, VEGF and NGF were measured 208

in a 200× magnification field using the ImageScope program (Aperio, Vista, CA, USA). Ten 209

fields of each slide were calculated and repeated three times for statistical analysis. Results are 210

expressed as the proportion (%) of positive immunoreactive area per total stained area. 211

To observe coexpression of HIF-1α with infiltrated inflammatory cells in the injured nerve, 212

we incubated the sections with rabbit polyclonal anti-HIF-1α (1:200, Santa Cruz Biotechnology, 213

CA, USA) and mouse monoclonal anti-monocytes/macrophages (ED1) (1:200, Millipore, CA, 214

USA) overnight at 4°C under gentle agitation. Sections were then incubated with the respective 215

secondary antibodies (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA), 216

goat anti-rabbit IgG fluorescein-conjugated (FITC, 1:1000) and goat anti-mouse IgG 217

rhodamine-conjugated (TRITC, 1:1000) secondary antibodies for 2 hours at room temperature. 218

Following washing with phosphate buffer three times, sections were incubated with a 219

streptavidin-horseradish peroxidase conjugate (Jackson ImmunoResearch Laboratories, Inc., 220

West Grove, PA, USA). Finally, the sections were washed three times in PBS and then 221

counterstained with 4′,6-diamidino-2-phenylindole (DAPI, Molecular Probes, Invitrogen 222

Corporation, Carlsbad, CA, USA) to reveal cell nuclei. Images were obtained using a 223

conventional fluorescence microscope (Fluoview X; Olympus, Tokyo, Japan). All of 224

quantitative image analyses were assessed by two independent observers who were blinded to 225

the origin of the sections to avoid bias from interobserver variability. 226

227

Enzyme-Linked Immunosorbent Assay

228

The amounts of TNF-α, IL-1β and BDNF concentrations in the supernatants were 229

determined using the DuoSet® ELISA Development kit (R&D Systems, Minneapolis, MN, 230

USA). Nerve extracts were incubated in 96-well plates coated with mouse anti-rat TNF-α and 231

goat anti-rat IL-1β. After washing at each step, biotinylated anti-rat TNF-α and anti-rat IL-1β 232

7 and then streptavidin-HRP were added and incubated in accordance with the manufacturer's 233

instructions. After washing, a NeA-Blue (Tetramethylbenzidine) Substrate solution (Clinical 234

Science Products, Inc., Mansfield, MA, USA) was added to each well. The enzyme reaction 235

was terminated by adding stop solution (2N H2SO4). The levels of TNF-α and IL-1β were

236

assessed by a reader (Thermo Scientific Multiskan EX, Finland) using a 450 nm filter and 237

normalized with an abundance of standard solution. Data were then analyzed using Ascent 238

Software (Thermo Scientific Ascent Software, Finland) and a four-parameter logistics curve-fit. 239

Data are expressed in pg/mg protein of duplicate samples. 240

241

Western Blot Analysis

242

Protein determination was performed by modified Lowry protein assays. Equal amounts 243

of protein were loaded and separated in 10% Tris-Tricine SDS-PAGE gels. The resolved 244

proteins were transferred onto PVDF membranes ((Millipore, Bedford, MA, USA). The 245

membranes were blocked in 5% non-fat milk for 1 hour at room temperature, and incubated 246

overnight at 4 °C with mouse monoclonal anti-HIF-1α (1:500, Novus Biologicals, CA, USA), 247

rabbit polyclonal anti-VEGF antibody (1:2500, Abbiotec, CA, USA), and rabbit monoclonal 248

anti-NGF-β (1:2500, Millipore, CA. USA) primary antibody. The blots were then incubated 249

with the horseradish peroxidase-conjugated goat anti-mouse and anti-rabbit IgG secondary 250

antibody (1:20000, Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA) for 1 251

hour at room temperature. Signals were finally visualized using enhanced chemiluminescence 252

detection system (Fujifilm LAS-3000 Imager, Tokyo, Japan) and the blots were exposed to 253

X-ray films. All Western blot analyses were performed at least three times, and consistent 254

results were obtained. Immunoreactive bands were analysed using a computer-based 255

densitometry Gel-Pro Analyzer (version 6.0, Media Cybernetics, Inc. USA). Grey levels, 256

obtained by densitometric analysis of immunoreactive bands, were normalized on β-actin. 257

258

Statistical Analysis

259

Results were averaged for each group and values were expressed as mean ± S.E.M. The 260

data obtained from MPWT, SFI, TFI and PFI were analyzed using mixed-design, two-way 261

repeated-measures ANOVA performed with group as a between-subjects factor and time as a 262

within-subjects factor. The Bonferroni adjustment was examined post hoc for multiple 263

comparisons at individual time points between groups. One-way ANOVA was performed for 264

comparison of individual group means for assessing parametric results of histopathology and 265

immunoassay. The Dunnett test was performed for multiple comparisons between experimental 266

and control groups at the post-tr time point. A P value of < .05 was considered statistically 267

significant. All data were analyzed using SPSS version 10.0 for Windows (SPSS Inc., IL, 268 USA). 269 270

RESULTS

271 272

Effects of Low-Level Laser Therapy on Mechanical Allodynia

273

After surgery, there were significant differences in MPWT among time points in each 274

group (P < .0001). MPWT was significantly decreased at post-op and 7d post-op conditions in 275

animals that received CCI when compared with that of the pre-op condition (both were P < 276

0.001). In animals that received sham-operated CCI, MPWT of post-op compared to that of 277

pre-op condition was significantly decreased (P < 0.0001), whereas there was no significant 278

difference between the 7d post-op and pre-op condition (P=0.36). There were also significant 279

8 differences among the four groups at each time point (all were P < 0.0001, Figure 2A).

