Electroacupuncture improves allodynia and central sensitization via modulation of microglial activation associated P2X4R and inflammation in a rat model of migraine

Abstract

Background: Recent studies have demonstrated that activated microglia were involved in the pathogenesis of central sensitization characterized by cutaneous allodynia in migraine. Activation of microglia is accompanied by increased expression of its receptors and release of inflammatory mediators. Acupuncture and its developed electroacupuncture (EA) have been recommended as an alternative therapy for migraine and are widely used for relieving migraine-associated pain. However, it remains rare studies that show whether EA exerts anti-migraine effects via inhibiting microglial activation related to a release of microglial receptors and the inflammatory pathway. Therefore, this study aimed to investigate EA’ ability to ameliorate central sensitization via modulation of microglial activation, microglial receptor, and inflammatory response using a rat model of migraine induced by repeated epidural chemical stimulation. Methods: In the present study, a rat model of migraine was established by epidural repeated inflammatory soup (IS) stimulation and treated with EA at Fengchi (GB20) and Yanglingquan (GB34) and acupuncture at sham-acupoints. Pain hypersensitivity was further determined by measuring the mechanical withdrawal threshold using the von-Frey filament. The changes in c-Fos and ionized calcium binding adaptor molecule 1 (Ibal-1) labeled microglia in the trigeminal nucleus caudalis (TNC) were examined by immunflurescence to assess the central sensitization and whether accompanied with microglia activation. In addition, the expression of Ibal-1, microglial purinoceptor P2X4, and its associated inflammatory signaling pathway mediators, including interleukin (IL)-1β, NOD-like receptor protein 3 (NLRP3), and Caspase-1 in the TNC were investigated by western blot and real-time polymerase chain reaction analysis. Results: Allodynia increased of c-Fos, and activated microglia were observed after repeated IS stimulation. EA alleviated the decrease in mechanical withdrawal thresholds, reduced the activation of c-Fos and microglia labeled with Ibal-1, downregulated the level of microglial purinoceptor P2X4, and limited the inflammatory response (NLRP3/Caspase-1/IL-1β signaling pathway) in the TNC of migraine rat model. Conclusions: Our results indicate that the anti-hyperalgesia effects of EA ameliorate central sensitization in IS-induced migraine by regulating microglial activation related to P2X4R and NLRP3/IL-1β inflammatory pathway.

Introduction

Cutaneous allodynia (CA) is defined as an abnormal perception of pain in response to non-harmful stimuli, is a common clinical presentation in migraine attacks, and is a crucial factor for migraine chronification, indicating central sensitization of the trigeminal vascular system. ¹ Central sensitization is the hyper-excited state of central neurons caused by repeated and continuous stimulation ¹ and is widely regarded as one of the main pathological mechanisms of migraine chronification. It is manifested by increased spontaneous activity of central neurons, reduced activation threshold, and enlarged receptive field of nociceptive neurons, ² principally occurring in the trigeminal nucleus caudalis (TNC) area. ³ Emerging studies suggest that microglial activation and inflammatory cytokines release could be involved in this process. ⁴ Microglial over-expression and inflammation, as the prime candidate, plays an important role in the modulation of migraine and are responsible for the progression toward chronification. ⁵ The activated microglia in the TNC region participates in the crosstalk with neurons by releasing of inflammatory mediators, including IL-1β, directly or indirectly affecting the establishment of neuronal hyper-excitability and promoting central sensitization, resulting in hyperalgesia and CA.^6,7 Furthermore, once microglia are activated, various microglia signaling molecules including its surface receptors are upregulated. ⁴ Previous studies have shown that microglia and its purinergic receptors, such as P2X4R, are obviously upregulated in the TNC of the migraine model rats. ⁸ Overall, these factors together promote central sensitization with CA as a symptomatic manifestation. However, more importantly, this allodynia can be alleviated by inhibiting the activation of microglia, which further affects the over-excitability of TNC neurons by altering microglial morphology and number and limiting subsequent associated inflammation. ⁶

Due to the high rate of disability associated with the chronic progression of migraine, ⁹ the main focus of current treatment strategies is to prevent the process of migraine chronification. Unfortunately, there are no modern medical means to inhibit or delay central sensitization to alleviate pain and thus inhibit the process of migraine chronification until now. Therefore, it is of great significance to search for valuable alternative therapy and carry out mechanism research. As an alternative and complementary therapy, acupuncture and its developed electroacupuncture (EA) have been widely used for prophylaxis and treatment of chronic pain and migraine worldwide, and a large number of studies provide evidence for their effectiveness in treating migraine. ¹⁰ Several recent studies confirmed that the microglia and neuroinflammation might participate in EA analgesia in several pain models.^11,12 In recent years, clinical reports and animal experiments supported that EA improved the migraine headaches in patients and hyperalgesia of migraine animal models.^13,14 It is believed that EA could relieve migraine-associated hyperalgesia via an anti-inflammatory effect by inhibiting releases of inflammatory factors, mainly in the peripheral pathway. ¹⁵ A few recent studies on EA and central sensitization have demonstrated the role of EA on hyperalgesia by improving central sensitization (corresponding to reduced c-Fos expression).^14,16 However, data on the potential central mechanisms of how EA treats migraine remain scarce.

Therefore, based on the evidence mentioned above, we hypothesized that EA could alleviated allodynia by inhibiting microglial activation, which is responsible for central sensitization, and decreasing microglial purinoceptor P2X4 and microglia-associated inflammatory response pathway in a rat model of migraine. To test this hypothesis, we developed a migraine rat model using a method similar to previously reported.^17,18 Next, changes in c-Fos and microglia labelled with Ibal-1 in the TNC were observed and quantified by immunofluorescence to investigate whether improving central sensitization by EA treatment was accompanied by inhibition of activated microglia. We further detected the underlying influence of EA on the expressions of Ibal-1, P2X4R, and associated inflammatory pathway mediators, including interleukin (IL-1β), NOD-like receptor protein 3 (NLRP3), and Caspase-1 in the TNC by western blotting and real-time polymerase chain reaction (PT-PCR) analysis, aiming to evaluate the anti-migraine effects of EA focusing on microglia and associated inflammation.

