2 signaling in a mouse model of complex regional pain syndrome

Accepted Manuscript Astrocyte contributes to pain development via MMP2-JNK1/2 signaling in a mouse model of complex regional pain syndrome

Guogang Tian, Xin Luo, Chaoliang Tang, Xiang Cheng, Sookja Kim Chung, Zhengyuan Xia, Chi Wai Cheung, Qulian Guo PII: DOI: Reference:

S0024-3205(16)30683-X doi: 10.1016/j.lfs.2016.11.030 LFS 15096

To appear in:

Life Sciences

Received date: Revised date: Accepted date:

18 October 2016 23 November 2016 30 November 2016

Please cite this article as: Guogang Tian, Xin Luo, Chaoliang Tang, Xiang Cheng, Sookja Kim Chung, Zhengyuan Xia, Chi Wai Cheung, Qulian Guo , Astrocyte contributes to pain development via MMP2-JNK1/2 signaling in a mouse model of complex regional pain syndrome. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Lfs(2016), doi: 10.1016/j.lfs.2016.11.030

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Astrocyte contributes to pain development via MMP2-JNK1/2 signaling in a mouse model of complex regional pain syndrome Guogang Tian1;2, Xin Luo3;6, Chaoliang Tang3;6, Xiang Cheng2, Sookja Kim Chung4;5;6,

PT

Zhengyuan Xia3, Chi Wai Cheung3;5;6#, Qulian Guo1#

1

RI

Department of Anesthesiology, Xiangya Hospital of Central South University, Changsha, China;

2

SC

Department of Anesthesiology and Pain Medicine, Affiliated Haikou Hospital of Xiangya

Medical School, Central South University, Haikou, China; 3

NU

Department of Anaesthesiology, The University of Hong Kong, HKSAR, China.

4

MA

Department of Anatomy, The University of Hong Kong, HKSAR, China;

5

Research Center of Heart, Brain, Hormone and Healthy Aging, The University of Hong Kong,

Laboratory and Clinical Research Institute for Pain, The University of Hong Kong, HKSAR,

PT E

6

D

HKSAR, China.

CE

China.

AC

Correspondence to: Dr. Chi Wai Cheung,

Department of Anaesthesiology, The University of Hong Kong, Room 424, 4/F, Block K, Queen Mary Hospital, 102 Pokfulam, Hong Kong E-mail: [email protected] Tel: (852) 2255-3303 Fax: (852) 2855-1654 1

ACCEPTED MANUSCRIPT

Or to: Prof. Qulian Guo Department of Anesthesiology, Xiangya Hospital, Central South University, 87 Xiangya Road,

PT

Changsha City, 410008, Hunan Province, China.

RI

E-mail: [email protected]

SC

Tel: (86) 0731- 84327412

AC

CE

PT E

D

MA

NU

Fax: (86) 0731- 84327412

2

ACCEPTED MANUSCRIPT

Abstract Background: The activation of spinal glial cells (astrocyte and microglia) is reported in patient with complex regional pain syndrome (CRPS). However, the roles of spinal glial activities in the pathophysiology of CRPS are unclear. Here, we explored the roles of spinal astrocyte and

PT

microglia and the molecular mechanisms underlying CRPS using a mouse model of chronic

RI

post-ischemia pain (CPIP).

SC

Results: CPIP injury increased the level of glial fibrillary acidic protein (GFAP, reactive astrocyte biomarker), but had no significant impact on ionized calcium binding adaptor molecule 1 (IBA1,

NU

reactive microglia biomarker), in the ipsilateral dorsal horn on post-injury day (PID) 3 when the

MA

pain threshold started to reduce significantly. Astrocytic inhibition with fluorocitrate but not microglial inhibition with minocycline attenuated the development of allodynia in CPIP-injured

D

mice, which was concomitant with increased spinal levels of phosphorylated c-jun N-terminal

inhibitor)

prevented

PT E

kinase 1/2 (pJNK1/2) on PID 3. Furthermore, the intrathecal administration of SP600125 (JNK the

development

of

allodynia

in

CPIP-injured

mice.

Double

CE

immunofluorescence staining showed that pJNK1/2 was mainly co-localized with GFAP.

AC

Subsequently, increased levels of pJNK1/2 were reversed by intrathecal fluorocitrate. Furthermore, the level of spinal matrix metalloproteinase-2 (MMP2) was increased and mainly expressed in NeuN (neuron biomarker) on PID 3 in the CPIP-injured mice, while intrathecal APR 100 (MMP2 inhibitor) delayed the development of allodynia and decreased spinal levels of GFAP and pJNK1/2 on PID 3. Conclusion: This study shows that activation of astrocyte MMP2/JNK1/2 signaling pathway contributes to the pathogenesis of pain hypersensitivity in the CPIP model. 3

ACCEPTED MANUSCRIPT

Key words Chronic post-ischemia pain; Spinal cord; Astrocyte; c-jun N-terminal kinase; Matrix

AC

CE

PT E

D

MA

NU

SC

RI

PT

metalloproteinase-2

4

ACCEPTED MANUSCRIPT

Introduction Complex regional pain syndrome (CRPS) is a multifactorial pain disorder, with syndromes similar to the abnormalities in the somatosensory, autonomy and motor system, and the aetiology and pathogenesis of CRPS remains largely unknown. CRPS are classified into two subtypes: type-I

PT

without a loss in the major nerve and type-II with nerve lesion [5]. The chronic post-ischemia pain

RI

(CPIP) model is an animal model developed for the need of CRPS study [7]. In this model, the

SC

injury caused early hyperemia and edema followed by chronic pain, which mimicked two essential features of CRPS-I in human. Therefore, the CPIP model has now been used as a valuable

NU

approach to explore the pathophysiology of CRPS [7]. Using this model, it is suggested that the

MA

central mechanisms might contribute to the pathogenesis of CPIP [8; 20; 31].