280

At the post-tr time point, there was a significant difference in MPWT compared with that 281

of the 7d post-op condition in CL group (P < .0001), but there were no significant differences 282

compared with values obtained in the CsL (P=0.59), sCL (P=0.22) and sCsL (P=0.98) groups. 283

The significant differences in MPWT were shown among CL, CsL, sCL and sCsL groups after 284

treatments (P < .0001). Significantly higher MPWT existed after LLLT treatment in CL group 285

compared with those in CsL groups after sham-irradiated LLLT treatment (P < .0001). 286

However, no significant difference was observed between sCL and sCsL groups (P=0.98). 287

288

Effects of Low-Level Laser Therapy on Functional Recovery

289

After surgery, there were significant differences in SFI, TFI and PFI among time points in 290

each group. SFI, TFI and PFI values were around 0 at pre-op condition and decreased 291

significantly after surgery in all groups (P < .001). SFI and TFI were still significantly 292

decreased at 7d post-op condition in animals that received CCI when compared with those of 293

post-op (SFI: P=0.83; TFI: P=0.99), but PFI showed significant recovery (P < .0001). 294

However, in sham-operated CCI animals at 7d post-op condition, PFI values significantly 295

recovered and approached that of the pre-surgery condition (P = 0.99), and SFI and TFI were 296

significant increased compared with those of post-op conditions (both were P < .0001, Figure 297

2B-D). 298

At the post-tr time point, SFI, TFI and PFI values were significantly higher when 299

compared with those of 7d post-op in CL group (SFI: P=0.001; TFI: P=0.003; PFI=0.03), but 300

no significant differences were found in CsL (SFI: P=1.0; TFI: P=0.73; PFI: P=1.0). SFI, TFI 301

and PFI values in sCL and sCsL groups showed no significant difference from pre-op level (all 302

were P > .05). Significant differences in SFI, TFI and PFI were shown among CL, CsL, sCL 303

and sCsL groups (all were P < .0001). Significantly higher values of SFI, TFI and PFI existed 304

after LLLT treatment in CL group compared with those of sham-irradiation treatment in CsL 305

groups (SFI: P=0.001; TFI: P=0.004; PFI: P=0.002). 306

307

Effects of Low-Level Laser Therapy on

Inflammation

and Cytokines

308

The results of H&E study showed there was pronounced infiltration of immune cells at 309

the site of CCI injury as compared with the site of sham-operated CCI (Figure 3A, 3B, 3C, 3D). 310

The percentages of nuclei in nerve contents were significantly different among the four groups 311

(P < .0001). The percentage of nuclei was significantly decreased and showed less 312

inflammation and cell infiltration in CL groups when compared with CsL group (Figure 3G). 313

Similar results were found for ED1 immunoreactivity which showed significant increases in 314

CsL group, but was reduced in CL group (Figure 3E, 3F and 3H). 315

TNF-α and IL-1β of the sciatic nerve contents were significantly different among the four 316

groups (both were P < .0001). There were significantly higher levels of TNF-α and IL-1β in 317

CsL groups in comparison with those of sCsL and sCL groups (both were P < .0001). No 318

significant differences were observed between sCL and sCsL groups (P=1.0). There was a 319

significant reduction of these cytokines in the CL group when compared with CsL groups (P 320

< .0001), but no significant difference was found when compared with those of sCL (TNF-α: 321

P=0.29; IL-1β: P=0.39) or sCsL (TNF-α: P=0.33; IL-1β: P=0.28) groups (Figure 4).

322 323

Effects of Low-Level Laser Therapy on HIF-1

α

324

The expressions of HIF-1α immunoreactivity in sciatic nerves were significantly different 325

among the four groups (P < .0001). The results showed there were sparse HIF-1α-positive cells 326

in sCL and sCsL groups (Figure 5A, 5B) and no significant differences were found among 327

9

these groups (both were P > .05, Figure 5G). In the CsL group, overexpression of HIF-1α 328

immunoreactivity was observed and localized in both the nucleus and cytoplasm of the injured 329

samples at higher-power magnification (Figure 5C). The accumulation of HIF-1α-positive cells 330

was decreased significantly in CL group when compared with CsL group (P=0.006, Figure 5D). 331

Double staining with HIF-1α and ED1 showed the ED1 immunoreactive cells which were 332

morphologically consistent with macrophages, mainly by inflammatory infiltration of the 333

inflamed nerve coexpressed by the specific HIF-1α immunoreactivity. The number of double 334

positive cells was decreased in CL groups when compared with those in CsL group (Figure 5E 335

and 5J). The observed HIF-1α expressions were further supported at the protein level assay by

336

Western blotting. The levels of HIF-1α in sciatic nerve was shown as gray density percentages 337

(normalized on β-actin) in the form of a representative Western blotting (Figure 6H). The 338

protein levels of HIF-1α in sciatic nerve contents were significantly different among the four 339

groups (P < .0001). No significant differences were observed between sCL and sCsL groups (P 340

> .05). Significantly higher levels of HIF-1α level were found in CsL groups in comparison 341

with those of CL, sCsL and sCL groups (all were P < .0001). The protein levels of HIF-1α was 342

significantly decreased in CL group in comparison with CsL groups (P=0.006) and 343

approximated the levels of sCL control group (P=0.064). 344

345

Effects of Low-Level Laser Therapy on VEGF, NGF and Schwann Cells

346

At day 14 after CCI, the constitutive expressions of VEGF and NGF in sciatic nerves 347

were significantly different among the four groups (VEGF: P < .0001; NGF: P=0.003). There 348

were no significant differences of VEGF and NGF expression between sCL and sCsL groups

