EA exerts analgesic effects by activating AMPK in the inflammatory skin tissue

After daily treatment, the pain behavior test was used for assessing EA efficacy. Results showed that on the first day after modeling, compared with the Control group, the mechanical and thermal pain latency of the affected hind paw of mice in CFA, CFA + EA and CFA + sham EA groups were significantly lower than those in the Control group (Fig. 1A, B, P < 0.05), which proved that the CFA modeling was successful and comparable. Compared with the Control group, mice in the CFA group had significantly lower mechanical pain threshold (Fig. 1A, P < 0.05) and shorter thermal pain latency (Fig. 1B, P < 0.05); In comparison with the CFA group, mice in the EA group but not in the sham EA group demonstrated significantly elevated mechanical pain threshold (Fig. 1A, P < 0.05) and extended thermal pain latency (Fig. 1B, P < 0.05). Moreover, the pain threshold progressively increased in correlation with the duration of EA treatment (Fig. 1). These findings suggested that EA has an analgesic effect.

Fig. 1

Effect of EA on pain behavior and AMPK activation in inflamed skin tissue. A Mechanical pain threshold. B thermal pain latency. The data was shown as mean ± SEM (n = 10). Two-way ANOVA was used to analyze the data. C AMPK phosphorylation level in affected (ipsilateral) dorsal hindpaw inflamed skin tissue. D A statistical histogram of p-AMPK percentage in total AMPK.E statistical histogram of total AMPK percentage in β-actin. F AMPK phosphorylation level in normal skin tissue of the healthy side (contralateral) dorsal hindpaw.G A statistical histogram of p-AMPK percentage to total AMPK. H A statistical histogram of total AMPK percentage to β-actin. The data is shown as mean ± SEM (n = 4).One-way ANOVA was used to analyze the data. *P < 0.05 compared with Control group; #P < 0.05 compared with CFA group; &P < 0.05 compared with CFA + EA group

In order to determine whether EA affects the inflammatory response, we used H&E staining to investigate the pathological changes of the inflammed skin of the left dorsal foot. It revealed significant epidermal hyperplasia, significantly increased telangiectasia, and inflammatory cell infiltration in CFA, CFA + EA and CFA + sham EA groups (Supplementary Fig. 3 A). EA significantly improved the histologic changes caused by CFA. The skin thickness of the CFA group was significantly increased compared with the control group, which was decreased in the CFA + EA group ((Supplementary Fig. 3B, P < 0.01).

The phosphorylation level of AMPK was assessed on day 7 following the completion of EA. WB analysis of the ipsilateral dorsal hind paw, which was the affected side, revealed significant differences in the ratio of phosphorylated AMPK (p-AMPK) to total AMPK. Specifically, the p-AMPK at the α-subunit Thr172 site, normalized to total AMPK, exhibited a marked increase in the CFA group, CFA + EA group and CFA + sham EA group when compared to the control group (Fig. 1C, D, P < 0.05). However, the analysis did not reveal any significant alterations in the overall levels of AMPK proteins (Fig. 1E, P > 0.05). It indicated that CFA activated AMPK in inflamed skin tissue. Compared with the CFA group, EA but not sham EA significantly increased p-AMPK /AMPK percentage of EA group (Fig. 1C, D, P < 0.05). Compared with CFA + EA group, AMPK activation was lower in CFA + sham EA group (Fig. 1C, D, P < 0.05). Importantly, there was no significant difference in the expression level of p-AMPK in normal skin of the healthy (contralateral) side dorsal hindpaw among all groups (Fig. 1F–H, P > 0.05), indicating that EA specifically activated AMPK in inflamed skin tissue.

As a kind of non-specific immune defense cells, macrophages are widely found in blood vessel walls and loose connective tissues, and play a key role in CFA-induced inflammatory pain[29].To explore AMPK activation in macrophages during EA treatment of inflammatory pain, p-AMPK expression was observed by immunofluorescence double-labeling technique. Macrophages of mice in each group were mainly distributed in the dermis, and the cytoplasm showed light green fluorescence (Fig. 2A). p-AMPK positive cells were mainly distributed in all skin layers, and their nuclei and cytoplasm showed red fluorescence (Fig. 2A), while macrophages and p-AMPK double-labeled positive cells showed yellow fluorescence (Fig. 2A).

