Parameter Optimization Analysis of Prolonged Analgesia Effect of tDCS on Neuropathic Pain Rats

Background: Transcranial direct current stimulation (tDCS) is widely used to treat human nerve disorders and neuropathic pain by modulating the excitability of cortex. The effectiveness of tDCS is influenced by its stimulation parameters, but there have been no systematic studies to help guide the selection of different parameters. Objective: This study aims to assess the effects of tDCS of primary motor cortex (M1) on chronic neuropathic pain in rats and to test for the optimal parameter combinations for analgesia. Methods: Using the chronic neuropathic pain models of chronic constriction injury (CCI), we measured pain thresholds before and after anodal-tDCS (A-tDCS) using different parameter conditions, including stimulation intensity, stimulation time, intervention time and electrode located (ipsilateral or contralateral M1 of the ligated paw on male/female CCI models). Results: Following the application of A-tDCS over M1, we observed that the antinociceptive effects were depended on different parameters. First, we found that repetitive A-tDCS had a longer analgesic effect than single stimulus, and both ipsilateral-tDCS (ip-tDCS) and contralateral-tDCS (con-tDCS) produce a long-lasting analgesic effect on neuropathic pain. Second, the antinociceptive effects were intensity-dependent and time-dependent, high intensities worked better than low intensities and long stimulus durations worked better than short stimulus durations. Third, timing of the intervention after injury affected the stimulation outcome, early use of tDCS was an effective method to prevent the development of pain, and more frequent intervention induced more analgesia in CCI rats, finally, similar antinociceptive effects of con- and ip-tDCS were observed in both sexes of CCI rats. Conclusion: Optimized protocols of tDCS for treating antinociceptive effects were developed. These findings should be taken into consideration when using tDCS to produce analgesic effects in clinical applications.


