Electroacupuncture Attenuates Fibromyalgia Pain via Toll-like Receptor 4 in the Mouse Brain

Background: Fibromyalgia (FM) is characterized by complex pain symptoms lacking impersonal considerations in diagnosis and treatment evaluation, which often happens in women. Chronic and persistent widespread pain is the key symptom disturbing patients with FM, leading to depression, obesity, and sleep disturbances. Toll-like receptor 4 (TLR4) activation produces a harmful sensory input involved in central pain; this is the focus of this study. Electroacupuncture (EA) has beneficial effects in reducing FM pain, but its connection with TLR4 signaling is still unknown. Methods: Intermittent cold stress significantly induced mechanical and thermal pain. EA, but not sham EA, reliably attenuated mechanical and thermal hyperalgesia. The increased inflammatory mediators in FM mice were reduced in the EA group, but not in the sham group. Results: All TLR4 and related molecule levels increased in the FM mice’s hypothalamus, periaqueductal gray (PAG), and cerebellum. These increases could be attenuated by EA but not sham stimulation. Activation of TLR4 by lipopolysaccharide (LPS) significantly induced FM and can be further reversed by a TLR4 antagonist. Conclusions: These mechanisms provide evidence that the analgesic effect of EA is related to the TLR4 pathway. In addition, we showed that inflammation can activate the TLR4 pathway and provided new possible therapeutic targets for FM pain.


Introduction
Fibromyalgia (FM) is accompanied by chronic general musculoskeletal pain lasting over 3 months and is highly associated with sleep disturbances, fatigue, and cognitive and somatic symptoms. These symptoms may persist over several years and cause extensive healthcare use [1]. For clinical diagnosis, FM can be characterized by trigger pain with tenderness in 18 special points [2]. The prevalence of FM ranges from 0.2 to 4.7%; almost 90% of the cases occur in women, of which >50% were aged 40-60 years. The new FM diagnosis is based on widespread pain index (WPI) ≥ 7 and a symptom severity scale (SS) ≥ 5 or WPI 3-6 and SS ≥ 9 for over 3 months [3,4]. The standard treatment into four main groups: normal mice (Group 1: normal); fibromyalgia mice (Group 2: FM); 2 Hz EA group (Group 3: 2 Hz EA); and sham EA group (Group 4: sham EA).

FM Model and Bio-Plex ELISA
Mice were maintained in the laboratory at 24 ± 1 • C before the experiments. The intermittent FM process, which was not achieved in normal mice, required that two mice were accommodated in a flexible cage (13 cm × 18.8 cm × 29.5 cm) that was circumscribed with a stainless mesh. On day 0, the mice were placed in a cold chamber at 4 • C overnight (from 4:00 p.m. to 10:00 a.m.). The mice were subsequently moved to 24 • C for 30 min at 10:00 a.m. and further transported to the cold room for 30 min. The course continued till 4:00 p.m. The mice were then placed in the 4 • C cold room overnight. Normal mice were maintained at 24 ± 1 • C [12]. Mouse plasma was collected and confirmed through Q-view cytokine assays (Q-view, Los Altos, CA, USA).

EA Treatments
All mice were anesthetized with 5% isoflurane at the beginning and 1% isoflurane for maintenance. Stainless-steel acupuncture needles (1 inch, 36 G, YU KUANG, Taipei, Taiwan) were bilaterally implanted into the ST36 acupoints of mice under anesthesia. The mouse ST36 acupoint is located~3-4 mm deep and 1-2 mm lateral to the knee joint. In the EA group, electrical stimuli were delivered by a Trio 300 stimulator (Ito, Tokyo, Japan) by using 1 mA for 20 min at 2 Hz with a pulse width of 100 µs. Muscle twitching was observed during EA treatment as a de-qi sensation. Subsequent FM induction, EA management was performed from day 5 to 7. The same protocol was applied in the sham group but without electrical stimulation [12].

