2.1 Animals

All experiments mice were adult male C57BL/6J mice (22–25 g, 8–10 w). The mice were provided and raised by the Experimental Animal Center of Zhejiang Chinese Medical University. Ventilation and air filtration units were available for the study. The mice were raised under a 12-h light/dark cycle (light cycle 8:00 a.m.–8:00 p.m.) at room temperature of 23–25°C and humidity of 40%–60%. Each cage housed 4 male mice with corn cob bedding at the bottom. The mice were fed with standard pellet chow and water ad libitum. The experimental operations complied with the experimental ethics requirements of the Experimental Animal Management and Ethics Committee of Zhejiang Chinese Medical University (IACUC-20210118-07) and were conducted following the relevant provisions of the Regulations on the Administration of Laboratory Animals of the People's Republic of China and the spirit of humanitarianism.

2.2 von Frey filament test

The mice were placed on wire mesh and covered with translucent plexiglass cover. They were allowed to acclimate to the environment for 1 h. After the mice were quiet, the mice were stimulated with von Frey filaments in the plantar surface of the left hind paw until the von Frey filaments bent into an S-shape and held for 6–8 s. A positive reaction was counted when mice retracted, licked, or flinched their claws quickly. Stimulation intervals for von Frey filaments should be at least 1 min each time. The force of the von Frey filament with three positive reactions out of five tests was recorded as PWTs.¹⁹ PWTs were measured at baseline, day 7, and day 14 after modeling.²⁰

2.3 Chronic neuropathic pain mouse model

SNI surgery was used to establish a chronic neuropathic pain model.²¹ Mice were anesthetized with 0.3% pentobarbital sodium (60 mg/kg, ip). The hair of the left hind limb was removed. The skin was disinfected with iodophor and 75% ethanol. An incision was made on the skin above the midpoint between the tibial head and the greater trochanter of the femur. Then, the muscles were bluntly dissected to expose the sciatic nerve. The peroneal nerve and common peroneal nerve were tightly ligated with non-absorbent 6–0 sutures and the nerves were transected. A section of 2–3 mm was removed distal to the ligature, leaving the tibial nerve intact. The incision was closed in layers and disinfected with iodophor. The sham group underwent the same surgery without the ligation or severance of the nerve.

2.4 Viral injection

Mice were anesthetized with 0.3% sodium pentobarbital (60 mg/kg, ip). The mice were fixed on a stereotaxic frame (RWD, 68025, Shenzhen, China). A glass microelectrode for injection was connected to an infusion pump (WPI, UMC4, Sarasota, FL, United States). A volume of 160 nL of virus was administered into the BLA at a rate of 100 nL/min (BLA: anterior–posterior: −1.36 mm; mediolateral: ±3.20 mm; dorsoventral: −4.25 mm). A volume of 90 nL of virus was injected into the rACC at a rate of 60 nL/min (rACC: anterior–posterior: +1.35 mm; mediolateral: ±0.25 mm; dorsoventral: −0.85 mm). A dental drill (WPI, OmniDrill35, Sarasota, FL, United States) was used to create the cranial openings. At the end of the infusion, the glass microelectrode was left at the injection site for 10 min to prevent the virus from overflowing. The BLA coordinates and rACC coordinates were determined based on Paxinos and Franklin's The Mouse Brain in Stereotaxic Coordinates (Fourth version).

2.5 Tracer virus injection strategy

Mice were injected with anterograde tracer virus (AAV2/9-CaMKIIα-EGFP, 5.34 × 10¹² vg/mL, BrainVTA, China, PT-0290) in the right BLA. The virus (AAV2/R-CaMKIIα-EGFP, 5.24 × 10¹² vg/mL, BrainVTA, PT-0290) was injected into the right rACC for retrograde monosynaptic tracking. After 2 weeks, the tracer virus successfully transfected neurons in BLA and rACC. Then, the mice were anesthetized with 0.3% sodium pentobarbital (60 mg/kg, ip) and perfused sequentially with saline (20 mL) and 4% paraformaldehyde (20 mL) via the cardiac. Finally, 20 μm brain slices were prepared for recording green fluorescent protein signals.