D

Increasing body of studies indicates that glial cells are involved in the abnormal pain perception in

PT E

animals under different pathological conditions [6], and in the patients suffering from chronic pain [22]. Damage to the tissue or the nerve causes the change of glial cells from “normal” status to

CE

“reactive” status. Subsequently, these reactive glial cells synthesize and release cytokines and

AC

chemokines, which contribute to the neuroinflammation and central mechanisms of pathological pain [6; 25]. Previously, it has been shown that the levels of proinflammatory cytokines (like IL-1β and IL-6) in the cerebrospinal fluid (CSF) were enhanced in patients with CRPS [1; 2], which implicated the involvement of glia-related neuroinflammation in the pathophysiology of CRPS. Recently, Valle et al. reported that spinal astrocyte and microglia were activated in a patient with CRPS by detecting the glial markers in the autopsy materials [9]. However, the roles of glial activation in the central mechanisms of CRPS are unclear, which might limit the development of 5

ACCEPTED MANUSCRIPT the glia-targeted therapy for CRPS in patients.

Recently, we found that chemokine axis CXCL12/CXCR4 axis contributes to the development of pathological pain, including CPIP [24; 25]. In mice CPIP model, it was found that spinal levels of

PT

CXCL12 and glial fibrillary acidic protein (GFAP) were increased and co-localized on

RI

post-operative day (PID) 3, implicating that astrocyte-dependent CXCL12 might play an

SC

important role in the early stage of CPIP pathology [23]. Therefore, it is highly possible that astrocyte might contribute to the development of CPIP. As the roles and mechanisms of glial

NU

activation in the pathogenesis of CPIP, potentially CRPS in human, remain largely unclear, the

AC

CE

PT E

D

MA

goal of this study is to testify this hypothesis using mice CPIP model.

6

ACCEPTED MANUSCRIPT

2. Materials and Methods 2.1. Animals Animal experiments were approved by the Committee on the Use of Live Animals in Teaching and Research (CULATR) (permit No. 2610-11) and performed following the guidelines for the

PT

care and use of laboratory animals as established by the Laboratory Animal Unit (LAU) at the

RI

University of Hong Kong. In this study, adult male C57BL/6 wild-type mice (28-30 grams) were

SC

used. Mice were housed at 23 ± 3C, with a 12-hour light/12-hour dark cycle (lights on at 07:00) and the humidity (25%-45%). Animals were offered free access to water and food (Lab Diet 5012

MA

NU

(0.5% phosphorus, 1.0% calcium and 3.3 IU/g of vitamin D3)).

2.2. CPIP model

D

The CPIP injury produced swelling and mechanical allodynia in the hindpaw, mimicking the

PT E

clinical characters in patients with CRPS type-I [7; 27]. Animals were anesthetized with gaseous isoflurane and O2. Durometer O-rings (O-rings West) were placed around right hindlimb to

AC

CE

produce ischemia in mice. Three hours later, the rings were removed to initiate reperfusion.

2.3. Hindpaw volume In this study, the volume of mice hindpaw was measured with the U-shaped volume sensor (IITC) following the manufacturer‟s protocol.

2.4. Study drugs Fluorocitrate (2 μg per day, Sigma #F9634), minocycline (10 μg per day, Sigma # M9511), 7

ACCEPTED MANUSCRIPT SP600125 (5 μg per day, Sigma #S5567) and APR 100 (10 μg per day, Santa Cruz #203522) were freshly prepared in 1% Dimethyl Sulphoxide (DMSO) (diluted in saline solution) on the day of the experiment. We took DMSO (1%, diluted in saline solution) as the vehicle in this study.

PT

2.5. Intrathecal injection

RI

In this study, the single intrathecal injection was performed daily at 1 hour before the operation

SC

and up to PID 3. Animals were anesthetized by inhalation anesthesia with isoflurane and O2. A microliter syringe (Hamilton) with a 30-gauge needle (BD) was applied to make a spinal cord

NU

puncture. A total volume of 5 μl of drug(s) was delivered to the subarachnoid space between the

MA

L3 and L5 lumbar spinal cord. Successful injection was indicated by a tail configuration of the “S”

D

type or tail swinging immediately following the administration.

PT E

2.6. von Frey test

The paw withdrawal threshold (PWT) of mice was assessed by von Frey test, and the protocol was

CE

mentioned in our previous research [24]. Animals were placed on a metal mesh floor with a

AC

transparent plastic dome for nearly 30 minutes before the experiment. During the experiment, a series of von Frey filaments (IITC) was applied to the plantar surface of mice hindpaws. When the force was sufficient to bend the filaments into an „„S‟‟ shape and the mouse withdrew the hind paw from the filament, displayed value of this force was taken as the PWT of mice.