Materials and methods

Experimental animals

Adult male Sprague-Dawley (SD) rats (200-250g) were obtained from the Experimental Animal Centre of Chongqing Medical University (Chongqing, China). All rats were housed under specific pathogen-free (SPF) laboratory conditions (temperature 20–25°C, humidity 50%–70%) with a 12-hour light/dark cycle allowing free access to food and water. The subsequent experiments were conducted after one week of habituation acclimation. All animal experimental procedures were approved by the Ethics Committee of Chongqing Medical University. Every effort was made to reduce the number of rats in our experiments, and all surgeries were performed under anesthesia, aiming to minimize animal suffering.

Experimental design

Experiment I

To assess the behavioral characteristics of migraine model induced by inflammatory soup (IS) injections and the effects of EA on rats, we divided 50 rats randomly into five groups (n = 10): Blank group, phosphate-buffered saline (PBS) group, IS group, Sham-EA (SEA) group and EA group. The Blank group received no epidural cannula implantation, chemical stimulation, or intervention, only routine fixation. The other 4 groups received epidural cannula implantation and postoperative chemical stimulation. Among them, the chemical stimulus of the IS group, SEA group, and EA group was IS, while the chemical stimulus of the PBS group was PBS, a total of 7 times. Then, the PBS group and IS group were performed with routine fixation without any other intervention after chemical stimulation. In the SEA group, acupuncture was applied at sham acupoints, and only the electrodes were connected without electric stimulation. In contrast, the EA group was given bilateral GB20 and GB34 EA stimulation, once a day until the end of the experiment. The pain thresholds in the periorbital and posterior paw areas were measured by von-Frey filament on day 0 (baseline) and every other day of subsequent experiment (day 1, day 3, day 5, and day 7) to reflect the changes of pain thresholds in each group. The time courses and procedures of the experiment are shown in Figure 1.

Figure 1.

Diagram of the experimental protocol and schedule. Seven days of recovery after epidural cannula implantation, recorded day 7 of the recovery period as day 0 of the experiment and measured the mechanical withdrawal threshold by the von-Frey filament. Rats except for those in the Blank group received corresponding epidural chemical stimulation (IS or PBS injection) every day from day 1 to day 7 for a total of seven sessions, and rats in the EA and SEA groups received treatment with electroacupuncture and sham-acupoint acupuncture after daily epidural IS stimulation. The changes in the mechanical withdrawal threshold were assessed every other day (day 1, day 3, day 5, and day 7). TNC tissues of rats were collected after the seventh intervention.

Experiment II

To test whether migraine-associated pain induced by IS and forming a central sensitization state was accompanied by the microglial activation and the effect of EA in a rat model of migraine, the marker of neuronal activation (c-Fos) and microglia labeled with ionized calcium-binding adaptor molecule 1 (Ibal-1) were detected by Immunofluorescence (IF) analysis. After the last behavioral test, rats were sacrificed after sequential perfusion with 0.9% saline and precooled 4% paraformaldehyde (PFA) under anesthesia. The brain tissues of rats containing TNC were stripped and post-fixed in vials of 4% PFA solution at 4°C overnight, dehydrated and paraffin-embedded, and used for subsequent sectioning and IF staining aiming to detect the fluorescence intensity of c-Fos and Ibal-1.

Experiment III

To examine microglial purinergic receptor P2X4 expression and associated neuroinflammation on IS-induced migraine-associated pain and the effect of EA, the levels of Ibal-1, P2X4R, IL-1β, NLRP3, and Caspase-1 proteins were detected by western blot analysis and the contents of P2X4R, NLRP3, and Caspase-1 mRNA were evaluated by real-time reverse trancriptase polymerase chain reaction (RT-PCR). After the rats were euthanized under anesthesia, the stripped TNC of rats was placed into cryopreservation tubes and rapidly frozen at −80°C for the subsequent western blot and RT-PCR analysis.

Epidural cannula implantation and repeated chemical stimulation

The surgical procedure of implanting epidural cannula implantation was performed as reported previously, ¹⁸ and all rats underwent cannula implantation except for those in the Blank group without any intervention. After anesthesia with 3% pentobarbital sodium (3 mL/kg, intraperitoneally), rats were restrained on a stereotaxic instrument, and then the skull of them was fully exposed. A 1 mm diameter cranial hole (1 mm posterior to the bregma, 1.5 mm lateral to the midline) was made with a dental drill to expose the dura mater carefully and avoid damage. Then, a stainless steel inner cannula (62,001, RWD Life Science, China) was inserted into the hole, and the surrounding area was sealed with a mixture of dental cement for fixing. After a matching cap (62,101, RWD Life Science, China) was inserted into the inner cannula to ensure its patency, the skin incision was closed with a silk suture. The rats were returned to their separate cages while awake and mobile. During the seven days of the postoperative recovery period, we observed the overall state of the rats and sterilized the incisions. The seventh day of the convalescence period was recorded as day 0 of the subsequent experiments.

Except for the Blank group, without any intervention, the other four groups received chemical stimulation once a day for a total of seven times through the polyethylene tube (62,329, RWD Life Science, China) connected with the microsyringe and the cannula on the epidural. According to previously described methods,^18,19 in addition to the PBS group which was used an equal amount of PBS instead of IS, the other three groups, including the IS group, EA group, and SEA group, were given daily repeated injections with 10 ul IS at a rate over 5 minutes aiming to stimulate the dura. IS was composed of 1 mM serotonin (H9523, Sigma, USA),1 mM histamine (H7375, Sigma, USA),1 mM bradykinin (A304378, Aladdin, China) and 0.1 mM prostaglandin (P5640, Sigma, USA) in PBS at PH 7.4, and the rats were able to move freely during the period of administration.

