A total of 126 male C57BL/6J wild-type mice weighing 20–25 g and aged 7–8 weeks old were used. The mice were purchased from Hangzhou Ziyuan Experimental Animal Technology Co., Ltd. (License No. SCXK (Zhe) 2019-0004, Hangzhou, Zhejiang, China). The mice were housed in the laboratory animal room of Anhui University of Chinese Medicine with alternating lighting (lights on at 08:00, off at 20:00). The vivarium environment was maintained at 25 $\pm$ 1 °C and 55 $\pm$ 5% relative humidity. The mice had free access to autoclaved rodent diet and sterile water. Experiments followed followed by cervical dislocation as a confirmatory measure per AVMA guidelines and were approved by the Animal Ethics Committee of Anhui University of Chinese Medicine (No. AHUCM-mouse-2022083).

All surgical procedures were conducted under aseptic conditions. The mice were fasted for 12 hours (h) before surgery. Prior to surgical intervention, the mice were anesthetized with 4% isoflurane (1 L/min, R510-22-10, RWD Life Technology Co., Shenzhen, China) in an induction chamber and maintained on 1% isoflurane. The status of the mice was carefully observed throughout the procedure. The mouse body temperature was maintained at 37 °C and subsequently rewarmed postoperatively using a regulated heating pad until full recovery. During virus injections and in vivo electrophysiological procedures, the eyes of the mice were coated with roxithromycin ophthalmic ointment (H20003077, Tiantai Pharmaceutical CO., Ltd, Sanming, Fujian, China) to prevent light-induced damage to the eyes.

After anesthetization, EA was performed by vertically inserting a unipolar stainless-steel acupuncture needle (0.25 $\times$ 13 mm, Suzhou Acupuncture Supplies Co., Ltd., Suzhou, Jiangsu, China) at a depth of 0.3 cm into the Shenmen (HT 7) acupoint, which is located in the forelimb of the mouse, close to the transverse wrist striation. Another needle was inserted at a depth of 0.3 cm into the Tongli (HT 5) acupoint, which is located 0.75 cm proximal to the HT7. Stimulation was delivered for 20 min daily at 2 Hz and 0.5 mA over 7 consecutive days using an EA device (HANS-200A/100B; HANS, Beijing, China), with needles at HT7 and HT5 connected to the anode and cathode, respectively. The sham electroacupuncture (sEA) treatment involved stimulating non-acupuncture points on the tails of mice under identical electroacupuncture parameter conditions. To control for the potential confounding effects of anesthesia, non-pretreated mice in the same experiment underwent inhalation of 1% isoflurane for an identical duration.

Under anesthesia (4% isoflurane, 1 L/min), the anterior thoracic region of the mouse was depilated using depilatory cream. After this, the muscle tissue was carefully separated to expose the third and fourth intercostal spaces. Mechanical ventilation was maintained throughout the procedure using an ALC-V8S ventilator (Shanghai Alcott Biotech Co., Ltd., Shanghai, China) with the following parameters: respiratory frequency, 120 breaths/min; inspiration-to-expiration ratio, 1:2; tidal volume, 1 mL. The MIRI model was established by ligation of the left anterior descending coronary artery (LAD) using a 10–0 silk suture. Following 30 min of occlusion, the suture was loosened to permit a reperfusion period of 2 h. In the Sham group, a left thoracic incision was made to expose the pericardium without ligating the LAD.

Under anesthesia, mice were secured within the brain stereotaxic apparatus (RWD Life Technology Co., Shenzhen, Guangzhou, China) using a nose clamp and ear bars. The fur on the mouse’s head was clipped and the scalp was cut to expose the skull. The cranial surface was wiped with a cotton swab soaked in saline to keep the skull clean. Stereotaxic coordinates were based on Paxinos and Franklin’s the Mouse Brain in Stereotaxic Coordinates [31]. For Crus Ⅰ bilateral injections, the coordinates from bregma were defined as –6.25 mm anteroposterior (AP), $\pm$ 2.50 mm mediolateral (ML), –2.00 mm dorsoventral (DV). The injection site over Crus Ⅰ was indicated with a marker, followed by careful creation of a cranial opening with a drill to minimize bleeding. Virus was loaded into a 10 µL micro-syringe (Gaoge, Shanghai, China). Injections were performed with a microinjection pump (KD Scientific Inc., Holliston, MA, USA) at a flow rate of 40 nL/min. The needle was left in place for an additional 10 min after injection and then the wound was sutured.