2.7. Immunohistochemistry Mice were deeply anesthetized with sodium pentobarbital and perfused with the phosphate 8

ACCEPTED MANUSCRIPT buffered saline (PBS), followed by 4% paraformaldehyde (PFA) in 0.1 M PBS via the cardiovascular system. L3-L5 lumbar spinal cords were collected from these mice and post-fixed in 4% PFA. Samples were dehydrated in 25% sucrose at 4 C overnight. Frozen samples in tissue freezing medium (Jung) were sliced longitudinally at 15 μm by a cryostat (Leica). Then sections

PT

were blocked by 4% goat serum in PBS containing 0.1% Triton X-100 (PBST) at room

RI

temperature for 2 hours and incubated with antibody(s) against GFAP (1:250, Abcam, #10062),

SC

ionized calcium binding adaptor molecule 1 (IBA1, 1:100, Abcam, #15690), NeuN (1:200, Abcam, #177487), phosphorylated c-jun N-terminal kinase (JNK) (1:50, Cell signaling, #4668) and/or

NU

matrix metalloproteinase-2 (MMP2, 1:200, Abcam, #37150) at 4 C overnight. Then, sections

MA

were washed by PBS and incubated with secondary antibody conjugated with Goat Anti-Rabbit IgG H&L (Alexa Fluor 488) (1:1000, Abcam, 150077) and/or Donkey Anti-Mouse IgG H&L

D

(Alexa Fluor 568) (1:1000, Abcam, 175472) for 2 hours at room temperature. DAPI (Vector) was

PT E

used to stain nuclear in sample sections. The immunoreactivity in these sections was captured with the confocal scanning microscope LSM 700 and LSM 710 (Zeiss), and the immunofluorescent

AC

CE

images were analyzed by Image-Pro Plus (Media Cybernetics).

2.8. Western blotting

Animals were euthanized with pentobarbital before the sample harvest. Ipsilateral L3-L5 lumbar spinal cord was quickly removed and homogenized in ice-cold RIPA lysis buffer. Protein samples were prepared following our previous protocol [24]. These samples were separated in 10% SDS-PAGE and transferred to PVDF membranes (Bio-Rad). The protein samples on the PVDF membrane were incubated with first antibodies against GFAP (1:2500, Abcam, #10062), IBA1 9

ACCEPTED MANUSCRIPT (1:2500, Abcam, #15690), phosphorylated JNK (1:1000, Cell signaling, #4668), total JNK (1:3000, Cell signaling, #9258), and GADPH (1:10000, Sigma, #G9545) as a loading control. The expression levels of targets were visualized with horseradish peroxidase (HRP)-conjugated secondary antibodies (Anti-mouse IgG, HRP-linked (1:2000, Cell signaling, #7076), Anti-rat IgG,

PT

HRP-linked (1:2000, Cell signaling, #7074)) and followed by the exposure to X-ray films (Kodak,

RI

USA). Gray scale of all bands in the scanned images (films) was determined by ImageJ software

SC

(NIH, USA) for the analysis.

NU

2.9. Statistical analysis

MA

The data were expressed as means ± SEM. The results from the immunohistochemical work and Western blotting test were analyzed with t-test. The results from the behavioral test were analyzed

AC

CE

PT E

D

with two-way ANOVA. In all cases, p < 0.05 was considered statistically significant.

10

ACCEPTED MANUSCRIPT

Results Mechanical allodynia and edema was induced in the CPIP model. As shown in Figure 1, the CPIP injury reduced PWT of the ipsilateral hindpaws from PID 3 and up to PID 15, as compared to the baseline (Figure 1A, p<0.001, n=6). Furthermore, the ipsilateral

PT

PWT increased on PID 15 as compared to that at PID 7 (Figure 1A, p<0.001, two-way ANOVA, F

RI

(5, 242) = 22.21, n=6). Moreover, the CPIP injury increased the volume of the ipsilateral hindpaw

SC

from 2 hours after the reperfusion and up to PID 1 comparing to the baseline (Figure 1B, p<0.001, two-way ANOVA, F (3, 32) = 13.86, n=5). Thus, the results indicated that the CPIP model was

NU

successfully established in this study. As the earliest start time for animal to develop mechanical

MA

allodynia was PID 3 in the CPIP model, we chose PID 3 as the time-point for the ensuring

D

mechanistic exploration.

PT E

Roles of astrocyte and microglia in the central mechanisms of pain hypersensitivity in the CPIP model.

CE

In the current study, we determined the spinal activation of glial cells in the CPIP model. In pain

AC

research, spinal GFAP and IBA1 are important anatomical indices of astrocyte activation [14] and microglia activation [30], respectively. Our immunohistochemical evidence showed that the levels of GFAP (Figure 2A and B, p<0.05, t test), but not IBA1 (Figure 2A and C, p>0.05, t test), were increased in the ipsilateral lumbar spinal cord on PID 3 in the CPIP model, comparing to that in sham-operated animals (n=3 in each group). There was not any change in GFAP or IBA1 in the contralateral dorsal horn (data not shown). Furthermore, the morphology of astrocyte was changed following the injury. Fluorocitrate is a selective astrocytic inhibitor at the low concentration, and 11

ACCEPTED MANUSCRIPT minocycline was a selective microglial inhibitor [3; 17; 33; 35]. As the molecular mechanisms for these two glial inhibitors remained unclear, they were just applied to study the effects of the glial activation on the pain processing, but not upstream mechanisms, in the current study. Here, the single intrathecal injection was performed daily at 1 hour before the operation and up to PID 3.