349

(both were P > .05). After CCI, the expressions of these factors in the injured sciatic nerve

350

were slightly increased in CsL group as shown in Figures 6A and 6D, but the difference was of 351

non-significant when compared with those of sCsL groups (NGF: P=0.9; VEGF: P=0.22). As

352

expected, our results demonstrated that there were significant increases of VEGF and NGF in

353

CL groups compared with those in CsL group (VEGF: P=0.009; NGF: P=0.002, Figure 6B, 6C,

354

6E and 6F).Furthermore, as demonstrated in Figure 6I and 6J, the observed VEGF and NGF

355

immunoreactive expressions could be further supported at the protein level by Western blotting.

356

The protein levels of VEGF and NGF in sciatic nerve contents were also significantly different 357

among the four groups (VEGF: P < .0001; NGF: P < .0001). No significant differences were 358

observed between sCL and sCsL groups (both were P=1.0). The protein levels of VEGF and 359

NGF in CsL group also showed a slight elevation over 14 days after CCI surgery but the 360

calculation was not significant when compared with those of sCsL groups (NGF: P=0.18; 361

VEGF: P=0.07). There were significant increases of levels of VEGF and NGF in CL group 362

when compared with those of CsL groups (VEGF: P=0.009; NGF: P=0.002). Using S100 363

immunohistochemistry for Schwann cells, the S100 expression was decreased in injured nerve 364

in CsL group (Figure 6G), but increased in CL group (Figure 6H). The S100 immunoreactivity 365

in sciatic nerve contents was also significantly different among the four groups (P < .0001). 366

There was a significant decrease in S100 expression in CsL group when compared with values 367

seen in CL (P=0.005), sCL (P=0.035) and sCsL (P=0.027) groups (Figure 6I). 368

369

DISCUSSION

370

371

In the current study, we demonstrated that 660nm-GaAlAs-LLLT at a dose of 9 J/cm2 372

significantly reduced neuropathic allodynia in CCI rats. Our results are similar to those of 373

previous reports demonstrating that Nd: YAG laser-applied rats that received soft tissue surgery 374

had significantly higher nociceptive thresholds of the hind paw compared with the controls on 375

10 the 7th postoperative day (Kara et al., 2010) and 830 nm-wavelength LLLT at doses of 4 and 8 376

J/cm2 over the surgical incision on the 3rd postoperative day was effective in reducing pain in 377

rats with sciatic nerve compression using catgut thread (Bertolini et al., 2011). In clinical 378

studies of carpal tunnel syndrome, there was a significant improvement in neuropathy-induced 379

pain and delay of nerve conduction in patients undergoing LLLT over the carpal tunnel area 380

(Elwakil et al., 2007) (Shooshtari et al., 2008). 381

Pain due to inflammation is characteristic of neuropathy (Sommer and Kress, 2004, 382

Sommer and Schäfers, 2004, Li et al., 2011, Liou et al., 2011). As previously described,

383

mediators released from infiltrated cells, such as TNF-α and IL-1, have been implicated

384

directly in neuropathic pain, chronic hyperalgesia, and allodynia (Wagner and Myers, 1996,

385

DeLeo et al., 1997). Based on our observations from CCI rats in this study, the infiltration cells 386

and the protein levels of TNF-α and IL-1β in damaged nerves were significantly increased in

387

the control group. It seems that the contribution of inflammation and pro-inflammatory

388

cytokines to neuropathic pain were predominantly observed in the late postinjury phases.Our

389

results are further supported by a recent study with CCI rat model which showed reduction of 390

MPWT was correlated with increases of TNF-α and IL-1β gene expression in sciatic nerve 391

(Okamoto et al., 2001). Our results also demonstrated the infiltration of inflamed cells and the

392

release of proinflammatory cytokines were significantly reduced after LLLT in comparison 393

with the sham-irradiated controls. This result is similar to findings of previous studies with a 394

rat model of carrageenan-induced inflammation (Albertini et al., 2008, Boschi et al., 2008). 395

Therefore, the alleviation of neuropathic pain treated with LLLT in this study was probably

396

due to the reduction of inflammation and pro-inflammatory cytokines of injured nerve tissue.

397

SFI, TFI and PFI described by Bain et al. (Bain et al., 1989) are well-established and are 398

useful techniques for quantitatively assessing a rat’s lower limb deficits and determining 399

lesion-induced changes in function in sciatic nerve and its muscular branches in the rat. 400

Therefore, footprints were obtained after CCI for evaluation of functional locomotor recovery 401

by means of the SFI, TFI and PFI in this study. Our results showed that the SFI, TFI and PFI 402

were significantly affected by CCI at proximal stump of sciatic nerve. Probably owing to 403

impairment of sciatic nerve function and pain induced by CCI, prints were found to be 404

abnormal with evidence of toe dragging and a more “slurred” print. The use of LLLT 405

significantly promoted functional recovery as evidenced by increases in the SFI, TFI and PFI. 406

These results are consistent with the findings of a previous study that demonstrated LLLT was 407

effective in promoting early functional recovery as indicated by the SFI (Barbosa et al., 2010). 408

A nerve constriction injury produces histopathologic changes similar to the manner in 409

which a ischemic nerve injury can produce hyperesthesia when it causes Wallerian 410

degeneration (Myers et al., 1993). These data suggest that the nerve ischemia itself may play 411

an important role in the development of the hyperesthesia and allodynia induced by nerve CCI 412