INTRODUCTION
Chronic neuropathic pain is a common and severely disabling state that typically develops when peripheral nerves are damaged due to surgery, bone compression in cancer, diabetes or infection (Dworkin et al., 2013). Recent studies indicate increased activation of descending modulatory circuits (both descending facilitation and inhibition) in chronic neuropathic pain syndrome, and these circuit changes reflect long-lasting changes in synaptic efficacy (Zhuo, 2008(Zhuo, , 2013). An increasing number of investigators hold the view that non-invasive brain stimulation techniques can be used to treat chronic neuropathic pain (O'Connell et al., 2011;Volz et al., 2012). The aim of brain stimulation in managing of pain is to reduce pain symptoms by altering activity in brain areas that are involved in processing painful stimuli (Nguyen et al., 2011;Mylius et al., 2012;Woods et al., 2016).
Repetitive transcranial magnetic stimulation (rTMS) and transcranial direct current stimulation (tDCS) are two typical and common techniques of non-invasive brain stimulation techniques (Woods et al., 2016). Relative to invasive stimulation, non-invasive stimulation requires no surgical procedure and is therefore easier and safer to administer. However, there are several advantages of tDCS over rTMS, such as lower cost, increased portability, and more convincing sham conditions (Zaghi et al., 2009). TDCS is considered as a neuromodulatory intervention for the brain (Nitsche et al., 2008) which modulates the membrane potential dependently by type of electrode's application. Anode-tDCS (A-tDCS) is able to facilitate the depolarization of neurons and increase the cortical excitability. However, cathode-tDCS (C-tDCS) hyperpolarizes the resting membrane potential and reduces the cortical excitability (Nitsche et al., 2008;Mylius et al., 2012). Numerous clinical studies have concluded that A-tDCS is an effective method for pain modulations, which is helpful at reducing both fibromyalgia and spinal cord injury related-pain Roizenblatt et al., 2007;Soler et al., 2010). There are also studies showing effects of A-tDCS in inflammatory and neuropathic pain in animals Spezia Adachi et al., 2012;Cioato et al., 2016;Filho et al., 2016).
Sensory-motor cortex has been reported to decrease pain sensation and to increase pain threshold (Xie et al., 2009;Ossipov et al., 2010). As a result, the primary motor cortex (M1) is regarded as the location for stimulation electrode placement in the vast majority of trials in patients (Ferrucci et al., 2015;Woods et al., 2016), but there was little research testing the location of the stimulation electrode in animal research, the stimulation electrodes of those studies were placed on the middle of the scalp of rat rather than on a particular location (Ferrucci et al., 2015;Woods et al., 2016). There are three location modes for electrodes as related to the region of pain: contralateral M1, bilateral M1 and ipsilateral M1 to the injured region. A-tDCS delivered to contralateral M1 and bilateral M1 have been reported to have an antinociceptive effect in a number of patients with chronic pain (Ngernyam, 2014;O'Neill et al., 2015;Woods et al., 2016), whereas few studies of tDCS of ipsilateral M1 have been reported.
Beside the location of the stimulation electrode, physical parameters and practical applications of tDCS are important factors in treating neuropathic pain in animal and clinical experiments (Nitsche et al., 2008;Nitsche and Paulus, 2011;Stagg and Nitsche, 2011). Therefore, tDCS protocols should state current intensity, electrode size, stimulation duration and other parameters to aid in assessing comparability among studies (Nitsche et al., 2008). Originally, tDCS was believed to follow simple rules: the more electric charge flowing through the electrode, the stronger the analgesic effect would be (Nitsche et al., 2008;Nitsche and Paulus, 2011). However, the safety and side effects of tDCS on participants should be precisely calculated, because high current density can cause tissue damage (Nitsche et al., 2008;Liebetanz et al., 2009). Another important parameter of tDCS is the intervention time, which involves the time when stimulation is administered in relation to the course of pain processing and how many times tDCS is delivered. According to previous findings, repetitive stimulation has been proven to enhance efficacy and prolong after-effects of tDCS during specific time intervals (Nitsche et al., 2008). In addition, because the course of neuropathic pain is fairly long and easy to be targeted repeatedly (Dworkin et al., 2013), it will be significant if the patient can be treated when the pain is onset with negligible side effects.
Therefore, a combination of tDCS stimulation parameters including current intensity as well as stimulation time, interval and position need to be studied systematically in clinical and animal experiments. However, due to the limitation of choosing suitable parameters in clinical studies, optimal protocol characteristics have not been explored systematically (Ngernyam, 2014). In this article, we conducted studies to explore the optimal physical parameters and practical applications of A-tDCS in treating chronic neuropathic pain in both male and female rats.

Experimental Animals
Experiments were carried out on adult Sprague-Dawley rats (8-10 weeks old weighing 220-250 g), which were purchased from the Center of Laboratory Animal, Third Military Medical University, Chongqing, China. The SD rats were housed in plastic cages with soft bedding under controlled temperature settings (24 ± 1 • C), humidity (60 ± 5%) and a 12-h light/dark cycle. The study, animal care and handling procedures were in strict accordance with the recommendations of International Association for the Study of Pain's ethical guidelines (Zimmermann, 1983), and the protocol was also approved by the Ethical Committee for Animal Research of Third Military Medical University.

Animal Model of Neuropathic Pain
A typical neuropathic pain model was established through chronic constriction injury (CCI) of the sciatic nerve (Bennett and Xie, 1988).The right sciatic nerve of rats was tied with four 4-0 chromic gut ligatures 1 mm apart under 4% chloral hydrateanesthesia (10 ml/kg, i.p.). The sutures were not tied so tight that blood flow was affected. The overlying muscle was sutured and the skin wound was sealed with topical antibiotics. Rats with right sciatic nerve exposed without chromic gut ligature served as sham CCI controls.