Intervention with EA and SEA

Rats in the EA and SEA groups were restrained in the rat fixating device for a total of seven corresponding interventions following a daily injection of IS. According to the experience of clinical practice and previous studies reported,^15,20 Fengchi acupoint (GB20) was utilized most frequently in the treatment of migraine as a main acupuncture point, while Yanglingquan acupoint (GB34) was observed to strengthen the effect of GB20 in migraine treatment as a distant acupuncture point. Multiple studies confirmed that combined GB20 and GB34 acupuncture was superior to GB20 acupuncture alone in the treatment of migraine because of the enhancement efficacy of GB34.^15,16 As previous studies described,^14,15 the bilateral GB20 (located 3 mm lateral to the center of a line joining the two ears) and GB34 (located in the depression anterior and distal to the head of the fibula) of rats in the EA group were scrubbed with 75% alcohol disinfectant after fixation and then were inserted to a depth of 2-3 mm using stainless-steel acupuncture needles (13 mm х 0.25 mm, Suzhou Medical Appliance Factory, China). Following the needles connected to the electrical stimulator (SDZ-II, Suzhou Medical Appliance Factory, China), rats received 20 minutes of EA stimulation with the intensity of 0.5-1 mA and amplitude-modulated wave current of 2/15 HZ. The rats in the SEA group were inserted bilaterally at a distant sham-acupoint (about 10 mm above the iliac crest) with acupuncture needles. Then the needles were only connected to the electrical stimulator without electricity. ¹⁶ Rats in the other three groups, including the Blank group, PBS group, and IS group, were similarly placed onto the fixed devices to restrict mobilization, but did not receive acupuncture and electrical stimulation.

Sensory sensitivity testing

To determine pain sensitivity, we employed the von-Frey filament to evaluate mechanical withdrawal thresholds for responses to mechanical stimuli in both the facial and hind-paw areas according to the up-and-down method, as previous studies described.^21,22 After rats had an adaptation period of at least 30 min in the test environment, they were placed alone in the dedicated elevated test frame covered with transparent perspex chambers and wire mesh at the bottom, allowing the filament to make complete contact with their face and paw. For facial mechanical withdrawal threshold measures, the periorbital region, including the caudal areas of the eyes to the midline, approximately, was stimulated with perpendicular pressure by the tip of a series of von-Frey filaments. For paw mechanical withdrawal threshold measures, the central area of the plantar surface of the hind paw was stimulated from the underside of the mesh stand. Each filament was applied for three seconds with an interval of at least one minute, and a positive response to three or more of the five stimulations was considered a positive response for that filament. A positive response of periorbital mechanical sensitivity and paw mechanical sensitivity was defined as retracting quickly of the head, scratching the face with the fore-paw, or vocalizing, and lifting from the mesh stand, shaking, or licking of the paw, respectively. ²² All rats were first tested for mechanical pain thresholds of baseline (day 0) and then measured on every other day of experimental intervention (day 1, day 3, day 5, and day 7).

Immunofluorescence staining and image analysis

Sections were blocked with 3% bovine serum albumin (Beyotime, China) for 30 min after the tissues were sliced coronally. They were then incubated overnight at 4°C with primary antibodies diluted in PBS: rabbit anti-c-Fos and rabbit anti-Ibal-1. After being washed with PBS the next day, sections were incubated at room temperature for 50 min with cy3-labeled goat anti-rabbit immunoglobulin (IgG) as the secondary antibody. Then, nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI) for 10 min at room temperature. After being sealed with anti-fluorescence quenching liquid, the sections were visualized and scanned using Pannoramic P-MIDI (3D HISTECH, Hungary) and Pannoramic Scanner software (3D HISTECH, Hungary) at 40x magnification, respectively, and the immunofluorescence images were finally analyzed by Image J software. To quantify the relative fluorescence intensity of c-Fos and Ibal-1 positive expressions in the TNC, we investigated three visual fields randomly per section selected from each group (n = 3 rats/group). The antibodies used in immunofluorescence are listed in Table 1.

Table 1.

Antibodies used in immunofluorescence staining and western blot.

Antibodies	Manufacturer	Catalogue number	Host	Dilution
For immunofluorescence staining
Ibal-1	Abcam, UK	Ab178847	Rabbit	1:100
c-Fos	Cell signaling, USA	2250T	Rabbit	1:200
Cy3 goat anti-rabbit IgG	AiFang, China	AFSA006	Goat	1:200
For western blot
Ibal-1	Abcam, UK	ab178847	Rabbit	1:1000
P2X4R	Bioss, China	bs-7690R	Rabbit	1:1000
NLRP3	Novus, USA	NBP2-12446SS	Rabbit	1:1000
Caspase-1	Novus, USA	14F468	Mouse	1:1000
IL-1β	Abcam, UK	ab178847	Rabbit	1:1000
β-actin	Zenbio, China	200068-8F10	Mouse	1:10000
Anti-rabbit IgG	Zenbio, China	511,203	Goat	1:10000
Anti-mouse IgG	Zenbio, China	511,203	Goat	1:10000

Western blot analysis

TNC tissues were lysed in radioimmunoprecipitation buffer containing phenylmethylsulfonyl fluoride and then were homogenized at least three times with an interval of 10 s each time. Next, protein concentrations in the supernatants were measured using the bicinchoninic acid (BCA) protein assay kit (Beyotime, China) after centrifugation at 12,000 r for 20 min. Equal amounts of protein samples were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinylidene difluoride (PVDF) membranes (Millipore, USA) after electrophoresis. Then, membranes were blocked with 5% fat-free milk in Tris-buffered saline containing 0.1% Tween 20 (TBST) for 2 h at room temperature, followed by incubation with the corresponding primary antibodies overnight at 4°C: rabbit anti-Ibal-1, rabbit anti-P2X4R, rabbit anti-NLRP3, mouse anti-Caspase-1, rabbit anti-IL-1β, and mouse anti-β-actin. The next day, the membranes were washed with TBST three times. Then, the HRP-labeled goat anti-rabbit and HRP-labeled goat anti-mouse secondary antibodies were incubated for 1 hour at room temperature, respectively. After washing with TBST three times, the protein bands were visualized with an imaging system (LI-COR, USA) using an ECL hypersensitive luminescence kit (4A Biotech, China), and the relative density of proteins was quantified using Image J software. β-actin was used as a loading control to normalize the level of target proteins. The antibodies used in the western blot are listed in Table 1.