(Myers et al., 1993). In response to ischemic damage in nerve, involvement of the 413

ischemia-related gene HIF-1 has been reported (Goldenberg-Cohen et al., 2009). HIF-1 has 414

dual effects and can induce either cell survival or cell death (Semenza, 2000). Accumulation of 415

HIF-1α protein and increase of HIF-1 activity have been found to exist following inflammation, 416

probably induced by pro-inflammatory cytokines, i.e., IL-1 and TNF-α (Hellwig-Burgel et al., 417

2005, Dehne and Brune, 2009, Chou et al., 2011). HIF-1 also existed in macrophage to 418

optimize its innate immunity, control pro-inflammatory gene expression and influence cell 419

migration (Dehne and Brune, 2009). Our previous findings showed pain and infiltration of 420

inflamed cells can be reduced by reducing HIF-1α protein accumulation in an arthritic animal 421

model (Chou et al., 2011). An in vitro study demonstrated that impaired neurons can be 422

rescued to promote neurogenesis by stabilizing HIF-1α (Milosevic et al., 2009). Therefore, 423

11 stabilization of HIF-1α protein expression as a regulator of gene expression in tissues is

424

required for the establishment of normal physiological systems (Semenza, 2000). The results of 425

this study demonstrated that the accumulation of HIF-1α in damaged nerve tissues was 426

prominent in response to CCI and were suppressed after LLLT. LLLT also reduced HIF-1α 427

expression in macrophages which coordinate chronic inflammation and immune responses. 428

Our results are consistent with a recent study which employed a mouse infection model to 429

investigate wound healing and demonstrated that untreated lesions showed high 430

immunoreactivity for HIF-1 , whereas little immunoreactivity could be detected in 431

laser-treated lesions (Ferreira et al., 2009). We postulate that this finding may help to explain 432

the ability of laser radiation to eliminate HIF-1α accumulation and then stabilize its activity, 433

thereby stimulating aerobic cell metabolism, accelerating tissue repair and promoting 434

functional recovery. 435

Vascular alterations of peripheral nerves occurring after injury are well described. 436

Angiogenesis is an essential component of nerve re-growth, and regeneration of the endoneural 437

vasculature precedes the outgrowth of axons from the proximal stump (Hoke, 2006, Webber 438

and Zochodne, 2010). It is thought that VEGF, a potent growth factor for angiogenesis, also 439

plays an important role in proliferation of Schwann cells, nerve repair and motor performance 440

(Hobson et al., 2000, Pereira Lopes et al., 2011). Increased angiogenesis primarily takes place 441

in metabolically altered or in injured peripheral nerves (Samii et al., 1999). Moreover, 442

stabilization of HIF-1α in a mouse with diabetes enhances wound healing and increases VEGF 443

production (Mace et al., 2007). Our findings demonstrated that CCI rats with sensory 444

neuropathy expressed VEGF in sciatic nerves. LLLT could further facilitate a prominent 445

increase of VEGF immunoreactivity compared with that obtained by sham-irradiation. This 446

effect was probably achieved through the stabilization of HIF-1α protein activity. In a study 447

which revealed similar findings to those of the present investigation it was shown that LLLI 448

could stimulate proliferation, increase VEGF secretion and facilitate myogenic differentiation 449

of bone marrow-derived mesenchymal stem cells (Hou et al., 2008), indicating that LLLT can 450

accelerate the healing process of tissues by stimulating VEGF. 451

NGF may act positively on the regeneration and growth of axonal processes to promote 452

the survival and integrity of sensory neurons and reverse distinct morphological and sensory 453

deficits and degeneration of myelin (Apfel et al., 1994). NGF also increases the levels of 454

VEGF in normal neural cells (Calza et al., 2001) and stimulates angiogenesis in animal models 455

under ischemic condition (Turrini et al., 2002). Local administration of anti-NGF serum can 456

block sprouting of collateral nerve fiber after sciatic nerve CCI in rats (Ro et al., 1998). 457

Improvement of sensory neuropathy and nerve fiber morphology could also be achieved by 458

application of NGF (Unger et al., 1998). In accordance with these previous findings, our results 459

showed that the elevation of NGF protein by LLLT was greater than that found in animals 460

treated with sham-irradiation. Moreover, in this study, an increase of S100 immunoreactivity 461

was also found after LLLT, indicating an increase in Schwann cells and these changes may be 462

attributed to improvement of functional motor status measured by SFI, TFI and PFI. Therefore, 463

improvement of neural function could also be achieved by application of LLLT which can 464

increase protein levels of NGF and VEGF to repair the myelin sheath in the injured nerve 465 tissues. 466 467

CONCLUSIONS

468 469

12 The aim of this study was to analyze the influence of injured nerve irradiation using a 470

660-nm Ga-Al-As diode laser on the neurorehabilitation of CCI sciatic nerves. The behavioral 471

evaluation of rats indicated that LLLT on CCI nerve tissues yielded much better recovery with 472

regard to motor function, pain behavior and histomorphometry than that achieved by 473

sham-irradiation. LLLT also reduced the protein levels of pro-inflammatory cytokines and 474

HIF-1α accumulation, and elevated levels of VEGF and NGF of the nerve tissue. These results 475

support our postulation that LLLT applied transcutaneously to the CCI nerve can suppress 476

inflammation-induced TNF-α, IL-1β and HIF-1α accumulation to control the neuropathic pain 477

and elevate the levels of VEGF and NGF in injured nerve thereby promoting functional 478

recovery and nerve regeneration. These results also indicate that the LLLT can modulate 479