Pain Behavior Test
After FM induction, either mechanical or thermal hyperalgesia was verified thrice from day 5 to day 7. Before behavior examinations, mice were relocated to the behavioral room and left to adjust to the room for ≥30 min. All tests were achieved at 24 ± 1 • C; the stimuli were tested when the mice were silent. Mechanical hyperalgesia was confirmed by applying force to stimulate trice the electronic von Frey filament (IITC Life Science Inc., Woodland Hills, CA, USA). Furthermore, the Hargreaves' test was applied to measure thermal hyperalgesia by analyzing the latency-to-thermal stimulation thrice by using Hargreaves' test IITC analgesiometer (IITC Life Sciences, SERIES8, Model 390G). In the Hargreaves' test, the pain threshold was measured via the latency of paw withdrawal time [12].

Intracerebroventricular Injection
Mice were anesthetized by using isoflurane with their heads fixed in a stereotaxic cannula, and the cannula was implanted at the ventricle site. The stereotaxic cannula was subsequently placed 0.5 mm in the anteroposterior axis, ±1 mm in the mediolateral axis, and −2.5 mm in the dorsoventral axis below the cortical surface. This cannula was a 23 gage, including a 2 mm stainless steel. In addition, it was immobile at the skull with a dental glue. Subsequently, the cannula was implanted and linked to a Hamilton syringe with a PE tube (PE10, Portex, Kent, UK). Totally, 5 µL of LPS/PBS (5 µL/ventricle) was injected over a period of 5 min by using a syringe pump (KD Scientific, Shanghai, China). After the injection, the cannula was left at the ventricle for an additional 2 min to allow the LPS to diffuse. A 25 µM TIRAP was used as an antagonist of TLR4.

Statistical Analysis
Statistical analysis was performed with SPSS statistical software. All results are shown as mean ± standard error (SEM). Shapiro-Wilk examination was used to test the normal distribution of the data. Statistical differences were investigated through repeated-measures ANOVA tests. The post hoc Tukey's test was performed after ANOVA. p < 0.05 indicated the threshold for statistical significance.

Electroacupuncture Attenuated Fibromyalgia-like Pain in Mice
We tested the therapeutic effects of EA in a mouse FM model to address EA function in the subacute pain model to verify whether EA can reduce FM. Before FM induction, all groups showed similar mechanical thresholds. Cold stress significantly induced mechanical pain ( Figure 1A, red column, D7: 1.32 ± 0.31 g, * p < 0.05, F(3, 36) = 24.41, n = 10), as shown by the von Frey test. Mechanical hyperalgesia was substantially alleviated by EA, but  Figure 1A, blue and green columns, D7: 4.12 ± 0.29 g and 1.31 ± 0.34 g, n = 10, respectively). This cold stress model reliably mimicked clinical symptoms, including thermal hyperalgesia. Next, we examined whether EA could alleviate thermal pain in FM mice. The Hargraves' test showed significant thermal hyperalgesia, as measured by the paw withdrawal latency after CSP induction ( Figure 1B, red column, D7: 3.18 ± 0.39 s, * p < 0.05, F(3, 36) = 30.02, n = 10). The hyperalgesic latency could be further reversed by EA but not by sham EA, suggesting an acupoint-specific effect ( Figure 1B, D7: 8.63 ± 0.39 s and 3.68 ± 0.38 s, n = 10, respectively).
We tested the therapeutic effects of EA in a mouse FM model to address EA function in the subacute pain model to verify whether EA can reduce FM. Before FM induction, all groups showed similar mechanical thresholds. Cold stress significantly induced mechanical pain ( Figure 1A, red column, D7: 1.32 ± 0.31 g, * p < 0.05, F(3, 36) = 24.41, n = 10), as shown by the von Frey test. Mechanical hyperalgesia was substantially alleviated by EA, but not sham EA ( Figure 1A, blue and green columns, D7: 4.12 ± 0.29 g and 1.31 ± 0.34 g, n = 10, respectively). This cold stress model reliably mimicked clinical symptoms, including thermal hyperalgesia. Next, we examined whether EA could alleviate thermal pain in FM mice. The Hargraves' test showed significant thermal hyperalgesia, as measured by the paw withdrawal latency after CSP induction ( Figure 1B, red column, D7: 3.18 ± 0.39 s, * p < 0.05, F(3, 36) = 30.02, n = 10). The hyperalgesic latency could be further reversed by EA but not by sham EA, suggesting an acupoint-specific effect ( Figure 1B, D7: 8.63 ± 0.39 s and 3.68 ± 0.38 s, n = 10, respectively).