PT

Our behavioral study results showed that intrathecal fluorocitrate (2 μg per day), but not

RI

minocycline (10 μg per day), increased the PWT in the ipsilateral hindpaws of CPIP injured mice

SC

from PID 2 to 4 (Figure 2D, p<0.001, two-way ANOVA, F (12, 464) = 2.623). These glial inhibitors did not affect the PWT in the contralateral hindpaw in the CPIP model (Figure 2E,

NU

p>0.05, n=7-8 in each group). Taken together, our immunohistochemical and pharmacological

MA

evidence showed that astrocyte, but not microglia, contributed to the development of mechanical

D

allodynia in the CPIP model.

PT E

Roles of JNK in the central mechanisms for pain hypersensitivity in the CPIP model. Spinal JNK signaling was considered an astrocyte-specific pain pathway, and was activated in

CE

various pathological pain states [18]. However, the role of JNK in the central mechanisms of CPIP,

AC

potentially CRPS in human, remained unclear. Therefore, we determined the distribution and phosphorylation/activation of JNK in the spinal cord of CPIP-injured mice. Western blotting analysis showed that two isoforms of JNK, JNK1 and JNK2, were activated in the ipsilateral lumbar spinal cord on PID 3 (Figure 3A, p<0.05, t test). Furthermore, the role of spinal JNK in the central mechanisms of CPIP was substantiated by our pharmacological evidence, in which intrathecal JNK inhibitor (SP600125, 5 μg per day) partially prevented the development of mechanical allodynia in CPIP injured mice (Figure 3B, p<0.001, two-way ANOVA, F (3, 172) = 12

ACCEPTED MANUSCRIPT 39.9, n=7-8 in each group). The immunohistochemical evidence showed that spinal JNK1/2 was strongly activated on PID 3 and mainly co-localized with astrocytic biomarker GFAP (Figure 3C and D). Moreover, intrathecal fluorocitrate (2 μg per day) decreased the phosphorylation level of JNK1 (p<0.05) and JNK2 (p<0.001) (the ratio of phosphorylated JNK and total JNK) in the

PT

ipsilateral lumbar spinal cord in CPIP-injured mice (Figure 3E, t test). Taken together, our

RI

evidence showed JNK1/2 signaling contributed to the pathogenesis of pain hypersensitivity and

SC

was the intracellular signaling downstream of astrocyte activation in the CPIP model.

NU

Roles of MMP2 in the central mechanisms for pain hypersensitivity in the CPIP model.

MA

Matric metalloproteases (MMPs) were considered emerging pain modulators for their roles in regulating the neuroinflammation [19]. However, the role of MMP2 in the pathogenesis of CPIP

D

remained unknown. In this study, the immunohistochemical evidence showed that the spinal level

PT E

of MMP2 was increased on PID 3 (Figure 4A). It was further found that MMP2 was mainly co-localized with NeuN, but not GFAP, on PID 3 (Figure 4B and C). The role of MMP2 in the

CE

central mechanisms of CPIP was substantiated by our pharmacological evidence, in which

AC

intrathecal MMP2-specific inhibitor (APR 100, 10 μg per day) partially prevented the development of mechanical allodynia in CPIP-injured mice (Figure 4D, p<0.001, two-way ANOVA, F (3, 170) = 32.29). MMP2 was an upstream factor accounting for the astrocyte activation in a neuropathic pain model [19]. Therefore, we proposed that MMP2 would contribute to the activation of astrocyte and JNK1/2 in the CPIP model. Our immunohistochemical evidence showed that intrathecal APR 100 (10 μg per day) decreased the levels of GFAP (Figure 4E) and the phosphorylated JNK1/2 (Figure 4F, p<0.05, t test, n=3 in each group) in the ipsilateral lumbar 13

ACCEPTED MANUSCRIPT spinal cord of CPIP-injured mice on PID 3. Taken together, our evidence showed MMP2 activity contributed to the pathogenesis of pain hypersensitivity and was an upstream mechanism of

AC

CE

PT E

D

MA

NU

SC

RI

PT

astrocyte activation in the CPIP model.

14

ACCEPTED MANUSCRIPT

Discussion In the present study, we found that spinal astrocyte, but not microglia, was activated at the early stage of chronic pain development in the CPIP model, which was substantiated by the fact that the intrathecal inhibition of astrocyte activation, but not that of microglial activation, reversed CPIP

PT

injury-induced mechanical allodynia. Glial cells, including astrocyte and microglia, contribute to

RI

the central mechanisms of pathological pain through glial-neuronal and glial-glial crosstalk [4].

SC

Spinal astrocyte and microglia were activated and exerted pain-related function in the various pathological pain states, including neuropathy, inflammation and cancer [6; 26]. Using a novel

NU

imaging technique, Loggia et al. found that the levels of translocator protein (TSPO, biomarker of

MA

glial activation) were increased in the multiple brain regions (like thalamus and putative somatosensory representations of the lumbar spine) of patients with chronic low back pain, which

D

firstly provided the evidence of the glial activation under chronic pain condition in human [22].