Under deep anesthesia, mice were secured in a stereotaxic instrument to perform craniotomy. A fine-wire electrode was implanted into the Crus Ⅰ, and neural signal data were acquired using the Plexon multichannel neural signal recording system in digital form. The raw data were processed using Offline Sorter and Neuro Explorer software for sorting units, firing rates analysis, waveform comparison and spectral analysis.

The real-time raw data were detected after a 300 Hz low-cut filter, which was three standard deviations above the noise amplitude. Suspected individual units were isolated off-line by manual cluster cutting of various spike waveform parameters. Only units with sufficient separation and no inter-spike interval 2 ms were included in the data analysis (L-ratio 0.2, isolation distance $>$15) [32]. Neurons were categorized into wide spiking (WS) and narrow spiking (NS) neurons by the K-means algorithm. The NS group was subdivided into putative fast-spiking parvalbumin (FS-PV Ins) ($>$10 Hz) and non-fast-spiking (non-FS) NS neurons.

Mice were bilaterally injected with a virus mixture 200 nL of recombinant adeno-associated virus-cellular immediate-early gene Fos promoter-tetracycline transactivator (rAAV-c-fos-tTA) and 200 nL of the recombinant adeno-associated virus-tetracycline response element-tight promoter–mCherry (rAAV-TRE-tight-mCherry) into Crus Ⅰ. The mice were then housed individually for 3 weeks and fed Dox water (Dox, Spark jade, Cas: 24390-14-5, Jinan, Shandong, China). For neuron activity-dependent labeling, the mice were fed with regular water and received bilateral EA stimulation 20 min/day for 3 days. Dox diets were resumed until the mice were perfused for histological analysis.

Electrocardiography (ECG) signals were recorded in mice using limb-lead electrodes connected a PowerLab Standard Ⅱ leads (PowerLab 8, AD Instruments, Sydney, Australia). ST-segment displacement values and low-frequency to high-frequency ratio (LF/HF) were derived from ECG traces using LabChart V8.1.19 software (AD Instruments, Sydney, Australia). Data segments with at minimum duration of 5 min were included for analysis.

Under anesthesia, cardiac ultrasound was performed on mice using a high-frequency ultrasound with a dedicated small animal imaging system (VINNO6 LAB, Suzhou, Jiangsu, China). Two-dimensional and M-mode echocardiograms were acquired from the parasternal long-axis view. Echocardiographic parameters, including left ventricular ejection fraction (LVEF) and left ventricular fractional shortening (LVFS), were measured over three consecutive cardiac cycles. Throughout the procedure, body temperature was regulated within the range of 36–37.5 °C.

The coronary artery was re-occluded at the end of reperfusion. Mice were euthanized via terminal overdose of isoflurane anesthesia (5% concentration for $\ge$5 minutes post-respiratory cessation), followed by cervical dislocation as a confirmatory measure per AVMA guidelines. Evans blue dye (0.5%) was injected into the apex to stain normal area. The heart was frozen at –80°C for 30 min and cut into 1 mm transverse slices. The slices were immersed in 1% triphenyl-tetrazolium chloride solution (TTC, Solarbio, Beijing, China) and incubated at 37 °C in the dark for 15 min. Prior to photography, heart tissue sections were set in 4% paraformaldehyde fixative overnight. The infarct area (IA), the area-at-risk (AAR) and the area of left and right ventricle (LV + RV) were analyzed using ImageJ software (Fiji version, NIH, Bethesda, MD, USA). The infarct volume is presented as the ratio of IA/AAR.