HIF-1α activity and may represent a novel therapeutic approach as a clinically applicable 480

modality for improvement of tissue hypoxia/ischemia in nerve entrapment neuropathy as well 481

as for acceleration of the reinnervation rate of regenerated nerves, which may lead to sufficient 482

morphologic and functional recovery of the peripheral nerve. 483

484

ACKNOWLEDGEMENTS

485

The authors gratefully acknowledge the pathological and technical expertise of Mr. 486

Shih-Chung Chen in this study. This study was supported by a grant from China Medical 487

University (CMU99-TC-23), Taiwan. 488

489 490

13

REFERENCES

491 492

Albertini R, Villaverde AB, Aimbire F, Bjordal J, Brugnera A, Mittmann J, Silva JA, Costa M 493

(Cytokine mRNA expression is decreased in the subplantar muscle of rat paw subjected to 494

carrageenan-induced inflammation after low-level laser therapy. Photomed Laser Surg 495

26:19-24.2008). 496

Albertini R, Villaverde AB, Aimbire F, Salgado MA, Bjordal JM, Alves LP, Munin E, Costa 497

MS (Anti-inflammatory effects of low-level laser therapy (LLLT) with two different red 498

wavelengths (660 nm and 684 nm) in carrageenan-induced rat paw edema. J Photochem 499

Photobiol B 89:50-55.2007). 500

Apfel SC, Arezzo JC, Brownlee M, Federoff H, Kessler JA (Nerve growth factor 501

administration protects against experimental diabetic sensory neuropathy. Brain Res 502

634:7-12.1994). 503

Bain JR, Mackinnon SE, Hunter DA (Functional evaluation of complete sciatic, peroneal, and 504

posterior tibial nerve lesions in the rat. Plast Reconstr Surg 83:129-138.1989). 505

Barbosa RI, Marcolino AM, de Jesus Guirro RR, Mazzer N, Barbieri CH, de Cassia Registro 506

Fonseca M (Comparative effects of wavelengths of low-power laser in regeneration of sciatic 507

nerve in rats following crushing lesion. Lasers Med Sci 25:423-430.2010). 508

Baron R (Peripheral neuropathic pain: from mechanisms to symptoms. Clin J Pain 509

16:S12-20.2000). 510

Bennett GJ, Xie YK (A peripheral mononeuropathy in rat that produces disorders of pain 511

sensation like those seen in man. Pain 33:87-107.1988). 512

Bertolini GR, Artifon EL, Silva TS, Cunha DM, Vigo PR (Low-level laser therapy, at 830 nm, 513

for pain reduction in experimental model of rats with sciatica. Arq Neuropsiquiatr 514

69:356-359.2011). 515

Boschi ES, Leite CE, Saciura VC, Caberlon E, Lunardelli A, Bitencourt S, Melo DA, Oliveira 516

JR (Anti-Inflammatory effects of low-level laser therapy (660 nm) in the early phase in 517

carrageenan-induced pleurisy in rat. Lasers Surg Med 40:500-508.2008). 518

Byrnes KR, Wu X, Waynant RW, Ilev IK, Anders JJ (Low power laser irradiation alters gene 519

expression of olfactory ensheathing cells in vitro. Lasers Surg Med 37:161-171.2005). 520

Calza L, Giardino L, Giuliani A, Aloe L, Levi-Montalcini R (Nerve growth factor control of 521

neuronal expression of angiogenetic and vasoactive factors. Proc Natl Acad Sci U S A 522

98:4160-4165.2001). 523

Chou LW, Wang J, Chang PL, Hsieh YL (Hyaluronan modulates accumulation of 524

hypoxia-inducible factor-1 alpha, inducible nitric oxide synthase, and matrix 525

metalloproteinase-3 in the synovium of rat adjuvant-induced arthritis model. Arthritis Res Ther 526

13:R90.2011). 527

Dehne N, Brune B (HIF-1 in the inflammatory microenvironment. Exp Cell Res 528

315:1791-1797.2009). 529

DeLeo JA, Colburn RW, Rickman AJ (Cytokine and growth factor immunohistochemical 530

spinal profiles in two animal models of mononeuropathy. Brain Res 759:50-57.1997). 531

Elwakil TF, Elazzazi A, Shokeir H (Treatment of carpal tunnel syndrome by low-level laser 532

versus open carpal tunnel release. Lasers Med Sci 22:265-270.2007). 533

Ferreira MC, Gameiro J, Nagib PR, Brito VN, Vasconcellos Eda C, Verinaud L (Effect of low 534

intensity helium-neon (HeNe) laser irradiation on experimental paracoccidioidomycotic wound 535

healing dynamics. Photochem Photobiol 85:227-233.2009). 536

Fraisl P, Aragones J, Carmeliet P (Inhibition of oxygen sensors as a therapeutic strategy for 537

ischaemic and inflammatory disease. Nat Rev Drug Discov 8:139-152.2009). 538

14 Gigo-Benato D, Geuna S, Rochkind S (Phototherapy for enhancing peripheral nerve repair: a 539

review of the literature. Muscle Nerve 31:694-701.2005). 540

Gigo-Benato D, Russo TL, Tanaka EH, Assis L, Salvini TF, Parizotto NA (Effects of 660 and 541