2.6 Chemogenetic method

The BLA^CaMKII-rACC neural circuit was manipulated via the chemogenetic method to determine whether the neural circuit modulated mechanical allodynia and anxiety-like behaviors. The mice were injected rAAV2/9-CaMKIIα-DIO-hM4Di-mCherry-WPRE-pA (3.38 × 10¹² vg/mL, BrainVTA, China, PT-1143), rAAV2/9-CaMKIIα-DIO-hM3Dq-mCherry-WPRE-pA (3.04 × 10¹² vg/mL, BrainVTA, China, PT-1144) or rAAV2/9-CaMKIIα-DIO-mCherry-WPRE-pA (5.35 × 10¹² vg/mL, BrainVTA, China, PT-1167) into the BLA. rAAV2/R-CaMKIIα-Cre (6.65 × 10¹² vg/mL, BrainVTA, China, PT-0220) was injected into the rACC. Clozapine-N-oxide (CNO) (0.2 mg/mL, dissolved in dimethyl sulfoxide and diluted with 0.9% saline, 2 mg/kg, ip) was intraperitoneally injected on days 8, 10, 12, 14, and 16 after SNI/sham surgery. Among them, rAAV2/9-CaMKIIα-DIO-hM4Di-mCherry-WPRE-pA and rAAV2/R-CaMKIIα-Cre were injected bilaterally to inhibit the BLA^CaMKII-rACC neural circuit. rAAV2/9-CaMKIIα-DIO-hM3Dq-mCherry-WPRE-pA and rAAV2/R-CaMKIIα-Cre were injected into the right side to activate the BLA^CaMKII-rACC neural circuit. The rAAV2/9-CaMKIIα-DIO-mCherry-WPRE-pA is a control virus without any activating or inhibitory effect.

2.7 EA treatment

EA treatment was conducted on the 8th, 10th, 12th, 14th, and 16th days after SNI surgery. In addition to the EA group, the sham, SNI, and SNI-sham EA groups were all fixed loosely. Bilateral Zusanli (ST36) and Sanyinjiao (SP6) acupoints were selected. 0.16 × 7 mm acupuncture needles were directly inserted into the acupoints for 5 mm in the EA group. The two needles on the same side were connected to the Hans acupoint nerve stimulator (HANS-200A, Beijing Hua Wei Industrial Development Co). The intensity of EA was 0.3 mA, the frequency was 2 Hz and the stimulation time was 30 min. In the SNI-sham EA group, electrodes, but no current, were attached to the acupuncture needles.

2.8 Elevated plus maze test (EPMT)

The EPMT is the most commonly used test for studying general anxiety-like behavior.²² Mice were subjected to EPMT on day 14 after sham/SNI surgery. The day before EPMT, the mice were transferred to the behavioral room to acclimatize. The room temperature was 23–25°C and the humidity was 40%–60%. The elevated plus maze (EPM) was composed of two open arms (30 × 6 cm), two closed arms (30 × 6 × 15 cm), and a center area (6 × 6 cm) with a height of 35 cm. The open arms and closed arms are crossed vertically. In the beginning, each mouse was placed in the center area with its head facing the open arm. After the experiment, the urine and feces of the mice were cleaned up. To remove the odor left by the last animal, the EPM will be cleaned with 75% ethanol and double-distilled water. The mice's behaviors during the test were recorded on a video tracking system (ANY-maze V6.14, Stoelting, USA). The video was recorded for 5 min and 30 s, with the first 30 s being the acclimatization phase.

2.9 Open field test (OFT)

The OFT was conducted on day 16 after surgery to determine whether mice displayed anxiety-like behaviors.²³ The day before OFT, the mice were relocated to the behavioral room. The open field (OF) was a 40 × 40 × 40 cm³ uncovered cube. The bottom of the cube was divided equally into 16 squares of identical area. The outer 12 square areas were defined as the peripheral area. The 4 square areas in the middle are defined as the central zone (with a total area of 20 × 20 cm²). The mice were placed in the center zone at the beginning. Finally, the mice's urine and feces were cleaned up. The interior of the OF was cleaned with 75% ethanol and double-distilled water in turn. The mice's behaviors during testing were recorded on a video tracking system (ANY-maze V6.14, Stoelting, USA). The duration of the video is 5 min and 30 s.