RT-PCR analysis

Total RNA was extracted from TNC tissues using an RNAiso Plus reagent (9108, Takara, Japan), and then the RNA concentration was quantified using a NanoDrop spectrophotometer (Thermo, USA). Next, cDNAs were synthesized in the T100^TM thermoCycler (Bio-Rad, USA) using a PrimeScript^TM RT reagent kit with gDNA Eraser (RR047 A, Takara, Japan). Finally, RT-PCR was performed with TB Green® Premix Ex Taq^TM II (RR820, Takara, Japan) on a CFX96 touch thermocycler (Bio-Rad, USA) with PCR cycle conditions as follows: first 95°C for 2 min, then 95°C for 5 s, 60°C for 30 s for 40 cycles. β-actin was used as a non-regulated reference gene for normalization, and the different mRNA expression levels were calculated using the 2^-△△CT method. The gene-specific primers used in RT-PCR were obtained from Sangon Biotech (Shanghai, China), and specific sequences are listed in Table 2.

Table 2.

Sequences and product length of primers used.

Gene	Primers sequence (5′-3′)	Product (bp)
P2X4R	Forward GTGACAACCAAAGCCAAAGG	185
	Reverse ATGCTGGTCTTATCAGGAATCTC
NLRP3	Forward AGCCTTGAAGAGGAGTGGATAG	177
	Reverse CTGGGTGTAGCGTCTGTTGAG
Caspase-1	Forward GCATTTCCTGGACCGAGTG	124
	Reverse GGGCAAGACGTGTACGAGTG
β-actin	Forward ACGGTCAGGTCATCACTATCG	155
	Reverse GGCATAGAGGTCTTTACGGATG

Statistical analysis

Statistical analysis and graphical representations were performed with GraphPad Prism 8.0. Data in this article were tested for normality before statistical analysis. One-way analysis of variance (ANOVA) followed by Tukey’s post-hoc test was used to compare multiple groups. Nonparametric distribution data sets were analyzed using the Kruskal-Wallis test, followed by Dunn’s multiple comparisons test. All data are presented as the mean ± standard deviation (SD). The results were considered significant at p < 0.05 in statistical comparisons.

Results

EA ameliorates mechanical allodynia induced by repeated IS stimulation

To investigate the changes in hyperalgesia after repeated IS stimulation and the effects of EA, facial and paw mechanical withdrawal thresholds were measured before the trial (day 0) and every other day during the experiments (day 1, day 3, day 5, and day 7) to evaluate the behavioral changes of different five groups. There was no significant difference in the mechanical threshold between the five groups (p > 0.05; Figure 2), both basal threshold on day 0 and after the first chemical stimulation and corresponding intervention on day 1. After the third infusion, the statistical difference appeared and became more significant with the increasing number of injections, both facial and paw withdrawal thresholds.

Figure 2.

The paw mechanical withdrawal threshold for each group at all time points. Repeated IS injections induced an increase in mechanical hyperalgesia while partly reversing the effects of IS on hyperalgesia by treatment with EA. Data are presented as mean ± SD (n = 8 rats/group). The blank column represents the Blank group. *p < 0.05, **p < 0.01, ***p < 0.001, Blank versus IS. ⏀p < 0.05, ⏀⏀p < 0.01, PBS versus IS. #p < 0.05, ##p < 0.01, ###p < 0.001, Blank versus SEA. ++p < 0.01, +++p < 0.001, PBS versus SEA. ^p < 0.05, IS versus EA. †p < 0.05, SEA versus EA.

As described in our previous study, ²³ for facial mechanical withdrawal threshold, it was observed that rats injected with IS were gradually decreased from day 3 to day 7 compared with the Blank group and PBS group. In contrast, the post-treatment mechanical threshold by EA was increased compared with the IS group with IS stimulation alone. In contrast, no significant differences in facial mechanical thresholds were observed between the IS and SEA groups at any time. For the paw mechanical withdrawal threshold, repeated IS injections induced a decrease in rats on day 3 compared with those of the Blank group (p < 0.05; Figure 2) and PBS group (p < 0.05; Figure 2). It was found that repeated IS injections induced a time-dependent increase in hyperalgesia. Especially, the administration of IS led to the significant aggravation of the hind-paw threshold on the fifth day compared with the Blank group (p < 0.01; Figure 2) and PBS group (p < 0.01; Figure 2), which was most obvious on the seventh day (Blank vs IS: p < 0.001, PBS vs IS: p < 0.01; Figure 2). Interestingly enough, there was no obvious statistical difference between those in the Blank group and PBS group (all p > 0.05; Figure 2), which provided an evidence that repeated PBS infusions did not establish migraine model as IS. Importantly, compared with the IS group, paw mechanical threshold showed an ascending trend by treatment with EA on the same measured time point (all p < 0.05; Figure 2), indicating that EA could attenuate the pain hypersensitivity while acupuncture at sham-acupoints did not improve migraine-associated hyperalgesia (all p > 0.05; Figure 2).

Thus, these data suggest that EA function contributed to the improving of pain hypersensitivity in migraine induced by repeated IS stimulation, while SEA did not play a similar role. The paw mechanical withdrawal thresholds for each group at all time points are shown in Figure 2.

EA suppresses IS-induced c-Fos expression in the TNC

c-Fos is closely related to the pathogenesis of migraine, which is considered a valuable indicator of neuronal activation in nociceptive pathways and a key mediator of central sensitization.^24,25 To investigate the effects of EA on central sensitization of migraine, we measured alterations in c-Fos expression in the TNC in different groups by IF staining, which is also a common means to mark c-Fos in studies of migraine models. ²⁶ As shown in Figures 3(a) and (b), the relative fluorescence intensity of c-Fos immunoreactive cells of TNC in the IS group was increased compared with that in the Blank group (p < 0.05) and PBS group (p < 0.01). However, no distinct difference was observed between the Blank group and PBS group (p > 0.05), indicating that repeated infusions of PBS failed to cause migraine-associated neuronal activation and central sensitization as in the IS group. In contrast, the increased immunofluorescence intensity of c-Fos induced by IS was partly abolished using EA intervention (p < 0.05), whereas the SEA group did not significantly differ from those in the IS group (p > 0.05). Altogether, these results suggested that EA treatment could alleviate IS-induced central sensitization. Representative fluorescence images of c-Fos immunopositive cells in the TNC are shown in Figure 3(a), and quantitative analysis of the relative fluorescence intensity of c-Fos positive cells in the TNC for each group is in Figure 3(b).