780 nm low-level laser therapy on neuromuscular recovery after crush injury in rat sciatic 542

nerve. Lasers Surg Med 42:673-682.2010). 543

Goldenberg-Cohen N, Dadon-Bar-El S, Hasanreisoglu M, Avraham-Lubin BC, 544

Dratviman-Storobinsky O, Cohen Y, Weinberger D (Possible neuroprotective effect of 545

brimonidine in a mouse model of ischaemic optic neuropathy. Clin Experiment Ophthalmol 546

37:718-729.2009). 547

Hellwig-Burgel T, Stiehl DP, Wagner AE, Metzen E, Jelkmann W (Review: hypoxia-inducible 548

factor-1 (HIF-1): a novel transcription factor in immune reactions. J Interferon Cytokine Res 549

25:297-310.2005). 550

Hobson MI, Green CJ, Terenghi G (VEGF enhances intraneural angiogenesis and improves 551

nerve regeneration after axotomy. J Anat 197 Pt 4:591-605.2000). 552

Hoke A (Neuroprotection in the peripheral nervous system: rationale for more effective 553

therapies. Arch Neurol 63:1681-1685.2006). 554

Hou JF, Zhang H, Yuan X, Li J, Wei YJ, Hu SS (In vitro effects of low-level laser irradiation 555

for bone marrow mesenchymal stem cells: proliferation, growth factors secretion and 556

myogenic differentiation. Lasers Surg Med 40:726-733.2008). 557

Kara C, Suleyman H, Tezel A, Orbak R, Cadirci E, Polat B, Kara I (Evaluation of pain levels 558

after Nd: YAG laser and scalpel incisions: an experimental study in rats. Photomed Laser Surg 559

28:635-638.2010). 560

Kingery WS, Castellote JM, Wang EE (A loose ligature-induced mononeuropathy produces 561

hyperalgesias mediated by both the injured sciatic nerve and the adjacent saphenous nerve. 562

Pain 55:297-304.1993). 563

Li F, Fang L, Huang S, Yang Z, Nandi J, Thomas S, Chen C, Camporesi E (Hyperbaric 564

Oxygenation Therapy Alleviates Chronic Constrictive Injury-Induced Neuropathic Pain and 565

Reduces Tumor Necrosis Factor-Alpha Production. Anesth Analg.2011). 566

Liou JT, Liu FC, Mao CC, Lai YS, Day YJ (Inflammation confers dual effects on nociceptive 567

processing in chronic neuropathic pain model. Anesthesiology 114:660-672.2011). 568

Mace KA, Yu DH, Paydar KZ, Boudreau N, Young DM (Sustained expression of Hif-1alpha in 569

the diabetic environment promotes angiogenesis and cutaneous wound repair. Wound Repair 570

Regen 15:636-645.2007). 571

Milosevic J, Adler I, Manaenko A, Schwarz SC, Walkinshaw G, Arend M, Flippin LA, Storch 572

A, Schwarz J (Non-hypoxic stabilization of hypoxia-inducible factor alpha (HIF-alpha): 573

relevance in neural progenitor/stem cells. Neurotox Res 15:367-380.2009). 574

Myers RR, Yamamoto T, Yaksh TL, Powell HC (The role of focal nerve ischemia and 575

Wallerian degeneration in peripheral nerve injury producing hyperesthesia. Anesthesiology 576

78:308-316.1993). 577

Okamoto K, Martin DP, Schmelzer JD, Mitsui Y, Low PA (Pro- and anti-inflammatory 578

cytokine gene expression in rat sciatic nerve chronic constriction injury model of neuropathic 579

pain. Exp Neurol 169:386-391.2001). 580

Pereira Lopes FR, Lisboa BC, Frattini F, Almeida FM, Tomaz MA, Matsumoto PK, Langone F, 581

Lora S, Melo PA, Borojevic R, Han SW, Martinez AM (Enhancement of sciatic-nerve 582

regeneration after VEGF gene therapy. Neuropathol Appl Neurobiol.2011). 583

Ro LS, Chen ST, Tang LM (Extent of collateral sprouting of intact nerve fibers in rats depends 584

on the local availability of nerve growth factor. J Formos Med Assoc 97:247-251.1998). 585

Rochkind S (Phototherapy in peripheral nerve regeneration: From basic science to clinical 586

15 study. Neurosurg Focus 26:E8.2009).

587

Rochkind S, Geuna S, Shainberg A (Chapter 25: Phototherapy in peripheral nerve injury: 588

effects on muscle preservation and nerve regeneration. Int Rev Neurobiol 87:445-464.2009). 589

Samii A, Unger J, Lange W (Vascular endothelial growth factor expression in peripheral nerves 590

and dorsal root ganglia in diabetic neuropathy in rats. Neurosci Lett 262:159-162.1999). 591

Semenza GL (HIF-1: mediator of physiological and pathophysiological responses to hypoxia. J 592

Appl Physiol 88:1474-1480.2000). 593

Shooshtari SM, Badiee V, Taghizadeh SH, Nematollahi AH, Amanollahi AH, Grami MT (The 594

effects of low level laser in clinical outcome and neurophysiological results of carpal tunnel 595

syndrome. Electromyogr Clin Neurophysiol 48:229-231.2008). 596

Sommer C, Kress M (Recent findings on how proinflammatory cytokines cause pain: 597

peripheral mechanisms in inflammatory and neuropathic hyperalgesia. Neurosci Lett 598

361:184-187.2004). 599

Sommer C, Schäfers M (Mechanisms of neuropathic pain: the role of cytokines. Drug 600

Discovery Today: Disease Mechanisms 1:441-448.2004). 601

Takeda K, Ichiki T, Narabayashi E, Inanaga K, Miyazaki R, Hashimoto T, Matsuura H, Ikeda J, 602