2.10 Immunofluorescence staining

Approximately 90 min after OFT, mice were deeply anesthetized with sodium pentobarbital (60 mg/kg, ip). The mice were perfused transcardially with 0.9% saline and 4% (w/v) paraformaldehyde sequentially. After perfusion, the mice brains were removed and placed in 4% (w/v) paraformaldehyde for 24 h and then subjected to gradient dehydration with 15% and 30% (w/v) sucrose. The brains were sliced into frozen coronal sections of 20 μm thickness using the freezing microtome (CryoStar NX50 HOP, Thermo Fisher Scientific).

The slices were mounted on gelatin-coated glass. The slices were rewarmed at 37°C for 1 h, and they were washed with Tris Buffered Saline with Tween-20 (TBST) on a shaker 6 times for 10 min each. The slices were blocked with 10% donkey serum (with 0.3% Triton X-100) for 1 h at 37°C and incubated with the primary antibodies for 20 h at 4°C. The primary antibodies included anti-CaMKII (mouse, 1:200, Abcam, ab22609) and anti-cFos (rabbit, 1:500, Abcam, ab190289). The slices were then rewarmed again at 37°C for 1 h, washed 6 times, and incubated with Alexa Fluor 647-conjugated secondary antibody (1:1000, Abcam, ab150111) and Alexa Fluor 488-conjugated secondary antibody (1:800, Jackson ImmunoResearch Labs, 711-545-152) for 1 h at 37°C. The slices were washed again with TBST, followed by incubation with 4′,6-diamidino-2-phenylindole (DAPI, Abcam, USA). The fluorescence signal was then observed using a ZEISS digital scanner (ApoTome.2, ZEISS, Germany).

2.11 Statistical analysis

All analyses and graphs were calculated and plotted using GraphPad Prism 9. All results are expressed as mean ± standard errors of the mean (SEM). The normality of the distribution of continuous variables was assessed using the Shapiro–Wilk normality test. The PWTs between different groups were tested with two-way repeated-measures Analysis of Variance (ANOVA) with Tukey's post-hoc test. For normally distributed data, Student's t-test (two-tailed) and one-way ANOVA followed by Tukey's post-hoc test were used to compare means of two and multiple groups, respectively. Additionally, the Mann–Whitney U-test and Kruskal–Wallis were employed to compare two or more groups of non-normally distributed data, respectively. A significance level of p < 0.05 was considered statistically significant.

3.1 2 Hz EA reduced mechanical allodynia and anxiety-like behaviors in SNI mice

Research findings indicate that EA effectively alleviates mechanical allodynia and anxiety-like behaviors.^24-26 Therefore, we aim to further investigate the effects of 2 Hz EA on mechanical allodynia and anxiety-like behaviors in neuropathic pain mice (Figure 1A). We established a chronic neuropathic pain mouse model by employing SNI injury on C57BL/6J mice (Figure 1B), following previously reported protocols.²⁷ Similar to previous studies, we chose bilateral Zusanli (ST36) and Sanyinjiao (SP6) as the sites for EA stimulation (Figure 1C).^{25, 26}

Compared to the sham mice, the PWTs in the SNI mice decreased (Figure 1D). The SNI mice spent less time in the open arm and central zone (Figure 1E,F). The result suggests that SNI surgery can induce neuropathic pain and anxiety-like behaviors in mice.

After EA intervention, the PWTs of the SNI-EA mice increased compared to SNI mice and SNI-sham EA mice (Figure 1D). SNI-EA mice exhibited an increase in open-arm and central zone time (Figure 1E,F). The results suggest that 2 Hz EA is effective at reducing mechanical allodynia and anxiety-like behaviors in SNI mice. At the same time, there were no statistically significant differences in the total distance traveled among the 4 groups (Figure 1G), which suggests the locomotor activity of the 4 groups was not affected. Figure 1H,I show the trajectory maps and heat maps of the 4 groups. These results underscore the efficacy of the SNI mice in inducing mechanical allodynia and anxiety-like behaviors. Moreover, 2 Hz EA significantly mitigated chronic neuropathic pain and anxiety-like behaviors in SNI mice.