PT E

Moreover, glial cells are activated in a pathology-dependent pattern. In some animal models of the neuropathy (like spinal and sciatic neuropathy) and the inflammation (like by complete Freund‟s

CE

adjuvant and formalin), microglia was responsible for the development of chronic pain while

AC

astrocyte for the maintenance [6; 26]. However, only the activation of astrocyte, but not microglia, was observed in chemotherapy-induced peripheral neuropathy [29]. In CRPS patients, the CSF levels of IL-6 and GFAP levels were increased [1; 2]. In the CPIP model, spinal levels of TNF-α, IL-1β and IL-6 were increased following ischemia-induced injury [23]. As glia is the major source of cytokines and chemokines in the CNS [6; 26], these findings implicated the involvement of glia-related neuroinflammation in the pathogenesis of CRPS. In the autopsy materials of a patient suffered from CRPS for 17 years, the activation of astrocyte was identified by observing the 15

ACCEPTED MANUSCRIPT elevated levels of GFAP [9]. In the CPIP model, it was found that astrocyte-dependent CXCL12 release contribute to the development of mechanical allodynia and neuroinflammation. Therefore, our findings firstly reported that reactive astrocyte, but not reactive microglia, contributed to the central mechanisms of pain hypersensitivity at the early stage of pathology in the CPIP model.

PT

These findings further implicated that astrocyte would be the potential cellular target for the

RI

pre-emptive treatment of CRPS. Notably, microglia activity was also identified by observing the

SC

elevated levels of CD-68 (biomarker of microglia) in the autopsy materials of CRPS patient [9]. In the current experiment setting, we found that microglia might not be involved in the development

NU

of CPIP. However, microglial mechanisms might be involved in the maintanence of CPIP or CPRS,

MA

which needs further exploration.

D

Here, we firstly reported that JNK1/2 contributed to the central mechanisms of pain

PT E

hypersensitivity and was the intracellular signaling molecule downstream of the astrocyte activation in the CPIP model. JNK is a stress-activated membrane of MAP kinase family, and JNK

CE

1 and JNK 2 contribute to the pathological pain processing via an astrocyte-specific manner [12;

AC

18] (in inflammatory [13], neuropathic [36] and cancer pain [11]). In the current study, our results showed that phosphorylated JNK1/2 was increased and mainly localized in spinal astrocyte in the CPIP model. Previously, increased spinal levels of TNF-α, IL-1β, IL-6, CXCL12 and GFAP in CPIP-injured mice indicated the role of astrocyte-induced neuroinflammation in the development of CPIP [23]. JNK signaling regulates astrocyte gene transcription in pain states via the activation of the transcription factor c-Jun or other factors [12]. Therefore, it is highly possible that JNK1/2 play an essential role in astrocyte-dependent cytokines and chemokines release in the CPIP model, 16

ACCEPTED MANUSCRIPT which needs further exploration.

Here, we found that the spinal expression of MMP2 was increased following the CPIP injury and that intrathecal MMP2 inhibitor prevented CPIP injury-induced allodynia and reversed the

PT

increases of GFAP and pJNK1/2 in the CPIP model. Multiple upstream factors accounted for the

RI

activation of spinal glial cells (such as microglia and astrocyte), such as substance P, calcitonin

SC

gene-related peptide, nitrite oxide and ATP [6]. However, the candidates that selectively trigger the spinal astrocyte activation remain poorly known [19]. MMPs are well known for their proteolytic

NU

actions in remodeling the extracellular matrix [28] and producing active cytokines (like IL-6) [32].

MA

Under neuropathic pain states, microglial activation contributed to the development of, while astrocytic activity contributed to the maintenance of the persistent pain [6]. Among the MMP

D

members, MMP9 evokes the spinal microglial activation selectively [19]. MMP2 was mainly

PT E

expressed in soma of spinal neurons in naïve animals, whereas the level of MMP2 is upregulated in astrocytes at the late stage of neuropathic pain [19]. Therefore, these findings implicated that

CE

MMP2 activated astrocyte and contributed to the development of CPIP. Our study firstly found

AC

that MMP2 was increased and mainly expressed by soma of spinal neurons on PID 3, which implicated that MMP2 was released via a neuron-dependent manner at the early stage of CPIP. This finding would provide insight into MMP2 and CPIP for future studies.

Our current study proposed the novel cellular (astrocyte) and molecular (JNK1/2 and MMP2) targets for the future research regarding the mechanism of CPIP, potentially CRPS in human. Currently, some central mechanisms-targeted therapeutic strategies have been used to manage 17

ACCEPTED MANUSCRIPT CRPS, such as opioid, NMDA receptor antagonists (like ketamine and dextromethorphan) and the epidural spinal cord stimulation [15]. However, none of them have been shown to be effective enough after the long-term use, due to decreased efficacy or adverse effects. Therefore, there is urgent need for the development of new treatment for CRPS. Numerous studies indicate that glial

PT

cells might be the ideal targets for the clinical pain therapy. However, in the first clinical trial, the

RI

glial inhibitor propentofylline failed to attenuate postherpetic neuralgia as compared to the placebo

SC

treatment [21]. The results of this trial remind us that the evidence of the glial activation and the temporal appearance of activated astrocyte and microglia in the pain models should be considered

NU

before designing the clinical trial with glial inhibitor [16; 34]. Therefore, our study provided the

MA

evidence of the astrocytic activation in the CPIP model. Theoretically, JNK could be taken as a potential target for analgesia, however, study regarding clinical trial with JNK inhibitor has not

D

been reported yet. In this study, our findings revealed the essential role of JNK in the pathogenesis

PT E

of CPIP. Furthermore, a clinical trial with CPL 7075 (an MMP inhibitor) is under way to study its effects on neuropathic pain in human [10]. In this study, our findings further indicated the

AC

CE

potential value of MMP2 inhibitor in the management of CRPS.