Fig. 2

Effect of EA on AMPK activation in macrophages in inflamed skin tissue. A Represents the double-labeled fluorescence results of macrophages (CD68 +) and p-AMPK in inflamed skin tissue. Macrophages (CD68 +) and p-AMPK had double labeling. The scale bars are 50 μm and 10 μm (Zoom in). B A statistical histogram of the percentage of positive area of double-labeled cells/macrophages. The data is shown as mean ± SEM (n = 4). One-way ANOVA was used to analyze the data. *P < 0.05 compared with Control group; #P < 0.05 compared with the CFA group; &P < 0.05 compared with the CFA + EA group

Compared with the Control group, the percentage of positive area of macrophages expressing p-AMPK in the inflamed skin tissue of the CFA group, CFA + EA group and CFA + sham EA group was significantly upregulated (Fig. 2B, P < 0.05). The results showed that CFA-induced inflammatory pain significantly increased the expression of p-AMPK in macrophages that had infiltrated the inflamed skin tissue. This increase in p-AMPK expression corresponds to an enhanced activation of the AMPK pathway. In addition, compared with the CFA group, the expression of p-AMPK in macrophages of inflamed skin tissue in CFA + EA group was significantly upregulated (Fig. 2B, P < 0.05), indicating that EA could further promote AMPK activation in macrophages (Fig. 2B, P < 0.05). There was no significant difference in the percentage of positive area of double-labeled cells in the CFA + sham EA group compared with the CFA group (P > 0.05), but it was significantly lower than that in the CFA + EA group, suggesting that sham EA could not further promote AMPK activation in macrophage. The data indicate that inflammatory pain induced by CFA is associated with an increased activation of AMPK in macrophages that have infiltrated the inflamed skin. Furthermore, our findings suggested that EA may augment AMPK activation in macrophages of the affected side's inflamed skin, potentially contributing to the analgesic effects observed in our study.

On the 2–7 days after CFA modeling, mice were intraperitoneally injected with AMPK inhibitor Compound C 30 min before EA treatment assess the potential impact of AMPK inhibitor on the analgesic effects of EA. In comparison to the CFA group, the CFA + Veh. + EA group demonstrated a significant elevation in both mechanical and thermal pain latency (Fig. 3A, B, P < 0.05). Conversely, the CFA + CC + EA group exhibited no significant alterations in either mechanical or thermal pain latency (Fig. 3A, B, P > 0.05). Furthermore, when contrasted with the CFA + Veh. + EA group, there was a marked reduction in both mechanical and thermal pain latency in CFA + CC + EA group (Fig. 3A, B, P < 0.05). These results suggested that the intraperitoneal administration of Compound C attenuated the analgesic effects of EA on mechanical allodynia and thermal hyperalgesia induced by CFA.

Fig. 3

AMPK inhibitor Compound C reversed the analgesic effect of EA and the EA-mediated AMPK activation in inflamed skin tissue. A Mechanical pain threshold. B thermal pain latency. The data was shown as mean ± SEM (n = 8). Two-way ANOVA was used to analyze the data. *P < 0.05 compared with Control group; #P < 0.05 compared with CFA group; &P < 0.05 compared with CFA + EA group. C The gel representation of p-AMPK and AMPK in WB. D A statistical histogram of the percentage of p-AMPK in total AMPK. E A statistical histogram of the percentage of total AMPK in β-actin. The data is shown as mean ± SEM (n = 4). One-way ANOVA was used to analyze the data. *P < 0.05 compared with Control group; #P < 0.05 compared with CFA group; &P < 0.05 compared with CFA + Veh. + EA group

On the 7th day post-modeling, the affected skin tissue was analyzed using WB. The findings revealed that, in comparison to the Control group, the percentage of p-AMPK /AMPK was significantly elevated in the CFA, CFA + Veh. + EA, and CFA + CC + EA groups (Fig. 3C, D, P < 0.05). However, there was no significant change in the total AMPK protein levels (Fig. 3E, P > 0.05). When compared to the CFA group, the percentage of p-AMPK/AMPK in the CFA + Veh. + EA group was significantly higher (Fig. 3C, D, P < 0.05), whereas the percentage in the CFA + CC + EA group remained unchanged (Fig. 3C, D, P > 0.05). Notably, the p-AMPK /AMPK percentage in the CFA + CC + EA group was significantly reduced compared to the CFA + Veh. + EA group, indicating that the intraperitoneal administration of Compound C counteracted the activation of AMPK in the inflamed skin tissue.