3.2 The effects of EA on mechanical allodynia and anxiety in SNI mice may be linked to BLA CaMKII neurons activity

Research suggests BLA mediates neuropathic pain.^{11, 28, 29} Our preliminary research indicates BLA is also associated with anxiety-like behavior induced by SNI.¹³ Modulating BLA influences anxiety-like behaviors in neuropathic pain mice.¹⁰ Therefore, we aim to further explore whether the analgesic and anti-anxiety effects of EA on SNI mice are associated with the BLA.

cFos is a marker for neuronal activity. When neurons are activated, the expression of cFos is upregulated.³⁰ As shown in Figure 2A,B, compared to the sham group, the co-expression rate of cFos in BLA CaMKII neurons is reduced in SNI mice, suggesting that the mechanical allodynia and anxiety-like behavior in SNI mice may be associated with decreased activity of BLA CaMKII neurons. Meanwhile, compared to the SNI group, the co-expression rate of cFos in BLA CaMKII neurons is higher in the SNI-EA group, indicating that the analgesic and anti-anxiety effects of EA may be linked to the activity of BLA CaMKII neurons.

3.3 CaMKII neurons project from BLA to rACC

To investigate whether BLA is structurally related to rACC, an anterograde tracing virus (AAV2/9-CaMKIIα-EGFP) was injected into the right BLA (Figure 3A). After 2 weeks, CaMKII neurons that were transfected and labeled by the virus were observed in the right BLA (Figure 3B). In addition, nerve fibers transfected and labeled by the virus could be seen in the right rACC (Figure 3C). Furthermore, we injected retrograde tracer viruses (AAV2/R-CaMKIIα-EGFP) into the right rACC (Figure 3D). Two weeks after injection, transfected CaMKII neurons could be visualized in the right rACC (Figure 3E). The CaMKII neurons transfected by the virus could be observed in the right BLA (Figure 3F). Therefore, CaMKII neurons in the BLA project to the rACC.

3.4 Inhibition of BLA^CaMKII-rACC induced mechanical allodynia and anxiety-like behaviors in sham mice

Our previous research demonstrated the correlation between the activity of BLA CaMKII neurons and chronic neuropathic pain and anxiety-like behaviors and the existence of the BLA^CaMKII-rACC. In addition, rACC is also implicated in chronic neuropathic pain and pain-related negative emotions.^{31, 32} However, the role of the BLA^CaMKII-rACC in chronic neuropathic pain and anxiety-like behaviors remains unclear. Therefore, we used chemogenetic methods in sham mice to manipulate the BLA^CaMKII-rACC neural circuit. We observed alterations in PWTs, open arm time, and central time to investigate if BLA^CaMKII-rACC plays a role in mechanical allodynia and anxiety-like behaviors.

In sham-hM4D-CNO group, bilateral injections of rAAV2/9-CaMKIIα-DIO-hM4Di-mCherry-WPRE-pA were administered into the BLA, while bilateral injections of rAAV2/R-CaMKIIα-Cre were delivered into the rACC to inhibit bilateral BLA^CaMKII-rACC. In sham-mcherry-CNO group, the BLA was injected bilaterally with AAV2/9-CaMKIIα-DIO-mCherry-WPRE-pA, the rACC was injected bilaterally with rAAV2/R-CaMKIIα-Cre to be used as a control group (Figure 4A,B). After 2 weeks, viral expression was observed in BLA (Figure 4C). Approximately 56.0% of the BLA-projecting neurons labeled with mCherry were immunoreactive for CaMKII (Figure 4D,E). To determine whether the hM4D virus inhibits the activity of BLA CaMKII neurons projecting to rACC, we compared the percentage of colocalization of the virus-labeled neurons with cFos between the sham-mCherry-CNO and sham-hM4D-CNO groups (Figure 4F). Results showed that the hM4D virus reduced the activity of BLA CaMKII neurons projecting to rACC from 17.9% to 11.9% (Figure 4G).

In comparison to the sham-mCherry-CNO group, the sham-hM4D-CNO group exhibited significantly lower PWTs (Figure 4H), suggesting that inhibition of BLA^CaMKII-rACC induced mechanical allodynia in sham mice. Compared to the sham-mCherry-CNO group, the sham-hM4D-CNO group spent less time in the open arm and the center zone (Figure 4I,J). There was no notable variance observed in the total distance traveled (Figure 4K), suggesting similar locomotor abilities between the 2 groups. Figure 4L,M depict the representative motion trajectories and heat maps in EMPT and OFT, respectively. Hence, the results suggest inhibition of BLA^CaMKII-rACC-induced mechanical allodynia and anxiety-like behaviors in sham mice.