Under deep anesthesia, mice were perfused through the heart with chilled 0.9% saline and 4% paraformaldehyde (PFA, G1101-500ML, Servicebio, Wuhan, Hubei, China) solution. The mouse brains were carefully dissected from the cranial cavity, post-fixed in 4% PFA for 3 days, and subsequently cryoprotected in a 30% sucrose solution. Using a cryostat microtome (Leica CM1900, Wetzlar, Germany), brains were coronally sectioned at 40-µm and rinsed in phosphate-buffered saline (PBS). Following permeabilization and blocking with 0.5% TritonX-100 (BS084-500ml, Biosharp, Beijing, China) and 5% bovine serum albumin (BSA, ED0017-B, Spark jade, Jinan, Shandong, China) in PBS for 1.5 h, the sections were incubated overnight at 4 ℃ in a base solution containing 0.3% Triton X-100 and 3% BSA, to which the primary antibody (anti-Calbindin, ab108404, Abcam, Waltham, MA, USA; 1:500). The next day, the sections were incubated with the secondary antibody (Donkey anti-rabbit IgG HL, ab150061, Abcam, 1:500) in a base solution containing 0.3% Triton X-100 and 3% BSA for 2 h at room temperature under light-protected conditions. After being washed three times with PBS, the sections were mounted using an antifade medium with 4′,6-diamidino-2-phenylindole (DAPI, EE0011-A, Spark jade, Jinan, Shandong, China). Upon completion of the staining process, three randomly selected images were captured.

Myocardial tissue was harvested after surgery, rinsed with saline, fixed in 4% PFA overnight. Then, cardiac tissues were dehydrated in an ethanol gradient at room temperature, cleared of xylene, embedded in paraffin and cut into 5 µm slices. The slices were stained with hematoxylin-eosin (HE, BA4041, BaSO, Shenzhen, Guangdong, China) for observation of their morphological structure. The slices were stained with Weigert ferrohematoxylin, Ponceau acid fuchsin, Marson blue solution, and Aniline blue in order to observe the degree of myocardial fibrosis.

At the conclusion of the experiment, blood was obtained via the orbital sinus from mice under deep anesthesia. After collection, the blood was incubated at 4 °C for 30 min and centrifuged at 3500 rpm for 15 min. The serum was collected, quantified, and assessed according to assay-kit instructions (AiFang biological, Changsha, Hunan, China). Levels of norepinephrine (NE) were measured to assess sympathetic nerve activity, and levels of creatine kinase isoenzyme MB (CK-MB) and cardiac-specific troponin T (cTnT) were measured to determine the extent of myocardial injury.

All statistical analyses were conducted using GraphPad Software version 8.0 (GraphPad Software, San Diego, CA, USA). Data are presented as mean $\pm$ standard error of mean (SEM). Differences among groups were assessed by one-way analysis of variance (ANOVA) followed by Tukey’s post hoc test for multiple comparison, with a significance threshold set at p $\le$ 0.05. All experimental data are available in the Supplementary Material-Datasets.

To determine the protective effect of EA-pre on MIRI, cardiac function was compared among four groups: Sham, MIRI, MIRI+EA-pre, and MIRI+sham EA-pre (sEA-pre). The MIRI mouse model was induced by 30 min LAD ligation followed by 120 min reperfusion [33, 34] (Fig. 1A). The ST segment of the ECG was elevated in myocardial ischemia, and partially restored after 2 h of reperfusion (Fig. 1B). LF/HF is often taken to be an indicator of sympathetic-vagal balance [35, 36]. ECG analysis indicated a reduction in ST segment deviation and a lower LF/HF ratio in the MIRI+EA-pre group relative to the MIRI group. Conversely, no significant differences were observed between the MIRI+sEA-pre group and the MIRI group (Fig. 1C,D). In addition, we conducted echocardiographic experiments to evaluate cardiac function (Fig. 1E). The MIRI procedure resulted in significant reduction in LVEF and LVFS, as well as increases in left ventricular end-systolic internal diameters (LVIDs) and left ventricular end-diastolic internal diameters (LVIDd). Cardiac function was markedly improved in the MIRI+EA-pre group compared to the MIRI group, whereas no significant difference was observed between the MIRI+sEA-pre group and MIRI group (Fig. 1F). To evaluate the direct cardiac effects of EA-pre, ECG, and echocardiography were performed on both the Sham group and Sham+EA-pre group. Results revealed no significant differences between the groups, indicating that EA-pre administration did not compromise cardiac function (Supplementary Fig. 1).