EA promotes the expression of β-END in inflamed skin tissue by activating AMPK

To determine whether EA promotes the expression of β-END in the inflamed skin tissue, WB test was used to detect its protein level. The results showed that, compared with the Control group, β-END expression in the CFA group, CFA + EA group and CFA + sham EA group was significantly increased (Fig. 4A, B, P < 0.05). Compared with the CFA group, the β-END expression in the CFA + EA group was significantly increased (Fig. 4A, B, P < 0.05). Compared with the CFA group, there was no significant change in the CFA + sham EA group (Fig. 4A, B, P > 0.05). The expression of β-END in the CFA + sham EA group was significantly lower than that in CFA + EA group (Fig. 4A, B, P < 0.05). The above experimental results showed that the inflammatory response leads to an increase in tissue β-END expression, and EA promotes this response.

Fig. 4

The effect of AMPK inhibitor Compound C reversed the effect of EA on promoting β-END expression in inflamed skin tissue. A Gel representation of β-END expression levels in skin tissue in Control, CFA, CFA + EA and CFA + sham EA group. B A statistical histogram of β-END expression levels in skin tissue. The data is shown as mean ± SEM (n = 4). C Gel representation of β-END expression levels in inflamed skin tissue of Control, CFA, CFA + Veh. + EA and CFA + CC + EA group. D A statistical histogram of β-END expression level in skin tissue. The data is shown as mean ± SEM (n = 4). One-way ANOVA was used for all analysis. *P < 0.05 compared with Control group; #P < 0.05 compared with CFA group; &P < 0.05 compared with the CFA + EA or CFA + Veh. + EA group

To further clarify that EA enhanced expression of the analgesic substance β-END in the macrophages of inflamed skin tissue, we tested the percentage of the β-END expressing macrophages. The results showed that macrophages (CD68 +) and β-END positive cells were mainly distributed in the dermis, and the cytoplasm of macrophages showed light green fluorescence, the cytoplasm of β-END positive cells showed red fluorescence, and the double-labeled positive cells showed yellow fluorescence (Fig. 5A). Compared with the Control group, the percentage of the β-END expressing macrophages in CFA group was significantly upregulated (Fig. 5B, P < 0.05). Compared with CFA group, the percentage of the β-END expressing macrophages in CFA + EA group was significantly upregulated (Fig. 5B, P < 0.05). Compared with the CFA group, the percentage of the β-END expressing macrophages in the CFA + sham EA group had no significant change (Fig. 5B, P > 0.05). The percentage of the β-END expressing macrophages in CFA + sham EA group was significantly lower than that in the CFA + EA group (Fig. 5B, P < 0.05). The above results suggested that EA can up-regulate the expression of β-END in macrophages infiltrated in inflamed skin tissue.

Fig. 5

Effects of EA on macrophages and β-END expression in the inflamed skin tissue. A Double labeling of β-END and macrophage (CD68 +) in skin tissue with scale bars of 50 µm and 10 µm (Zoom in). B A statistical histogram of the percentage of positive area of macrophages (CD68 +) and β-END double-labeled cells/macrophages (CD68 single label) in the skin tissue of WT mice. The data is shown as mean ± SEM (n = 4). One-way ANOVA was used for analysis. *P < 0.05 compared with Control group; #P < 0.05 compared with the CFA group; &P < 0.05 compared with the CFA + EA group

To confirm whether EA promotes the expression of β-END by activating AMPK in the inflamed skin tissue, WB experiment was used to investigate whether AMPK inhibitor Compound C could reverse the effect of EA on the expression of β-END. Compared with the Control group, the expression level of β-END in the CFA group and CFA + Veh. + EA group was significantly increased (Fig. 4C, D, P < 0.05). Compared with the CFA group, the expression level of β-END in the CFA + Veh. + EA group was significantly increased (Fig. 4C, D, P < 0.05). Compared with the CFA + Veh. + EA group, the expression level of β-END in CFA + CC + EA group was significantly decreased (Fig. 4C, D, P < 0.05). There was no significant difference in β-END expression level between CFA + CC + EA and CFA group (Fig. 4 C, D, P > 0.05). These results indicated that inhibiting AMPK reversed the EA induced up-regulation of β-END expression in inflamed skin tissue. It is suggested that EA may promote the expression of β-END in inflamed skin tissue by activating AMPK, thus exerting analgesic effect, which may be the downstream of EA mediated AMPK activation.