Transcranial Direct Current Stimulation
We improved the previously implanted electrode protocol by optimizing the internal structural stability and by decreasing the contact impedance of the electrode (Liebetanz et al., 2006;Yu et al., 2015). Three days before CCI or sham CCI surgery, a saline soaked sponge was placed at the end of a plastic tube (inner diameter: 2 mm; length: 1 cm; Figures 1A,B). A copper wire was inserted into the sponge and held in place inside the tube with polyacrylate adhesive. Next, the tube was fixed with glass ionomer cement onto the cranium over M1 as a stimulation electrode using a stereotaxic apparatus; ipsilateral tDCS (ip-tDCS) refers to a stimulation electrode being fixed overlying the ipsilateral M1 to the ligated hind paw, and contralateral tDCS (con-tDCS) refers to stimulation electrode fixed overlying the contralateral M1 to the ligated hind paw ( Figure 1A). A far larger conventional sponge electrode (10 cm 2 ) was placed on the ventral thorax with a corset and served as a reference electrode ( Figure 1B).
As described below, constant current was applied via the stimulation electrode as A-tDCS at the schedule times (1st, 7th or 14th days after CCI; Figure 1C). For sham tDCS, the stimulation electrode was placed in the same positions as for real stimulation, but stimulation duration of 10 s was used as described above (Yu et al., 2015). Meanwhile, at the beginning and end of tDCS, current was ramped up and down for 10 s to prevent damage to the brain tissues by suddenly changing current (Bindman et al., 1964;Yu et al., 2015). Saline was injected continuously into the sponge through the hole that we left on the top of the tube during the DC stimulation in order to reduce contact impedance when stimulating (Yu et al., 2015). Notably, animals were not anesthetized during tDCS or sham tDCS.

Radiant Heat Test
The radiant heat test was carried out to estimate the thermal withdrawal latency (TWL; Hargreaves et al., 1988).
After an adaptation period of 30 min, the rats were placed into the test cage with a glass plate under which a light was located; 52 ± 0.2 • C radiant heat was applied to the plantar surface of the right hind-paw. The latency period was recorded in response to the thermal hyperalgesia by lifting hind-paw licking, flicking or commences jumping. To avoid tissue injury, the cut-off limit was set at 60 s (Hargreaves et al., 1988). Each hind-paw was measured for three times alternately at a 5 min interval. The mean was recorded as TWL.
FIGURE 1 | (A) Schematic diagram displaying the location of the stimulation electrode in the rat brain model. (B) Electrode configurations in rat. Two electrodes were applied during transcranial direct current stimulation (tDCS). The stimulation electrode was fixed onto the skull and a reference electrode was attached onto the thorax. (C) Schematic diagram of the experiment protocol schedule. 5CCT-1/5CCT-7/5CCT-14 means five times daily contralateral-tDCS (con-tDCS) started on day 1, 7, or 14 following chronic constriction injury (CCI), respectively; 10CCT-1 means 10 times con-tDCS at 2 weeks. Test point was the day to test the mechanical allodynia and thermal hyperalgesia on the hind paw of CCI.

Von Frey Filaments Test
The von Frey filaments test using an up-down method was performed to estimate the 50% system mechanical withdrawal threshold (MWT) with bending forces at a range of 0.3-20.3 g von Frey hairs (vFh; Chaplan et al., 1994). Each rat was placed inside a transparent acrylic cage (18 cm × 12 cm × 12 cm) with wire mesh floor with 60 min of acclimatization. The test was initiated with 4.10 g vFh. The filament was applied to the ventral surface of each right hind-paw for 4-6 s, hind paw withdrawal was considered as a positive response. When a positive result was noted, then the filament was decremented by one step size. If a negative result occurs, the filament was increased. The test continues until four measurements have been made after the first change in direction.

Statistical Analysis
Analyses were done with the SPSS software package (version 19). All data are expressed as the mean ± SD. The pain thresholds were evaluated by one way (Figure 2) or two-way (except Figure 2) repeated measure analysis of variance (ANOVA), when significant differences were observed, a post hoc test was made via Tukey's test. In all cases, p < 0.05 was considered to be statistically significant.