Conclusions

In summary, this study shows that spinal astrocyte, but not microglia, contributes to the pathogenesis of pain hypersensitivity in the CPIP model, and that the JNK pathway contributes to the development of CPIP injury-induced allodynia and is the intracellular signaling molecule downstream of the astrocyte activation. Moreover, the MMP2 activation contributes to the central mechanisms of pain hypersensitivity and triggers the astrocyte/JNK1/2 signaling in the CPIP 18

ACCEPTED MANUSCRIPT model. Therefore, this study firstly reports the role of MMP2/astrocyte/JNK1/2 in the pathophysiology of CPIP, which might provide a new mechanistic insight for future research on

AC

CE

PT E

D

MA

NU

SC

RI

PT

CRPS.

19

ACCEPTED MANUSCRIPT

Abbreviations CRPS: complex regional pain syndrome; CPIP: chronic post-ischemia pain; GFAP: glial fibrillary acidic protein; IBA1: ionized calcium binding adaptor molecule 1; PID: post-injury day; JNK: c-jun N-terminal kinase; MMP2: matrix metalloproteinase-2; CSF: cerebrospinal fluid; PWT: paw

AC

CE

PT E

D

MA

NU

SC

RI

PT

withdrawal threshold; NFκB: Nuclear factor-κB; TSPO: translocator protein.

20

ACCEPTED MANUSCRIPT

Conflict of interest The authors declare that there is no conflict of interest.

Authors’ contributions

PT

GGT, XL, ZYX, CWC and QLG have made substantial contributions to the conception and design;

RI

GGT, XL and CLT have made substantial contributions to the acquisition of data; GGT, XL, CLT,

SC

ZYX, SKC, CWC and QLG have made substantial contributions to the analysis and interpretation of data; All authors have been involved in drafting the manuscript or revising it critically for

NU

important intellectual content; All authors have given final approval of the version to be submitted.

MA

Z. Xia, CW Cheung and QL Guo share senior authorship.

D

Acknowledgement

PT E

This study was supported by Department fund, Department of Anesthesiology, The University of Hong Kong, HKSAR, China, and The China national key clinical specialty projects (2011-872),

CE

Affiliated Haikou Hospital of Xiangya Medical School, Central South University, Haikou, China.

AC

The authors thank Dr. Liting Sun, Department of Anaesthesiology, The University of Hong Kong, for the help in immunohistochemistry work. The authors acknowledge the technical assistance of the University of Hong Kong Li Ka Shing Faculty of Medicine Faculty Core Facility Centre and the editorial assistance of the Vanscholar Editors Co. Ltd, Canada.

21

ACCEPTED MANUSCRIPT

Reference [1] Alexander GM, Perreault MJ, Reichenberger ER, Schwartzman RJ. Changes in immune and glial markers in the CSF of patients with Complex Regional Pain Syndrome. Brain Behav Immun 2007;21(5):668-676.

fluid

levels

of

pro-inflammatory

cytokines

in

CRPS.

Pain

RI

cerebrospinal

PT

[2] Alexander GM, van Rijn MA, van Hilten JJ, Perreault MJ, Schwartzman RJ. Changes in

SC

2005;116(3):213-219.

[3] Barragan-Iglesias P, Pineda-Farias JB, Cervantes-Duran C, Bravo-Hernandez M,

NU

Rocha-Gonzalez HI, Murbartian J, Granados-Soto V. Role of spinal P2Y(6) and P2Y(11)

MA

receptors in neuropathic pain in rats: possible involvement of glial cells. Mol Pain 2014;10:11.

D

[4] Basbaum AI, Bautista DM, Scherrer G, Julius D. Cellular and Molecular Mechanisms of Pain.

PT E

Cell 2009;139(2):267-284.

[5] Borchers AT, Gershwin ME. Complex regional pain syndrome: A comprehensive and critical

CE

reviewd. Autoimmun Rev 2014;13(3):242-265.

AC

[6] Cao H, Zhang YQ. Spinal glial activation contributes to pathological pain states. Neurosci Biobehav Rev 2008;32(5):972-983. [7] Coderre TJ, Xanthos DN, Francis L, Bennett GJ. Chronic post-ischemia pain (CPIP): a novel animal model of complex regional pain syndrome-Type I (CRPS-1; reflex sympathetic dystrophy) produced by prolonged hindpaw ischemia and reperfusion in the rat. Pain 2004;112(1-2):94-105. [8] de Mos M, Laferriere A, Millecamps M, Pilkington M, Sturkenboom M, Huygen F, Coderre TJ. 22

ACCEPTED MANUSCRIPT Role of NF kappa B in an Animal Model of Complex Regional Pain Syndrome-type I (CRPS-I). J Pain 2009;10(11):1161-1169. [9] Del Valle L, Schwartzman RJ, Alexander G. Spinal cord histopathological alterations in a patient with longstanding complex regional pain syndrome. Brain Behav Immun

PT

2009;23(1):85-91.

inhibitors

in

neuropathic

pain.

Expert

Opin

Investig

Drugs

SC

metalloprotease

RI

[10] Dev R, Srivastava PK, Iyer JP, Dastidar SG, Ray A. Therapeutic potential of matrix

2010;19(4):455-468.