EA significantly upregulated the content of endocannabinoids and the expression of CB2 receptor in inflammatory skin tissue

Chromatography was used to analyze the endocannabinoids 2-AG and AEA in skin tissue (Figs. 6, 7, P < 0.05). The results showed that compared with the Control group, the levels of 2-AG and AEA in the inflamed skin tissue of the CFA group, CFA + EA group and CFA + sham EA group were all upregulated (Figs. 6E, 7E, P < 0.05), indicating that the CFA can upregulate the contents of endocannabinoids in the inflamed skin tissue. Compared with the CFA group, the levels of 2-AG and AEA in the inflamed skin tissue of the CFA + EA group were upregulated (Figs. 6E, 7E, P < 0.05), indicating that EA further upregulated the levels of endocannabinoid in the inflamed skin tissue. Compared with the CFA group, the contents of 2-AG and AEA in inflamed skin tissue of the CFA + sham EA group had no significant change (Figs. 6E, 7E, P > 0.05). Compared with the CFA + EA group, the CFA + sham EA group had significantly lower levels of 2-AG and AEA in inflamed skin tissue (Figs. 6E, 7E, P < 0.05), indicating that sham EA could not further up-regulate the contents of endocannabinoids in inflamed skin tissue.

Fig. 6

Effect of EA on 2-AG levels in the inflammatory skin tissue of the affected side of the dorsal hindpaw in mice. A–D The chromatogram of 2-AG in the skin tissue of affected side with dorsal hindpaw inflammation in the Control group, CFA group, CFA + EA group, and CFA + sham EA group. The arrow indicates the 2-AG peak. E A statistical histogram of the 2-AG levels. Data were expressed as mean ± SEM (n = 8). One-way ANOVA was used to analyze the data. *P < 0.05 compared with the Control group; #P < 0.05 compared with CFA group; &P < 0.05 compared with the CFA + EA group

Fig. 7

Effect of EA on AEA levels in the inflammatory skin tissue of the affected side of the dorsal hindpaw in mice. A–D The chromatogram of AEA in the skin tissue of affected side with dorsal hindpaw inflammation in the Control group, CFA group, CFA + EA group, and CFA + sham EA group. The arrow indicates the AEA peak. E A statistical histogram of the AEA levels. The data is shown as mean ± SEM (n = 8). One-way ANOVA was used to analyze the data. *P < 0.05 compared with the Control group; #P < 0.05 compared with the CFA group; &P < 0.05 compared with CFA + EA group

Concurrently, WB analysis was conducted to assess CB2R expression in the skin tissue. The WB analysis indicated variable expression of CB2R protein among the Control, CFA, CFA + EA and CFA + sham EA group (Fig. 8A, P < 0.05). The percentage of gray value of CB2R and GAPDH in the above four groups was evaluated, and it was found that there was no significant change in CB2R protein expression in the CFA group compared with the Control group (Fig. 8B, P > 0.05). However, the expression level of CB2R protein in inflamed skin tissue was significantly increased in the CFA + EA group and CFA + sham EA group (Fig. 8B, P < 0.05). Compared with the CFA group, the expression level of CB2R protein in the inflamed skin tissue in the CFA + EA group was significantly increased (Fig. 8B, P < 0.05), while there was no significant change in the CFA + sham EA group (Fig. 8B, P > 0.05). The expression of CB2R protein in the CFA + sham EA group was significantly lower than that in the CFA + EA group (Fig. 8B, P < 0.05).

Fig. 8

Effect of EA on the expression of CB2R and CB2R KO reversed the analgesic effect of EA and inhibited EA induced AMPK activation in inflamed skin tissue. A A gel representation of the effect of EA on the expression of CB2R. B Statistical histograms of relative quantification of CB2R protein of Wild type (WT) mice. The data is shown as mean ± SEM (n = 3). C The mechanical pain threshold of the affected side of CB2R KO mice. D The thermal pain latency of the affected side of CB2R KO mice, and the data is shown as mean ± SEM (n = 10). *P < 0.05 compared with the Control group. E A gel representation of the effect of EA on the level of AMPK phosphorylation in local inflamed skin tissue of CB2R KO mice. F A statistical histogram of the percentage of p-AMPK in total AMPK. G A statistical histogram of the percentage of total AMPK in β-actin. The data is shown as mean ± SEM (n = 4). One-way ANOVA was used to analyze the data of B, F, G. Two-way ANOVA was used to analyze the data of C, D. *P < 0.05 compared with Control group; #P < 0.05 compared with CFA group; &P < 0.05 compared with CFA + EA group

These results indicate that the expression level of CB2R protein and the endocannabinoids 2-AG and AEA in the affected skin tissue can be significantly upregulated when EA exerts analgesic effect.