Fig. 1.

EA-pre alleviated MIRI in cardiac function. (A) Experimental timeline. (B) Typical ECG curves were compared across all groups. (C) Comparison of ST deviation in different groups. (D) Comparison of LF/HF ratios between groups. (E) Comparison of representative M-mode echocardiograms between groups. (F) Quantitative analysis of echocardiographic measurements. (C,D,F) Data are presented as mean $\pm$ SEM; n = 6/group; *p 0.05; **p 0.01; ***p 0.001; according to one-way ANOVA with Tukey’s post-test. EA-pre, Electroacupuncture pretreatment; MIRI, myocardial ischemia-reperfusion injury; ECG, Electrocardiography; LF/HF, low-frequency to high-frequency ratio; SEM, standard error of mean; ANOVA, analysis of variance; LAD, left anterior descending coronary artery.

Evans blue/TTC double staining showed similar AAR across all groups. It was interesting that the ratio of IA/AAR was significantly higher in the MIRI group than in the MIRI+EA-pre group, indicating greater myocardial damage in the absence of EA-pre (Fig. 2A,B). HE staining and Masson staining revealed elevated myocardial-tissue damage in the MIRI group relative to the sham group. This damage included uneven arrangement of cardiomyocytes, severe cellular edema, myocardial-fiber breakage, and a large infiltration of inflammatory cells, as a previous study reported [37]. Myocardial tissue damage was notably attenuated in the MIRI+EA-pre group (Fig. 2C,D). Correspondingly, the MIRI procedure markedly increased the levels of myocardial injury biomarkers cTnT and CK-MB [38]. Additionally, NE serves as a sensitive indicator of early myocardial sympathetic excitation, and large amounts of NE are released from sympathetic nerve endings during myocardial ischemia [39]. Consistently, the MIRI+EA-pre group exhibited a significant reduction in the expression of myocardial injury markers (Fig. 2E–G). These findings indicated that EA-pre potently alleviated the severity of symptoms, possibly by modulating sympathetic nerve activity after MIRI.

Fig. 2.

EA-pre attenuated the extent of cardiac tissue damage in MIRI mice. (A) Representative images of mouse hearts double-stained with TTC and Evans blue for determining infarcted areas, scale bar, 1 mm. (B) Ratios of infarcted area to area at risk and percentage of area at risk. n = 6/group, one-way ANOVA with Tukey’s post-test. (C,D) HE and Masson staining of myocardial tissue. Scale bar, 50, 25 µm. (E–G) Serum levels of CK-MB, cTnT, and NE across groups. Data are presented as mean $\pm$ SEM; n = 6/group; *p 0.05; **p 0.01; ***p 0.001; ns, not significant; according to one-way ANOVA with Tukey’s post-test. CK-MB, creatine kinase isoenzyme MB; cTnT, cardiac-specific troponin T; HE, hematoxylin-eosin; NE, norepinephrine; TTC, triphenyl-tetrazolium chloride; IA, Infarct Area; AAR, area-at-risk.

To investigate changes in Crus Ⅰ neuron activity in MIRI mice, in vivo electrophysiological recordings of Crus Ⅰ neuron firing were performed (Fig. 3A). First, well-separated neurons were classified by their spiking characteristics into WS (trough to peak duration 360.3105 $\pm$ 4.9505 ms, n = 236) presumed to be excitatory pyramidal neurons (EPs) and NS (trough to peak duration 137.4024 $\pm$ 2.2049 ms, n = 221) presumed to be inhibitory interneurons (INs) [40]. The putative INs were further categorized into fast-firing inhibitory interneurons (F-INs) (with an average firing rate $>$10 Hz) and non-F-INs based on their neuronal firing rates [41] (Fig. 3B,C). No significant differences were observed in the firing frequency of Crus Ⅰ EPs among the three groups, whereas the firing frequency of F-INs differed significantly (Fig. 3D). After MIRI surgery, the firing frequency of Crus Ⅰ neurons increased. The neural firing frequency was notably reduced in the MIRI+EA-pre group compared to the MIRI group (Fig. 3E,F).