Figure 3.

(a) Representative fluorescence images of c-Fos immunopositive cells in the TNC (40×, Scale bar = 50 μm) for each group: (a) Blank; (b) PBS; (c) IS; (d) SEA; (e) EA. (b) Quantitative analysis of each group’s relative fluorescence intensity of c-Fos positive cells in the TNC. Repeated IS injections induced an increase in the relative fluorescence intensity of c-Fos-positive cells, while it decreased with treatment with EA. Data are presented as mean ± SD (n = 3 rats/group). *p < 0.05, Blank versus IS. ⏀⏀p < 0.01, PBS versus IS. #p < 0.05, Blank versus SEA. ++p < 0.01, PBS versus SEA. ^p < 0.05, IS versus EA. †p < 0.05, SEA versus EA. (c) and (d) Quantitative analysis of the relative fluorescence intensity and number of Ibal-1 positive cells in the TNC for each group. Repeated IS injections induced an increase in the relative fluorescence intensity and number of Ibal-1-positive cells, while decreased by treatment with EA. Data are presented as mean ± SD (n = 3 rats/group). The blank column represents the Blank group. *p < 0.05, **p < 0.01, Blank versus IS. ⏀p < 0.05, ⏀⏀p < 0.01, PBS versus IS. ##p < 0.01, Blank versus SEA. ++p < 0.01, PBS versus SEA. ^p < 0.05, IS versus EA. †p < 0.05, SEA versus EA.

EA attenuates IS-induced microglial activation in the TNC

Multiple previous studies reported that activated microglia participated in migraine-associated central sensitization mainly featured hypertrophic somata and a shortened process.^22,27 To explore whether c-Fos mediated central sensitization was accompanied by microglial activation in migraine rat models and the effects of treatment with EA, the changes in microglia were evaluated by IF labeling Ibal-1. The immunofluorescence results of Ibal-1 were consistent with the c-Fos, as shown in Figure 3(c) and (d), Ibal-1 labeled microglia were observed sparsely in the TNC in the Blank group and PBS group, but the relative fluorescence intensity and number of Ibal-1 labeled cells were significantly higher than that in the Blank group (p < 0.01; Figure 3(c), p < 0.05; Figure 3(d)) and PBS group (p < 0.01; Figure 3(c), p < 0.05; Figure 3(d)) after repeated infusions of IS and we also observed activated microglia marked hypertrophic somata. Compared to that in the IS group, treatment with EA (p < 0.05; Figure 3(c), p < 0.05; Figure 3(d)) suppressed the IS-induced activation of microglia in the TNC, but acupuncture at the sham-acupoint did not significantly influence the activation of microglia in the TNC like EA (p < 0.05; Figure 3(c), p < 0.05; Figure 3(d)). Therefore, these results demonstrated that microglial activation plays a critical role in the TNC in the IS-induced migraine-associated central sensitization and EA could partly inhibit activated microglia to be involved in the process of anti-migraine. Quantitative analysis of the relative fluorescence intensity and number of Ibal-1 positive cells are shown in the TNC for each group in Figure 3(c) and (d). Representative fluorescence images of Ibal-1 immunopositive cells in the TNC are shown in Figure 4.

Figure 4.

Representative fluorescence images of Ibal-1 immunopositive cells in the TNC (40×, Scale bar = 50 μm) for each group: (a) Blank; (b) PBS; (c) IS; (d) SEA; (e) EA. Repeated IS injections induced activated microglia featured by hypertrophic somata and shortened process, while alleviated by treatment with EA.

EA depresses the expression of Ibal-1 labeled microglia and microglial receptor P2X4

To further determine whether the IS-induced migraine-associated central sensitization corresponds to microglia activation, we examined the Ibal-1 protein level using western blot analysis. The changes in the protein expression of Ibal-1 were consistent with the IF results, that is, the protein levels of Ibal-1 were increased after recurrent IS injections compared to that in the Blank group (p < 0.0001; Figure 5) and PBS group (p < 0.01; Figure 5). Similarly, the protein level of Ibal-1 in the TNC of migraine rats was significantly reversed by EA treatment (p < 0.01; Figure 5), while treatment with SEA (p > 0.05; Figure 5) failed to decrease the expression of Ibal-1.

Figure 5.

Levels of Ibal-1 and P2X4R protein expression in the TNC assessed by western blot analysis for each group. (a) Representative immunoblots and densitometry analysis of Ibal-1; (b) Representative immunoblots and densitometry analysis of P2X4R. Repeated IS injections induced an increase in the protein expression of Ibal-1 and P2X4R, while decreased by treatment with EA. Data are presented as mean ± SD (n = 6 rats/group). The blank column represents the Blank group. ****p < 0.0001, Blank versus IS. ⏀⏀p < 0.01, ⏀⏀⏀p < 0.001, PBS versus IS. ###p < 0.001, ####p < 0.0001, Blank versus SEA. ++p < 0.01, +++p < 0.001, PBS versus SEA. ^^p < 0.01, ^^^p < 0.001, IS versus EA. ††p < 0.01, †††p < 0.001, SEA versus EA.