TLR4 Signaling in the Cold-Stress-Induced Pain Mouse Cerebellum Was Attenuated by Electroacupuncture but Not Sham Electroacupuncture
Because the cerebellum plays a critical role in FM pain, we evaluated Iba1 and TLR4 signaling pathways in the mouse cerebellum after FM. Our data showed that Iba1 and TLR4 expression significantly increased after FM induction in the cerebellum lobule VI

Activation of TLR4 by LPS Injection Mimics CSP and Further Reversed by TLR4 Antagonist TIRAP
Recent articles showed that alterations in higher brain areas are essential for developing chronic pain [12,28]. Therefore, we hypothesized that activation of central TLR4 is involved in the central sensitization of chronic pain. Intracerebral ventricle infusion of LPS, which is an agonist of TLR4, significantly induced mechanical hyperalgesia (Figure Furthermore, we tried to identify whether TLR4 and Iba1 levels were altered and distributed in the mouse cerebellum, a brain area involved in patients with FM. Our data indicated that TLR4 was distributed in the CVI and VII in normal mice. TLR4 levels further increased after FM induction in the mouse CVI and VII (Figures 4 and 5C, green color, n = 4) and were then reversed by EA, but not sham EA (Figures 4 and 5C, green, n = 4). Similarly, Iba1 increased with FM and was attenuated in EA, but not in sham EA (Figures 4 and 5C, red color, n = 4). Moreover, double-staining of TLR4 and Iba1 were observed in the mice CVI and VII (Figures 4 and 5C, yellow color, n = 4); a pattern abrogated by EA but not sham treatment.

Discussion
FM is a complicated symptom with chronic pain, fatigue, sleep disturbance, obesity, and depression affecting the whole body. FM is often resistant to opioid drugs or nonsteroidal anti-inflammatory drugs [28]. Additionally, antidepressants SNRI and GABAergic medications (pregabalin) are the most commonly used drugs with several side effects. Accordingly, EA has a beneficial effect and chance to relieve FM. Vas et al. reported that acupuncture is an effective treatment for FM [29]. Langhorst et al. also verified that acupuncture treatment for FM is greater than medicines. Acupuncture combined with medicines and exercise noticeably increases pain thresholds [30]. Neyama and Ueda initiated an ideal FM model named the ICS model, which mimics similar symptoms in patients with FM. Gabapentin had a significant antiallodynic effect, but not in the morphinetreated group. Patients with FM also have characteristic emotional problems, including depression. Our recent publication indicated that mice subjected to the ICS process presented depression symptoms as in the human clinical presentation [12]. Our previous data suggest that acid saline injection-induced chronic pain and depression comorbidity produce changes in the cerebellum lobules VI, VII, and VIII. EA significantly ameliorated

