Differential Proteomic Analysis of the Spinal Cord in Bone Cancer Pain Rats by Two-Dimensional Gel Electrophoresis

Background: Bone cancer pain (BCP) is a common chronic pain that is caused by a primary or metastatic bone tumor. It is refractory to currently available clinical treatment owing to its complicated underlying mechanisms. Methods: In this study, we used proteomics approaches to investigate expressional changes of the rat spinal cord proteome from 7 to 21 d after inoculation. Proteins from the rat L4-6 spinal cord homogenates of BCP and Sham animals were fractionated by two-dimensional (2-DE) gel electrophoresis to produce a high-resolution map of the spinal cord soluble proteins. Proteins showing altered expression levels between BCP and Sham were selected. Results: A total of 60 spots were obtained, and isolated proteins were in-gel trypsin-digested and the resulting peptides were analyzed by matrix-assisted laser desorption/ionization-time of ight (MALDI-TOF) mass spectrometry. Using the mass spectrometric data, 34 differentially expressed proteins (DEPs) were identied. GO analysis of the identied proteins allowed us to explore the function of the represented proteins. Conclusions: Based on these results, the identied proteins may contribute to the maintenance of BCP, and may provided new or valuable information in the discovery of new therapeutic targets for BCP.

In this study, we rst implanted carcinoma cells into the tibial plateau of rats to establish a model of BCP. We used two-dimensional (2D) gel electrophoresis combined with matrix-assisted laser desorption ionization time-of-ight mass spectrometry (MALDI-TOF-MS) to perform a global analysis of the expressional changes in the rat spinal cord protein from 7 to 21 d after tumor injection. This study will provide valuable information in the way of understanding the pathogensis and discovering new therapeutic targets for BCP.

Cells and animals
The Walker 256 rat mammary gland carcinoma cells used in the experiment were provided by the Institute of Anesthesiology & Pain (IAP), Taihe Hospital. Healthy female Sprague-Dawley (SD) rats weighing 180-220 g were kindly provided by Institute of Laboratory Animal Science, Hubei University of Medicine (Hubei, China). All rats (n = 12-14 for each group) were kept in a control room under a constant temperature (22±2°C) and humidity (50±20%) with a rugular 12 h light/dark cycle. Rats had free access to standard diet and water. The experimental procedures and protocols in this study were approved by the Animal Care and Use Committee of Hubei University of Medicine ( Hubei, China) and were in accordance with the guidelines of the International Association for the Study of Pain (IASP).

Establishment of BCP model
Walker 256 cells (20 μL, 5×10 7 cells/mL) were injected into the tibial plateau of SD rats weighing 180-220 g to establish the BCP model. The rats were anesthetized with (3% for induction, 5% for maintenance) and placed on an operating table in the supine position. The right hind leg was slowly shaved and disinfected with 0.5% iodine. A long skin incision of about 0.5 cm was made at the right tibial plateau of rats. A 23-gauge needle was insert into the tibial plateau for making a drilled hole. A 20 μL volume of Walker 256 rat mammary cancer cells (5 × 10 7 cells/mL) or PBS was slowly injected into the intramedullary space of a rat's tibias plateau with a 25-μL glass microinjector. The syringe was left in the place for an additional 2 min to prevent the carcinoma cells from leaking out along the injection track, and then sealed with bone wax while the syringe was removed. The wound was carefullly sterilized and stitched up. After the operation, the rat was placed on a constant temperature resuscition blanket and put back into the cage after waking up.

Behavioral test of pain
The paw withdraw threshold (PWT) of rats were measured preoperatively and on postoperative days (PODs) 7, 14, and 21 with a dynamic plantar esthesiometer (Ugo Basile, Comerio, Italy), an automated version of the von Frey hair test. The measurements on each day were conducted from 8:00 to 11:00 a.m. in a quiet room. The rats were placed in a transparent plexiglass cage (20 cm × 25 cm × 15 cm) on an elevated mental mesh oor and were kept quiet for 30 min before testing. From bottom to top, Von Frey laments (0.5 mm diameter) stimulate the middle skin of the bottom of the right. The stimulus intensity was gradually increased from 0 g to 50 g within 20 s When the rats showed a contraction response, the intensity of the stimulus was stopped automatically. Recording the pressure value. Each rat was measured three times at an interval of 10 min, and the average of three values was used for data analysis.
Protein sample preparation From time points were set: 7, 14, and 21 day after inoculation, the rats (n = 3 -4 for each group) were intraperitoneally injected with 10% chloral hydrate (0.1-0.3 mL/100 g) for deep anesthesia at each time point. The lumbar spine was rapidly cut and the surrounding tissues were separated to obtain the L4-6 spinal cord. The obtained spinal cord tissue was placed into a 1.5 mL cryopreservation tube and then stored in a refrigerator at -80°C. The extracted spinal cord tissues were lysed in a lysis buffer (7 M urea, 2 M thiourea, 100 mM DTT, 4% CHAPS, 0.5 mM EDTA, 40 mM Tris, 2% NP40, 1% TritonX-100, 5 mM PMSF, and 2% pharmalyte). The lysates were homogenated rapidly for 3 min, placed at room temperature for 1 h, mixed intermittently during the period, and Finally, It was believed that Proteins presenting as spots with ≥ 1.5 fold and p < 0.05 changes (Student's t-test) in expression were considered as statistically signi cant altered proteins.