3.5 Activation of BLA^CaMKII-rACC attenuated mechanical allodynia and anxiety-like behaviors in SNI mice

Since the previous result shows the inhibition of BLA^CaMKII-rACC in sham mice can induce mechanical allodynia and anxiety-like behaviors, we next investigated whether mechanical allodynia and anxiety-like behaviors in SNI mice could be attenuated by activation of BLA^CaMKII-rACC (Figure 5A). First, rAAV2/9-CaMKIIα-DIO-hM3Dq-mCherry-WPRE-pA was injected into the right BLA, and AAV-CaMKIIα-Cre was injected into the right rACC to activate BLA^CaMKII-rACC (Figure 5B,C). Next, we examined the colocalization of CaMKII neurons with cFos (Figure 5D). It turned out that BLA CaMKII neurons projecting to rACC were significantly activated by the hM3D virus (Figure 5E).

Compared to the SNI-hM3D-Saline group, the PWTs of the SNI-hM3D-CNO group were higher (Figure 5F). The PWTs results suggest activation of BLA^CaMKII-rACC relieves mechanical allodynia in SNI mice. In addition, SNI-3D-CNO mice spent more time in the open arm and center zone compared with the SNI-hM3D-Saline group (Figure 5G,H). The total distance was not statistically different between the 2 groups (Figure 5I). Motion trajectory maps and heat maps of the 2 groups in EPMT and OFT are depicted in Figure 5J,K, respectively. Therefore, it can be concluded that activation of BLA^CaMKII-rACC can mitigate mechanical allodynia and anxiety-like behaviors in SNI mice.

3.6 Inhibition of BLA^CaMKII-rACC antagonized the analgesic and anti-anxiety effects of EA on SNI mice

The above results indicate that EA can alleviate mechanical allodynia and anxiety-like behaviors in SNI mice, possibly associated with the activity of BLA CaMKII neurons. Besides, BLA^CaMKII-rACC is involved in the mechanical allodynia and anxiety-like behaviors of both sham and SNI mice. However, it remains unclear if the analgesic and anti-anxiety effects of EA on SNI mice are linked to BLA^CaMKII-rACC.

To investigate whether the effects of EA are mediated through the BLA^CaMKII-rACC. In the SNI-hM4D-CNO-EA group, bilateral BLA was injected with rAAV2/9-CaMKIIα-DIO-hM4Di-mCherry-WPRE-pA, bilateral rACC was injected with rAAV2/R-CaMKIIα-Cre to inhibit the BLA^CaMKII-rACC. After administering CNO injection to SNI mice, EA intervention was performed 30 min later (Figure 6A,B), which allowed us to observe whether the effects of EA were influenced when the BLA^CaMKII-rACC was inhibited (Figure 6A,B).

Compared to the SNI-mCherry-CNO group, mice in the SNI-mCherry-CNO-EA group showed higher PWTs (Figure 6C), confirming the analgesic effect of EA. In comparison to the SNI-mCherry-CNO-EA group, the SNI-hM4D-CNO-EA group exhibited notably lower PWTs (Figure 6C). The PWTs result indicates the analgesic effect of EA is antagonized when the BLA^CaMKII-rACC is inhibited.

Compared to the SNI-mCherry-CNO group, the SNI-mCherry-CNO-EA group spent more time in the open arm and center zone (Figure 6D,E). The SNI-hM4D-CNO-EA group spent less time in both the open arm and center zone compared to the SNI-mCherry-CNO-EA group (Figure 6D,E). Meanwhile, the total distance traveled by the 3 groups was not statistically different (Figure 6F). Figure 6G,H show the trajectory map and heat map of the 3 groups in the EPMT and OFT, respectively.

Therefore, the above results indicate that when the BLA^CaMKII-rACC neural circuit is inhibited, the analgesic and anti-anxiety effects of EA are antagonized. This suggests that the BLA^CaMKII-rACC is involved in the analgesic and anti-anxiety mechanism of EA.