Discussion
FM is a complicated symptom with chronic pain, fatigue, sleep disturbance, obesity, and depression affecting the whole body. FM is often resistant to opioid drugs or nonsteroidal anti-inflammatory drugs [28]. Additionally, antidepressants SNRI and GABAergic medications (pregabalin) are the most commonly used drugs with several side effects. Accordingly, EA has a beneficial effect and chance to relieve FM. Vas et al. reported that acupuncture is an effective treatment for FM [29]. Langhorst et al. also verified that acupuncture treatment for FM is greater than medicines. Acupuncture combined with medicines and exercise noticeably increases pain thresholds [30]. Neyama and Ueda initiated an ideal FM model named the ICS model, which mimics similar symptoms in patients with FM. Gabapentin had a significant antiallodynic effect, but not in the morphine-treated group. Patients with FM also have characteristic emotional problems, including depression.
Our recent publication indicated that mice subjected to the ICS process presented depression symptoms as in the human clinical presentation [12]. Our previous data suggest that acid saline injection-induced chronic pain and depression comorbidity produce changes in the cerebellum lobules VI, VII, and VIII. EA significantly ameliorated these phenomena through action on TRPV1 and related molecular pathways [31].
TLR4 was reported to be expressed in both neuronal and glial cells in the CNS responding to inflammatory mediators. In inflammation or nerve injury, TLR4 was suggested to activate microglia and astrocytes to induce the release of inflammatory cytokines at the spinal cord level to develop inflammatory pain and neuropathic pain [32][33][34]. Microglia have been reported as non-neuronal cells that have TRPV1 receptors, and EA can inhibit the glial TRPV1 receptors to prevent S100B secretion to activate the receptor for advanced glycation end-products on the neuronal membrane. This results in the inhibition of inflammatory processes for pain relief. In our previous publication, we reported that EA can prevent neuronal TRPV1 and inactivate the down-regulating signaling pathways [35]. Our previous results showed that EA triggers endomorphin and adenosine release to attenuate inflammatory pain in the peripheral and central nervous systems of mice [21]. Xu et al. indicated that the blockage of the TLR4 signaling pathway could attenuate chronic constriction-injury-induced neuropathic pain by measuring mechanical withdrawal threshold and thermal withdrawal latency [36]. This study indicated that EA significantly attenuated FM accompanied by decreasing the TLR4/MyD88/ NF-κB pathway.
The study results proved that several inflammatory mediators increased in the circulating blood of FM mice and can be attenuated by EA, suggesting an anti-inflammatory effect. In the neuropathic pain model, mechanical or thermal hyperalgesia was abrogated after inhibition of TLR4. Ma and colleagues suggested that TLR4 increased in the neuropathic pain model [37]. Inhibition of inflammatory cytokines and TLR4 was reported to attenuate the pain threshold in neuropathic pain mice [38]. Our data indicated that TLR4 and downstream molecules increased in FM mice and can be further reduced after EA treatment. Furthermore, we found that TLR4 was colocalized with Iba1 and simultaneously increased after FM induction. EA, but not sham EA, significantly reduced these phenomena. A recent article showed that the mRNA and protein levels of TLR4, TNF-α, and IL-1β were simultaneously increased. Additionally, the expression of NF-κB was also found to increase in the spinal dorsal horn [39]. Furthermore, inhibition of TLR4 significantly reversed the hyperalgesia and inflammatory factors, indicating its crucial role in the neuropathic pain model [36]. The current results suggest that EA can attenuate the overexpression of TLR4, TNF-α, IL-1β, and NF-κB in FM mice [39,40]. This study provides evidence of how TLR4 and associated molecules participate in the brain regions of this murine fibromyalgia model. The novel discoveries are relevant for the treatment of fibromyalgia in clinical practice. Future investigations should be conducted through clinical trials.

Conclusions
In conclusion, the main outcomes in this study are that FM induction significantly activates pain, inflammation, and central sensitization in a mouse central pain model. Mechanical and thermal hyperalgesia were observed in the FM mice. Inflammation contains increased IL-1β, IL-2, IL-5, IL-6, IL-9, IL-12, IL-17A, TNF-α, IFN-γ, and MCP-1 in the mice plasma. Furthermore, we have confirmed that TLR4 and related molecules increased in the mice's hypothalamus, PAG, and cerebellum. Particularly, we reported TLR4 is an inflammation detector for FM. Additionally, inflammatory modulators were released to activate TLR4 for pain processing. All signals further activate the aforementioned brain regions for the pain signaling pathway, including MYD88 and TRAF6. EA potently suppressed these complicated molecular pathways in the FM mice brain (Figure 7).

Data Availability Statement:
The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest:
The authors declare no conflict of interest.

Data Availability Statement:
The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest:
The authors declare no conflict of interest.