MALDI-TOF-MS analysis and protein identi cation
The statistically signi cant protein spots were selected, excised manually from each 2-DE and the protein was digested as previously study described [7]. The excised protein spots were proteolyzed in-gel with trypsin and resulting tryptic peptides were subsequently extracted and dried under vacuum. The in-gel digested peptides were analyzed with MALDI-TOF-MS. Combined peptide mass ngerprinting (PMF) and tandem mass spectrometry (MS/MS) queries were searched against the NCBI database, using the MASCOT search engine from Matrix Science (http://www.matrixscience.com/) to identify the proteins. Proteins matching more than ve peptides and with a Mascot score higher than 60 were considered signi cant (p < 0.05) [8]. To identify the correct protein from a Mascot results list, MOWSE score and species had to be considered.

Bioinformatics analysis
To functionally annotate the correlated genes, gene ontology (GO) and pathway analysis, GI numbers of differentially identi ed proteins by 2D electrophoresis were uploaded to the Database for Annotation, Visualization and Integrated Discovery (DAVID) functional annotation tool (http://david.ncifcrf.gov/tools.jsp). The enrichment analysis was done using Fisher exact test. Functional categories with p < 0.05 were considered to signi cant after multiple term testing by Bonferroni correction. The protein-protein interactions (PPIs) network was predicted using Search Tool for the Retrieval of Interacting Genes (STRING; http://string-db.org) (version 10.0) online database. In the present study, PPIs network of DEPs were constructed using STRING database, and an interaction with a combined score 0.4 was considered statistically signi cant. Cytoscape (version 3.4.0) is an open source bioinformatics software platform for visualizing molecular interaction networks [9].

Statistical analysis
The GraphPad Prism 7 statistic software was used for data analysis calculations. The data are presented as mean ± standard error of the mean. One-way analysis of variance (ANOVA) was used to compare differences between groups. p < 0.05 were considered statistically signi cant.

Results
The PWT was decreased in rats with BCP The PWT of the rats were measured on the day 0, 7, 14, and 21 after inoculation. As Fig. 1 shows, there was no signi cant difference in the basic PWT for all rats before inoculation. The PWT was signi cantly lower in BCP rats compared to sham rats (p < 0.05) on days 7 to 21, While the pain threshold of the Sham group rats was relatively stable (Fig. 1). The PMF was obtained and analyzed using the MASCOT search engine against the NCBI database. The resulting protein was determined by comprehensively considering the corresponding experimental PI, Mr, the number of matched -peptides, and sequence coverage. Among 60 protein spots, 34 DEPs were successfully identi ed by PMF. The data of all identi ed protein spots were described in Table 1.

GO enrichment pathway and PPI network analysis of DEPs
To analyze the biological classi cation of the 34 DEPs, functional annotation clustering analyses were based on GO annotation category. Go enrichment analysis was conducted using DAVID online software according to three main annotation categories, namely biological process category, cellular component category and molecular function category. Among 34 proteins, 21 DEPs were clustered. As shown in Fig. 3, GO biological process category revealed that these DEPs were mainly clustered in actin lament organization, hippocampus development, oxidation-reduction process; GO cellular component category demonstrated that a majority of above proteins were linked to the extracellular exosome, myelin sheath, cytoplasm, neuron projection, growth cone; GO molecular function category displayed that those proteins were mainly enriched in cytoskeleton, actin binding, toxic substance binding.
To better illuminate the potential molecular mechanism of BCP, PPI network was analyzed using STRING database and visualized using Cytoscape online tool to evaluate the relationship among the DEPs. As diaplayed in Fig. 4, PPI revealed that remarkable interactions existed.