Miyata T, Sunagawa K (Inhibition of prolyl hydroxylase domain-containing protein suppressed 603

lipopolysaccharide-induced TNF-alpha expression. Arterioscler Thromb Vasc Biol 604

29:2132-2137.2009). 605

Turrini P, Gaetano C, Antonelli A, Capogrossi MC, Aloe L (Nerve growth factor induces 606

angiogenic activity in a mouse model of hindlimb ischemia. Neurosci Lett 323:109-112.2002). 607

Unger JW, Klitzsch T, Pera S, Reiter R (Nerve growth factor (NGF) and diabetic neuropathy in 608

the rat: morphological investigations of the sural nerve, dorsal root ganglion, and spinal cord. 609

Exp Neurol 153:23-34.1998). 610

Wagner R, Myers RR (Endoneurial injection of TNF-alpha produces neuropathic pain 611

behaviors. Neuroreport 7:2897-2901.1996). 612

Webber C, Zochodne D (The nerve regenerative microenvironment: early behavior and 613

partnership of axons and Schwann cells. Exp Neurol 223:51-59.2010). 614

Zimmermann M (Ethical guidelines for investigations of experimental pain in conscious 615

animals. Pain 16:109-110.1983). 616

Zimmermann M (Pathobiology of neuropathic pain. Eur J Pharmacol 429:23-37.2001). 617

618 619 620

16

Legends of Figures

621

Figure 1. Experimental design of the sequence of events for the entire course of the 622

experiment. Evaluations include measurements of mechanical paw withdrawal threshold 623

(MPWT), sciatic, tibial and peroneal functional indexes (SFI, TFI and PFI) at the periods 624

before surgery (pre-op), immediately after surgery (post-op), 7 days after surgery (7d post-op) 625

and after treatment (post-tr) in the chronic constriction injury (CCI) animals treated with LLLT 626

(CL group) and sham-irradiation (CsL group) as well as in the sham-operated CCI animals 627

treated with LLLT (sCL group) and sham-irradiation (sCsL group). After the final treatment, 628

the animals were sacrificed for histology, immunohistochemistry (IHC), immunofluorescence 629

(IFC), Western blotting (WB) and ELISA assays. Solid and dotted lines denote the CCI and 630

sham-operation on the animals sciatic nerve, respectively. Solid and dotted borders of columns 631

denote the LLLT and sham-irradiation on the animals’ sciatic nerve, respectively. 632

633

Figure 2. Assessments of mechanical allodynia and functional recovery. Data were 634

calculated before surgery (pre-op), immediately after surgery (post-op), 7 days after surgery 635

(7d post-op) and after treatment (post-tr) in the chronic constriction injury (CCI) animals 636

treated with LLLT (CL group) and sham-irradiation (CsL group) as well as in the 637

sham-operated CCI animals treated with LLLT (sCL group) and sham-irradiation (sCsL group). 638

Each value represents the mean ± SEM in mechanical paw withdrawal threshold (MPWT) (A), 639

sciatic, tibial and peroneal functional indexes (SFI, TFI and PFI) (B-D). There were no 640

significant differences in any of the data between sCL and sCsL groups. After LLLT, the 641

MPWT, SFI, TFI and PFI were significantly increased when compared with those that received 642

sham-irradiated LLLT. # indicates there were significant differences among the four groups (P 643

< .05). * indicates there was a significant differences between CL and CsL groups (P < .05). 644

645

Figure 3. Assessments of inflammation in sciatic nerves by H&E staining and ED1 646

immunohistochemistry. Representative sections of the sciatic nerves obtained from chronic 647

constriction injury (CCI) animals treated with LLLT (CL group) and sham-irradiation (CsL 648

group) as well as in the sham-operated CCI animals treated with LLLT (sCL group) and 649

sham-irradiation (sCsL group). A-D indicate H&E staining for histopathology of sciatic nerves. 650

In rats of sCL and sCsL groups, the nerve tissues show normal histological appearance (A, B). 651

In rats of CsL group, there was even greater and massive inflammatory cells infiltration in 652

injured nerves (C). However, in rats of CL group, there was less infiltration in the nerves and 653

less accumulation of inflamed cells (D). For ED1 immunohistochemistry, there was more ED1 654

immunoreactivity (DAB-brown) in CsL group (E) than that in CL group (F). The quantitative 655

analysis of H&E and immunostaining for inflamed cells and ED1 are showed in F and G, 656

respectively. # indicates a statistically significant difference (P < .05) when data for CsL group 657

were compared with those of CL, sCsL and sCL groups and * indicates a significant difference 658

(P < .05) when data for CL groups were compared with data from CsL, sCL, sCsL groups. A 659

scale bar indicates 100 μm. Original magnification was ×400. 660

661

Figure 4. Results of TNF-α and IL-1β protein levels in the sciatic nerve. The levels of 662

TNF-α (A) and IL-1β (B) proteins were measured by ELISA in the sciatic nerves removed 663

from the chronic constriction injury (CCI) animals treated with LLLT (CL group) and 664

sham-irradiation (CsL group) as well as in the sham-operated CCI animals treated with LLLT 665

(sCL group) and sham-irradiation (sCsL group). # indicates a statistically significant difference 666

(P < .05) between CsL group and sCsL and sCL groups. # indicates a significant difference (P 667