Changes of Pain Thresholds in CCI Rats
Significant mechanical allodynia and thermal hyperalgesia were elicited in surgical hind paw CCI rats (Figure 2). Two-way repeated measures ANOVA showed a significant main effect of group on the MWT (main effect of group F (2,30) = 663.718, p = 0.000) and TWL (F (2,30) = 828.857, p = 0.000). The values of MWT and TWL were decreased starting 1 day after CCI surgery, and the most severe stage appeared around day 14 (MWT: 1.03 ± 0.21 g, TWL: 8.61 ± 0.44 s) and then began . Control (CT); sham CCI (SC). All behavior tests were examined 1 day before the CCI surgery and on days 1, 3, 7, 10, 14, 21 and 28 after CCI surgery. Statistical significance was analyzed by two-way analysis of variance (ANOVA) followed by Tukey's post hoc test. The data are shown as mean ± SEM; * * p < 0.01, * * * p < 0.001 vs. CT group.
to recover on the following day (Figure 2). This trend was in line with our previous studies (Xiao et al., 2010;Ou et al., 2011;He et al., 2012). There was no significant difference in the thermal latency and mechanical threshold between control and sham CCI groups (p = 0.259 of TWL, p = 0.128 of MWT).

Effects of Single tDCS Treatment Disappeared within 24 h after Stimulation
We chose the 14th day after CCI to test the duration of antinociceptive effects after one time tDCS.
No difference was found in TWL and MWT among SC, SCSIT, SCSCT, SCIT and SCCT groups at all test points (one-way ANOVA, all p > 0.05). Sham tDCS (both SCSCT and SCSIT) had no impact on behaviors of Sham CCI rats, and the tDCS (both SCIT and SCCT) also did not have impact on behaviors of sham CCI rats (one-way ANOVA, all p > 0.05; Figure 3).
Compared with CCI, single con-tDCS (both CCT100 and CCT200) sharply increased TWL and MWT after one session of A-tDCS that ended 30 min before behavioral tests (oneway ANOVA, all p = 0.000). The effects were gradually diminished in the following 2 h (one-way ANOVA, TWL: p = 0.000 for CCT100, p = 0.000 for CCT200; MWT: p = 0.001 for CCT100, p = 0.000 for CCT200 vs. CCI). After 24 h, the antinociceptive effects were almost completely gone and there were no differences compared with the CCI group (one-way ANOVA, all p > 0.05). Moreover, we also observed a better recovery of pain in CCT200 than CCT100 (one-way ANOVA, TWL: p = 0.044 for 30 min, p = 0.004 for 2 h, p = 0.490 for 24 h; MWT: p = 0.037 for 30 min, p = 0.009 for 2 h, p = 0.523 for 24 h).
However, we did not observed changes in pain thresholds of sham tDCS (both CSIT and CSCT) in CCI rats (one-way ANOVA, all p > 0.05; Figure 3).

Both ip-and con-Repetitive tDCS Had Long-Term Antinociceptive Effects in CCI Rats
Repetitive A-tDCS had long-term antinociceptive effects, significant increases in TWL and MWT were observed not only during the stimulation process, but also 1 or more weeks following A-tDCS (Figures 4A,C). Five sessions of repetitive A-tDCS had similar antinociceptive effects in both CIT and CCT groups; similar effects were observed in surgical hind paw of CCI rats (two-way repeated measures ANOVA: TWL: F (3,40) = 480.888, p = 0.000; MWT: F (3,40) = 150.201, p = 0.000. Tukey's test of groups: all p = 0.000 vs. CCI). Moreover, the values of TWL and MWT in the CCT group were slightly higher than those in the CIT group at every test point, but the differences did not reach statistical significance (Tukey: TWL: p = 0.226, MWT: p = 0.051).
We also normalized the values of TWL and MWT (% of control) and observed the significant analgesia effect during and after tDCS in CIT and CCT groups (Figures 4B,D).

Intensity-Dependent Antinociceptive Effects of Repetitive tDCS in CCI Rats
After confirming the location of stimulation electrode in the above studies, we chose different stimulation intensities and simulation times to determine the most effective stimulation current.
Increasing stimulation current intensity resulted in an intensity-dependent increase in TWL of the ipsilateral In all three test phases, there were no difference between SC and SCSIT, SCSCT, SCIT, SCCT groups, and there was also no difference between CCI and CSCT, CSIT groups. However, there were significant differences between SC, SCSIT, SCSCT, SCIT, SCCT and CCI groups (all * p < 0.05 vs. CCI group). 30 min: the CIT and CCT groups (both 100 µA and 200 µA) differ from the CCI, CSIT and CSCT groups; 2 h: the CIT and CCT groups (both 100 µA and 200 µA) differ from the CCI, CSIT and CSCT groups; 24 h: there is no significant difference from CCI, CSIT, CSST, CIT and CCT groups ( # p < 0.05 compared with CCI group, and ∆ p < 0.05 compared with CIT100 or CCT100 group).