NU

[11] Gao YJ, Cheng JK, Zeng Q, Xu ZZ, Decosterd I, Xu XY, Ji RR. Selective inhibition of JNK

MA

with a peptide inhibitor attenuates pain hypersensitivity and tumor growth in a mouse skin cancer pain model. Exp Neurol 2009;219(1):146-155.

PT E

2008;437(3):180-183.

D

[12] Gao YJ, Ji RR. Activation of JNK pathway in persistent pain. Neurosci Lett

[13] Gao YJ, Xu ZZ, Liu YC, Wen YR, Decosterd I, Ji RR. The c-Jun N-terminal kinase 1 (JNK1)

CE

in spinal astrocytes is required for the maintenance of bilateral mechanical allodynia

AC

under a persistent inflammatory pain condition. Pain 2010;148(2):309-319. [14] Garrison CJ, Dougherty PM, Kajander KC, Carlton SM. Staining of glial fibrillary acidic protein (GFAP) in lumbar spinal-cord increases folowing a sciatic-nerve constriction injury. Brain Res 1991;565(1):1-7. [15] Gierthmuhlen J, Binder A, Baron R. Mechanism-based treatment in complex regional pain syndromes. Nat Rev Neurol 2014;10(9):518-528. [16] Hansen RR, Malcangio M. Astrocytes-Multitaskers in chronic pain. Eur J Pharmacol 23

ACCEPTED MANUSCRIPT 2013;716(1-3):120-128. [17] Ikeda H, Kiritoshi T, Murase K. Contribution of microglia and astrocytes to the central sensitization, inflammatory and neuropathic pain in the juvenile rat. Mol Pain 2012;8:10. [18] Ji RR, Kawasaki Y, Zhuang ZY, Wen YR, Decosterd I. Possible role of spinal astrocytes in

RI

bFGF/JNK pathway. Neuron Glia Biol 2006;2:259-269.

PT

maintaining chronic pain sensitization: review of current evidence with focus on

SC

[19] Kawasaki Y, Xu ZZ, Wang X, Park JY, Zhuang ZY, Tan PH, Gao YJ, Roy K, Corfas G, Lo EH, Ji RR. Distinct roles of matrix metalloproteases in the early- and late-phase

NU

development of neuropathic pain. Nat Med 2008;14(3):331-336.

MA

[20] Kwak KH, Lim DG, Baek WY. N-Acetyl-L-Cysteine Attenuates Ischemia/Reperfusion Injury-Induced Allodynia and N-Methyl-D-Aspartate Receptor Activation in Rats. Curr

D

Ther Res-Clin Exp 2011;72(5):216-227.

PT E

[21] Landry RP, Jacobs VL, Romero-Sandoval EA, DeLeo JA. Propentofylline, a CNS glial modulator does not decrease pain in post-herpetic neuralgia patients: In vitro evidence for

CE

differential responses in human and rodent microglia and macrophages. Exp Neurol

AC

2012;234(2):340-350. [22] Loggia ML, Chonde DB, Akeju O, Arabasz G, Catana C, Edwards RR, Hill E, Hsu S, Izquierdo-Garcia D, Ji RR, Riley M, Wasan AD, Zurcher NR, Albrecht DS, Vangel MG, Rosen BR, Napadow V, Hooker JM. Evidence for brain glial activation in chronic pain patients. Brain 2015;138:604-615. [23] Luo X, Tai WL, Sun L, Pan Z, Xia Z, Chung SK, Cheung CW. Crosstalk between astrocytic CXCL12 and microglial CXCR4 contributes to the development of neuropathic pain. Mol 24

ACCEPTED MANUSCRIPT Pain 2016;12. [24] Luo X, Tai WL, Sun L, Qiu Q, Xia Z, Chung SK, Cheung CW. Central Administration of C-X-C Chemokine Receptor Type 4 Antagonist Alleviates the Development and Maintenance of Peripheral Neuropathic Pain in Mice. PLoS One 2014;9(8):e104860.

PT

[25] Luo X, Wang X, Xia Z, Chung SK, Cheung CW. CXCL12/CXCR4 axis: an emerging

RI

neuromodulator in pathological pain. Rev Neurosci 2016;27(1):83-92.

SC

[26] Mika J, Zychowska M, Popiolek-Barczyk K, Rojewska E, Przewlocka B. Importance of glial activation in neuropathic pain. Eur J Pharmacol 2013;716(1-3):106-119.

NU

[27] Millecamps M, Laferriere A, Ragavendran JV, Stone LS, Coderre TJ. Role of peripheral

MA

endothelin receptors in an animal model of complex regional pain syndrome type 1 (CRPS- I). Pain 2010;151(1):174-183.

D

[28] Milligan ED, Watkins LR. Pathological and protective roles of glia in chronic pain. Nat Rev

PT E

Neurosci 2009;10(1):23-36.

[29] Robinson CR, Zhang H, Dougherty PM. Astrocytes, but not microglia, are activated in

CE

oxaliplatin and bortezomib-induced peripheral neuropathy in the rat. Neuroscience

AC

2014;274:308-317.