Discussion
BCP is a kind of chronic pain with unique and complicated mechanism, which is the most common symptom of pain for the patient with cancer. For a long time, it has seriously affect the life quality of patients with cancer and brought oppressive pain to patients and their families. Spinal cord is the primary central of pain information transmission and integration, and its sensitization is the main mechanism of chronic pain. Rat spinal cord contains thousands of proteins, which may be closely related to the development of BCP. Therefore, it may be a promising way to analyze the pro le of spinal cord protein alterations in order to try nding potential biomarker to elucidate the pathogenesis of BCP [10]. Nowadays, proteomic analysis is widely considered as a valuable strategy to conduct mechanistic studies [11]. In order to nd potential molecules and elucidate the regulatory mechanisms of BCP, Proteomic approaches were applied.
In this study, we rst implanted carcinoma cells into the tibial plateau of rats to establish a model of BCP. The PWT of rats ipsilateral paw was measured on the day 0, 7, 14 and 21 after inoculation. With the tumor growth, the PWT decreased obviously in BCP rats, it suggested that the bone cancer pain animal model was successfully established.
Proteomic method of 2-DE gels electrophoresis was used to analyze the rat lumbar spinal cord. For the results, a total of 60 DEPs were identi ed, including 32 downregulated proteins and 28 upregulated proteins. The 60 protein spots with signi cant differential expression were selected for MALDI-TOF-MS analysis. Among 60 protein spots, 34 DEPs were successfully identi ed by PMF.
We analyzed the differentially expressed proteins, and taking into consideration their function annotation, identi ed several proteins that may be related to the molecular mechanism of the BCP. For example, co lin1, tropomodulin-2 (Tmod2) and tropomyosin beta chain (Tpms), which are associated with actin lament organization function. Co lin1 is an actin-binding protein, the primary function of which is dynamic reorganization of actin cytoskeleton.
It plays an important role in regulating actin dynamic, affecting synaptic transmission and plasticity. A previous study reported that the expression of co lin1 is down-regulated in DRG and spinal cord of BCP, suggesting that co lin1 is involved in the formation and development of BCP [12]. Our results also showed co lin downregulation on the day 7 after inoculation in the BCP model. Actin laments have a wide range of functions in membrane morphology, cell movement, vesicle transport, neurite development, and synaptic communication, and are considered to play an important role in synaptic plasticity [13]. The state of actin (as monomers or polymeric strands) and the length of the laments are controlled by numerous actin-regulatory proteins. Tmod2 is a protein that binds to the slow-growing (pointed) ends of actin laments (F-actin), preventing both elongation and depolymerization. Tpm bind along the length of actin laments and prevents lament depolymerization. In fact, the pointed-end capping activity of Tmod2 is greatly enhanced by association with Tpm, which suggests that the two proteins function as a complex to stabilize the lament and regulate its length. In our results, Tmod2 was downregulated and Tpm was upregulated in the BCP model, suggesting that actin laments were activated and lose staility, which increased the growth rate of actin laments, and nally contributed to enhanced of information synaptic transmission of dendritc spines in neurons, inducing the occurrence of BCP. It has been reported knockout of Tmod2 increased long -term potentiation (LTP) in brain slices [14]. This fully suggests that Tmod2 is associated with synaptic information transmission and may be involved in the occurrence and development of BCP.
Glia maturation factor-β (GMFB) is considered to be a growth and differentiation factor for both glia and neurons, Which is expressed predominantly in the central nervous system (CNS) [15]. The major cells that express GMFB in the nervous system are glial cells, including astrocytes, Bergmann glia in the cerebellum, Schwann cells in the peripheral nervous system (PNS). It is commonly accepted that reactive oxygen species (ROS) are important byproducts of metabolism and participate in many pathophysiological processes. Recent studies have shown that ROS are involved in sensory transmission and regulation and are an important factor leading to carcinogenic pain.
As part of the cellular antioxidant system, Copper-zinc superoxide dismutase (Cu-Zn SOD) can scavence ROS, reduce oxidative stress at the spinal cord level and relieve bone cancer pain through its own antioxidant action and protective antioxidant action of prdx4 [16]. A previous study reported that overexpression of GMFB using an adenovirus vector carrying GMFB cDNA in C6 rat glioma cells led to the activation of NF-κB as well as an increase in Cu-Zn SOD expression [17]. In our study, GMFB is downregulated on the day 7 after surgery in the BCP model. It suggest that the production of reactive oxygen species increases in the BCP, while the stimulation of Cu-Zn SOD is weakened by the down-regulation of GMFB, and the expression of Cu-Zn SOD is reduced, which forms oxidative stress and stimulates the production of various harmful substances by the cells, causing the abnormal release of neurotransmitters as well as the abnormal activation of corresponding receptors, resulting in bone cancer pain.