Proper Intervention Time Enhanced the Long-Term Antinociceptive Effects of Repetitive tDCS in CCI Rats
Intervention time also played a role in the antinociceptive effects following repetitive tDCS, the after-effects were different depending on the time of intervention of tDCS (two-way repeated measures ANOVA: TWL: F (5,60) = 258.796, p = 0.000; MWT: F (5,60) = 160.171, p = 0.000; Figures 7A,C).
Compared to the CCI group, A-tDCS significantly increased the pain values to a high level in both CIT group (Tukey's test of groups: TWL and MWT: all p = 0.000) and CCT group (Tukey's test of groups: TWL and MWT: all p = 0.000). We also observed that the mechanical allodynia and thermal hyperalgesia of the CCT group were mildly increased compared to the CIT group, but the differences did not reach statistical significance (Tukey's test of groups: TWL: p = 0.074; MWT: p = 0.093; Figure 8). We also normalized the values of TWL and MWT (% of control) and observed the significant analgesia effect during and after tDCS (Figures 8B,D).

DISCUSSION
Clinical treatment of neuropathic pain is still a major challenge because of its long duration and difficulty in managing . As a non-invasive electrical stimulation treatment, tDCS technology has been used for many years in clinical settings (Gandiga et al., 2006;Brunoni et al., 2012). Radiological and clinical practice have proven that tDCS can play a role in the plasticity of CNS regulation and serve as a treatment for neuronal abnormalities Brunoni et al., 2012;DosSantos et al., 2014;Woods et al., 2016). However, as systematical basic research is lacking, clinical effects are inconsistent and critical stimulus parameters are uncertain (Lee et al., 2015;Dedoncker et al., 2016). In this article, we choose a rat CCI model as the neuropathic pain model, which is the most common pain model that could finely simulate the clinical chronic neuropathic pain as reported by previous studies (Xiao et al., 2010;Ou et al., 2011;He et al., 2012). The pain threshold of MWT and TWL were evaluated at the different times after CCI. Previous clinical studies chose the M1 cerebral cortex as the stimulus location because of its roles in the modulation of chronic pain, as it receives pain-related information from the thalamus and the somatosensory cortex (Nguyen et al., 2011;Mylius et al., 2012). Earlier findings indicated that the neural activity of contralateral M1 to an injured paw was increased after CCI (Ooi et al., 2006), and contralateral M1 stimulation has been used as a clinical treatment of chronic pain with the use of transdural motor cortex stimulation (MCS) and rTMS (Lefaucheur, 2006;Fontaine et al., 2009;Young et al., 2014). However, previous work has also reported that M1 stimulation with MCS or rTMS ipsilateral to injury could also significantly suppress pain-related responses in rats and human (Nahmias et al., 2009;Lucas et al., 2011;Viisanen et al., 2012), but there have been no extensive experimental study testing tDCS. In our study, the anode stimulation electrode was mounted contralateral or ipsilateral to the injured paw to measure the antinociceptive effect of tDCS, and our results demonstrated that both locations of A-tDCS led to significant decreases in pain. Our study supplements previous experiments in which the location of the electrode was not taken into consideration Filho et al., 2016). A possible mechanism of these results is supra-spinal antinociceptive activities via multiple parallel pathways (Pertovaara and Wei, 2003). Recent studies had found that M1 stimulation can also active the adjacent regions, including the periaqueductal gray, anterior cingulate cortex and amygdale (Nguyen et al., 2011). Further experiments have shown that rTMS enhances the corticospinal inhibitory system which might in turn mediate M1 stimulation-induced spinal antinociception (Rojas-Piloni et al., 2010;Dall Agnol et al., 2014). The transduction mechanism of antinociception release needs further study in the future.