In addition, we examined the levels of P2X4R in the TNC using western blot and RT-PCR analysis, as P2X4R is mainly involved in central sensitization of neuropathic pain acting as a microglial purinoceptor and activated microglia can induce and enhance its expression. ⁸ The results showed that statistically significant upregulation of microglial P2X4R in protein and mRNA levels were observed after the repeated IS administration compared with the Blank (p < 0.0001; Figure 5(a), p < 0.01; Figure 7(a)) and PBS group (p < 0.001; Figure 5(a), p < 0.01; Figure 7(a)). However, EA (p < 0.001; Figure 5(a), p < 0.05; Figure 7(a)) markedly downregulated IS-induced upregulation of P2X4R both in protein and mRNA levels compared with the IS group, whereas SEA did not exhibit similar effects because no significant difference was observed between the IS and SEA group (p > 0.05; Figures 5 and 7). These results indicated that EA could exert anti-migraine effects by attenuating activated microglia and its receptor P2X4 expression in the TNC of migraine rat models induced by repeated IS infusions.

EA attenuates microglia-associated neuroinflammation triggered by repeated IS stimulation

Neuroinflammation plays a crucial role in the pathological mechanism of migraine, and inhibiting neuroinflammation may contribute to improving the pathology of migraine. ²⁸ IL-1β, as a classical inflammatory factor, is mainly derived from microglia in the central nervous system (CNS). ²⁹ To validate changes of neuroinflammation in the TNC of migraine model rats induced by infusions of IS and the effects of EA, we performed a western blot analysis to detect the protein expression of IL-1β. The results showed that the protein levels of IL-1β were markedly upregulated after repeated injections of IS compared with the Blank group (p < 0.05; Figure 6(a)) and PBS group (p < 0.05; Figure 6(a)), while no significant difference was observed in the latter two groups (p > 0.05; Figure 6(a)). After treatment with EA (p < 0.01; Figure 6(a)), the IL-1β expression levels induced by IS were substantially reduced, whereas there was no significant decrease of IL-1β levels in the SEA group (p > 0.05; Figure 6(a)).

Figure 6.

Levels of IL-1β, NLRP3, and Caspase-1 protein expression in the TNC assessed by western blot analysis for each group. (a)Representative immunoblots and densitometry analysis of IL-1β; (b) Representative immunoblots and densitometry analysis of NLRP3; (c) Representative immunoblots and densitometry analysis of Caspase-1. Repeated IS injections increased in the protein expression of IL-1β, NLRP3, and Caspase-1, while decreased by treatment with EA. Data are presented as mean ± SD (n = 6 rats/group). The blank column represents the Blank group. *p < 0.05, ***p < 0.001, Blank versus IS. ⏀p < 0.05, ⏀⏀p < 0.01, ⏀⏀⏀p < 0.001, PBS versus IS. #p < 0.05, ###p < 0.001, Blank versus SEA. ++p < 0.01, +++p < 0.001, PBS versus SEA. ^p < 0.05, ^^p < 0.01, ^^^p < 0.001, IS versus EA. †p < 0.05, ††p < 0.01, SEA versus EA.

The maturation of IL-1β depends on NLRP3 inflammasome-mediated activation of Caspase-1, ³⁰ so we further measured the expression of NLRP3 and Caspase-1. Western blot analysis showed that the protein levels of NLRP3 and Caspase-1 increased significantly compared with the corresponding levels in the Blank group (p < 0.05; Figure 6(b), p < 0.001; Figure 6(c)) and PBS group (p < 0.01; Figure 6(b), p < 0.001; Figure 6(c)) after the repeated administration of IS, while the protein levels of IL-1β in the TNC were also decreased in the groups treated with EA (p < 0.05; Figure 6(b), p < 0.001; Figure 6(c)). However, there was no obvious difference between the SEA group and IS group (p > 0.05; Figure 6(b) and (c)).

To further investigate, we performed mRNA level verification. Consistent with the alterations in western bolt results, the mRNA levels of NLRP3 and Caspase-1 in the TNC were also upgraded in IS-induced migraine rat models compared to that in the Blank group (p < 0.01; Figure 7(b), p < 0.05; Figure 7(c)) and PBS group (p < 0.001; Figure 7(b), p < 0.05; Figure 7(c)). The treatment with EA (p < 0.05; Figure 7(b), p < 0.01; Figure 7(c)) showed an inhibitory effect on NLRP3 and Caspase-1, but there was no statistical significant treatment by SEA (p > 0.05; Figure 7(b) and (c)). Collectively, the results demonstrated that the IS-induced increase in inflammatory mediators associated with activated microglia could be regulated by EA.

Figure 7.

Levels of P2X4R, NLRP3, and Caspase-1 mRNA expression in the TNC assessed by RT-PCR analysis for each group. (a) P2X4R mRNA expression levels; (b) NLRP3 mRNA expression levels; (c) Caspase-1 mRNA expression levels. Repeated IS injections increased in the mRNA expression of P2X4R, NLRP3, and Caspase-1, while decreased by treatment with EA. The blank column represents the Blank group. Data are presented as mean ± SD (n = 3 rats/group). *p < 0.05, **p < 0.01, Blank versus IS. ⏀p < 0.05, ⏀⏀p < 0.01, ⏀⏀⏀p < 0.001, PBS versus IS. #p < 0.05, ##p < 0.01, Blank versus SEA. +p < 0.05, ++p < 0.01, +++p < 0.001, PBS versus SEA. ^p < 0.05, ^^p < 0.01, IS versus EA. †p < 0.05, ††p < 0.01, SEA versus EA.

Discussion

Although accumulating studies have been published on the role of microglia and inflammation in pain-associated diseases, rare studies have investigated whether EA inhibition of microglial activation and related release of its receptor and inflammatory mediators is one of the causes of anti-migraine effects. In the study, we established an IS-induced migraine rat model, followed by mechanical pain thresholds were tested to assess the allodynia. Central sensitization was evaluated by measuring c-Fos levels. Moreover, we found activated microglia accompanied by increased expression of its P2X4R and release of IL-1β, which is mediated by the NLRP3/Caspase-1 inflammatory pathway in the TNC. The morphology and levels of microglia labeled with Ibal-1 demonstrated the microglia activation, and the levels of IL-1β, NLRP3, and Caspase-1 indicated inflammatory response. Importantly, treatment with EA protected IS-induced hyperalgesia by inhibiting microglia activation, reducing the release of P2X4R and neuroinflammation, and attenuating aggravation of central sensitization.