< .05) between CL groups and CsL groups. 668

17 669

Figure 5. Results of HIF-1α expression in the sciatic nerve. Representative sections of the 670

sciatic nerves obtained from chronic constriction injury (CCI) animals treated with LLLT (CL 671

group) and sham-irradiation (CsL group) as well as in the sham-operated CCI animals treated 672

with LLLT (sCL group) and sham-irradiation (sCsL group). In rats of sCL and sCsL groups, 673

nerve tissue showed low HIF-1α expression (A, B). In rats of CsL group, there was even 674

greater and massive HIF-1α accumulation (DAB-brown) in injured nerves (C). But in rats of 675

CL group, there was less HIF-1α accumulation in nerves (D). Double staining with HIF-1α 676

(FITC-green), ED1 (TRITC-red) and DAPI (blue) by immunofluorescence showed there was 677

more co-expression of HIF-1α and ED1 (light red) in CsL groups (E) than that in CL groups 678

(F). The quantitative analysis of HIF-1α immunoreactivity for positive stained area is shown in 679

G. The protein levels of HIF-1α immunoblotting were significantly increased in CsL and 680

decreased in CL group (H). # indicates a statistically significant difference (P < .05) between 681

CsL group and sCsL and sCL groups. * indicates a significant difference (P < .05) for CL 682

compared with CsL groups. A scale bar indicates 100 μm. Original magnification was ×400. 683

684

Figure 6. Results of NGF, VEGF and S100 expressions in the sciatic nerve. Representative 685

sections of the sciatic nerves obtained from chronic constriction injury (CCI) animals treated 686

with LLLT (CL group) and sham-irradiation (CsL group) as well as in the sham-operated CCI 687

animals treated with LLLT (sCL group) and sham-irradiation (sCsL group). In rats of sCL and 688

sCsL groups, nerve tissue showed low NGF and VEGF expression (data not shown). In rats of 689

CsL group, there was slightly increased NGF (A) and VEGF (B) expression in injured nerves 690

compared with those in sham-operated CCI nerves. But in rats of CL group, the nerves 691

expressed more NGF and VEGF accumulation (D). For coexpression of ED1 and HIF-1α 692

immunofluorescence, there were more coexpressions (shown in light red) in CsL groups (E) 693

than those in CL groups (F). The quantitative analysis of HIF-1α immunoreactivity for positive 694

stained area is shown in G. The protein levels of HIF-1α immunoblotting showed a significant 695

increase in CsL and a decrease in CL group (H). # indicates a statistically significant difference 696

(P < .05) for CsL groups compared with CL, sCsL and sCL groups, and * indicates a 697

significant difference between CL group and CsL, sCsL and sCL groups (P < .05). A scale bar 698

indicates 100 μm. Original magnification was ×400. 699

700 701 702 703

0 1 7 14 CCI / sham-operation Laser / sham-irradiation Sacrificed Day

IHC

IFC

WB

7d post‐op

MPWT

SFI, TFI, PFI

pre‐op

MPWT

SFI, TFI, PFI

post‐op

MPWT

SFI, TFI, PFI

post‐tr

MPWT

SFI, TFI, PFI

0 1 7 14 Day

WB

ELISA

CL

C L

Group

CsL

sCL

sCsL

A

0

B

CsLCL

sCL sCsL

post‐tr pre‐op post‐op post‐op 7d

0 pre‐op post‐op post‐tr 7d post‐op 5 10 CsL sCsL W T ( g) _‐40 ‐20 sco re

#

_#

_#

*

15 20 MP W ‐80 ‐60

#

#

*

SFI s 25 post‐tr pre‐op post‐op post‐op 7d

0

‐100

#

#

#

0 pre‐op post‐op post‐tr 7d post‐op

C

D

‐40 ‐20 re ‐40 ‐20

*

re ‐80 ‐60

*

TFI sc o r ‐80 ‐60

#

#

*

PFI sc o r ‐100

#

#

#

*

‐100

#

A B

C D

E F

ED 1-po sit iv e s ta in ed a rea (% ) 0 10 20 30

sCsL sCL CsL

CL

0 2 4 6 8 10

sCsL sCL CsL

CL

A rea o f n uc lei (% )

G H

*

#

*

#

30000 30000 α

A B

(pg/ml) (pg/ml) 15000 20000 25000 15000 20000 25000 evels of TNF-α evels of IL-β

#

#

0 5000 10000 0 5000 10000 Protein l e Protein l e

*

*

sCsL

sCL

CsL

CL

sCsL

sCL

CsL

CL

A B

C D

E F

G H

0 0.1 0.2 0.3 0.4 0.5

sCsL

sCL

CSL

CL

0 5 10 15 20 25

sCsL

sCL

CsL

CL

Group

HIF-1α β-actin

sCsL sCL CsL CL

HI F -1 α / β -a c ti n ( fol d)

*

HI F -1 α -pos it iv e st ai n ed area (% )

*

#

#

0 5 10 15 20

sCsL

sCL

CsL

CL

0 5 10 15

sCsL

sCL

CsL

CL

0 20 40 60 80

sCsL

sCL

CsL

CL

E

S 100 -pos it iv e st ai n ed area (% ) VEG F -pos it iv e st ai n ed area (% ) NG F -pos it iv e st ai n ed area (% )

C

F

I

*

*

*

#

J K

0 0.2 0.4 0.6 0.8 1

sCsL

sCL

CSL

CL

VEG F / β -a c ti n (f ol d)

*

sCsL sCL CsL CL β-actin VEGF N G F / β -a c ti n ( fol d)

*

0 0.2 0.4 0.6 0.8 1

sCsL

sCL

CSL

CL

sCsL sCL CsL CL β-actin NGF

Group

Group

A

B

D

E

G

H