[30] Romero-Sandoval A, Chai N, Nutile-McMenemy N, Deleo JA. A comparison of spinal Iba1 and GFAP expression in rodent models of acute and chronic pain. Brain Res 2008;1219:116-126. [31] Ross-Huot MC, Laferriere A, Khorashadi M, Coderre TJ. Glycemia-dependent Nuclear Factor B Activation Contributes to Mechanical Allodynia in Rats with Chronic Postischemia Pain. Anesthesiology 2013;119(3):687-697. 25

ACCEPTED MANUSCRIPT [32] Schonbeck U, Mach F, Libby P. Generation of biologically active IL-1 beta by matrix metalloproteinases: A novel caspase-1-independent pathway of IL-1 beta processing. J Immunol 1998;161(7):3340-3346. [33] Shimizu K, Guo W, Wang H, Zou S, LaGraize SC, Iwata K, Wei F, Dubner R, Ren K.

PT

Differential involvement of trigeminal transition zone and laminated subnucleus caudalis

RI

in orofacial deep and cutaneous hyperalgesia: the effects of interleukin-10 and glial

SC

inhibitors. Mol Pain 2009;5:18.

[34] Watkins LR, Hutchinson MR, Johnson KW. Commentary on Landry et al.: "Propentofylline,

NU

a CNS glial modulator, does not decrease pain in post-herpetic neuralgia patients: In vitro

Neurol 2012;234(2):351-353.

MA

evidence for differential responses in human and rodent microglia and macrophages". Exp

D

[35] Wieseler-Frank J, Maier SF, Watkins LR. Central proinflammatory cytokines and pain

PT E

enhancement. Neurosignals 2005;14(4):166-174. [36] Zhuang ZY, Wen YR, Zhang DR, Borsello T, Bonny C, Strichartz GR, Decosterd I, Ji RR. A

CE

peptide c-Jun N-terminal kinase (JNK) inhibitor blocks mechanical allodynia after spinal

AC

nerve ligation: Respective roles of JNK activation in primary sensory neurons and spinal astrocytes

for

neuropathic

pain

development

2006;26(13):3551-3560.

26

and

maintenance.

J

Neurosci

ACCEPTED MANUSCRIPT Figure 1. Mechanical allodynia and edema was induced in the CPIP model. Effects of CPIP injury on the PWT (A) and the volume (B) in the ipsilateral and contralateral hindpaws of mice in the CPIP model were assessed. Results are means ± SEM (n =5-6). aaap <0.001 versus the baseline. #p

AC

CE

PT E

D

MA

NU

SC

RI

PT

<0.05 versus data on PID 7 in the same group.

27

ACCEPTED MANUSCRIPT Figure 2. Activation of astrocyte and microglia at the early stage in the CPIP model. In the spinal cord section, the immunohistochemical evidence showed the expression of GFAP, but not IBA1, was increased in CPIP-injured mice (A) on PID 3 and the data summary was shown in B and C. Original magnification: 200× for all the confocal images (Bar=100 μm). Intrathecal injection of

PT

fluorocitrate increased the ipsilateral PWT (D), but not contralateral PWT (E), of CPIP injured

RI

mice. Results are means ± SEM (n =3-8). ***p <0.001 and *p <0.05 versus the sham group or

SC

vehicle group. aaap <0.001, bbbp <0.001, cccp <0.001 and cp <0.05 versus the baseline from the same

AC

CE

PT E

D

MA

NU

group.

28

ACCEPTED MANUSCRIPT Figure 3. Activation of JNK1/2 at the early stage in the CPIP model. Spinal levels of phosphorylated JNK 1 and 2 were increased respectively relative to the levels of total JNK 1 and 2 (A). It was further shown that the intrathecal injection of SP600125 increased the PWT of CPIP-injured mice (B). In the spinal cord section, the immunohistochemical evidence showed the

PT

expression and distribution of phosphorylated JNK in mice receiving sham or CPIP operation (C)

RI

on PID 3 and the data summary was shown in D. Moreover, the intrathecal injection of

SC

fluorocitrate decreased the phosphorylation levels of JNK 1 and 2 on PID 3, relative to the levels of total JNK 1 and 2 (E). Original magnification: 200× for all the confocal images (Bar=100 μm).

AC

CE

PT E

D

MA

versus the baseline from the same group.

NU

Results are means ± SEM (n =3-8). ***p <0.001 and *p <0.05 versus control group.

29

bbb

p <0.001

ACCEPTED MANUSCRIPT Figure 4. Activities of MMP2 at the early stage in the CPIP model. In the spinal cord section, the expression of MMP2 was increased in CPIP-injured mice on PID 3 (A). High resolution images were shown in boxes and the co-expression was marked with arrows. MMP2 was mainly expressed in neurons (C), but not astrocyte (B), on PID 3. Intrathecal injection of APR 100

PT

increased ipsilateral PWT of CPIP injured mice (D). In the spinal cord section, the

RI

immunohistochemical evidence showed intrathecal APR 100 downregulated the levels of GFAP (E)

SC

and phosphorylated JNK1/2 (F) on PID 3 in CPIP-injured mice. Original magnification: 200× for all the confocal images (Bar=100 μm). Results are means ± SEM (n =3-8). ***p <0.001 and *p p <0.001,

AC

CE

PT E

D

MA

baseline from the same group.

aaa

bbb

NU

<0.05 versus sham group or vehicle group.

30

p <0.001 and ap <0.05 versus the

RI

PT

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

NU

SC

Figure 1

31

AC

CE

PT E

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Figure 2

32

AC

CE

PT E

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Figure 3

33

AC

Figure 4

CE

PT E

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

34