Therefore, GMFB is a valuable molecule for further functional research in BCP.
Calpain is a family of calcium-dependent intracellular cysteine protease, the activity of which is dependent on intracellular calcium levels. They participate in a variety of cellular processes such as signal transduction, cell proliferation, cell cycle progression, differentiation, apoptosis, membrance fusion, necrosis and platelet activation, etc [18]. Among them, Calpain is known to exert restrictive proteolysis on its speci c substrates, particularly IκBα, c-Jun, and c-Fos, all of which are essential molecular players in osteoclastogenesis [19]. Studies have suggested that inhibition of the activity of the protease calpain reduce the development of acute and chronic in ammation [20]. On the other hand, a previous study also reported that calpain inhibitor reduces BCP possibly through inhibition of osteoclastogenesis in rat BCP model [21]. However, Calpain was downregulated in our study. Therefore, the downregulation of Calpain may suggest activation of a otective mechanism in BCP model, and could reduce the extent of pain.
The N-myc downstream-regulated gene (NDRG) family consists of four members (NDRG1~4) that are differentially expressed in various organs and important cellular processes, such as cell proliferation, apoptosis, differentiation, development, and stress reponse [22]. As a member of the NDRG family, NDRG family member 2 (NDRG2) is predominantly expressed in astrocytes of the central nervous system (CNS). Its function in the CNS is mainly related to its role in astrocyte activity. The expression of NDRG2 is ubiquitous in astrocytes, and mostly colocalizes with the astrocyte marker glial brillary acidic protein (GFAP). NDRG2, which is involved in not only the regulation of apoptosis, blood-brain barrier integrity, glutamate clearance and other cellular metabolic processes, but also stress responses, is commonly regarded as a tumor suppressor [23,24]. NDRG2 has been found to be expressed in neurons, playing an important in neural differentiation, synapse formation, and axon survival [25]. A recent study revealed the correlations between Glutamatergic and GABAergic nerve terminals and NDRG2 immunopositive astrocytic processes [26]. The expression of the glutamate transporters glutamate aspartate transporter (GLAST) and glutamate transporter 1 (GLT-1) were increased by deletion of NDRG2 in glialtransmission.
A previous study demonstrated that NDRG2 is implicated in many neurological diseases including glioma, stroke, neurodegeneration and psychiatric disorders. However, the expression changes of NDRG2 in BCP have been rarely reported. Our study found that the expression of NDRG2 in BCP model was up-regulated, Considering the role of NDRG2 in astrocytes, NDRG2 may induce the activation of spinal astrocytes to participate in the occurrence and maintenance of bone cancer pain.
GO enrichment analysis revealed that the biological functions of the differentially expressed proteins relate mainly to actin lament organization, hippocampus development, oxidation-reduction process and give an insight into the underlying molecular mechanisms of the BCP. The functions of these proteins in biological processes should be further validated.

Conclusion
The comparative proteomics study is an effective platform to identify DEPs that may be implicated in the pathogenesis of BCP. Among 60 protein spots 34 DEPs were successfully identi ed by PMF. DEPs identi ed in the present study help us understand the molecular mechanisms underlying the carcinogenesis and progression of BCP, and provided candidate targets for diagnosis and treatment of BCP.

Declarations
Ethics approval and consent to participate This study was approved by the ethics Committee of Hubei University of Medicine.

Consent for publication
Not applicable.

Availability of supporting data
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The submission is not under review at any other publication, in whole or in part, and all the authors listed have approved the enclosed manuscript. All the authors declare no nancial con icts of interest and have ful lled the criteria of authorship for the manuscript.  Table 1 Identi cation of 34 proteins differentially expressed in the Sham and BCP groups  Fig. 2a, 2d, and 2g were Sham group. Fig. 2b, 2e, and 2h are up-regulated in BCP group compared with Sham group. Fig. 2c,   2f, and 2i were down-regulated in the BCP group compared with Sham group. DEPs were labeled in green number.
(B) Four representative DEPs circled in green were magni ed from two pairs of rat lumbar spinal cord samples.