EA can improve the central sensitization state of migraine

In the present study, we used repeated epidural IS injections to establish the migraine animal model, which is now generally recognized as reliable and widely used in research.^16,17 Allodynia, manifested by abnormal skin pain in the head and limbs, is a clinical feature of patients with chronic migraine ³¹ and an external manifestation of central sensitization in the pathogenesis of migraine chronification. ³² In the study, repeated IS injections produced a significant pain hypersensitivity, which was first observed after the third IS injection, while PBS injections did not evoke significant mechanical hyperalgesia like IS. The results of our study are consistent with the literature and those observed in migraine animal models that have been reported that repeated epidural administration of IS led to the CA effect.^17,32 The appearance of allodynia following IS injections might reflect the development of central sensitization, which may be mediated through abnormal neuronal excitability in the TNC and third-order neurons (thalamus). These pain hypersensitivity results, measured by periorbital and hind paw mechanical pain thresholds, correspond to the allodynia of the cephalic and extracephalic area in clinical patients with chronic migraine. Moreover, periorbital and hind paw mechanical thresholds were all improved in our model after the treatment with EA, which is consistent with the previous studies of anti-hyperalgesia effects of EA in the migraine model.^14,16 This allodynia was alleviated after the EA intervention, suggesting that EA effectively in processedanti-migraine chronification.

Numerous studies have revealed central sensitization’s crucial role in migraine chronification pathophysiology. ³³ Considering that central sensitization mainly occurred in the TNC area, ³ we accordingly focused on the TNC as the target region to observe changes in biomarkers. c-Fos has been considered a reliable marker mediating central sensitization in pain, and its expression levels reflect neuronal activity. ³⁴ Studies in different animal models of migraine demonstrated a distinct upregulation of c-Fos in the TNC.^8,16 Therefore, we examined c-Fos expression in the TNC to assess the corresponding central sensitization state and whether the pain hypersensitivity induced by repeated IS stimulation corresponded to an increase in c-Fos expression, and more importantly, further investigated the effect of EA on the central sensitization. Consistent with previous data,^8,16 we also found activating c-Fos in the TNC of migraine rat models in this study, and elevated levels corresponded to hyperalgesia, indicating abnormally excited neurons and a significant central sensitization state. In addition, in line with the previous study, ¹⁶ our finding showed that treatment with EA suppressed c-Fos levels, demonstrating that EA could attenuate central sensitization and partly relieve migraine rats from suffering allodynia. Conversely, the maintenance of central sensitization occurred when acupuncture at sham acupoints. These results of behavioral test and c-Fos expression indicated that rats could develop a central sensitization state after repeated IS stimulation. EA could contribute to ameliorating this status and thus participate in the suppression of migraine chronification.

Underlying molecular mechanisms of EA on anti-hyperalgesia in the migraine rat model

Activated microglia with its up-regulated receptor

Considerable studies on animal models of neuropathic pain have shown that activation of microglia and subsequent secretion of mediators play an important role in initiating or maintaining the hyperalgesic state.^7,35 The alterations in microglia are necessary for developing of central sensitization and chronic pain via interacting with neurons. ⁴ Previous data have demonstrated that microglia activation in the TNC of a migraine model was involved in central sensitization.^8,27 Changes in microglial morphology (proliferation, cell body hypertrophy, and branch shortening) and increased number in the TNC of previous nitroglycerin (NTG)-induced migraine model were observed. ²² In this study, we also found that repeated IS injections could activate microglia in the TNC, characterized by similar morphological changes and an increase in the relative immunofluorescence intensity and number of microglia. In addition, there is also a consistent change in the protein expression level of microglia labeled with Ibal-1 in our model.

Additionally, previous studies have shown that genetic knockout of microglia signaling molecules partially reverses neuropathic pain. ⁷ In the migraine model, the mechanical hyperalgesia and immunoreactivity staining of Ibal-1 were alleviated by using a microglia inhibitor. ⁸ Considering the effect of microglia in the pathogenesis of migraine chronification, the possibility of preventing migraine development and chronification by locking microglia over-expression increases. ⁵ Hence, we measured the changes in microglia after continuous treatment with EA, and found that reduced activated microglia and suppressed Ibal-1 levels, indicated that EA could regulate microglia’s activation and morphological changes in our model. From these data, we conclude that EA could inhibit microglial activation and thus regulate the development of IS-induced central sensitization.

Microglia interact with the brain micro-environment via purinergic ionotropic receptors signaling during injury and pathology. ³⁶ P2X4 receptor, as one of the most important members of the microglial purinoceptors, increased expression has been extensively reported in animal models of neuropathic pain. ³⁷ Recent data have found that microglia purinoceptor P2X4 was co-expressed with microglia in the TNC of migraine models, and the expression of P2X4R was increased, suggesting that the activation of P2X4R promoted the occurrence and progression of migraine. ⁸ In line with these results, our findings demonstrated the increased levels of P2X4R in the TNC after repeated IS injections. Since the changes in P2X4R were consistent with the upregulation of Ibal-1 and c-Fos in the TNC, and considering the co-expression of P2X4R with microglia, we believed that the changes in P2X4R in our migraine model also may be due to the activation of microglia. Involvement of P2X4R in central sensitization may also be due to microglia-neuron interactions since c-Fos was produced by neuron. Moreover, previous data confirmed that P2X4 inhibitor significantly reduced pain-related behaviors and microglial activation in neuropathic pain, ³⁸ and it was found to significantly alleviate c-Fos expression in the TNC of a migraine model accompanied by the improvement of the pain threshold. ⁸ Our above data demonstrated the reduction of activated microglia and the alleviation of pain behaviors by the treatment with EA. Therefore, it was reasonable to hypothesize that a decrease in the release of P2X4R may accompany this effect of EA. Indeed, P2X4R was decreased both protein and mRNA levels by the EA intervention, which tested our hypothesis.