We observed that the effects of A-tDCS on reducing hyperalgesia and allodynia depended on stimulus intensity and time. Stronger intensities or longer duration correlates with more charged input to the cortex (Nitsche and Paulus, 2011), and a better after-effect on relieving pain (Nitsche et al., 2008). In Series 2, we found the analgesic effect of single tDCS was diminished following 2 h, but repetitive tDCS prolongs the duration of analgesic effect (Series 3). We found that greater charge inputs given 1 week after original stimulation maintained high pain thresholds when with different intervention times were administered started on days 1, 7 and 14 following CCI surgery, respectively. Sham CCI (SC); CCI plus five sessions daily con-tDCS from day 1 after CCI (5CCT-1); CCI plus five sessions daily con-tDCS from day 7 after CCI (5CCT-7); CCI plus five sessions daily con-tDCS from day 14 after CCI (5CCT-14); CCI plus 10 sessions con-tDCS from day 1 after CCI (10CCT-1). All behavior tests were tested 1 day before the CCI surgery and at days 1, 3, 7, 10, 14, 21 and 28 after CCI surgery. Statistical significance was analyzed by two-way ANOVA followed by Tukey's post hoc test. * p < 0.05, * * p < 0.01, * * * p < 0.001 vs. CCI group.
re-assessed on days 21 and 28 following CCI. Therefore, to create more effective and long-lasting after-effects, we should increase the stimulation duration and the current intensity. In the last few decades, there has been consensus that increasing stimulation duration prolongs the occurrence and duration of after-effects in humans and animals (Nitsche et al., 2008), which we verified in our study. However, increasing of duration and intensity has limitations. Considering the safety of tDCS, we choose 200 µA as the highest intensity because in our preliminary experiments we observed transient tremors if the intensity was over 220 µA. On the other hand, we found that there were no statistical differences in pain thresholds between CCT200-30 and CCT200-20 groups (Figure 6), suggesting that the antinociceptive effects of tDCS may be saturated after 20 min of tDCS at 200 µA. As such, it is necessary to determine the best intervention time if the stimulation intensity and duration cannot be further increased to continue prolonging the after-effects of tDCS. CCI was first reported by Bennett and Xie (1988) and was regarded as a typical model in neuropathic pain research. The duration of neuropathic pain was divided into two parts (Figure 2). Part one was described as the ''progression period'' and referred to days 1 through 14 following CCI, peaking between days 10 and 14; part two was described as ''recovery period'' and referred to the period following day 14 in which the sensitivity to pain decreased daily. We chose two time points (1 and 7 days after surgery) in the progression period and one point in the recovery period (14 days after surgery) for delivering tDCS. There were substantial signs of recovery in each treatment group compared to the CCI group: (1) repetitive A-tDCS reversed the decreased thresholds observed in CCI rats at every time point; (2) early intervention prevented the sharp decrease in the pain threshold, but did not restore the threshold to its normal level; (3) as chronic pain worsened, the degree of antinociception following tDCS decreased and the aftereffect were not well maintained; (4) antinociceptive effects were present when tDCS was delivered throughout the progressed , 20 min) were administration started on day 7 following CCI surgery. CT (control); SC (sham CCI); CIT (CCI + ip-tDCS treatment); CCT (CCI + con-tDCS treatment). CIT and CCT treatments significantly increased the values of TWL and MWT. All behavior tests were tested 1 day before the CCI surgery and on days 1, 3, 7, 10, 14, 21 and 28 after CCI surgery. Statistical significance was analyzed by two-way ANOVA followed by Tukey's post hoc test. * p < 0.05, * * * p < 0.001 vs. CT group, ### p < 0.001 vs. CCI group. period; and (5) pain thresholds significantly improved when tDCS were given at the beginning of the recovery period (Figure 7).
To the best of our knowledge, we are the first to discover that intervention time is another key factor that influences the efficacy of stimulation. With the same intensity and duration, the efficacy of analgesia was prolonged and consolidated by choosing the right time points when stimulation is participated in the pain processing and extending the original session of tDCS.