Fig. 3.

Firing characteristics of Crus I neurons in MIRI mice. (A) Schematic of in vivo mouse electrophysiological experiments with Crus I electrodes. (B) Classification of recorded Crus I neurons into putative pyramidal neurons (orange triangles) and F-INs neurons (blue circles) based on peak duration and firing frequency. (C) Representative spike waveforms from EPs and F-INs neurons. (D) Comparison of firing frequencies in Crus I EPs and F-INs neurons across groups. (E) Sample traces of action potentials were recorded from Crus I EPs and F-INs neurons across various groups. (F) Comparison of Crus I energy spectrograms across different groups. (D) Data are presented as mean $\pm$ SEM; n = 6/group; *p 0.05; **p 0.01; according to one-way ANOVA with Tukey’s post-test. F-Ins, fast-firing inhibitory interneurons; EPs, excitatory pyramidal neurons.

To examine the impact of EA-pre on Crus Ⅰ, the Fos-TRAP labeling technique was used to identify the specific types of neurons in Crus Ⅰ that contribute to the attenuation of MIRI by EA-pre. Recombinant adeno-associated viruses (rAAV) carrying c-Fos: tetracycline-controlled transactivator (tTA) and tetracycline-responsive element (TRE), were injected into Crus Ⅰ (Fig. 4A). Comparison of mCherry expression/mm² of neurons across different groups was conducted to validate the Fos-TRAP reliability (Fig. 4B). The fluorescence results indicated that EA-labelled active neurons were regularly distributed in one layer (Fig. 4C). Dox could effectively repress the expression of the TRE promoter (Fig. 4D). Based on the neuronal morphology of the cerebellar cortex, we hypothesized that these EA-labelled active neurons were PC, which received modulation from interneurons [42, 43]. Collectively, these neurons contribute to the modulation of the cerebellar cortex [44].

Fig. 4.

Labeling of Crus Ⅰ neurons activated by electroacupuncture stimulation. (A,B) Schematic of Fos-TRAP—based labeling of Crus Ⅰ neurons. (C) Representative mCherry expression in Crus Ⅰ across groups, Scale bar, 100 µm, 50 µm. (D) Comparison of the number of mCherry-expressing neurons in Crus Ⅰ in each group. Data are presented as mean $\pm$ SEM; n = 6/group; *p 0.05; ***p 0.001; ns, not significant; according to one-way ANOVA with Tukey’s post-test; TRE3G, T​​etracycline-​​R​​esponsive ​​E​​lement, ​​3​​rd ​​G​​eneration; c-Fos, Finkel-Biskis-Jinkins O​​steosarcoma ​​S​​arcoma virus oncogene homolog; DOX, D​​oxycycline.

To investigate the role of Crus Ⅰ^PC in EA-pre mitigation of MIRI, a recombinase adenovirus carrying an L7-specific promoter (rAAV-L7-mCherry) was injected bilaterally to label PC in Crus Ⅰ. Immunofluorescence analysis revealed that PC were labelled successfully (Fig. 5A). Subsequently, we performed different interventions on Crus Ⅰ^PC in combination with CNO to control the activities of PC (Fig. 5B). ECG analysis indicated significantly elevated S-T segment deviation values and LF/HF ratio in the mCherry+MIRI group during MIRI, which were decreased with EA-pre intervention. Furthermore, the mCherry+MIRI group exhibited elevated levels of both parameters relative to the hM4Di+MIRI group, whereas the mCherry+EA-pre+MIRI group showed reduced values compared to the hM3Dq+EA-pre+MIRI group (Fig. 5C,D). Echocardiographic analysis indicated a pronounced reduction in cardiac function in the mCherry+MIRI group compared with the mCherry+Sham group. By contrast, the mCherry+EA-pre+MIRI group notably improved cardiac performance relative to the mCherry+MIRI group. Cardiac function demonstrated significant enhancement in the hM4Di+MIRI group relative to the mCherry+MIRI group, but substantial impairment was observed in the hM3Dq+EA-pre+MIRI group compared with the mCherry+EA-pre+MIRI group (Fig. 5E,F).