Inflammatory response

The neuroinflammation and activated neuroimmune are thought to mediate the central sensitization and the development of allodynia.^39,40 Recent studies have shown that activated microglia, which produce various inflammatory cytokines, promote neuroinflammation, and ultimately induce neuronal cell damage, which involves upregulation of P2X4R. ⁴¹ IL-1β is one of the most well-studied inflammatory cytokines primarily derived from microglia in the CNS, ²⁹ and increased levels of IL-1β are often observed upon CNS infection and brain injury. ⁴² Based on accumulating evidence, IL-1β is also involved in migraine pathology.^6,43,44 Supporting these findings, the expression of IL-1β in the TNC distinctly increased in our repeated IS-induced migraine model, indicating an inflammatory response in the TNC. Since IL-1β expressed by microglia binds to a specific receptor, IL-1R, primarily expressed in neurons, to regulate neuronal excitability, the mechanism of IL-1β-mediated central sensitization may be related to the microglia-neuron pathway. ⁴⁵ Previous data showed that blockade of IL-1β improved hyperalgesia and inhibited the increase in the c-Fos related to central sensitization. ⁶ Treatment with EA in our study caused the downregulation of IL-1β, which corresponded with relieved pain hypersensitivity and inhibited microglia, suggesting that EA could ameliorate hyperalgesia through inhibiting neuroinflammation mediated by activated microglia. These results are similar to the previous study, ¹⁵ which showed that IL-1β decreased to different degrees in the migraine model by treatment of EA, but different from the previous study, which mainly focused on the change of IL-1β in serum, that we focused on the changes of IL-1β in the CNS after EA intervention due to considering the relationship between neuroinflammation and central sensitization.

The maturation of IL-1β depends on the NLRP3 inflammasome, which mediates the activation of Caspase-1, which then cleaves pro-IL-1β to form mature IL-1β. ⁴⁶ Microglia are the prominent innate immune cells in the CNS that activate inflammasome. ⁴⁷ Since the changes in IL-1β were consistent with microglial activation in the TNC of our model, therefore, it was reasonable to consider that the NLRP3/Caspase-1 pathway also played a role in migraine and participated in hyperalgesia associated with central sensitization by involving in inflammatory responses. Previous studies have indicated that NLRP3 inflammasome in the cascade of events involved in migraine by promoting IL-1β maturation.^6,48 That is, NLRP3 was mainly expressed in microglia in the TNC of migraine models, and the activation of the NLRP3 in activated microglia led to the release of the downstream effector cytokine IL-1β and participated in the formation of central sensitization by NLRP3/Caspase-1/IL-1β inflammatory signaling pathway. Consistent with these results, we found that both the protein and mRNA levels, NLRP3 and Caspase-1, as important upstream mediators of the IL-1β, were significantly increased in the TNC after repeated IS stimulation, which further suggested that inflammatory mediators were involved in central sensitization. A previous study reported that blocking the NLRP3 inflammatory signaling pathway could attenuate mechanical allodynia, reduce the expression levels of c-Fos, and inhibit the microglia activation in migraine models. ⁶ Moreover, the previous data has shown that EA can significantly inhibit the NLRP3 signaling pathway in neurological diseases. ⁴⁹ A pain model showed that the EA induced anti-inflammatory responses to reduce pain through modulating microglial activation, indicating microglial activation and neuroinflammation might be an important mechanism for EA analgesia. ¹² In the study, intervention with EA partly reversed IS-induced increases in expression levels of NLRP3 and Caspase-1, reduced accumulation in the TNC, and promoted the limitation of inflammation. Moreover, we found that the downregulation of the NLRP3 inflammatory pathway was consistent with mechanical pain threshold, increased expression of c-Fos, and activated microglia decreased after continuous EA intervention. From these data, we conclude that microglial activation during IS injection involved the release of inflammatory mediators and that modulation of EA targeting to this process may be a mechanism for attenuating the central sensitization-associated abnormal allodynia.

Inflammasome signaling in the CNS is mainly attributed to microglia. ⁴⁷ Microglia use purinergic ionotropic receptors signaling to crosstalk with neurons for complex neural-glial cell interactions, ³⁶ and the purinoceptor P2X4 belongs to the priority receptor involved in NLRP3 inflammasome activation, ⁵⁰ and its prolonged activation leads to NLRP3 inflammasome assembly and IL-1β release, as well as microglial activation and proliferation. ⁵¹ Previous studies have shown that disrupting the ATP-driven interaction by knocking out P2X4R provoked an attenuation of cell death, dye uptake, and IL-1β release. ⁵² Another one demonstrated that the inhibitor of P2X4R markedly decreased the level of IL-1β. ³⁸ In a recent clinical study of chronic inflammation, the expression of these factors, including P2X4R, NLRP3, and Caspase-1 expression correlated with circulating markers of disease severity, were significantly elevated in chronic inflammation. ⁵³ Therefore, with microglia as the medium, due to the relationship between P2X4R and NLRP3, that is, P2X4R as a purinoceptor on microglia, its activation leads to the activation of inflammasome in microglia and the subsequent maturation and release of inflammatory factors, which then bind to specific inflammatory factor receptors expressed on the neuron, that together causes abnormal excitation of neurons and mediates central sensitization. In the study, we have only observed that EA can exert the subsequent anti-inflammatory and anti-central sensitization effects by inhibiting microglia activation. At present, we only observed this phenomenon, and the specific mechanism of the activation of microglia inhibited by treatment with EA needs to be further performed.

Conclusions

In summary, through a migraine rat model, we explored the relationship between central sensitization, activated microglia, associated purinoceptor P2X4, and the subsequent inflammatory mediators after repeated epidural IS stimulation, and more importantly the effects of EA on migraine were explored from the above fields. These results suggest that EA may ameliorate migraine-associated central sensitization and together exert anti-hyperalgesia effects by regulating the microglial activation that accompanies the release of its purinoceptor P2X4 and associated inflammatory mediators (NLRP3/Caspase-1/IL-1β). Our study may provide new evidence based on central mechanisms to enhance the anti-hyperalgesia effects of EA in migraine.