CB2R KO reverses the analgesic effect of EA and inhibits EA-induced AMPK activation in inflammatory skin macrophages

To substantiate the role of the CB2R in the analgesic effects of EA and its mediation of AMPK activation, we employed a genetic approach utilizing CB2R KO mice. This methodology allows for a direct assessment of CB2R’s contribution to the observed therapeutic responses. The identification of CB2R KO mice was detailed in the supplementary materials. Pain threshold test results showed that on the first day after CFA modeling, the mechanical pain threshold and thermal pain latency of mice in CFA group, CFA + EA group and CFA + sham EA group were significantly lower than those in the Control group (Fig. 8C, D, P < 0.05), indicating that CB2R KO mice with inflammatory pain were successfully induced by CFA. Compared with the CFA group, EA no longer relieved mechanical allodynia and thermal hyperalgesia of modeling mice (Fig. 8C, D, P > 0.05), suggesting that CB2R KO significantly inhibited the analgesic effect of EA.

WB experiment results showed that compared with the Control group, the activation level of AMPK in the inflamed skin tissue in the CFA group, CFA + EA group and CFA + sham EA group was increased (Fig. 8E, F, P < 0.05), while the total AMPK protein level was not significantly changed (Fig. 8G, P > 0.05). Compared with the CFA group and CFA + sham EA group, there was no significant change in AMPK activation in the CFA + EA group (Fig. 8E, F, P > 0.05), suggesting that CB2R KO significantly inhibited the EA-mediated activation of AMPK in inflamed skin tissue.

Furthermore, we investigated whether CB2R participated in EA-mediated activation of AMPK in macrophages infiltrated in inflamed skin tissue (Fig. 9A). After systemic knockout of CB2R, the expression of p-AMPK in macrophages of inflamed skin tissue in the CFA group was not significantly upregulated compared with that in Control group (Fig. 9B, P > 0.05). In addition, the percentage of p-AMPK in macrophages of the CFA + EA group was significantly increased compared with that in Control group (Fig. 9B, P < 0.05), but there was no significant difference compared with CFA group (Fig. 9B, P > 0.05). It is suggested that the systemic knockout of CB2R can inhibit EA-mediated AMPK activation in macrophages. Compared with the CFA + EA group, the percentage of double-labeled cells in the CFA + sham EA group significantly decreased (Fig. 9B, P < 0.05), suggesting that after CB2R KO, EA still had a certain effect on promoting AMPK activation in the inflamed skin tissue.

Fig. 9

CB2R KO inhibits the activation of AMPK by EA in macrophages of inflamed skin tissue. A The double-labeled fluorescence results of macrophages (CD68 +) and p-AMPK in the inflamed skin tissue of CB2R KO mice. The macrophages (CD68 +) and p-AMPK had double labeling. The scale bars are 50 μm and 10 μm (Zoom in). B A statistical histogram of the percentage of positive area of double-labeled cells/macrophages. The data is shown as mean ± SEM (n = 4). One-way ANOVA was used to analyze the data. *P < 0.05 compared with Control group; #P < 0.05 compared with the CFA group; &P < 0.05 compared with CFA + EA group. C A statistical histogram of the percentage of positive area of double-labeled cells/macrophages between WT and CB2R KO groups. The data is shown as mean ± SEM (n = 4). T-test analysis was used to analyze the data. *P < 0.05 compared with WT mice

To investigate the difference in AMPK activation in macrophages of inflamed skin tissue after CB2R KO in each group of mice, we compared the percentage of p-AMPK expressing macrophages between WT (Data from Fig. 2) and CB2R KO mice. The results showed that the percentage of p-AMPK expressing macrophages in the CFA group, CFA + EA group and CFA + sham EA group with CB2R KO was significantly lower than that of WT mice (Fig. 9C, P < 0.05). There was no significant difference in the percentage of p-AMPK expressing macrophages between the WT and CB2R KO Control groups (Fig. 9C, P > 0.05). These results further suggested that systemic CB2R KO significantly inhibited the activation of AMPK by CFA and EA in macrophages of inflamed skin tissue.