Fig. 5.

Chemogenetic manipulation of Crus Ⅰ^𝐏𝐂 neuronal activity in MIRI. (A) Schematic diagram of protocol and injection site for viral injections. Scale bar, 200 µm. (B) Experimental strategy for chemogenetic regulation of Crus Ⅰ^PC. (C) Group comparisons of ST deviation. (D) Intergroup differences in LF/HF ratio. (E) Representative M-mode echocardiograms for each group. (F) Echocardiographic measurements across groups. Data are shown as mean $\pm$ SEM; n = 6/group; *p 0.05; **p 0.01; ***p 0.001; according to one-way ANOVA with Tukey’s post. Crus Ⅰ^PC, Crus Ⅰ Purkinje cells; hM4Di, h​​uman ​​M​​uscarinic receptor subtype ​​4​​, ​​D​​esigner ​​i​​nhibitory; hM3Dq, h​​uman ​​M​​uscarinic receptor subtype ​​3​​, ​​D​​esigner ​​q​​-coupled.

Consistent with previous findings, Evans Blue/TTC double staining revealed that EA-pre effectively reduced the IA/AAR ratio in mice subjected to MIRI. Specifically, the hM4Di+MIRI group exhibited a lower IA/AAR ratio than did the mCherry+MIRI group, and the mCherry+EA-pre+MIRI group also showed a lower IA/AAR ratio than did the hM3Dq+EA-pre+MIRI group (Fig. 6A,B). HE staining and Masson staining results also verified these findings (Fig. 6C,D). ELISA results revealed markedly increased levels of CK-MB, cTnT, and NE in the mCherry+MIRI group relative to the mCherry+Sham, mCherry+EA-pre+MIRI, and the hM4Di+MIRI groups. The hM3Dq+EA-pre+MIRI group showed significantly higher CK-MB, CTnT, and NE than the mCherry+EA-pre+MIRI group (Fig. 6E,F,G). In conclusion, our findings indicated that EA-pre enhanced cardiac function in MIRI mice by mitigating sympathetic nerve injury, achieved through the inhibition of Crus Ⅰ^PC.

Fig. 6.

Effects of chemogenetic manipulation of neuronal activity in Crus I^𝐏𝐂 on cardiac tissue damage in MIRI mice. (A) Representative images of mouse heart double-stained with TTC and Evans blue for determining infarcted areas, scale bar, 1 mm. (B) Ratios of infarcted area to area at risk and percentage of area at risk. n = 6/group, one-way ANOVA with Tukey’s post-test. (C,D) HE and Masson staining of myocardial tissue. Scale bar, 50, 25 µm. (E,F,G) Serum levels of CK-MB, cTnT, and NE across groups. Data are presented as mean $\pm$ SEM; n = 6/group; *p 0.05; **p 0.01; ***p 0.001; according to one-way ANOVA with Tukey’s post-test.

First, this study used only male C57 mice, and the conditions for expressing the recombinant adenovirus with the L7 promoter are limited. Future studies will consider using L7-cre mice. Second, cerebellar cortex neurons exhibit complex types with distinct morphological features. Their projection patterns, transmitter types, and firing properties require further investigation. Last, the present study tentatively suggested that Crus Ⅰ is involved in the alleviation of MIRI by EA-pre. However, the role of Crus Ⅰ in modulating deep cerebellar nuclei in MIRI was not further investigated. This aspect will receive prioritized attention in our subsequent studies.