Clinically, A-tDCS is widely used in chronic pain treatment, an intensity of 2 mA and duration of 20 min with five or seven daily repetitions are usually chosen, parameters that have proven effects (Mori et al., 2010;Ngernyam, 2014). However, other researchers have also been obtained in which A-tDCS was ineffective in the treatment of chronic pain using the same conditions (Ihle et al., 2014;Nardone et al., 2014) The patient described in this study had stable chronic pain for at least 6 months with a high VAS scores before the stimulation, however, intervention time was not considered during therapy. A proper intervention time, the number of session and the length of stimulus duration should be considered and may be dependent on different states and causes of illness (Ihle et al., 2014). Our results suggest that clinicians should consider personalized treatment in patients with chronic neuropathic pain, pay attention to specific stimulus parameters and disease characteristics.
Furthermore, brain lesions were reported when the current density was greater than 142.9 A/m 2 in rat experiments. Our studies used a maximum current density of 63.69 A/m 2 (200 µA/3.14 mm 2 ) which is not associated with any tissue injure after tDCS (Liebetanz et al., 2009). Futhermore, we examined the effect of tDCS in female rats. Previous work found gender-related differences in utilitarian behavior after tDCS with greater effects in females as compared to males (Chaieb et al., 2008). For CCI rats, the chronic nociceptive processing was similar in both sexes, but male rats responded more quickly than females to a thermal nociceptive stimulus, and the stimulus elicited less robust thermal hyperalgesic symptoms in males than in females (Tall et al., 2001). In our study, we also found hyperalgesia and allodynia in female CCI rats. Both ip-and con-tDCS had the similar antinociceptive effects in female CCI rats. However, the hormonal fluctuate could be interfere in the nociceptive response (Tall et al., 2001) and the mechanism of gender difference needs further study in the future.
There is increasing evidence that the after-effects of tDCS are not only driven by the regulation of inhibition and excitation, but also by the modification of synapses (Nguyen et al., 2011;Stagg and Nitsche, 2011). Another view is that tDCS does not only elicit that rapid depolarization required to produce action potentials in neurons, but also may produces long-lasting changes in cortical excitability and activity (Mylius et al., 2012). Consistent with this, another key hypothesis holds that chronic pain is likely to employ highly selective synaptic connections and molecular signaling pathways within pain-related cortical areas (Zhuo, 2008(Zhuo, , 2013, resulting in cortical plasticity in both the descending and ascending systems. In the rat with peripheral injured, bilateral M1 receives pain-related information from the thalamus and the somatosensory cortex that maps to the injured paw (Xie et al., 2009;Ossipov et al., 2010). Stimulation of M1 might induce plasticity changes and reorganizations in the expression of neurotransmitter receptors (Lefaucheur et al., 2010;Stagg and Nitsche, 2011) which might include tonic activation of NMDA receptors (Pertovaara and Wei, 2003;Nguyen et al., 2011) and an enhanced anti-hypersensitivity effect in dopamine receptors (Viisanen et al., 2012). We observed increased NMDA receptors in bilateral M1 regions in CCI rats after repetitive tDCS (unpublished results). Therefore, the changes in NMDA receptors after tDCS might decrease the function of brain areas related to pain management through long-term potentiation (LTP) synaptic efficacy, thereby inducing cortical reorganization and CNS network processing. These affects are likely to reintroduce an optimal excitation/inhibition balance that allows for optimal homeostatic plasticity (Nitsche and Paulus, 2011;Stagg and Nitsche, 2011;Krause et al., 2013;Ngernyam, 2014).
The present study demonstrates the antinociceptive effect of tDCS in the male and female CCI rats. Both ip-DCS and con-tDCS produce a long-lasting analgesic effect on neuropathic pain, and the optimal stimulation parameters of tDCS are future studied. These dates may be helpful for the clinical applications of tDCS in pain control. More investigations on the synaptic mechanisms of tDCS should be conducted in the future.

AUTHOR CONTRIBUTIONS
H-ZR and H-ZW conceived and designed the study. H-ZW wrote the manuscript. S-HG, Y-DZ and W-JH carried out the animals' experiments. H-ZW and X-LT participated in the treatment of tDCS. H-ZW and S-HG participated in data analyses and arranged the figures. All authors have read and approved the final manuscript.