Electroacupuncture alleviates diabetic peripheral neuropathy through modulating mitochondrial biogenesis and suppressing oxidative stress

Abstract

BACKGROUND

Peripheral neuropathy caused by diabetes is closely related to the vicious cycle of oxidative stress and mitochondrial dysfunction resulting from metabolic abnormalities. The effects mediated by the silent information regulator type 2 homolog-1 (SIRT1)/peroxisome proliferator-activated receptor-gamma coactivator-1α (PGC-1α) axis present new opportunities for the treatment of type 2 diabetic peripheral neuropathy (T2DPN), potentially breaking this harmful cycle.

AIM

To validate the effectiveness of electroacupuncture (EA) in the treatment of T2DPN and investigate its potential mechanism based on the SIRT1/PGC-1α axis.

METHODS

The effects of EA were evaluated through assessments of metabolic changes, morphological observations, and functional examinations of the sciatic nerve, along with measurements of inflammation and oxidative stress. Proteins related to the SIRT1/PGC-1α axis, involved in the regulation of mitochondrial biogenesis and antioxidative stress, were detected in the sciatic nerve using Western blotting to explain the underlying mechanism. A counterevidence group was created by injecting a SIRT1 inhibitor during EA intervention to support the hypothesis.

RESULTS

In addition to diabetes-related metabolic changes, T2DPN rats showed significant reductions in pain threshold after 9 weeks, suggesting abnormal peripheral nerve function. EA treatment partially restored metabolic control and reduced nerve damage in T2DPN rats. The SIRT1/PGC-1α axis, which was downregulated in the model group, was upregulated by EA intervention. The endogenous antioxidant system related to the SIRT1/PGC-1α axis, previously inhibited in diabetic rats, was reactivated. A similar trend was observed in inflammatory markers. When SIRT1 was inhibited in diabetic rats, these beneficial effects were abolished.

CONCLUSION

EA can alleviate the symptoms of T2DNP in experimental rats, and its effects may be related to the mitochondrial biogenesis and endogenous antioxidant system mediated by the SIRT1/PGC-1α axis.

Key Words: Electroacupuncture; Type 2 diabetic peripheral neuropathy; Silent matching type information regulation 2 homolog-1/peroxisome proliferator-activated receptor-gamma coactivator-1α axis; Mitochondria biogenesis; Oxidative stress

Core Tip: Diabetic peripheral neuropathy (DPN) is thought to be linked to the vicious cycle of oxidative stress and mitochondrial dysfunction triggered by metabolic abnormalities. Electroacupuncture is commonly used as an adjuvant therapy for DPN in clinical practice, though its precise mechanism remains unclear. A rat model of DPN was established by combining a high-fat diet with streptozotocin injection. The final results suggest that the beneficial effects of electroacupuncture on DPN rats may be related to the regulation of the silent information regulator type 2 homolog-1/peroxisome proliferator-activated receptor-gamma coactivator-1α axis, which activates mitochondrial biogenesis and reduces oxidative stress and protects intact mitochondria.

INTRODUCTION

Diabetic peripheral neuropathy (DPN) is one of the most common complications of diabetes, characterized by symmetrical numbness and pain in the extremities[1]. It is considered a primary cause of diabetic foot[2,3]. Without proper treatment, the progression of pathological sensations such as numbness and pain can increase the risk of foot ulcers, lower limb amputation, and even death[4]. Current drugs targeting the pathological mechanisms, such as aldose reductase inhibitors and protein kinase C inhibitors, play a limited role in slowing the progression of DPN and are often associated with severe adverse effects[5-7]. Therefore, the need for potent and safe clinical treatments remains unmet.

Impaired glucose metabolism in diabetic patients leads to the diversion of glucose and glycolysis intermediates into alternative metabolic or non-metabolic pathways, producing reactive oxygen species (ROS) that exceed the regulatory capacity of the body[8]. It leads to oxidative stress, defined as a state of redox imbalance. Oxidative stress is generally considered an uncontrolled pathological process mediated by reactive free radicals. Excessive ROS can attack biological molecules (DNA, RNA, lipids, and proteins) and physiological processes (such as nucleic acid oxidation and lipid peroxidation)[9,10]. Mitochondria are the primary source of cellular ROS[11], and because mitochondrial lipids, proteins, and DNA are close to the ROS production sites, they are particularly vulnerable to ROS attacks[12]. This results in decreased activity of the mitochondrial electron transport chain (ETC) and reduced adenosine triphosphate production[13]. The morphology of peripheral nerves is highly polarized, with elongated axons, especially the small unmyelinated nerve fibers at their ends, relying heavily on locally anchored mitochondria for energy to maintain normal neurophysiological activities. However, peripheral nerves are fragile in terms of metabolism and structure, making them highly susceptible to oxidative stress damage, which results in mitochondrial dysfunction[6]. Mitochondrial dysfunction and oxidative stress reinforce each other in DPN, creating a vicious cycle. It leads to energy deficiency and worsens oxidative damage. Under conditions of mitochondrial degeneration and oxidative stress injury, a treatment strategy aimed at promoting the regeneration of healthy mitochondria and activating antioxidant defense mechanisms could theoretically alleviate energy deficiency and oxidative damage in the peripheral nerves of DPN patients, thus offering neuroprotection.

Silent information regulator type 2 homolog-1 (SIRT1) is a nutrient-dependent deacetylase[14], whose activity is regulated by the ratio of NAD+/NADH[15]. It is highly sensitive to changes in the energy demands of the body and can deacetylate various proteins and genes involved in the regulation of mitochondrial homeostasis and the activation of antioxidant pathways. Peroxisome proliferator-activated receptor-gamma coactivator-1α (PGC-1α), a transcriptional coactivator downstream of SIRT1, is considered a major regulator of mitochondrial biogenesis in many tissues. It enhances the expression of mitochondrial transcription factor A (TFAM) by activating downstream nuclear respiratory factor 1 (Nrf1)[16]. TFAM is directly correlated with mitochondrial DNA levels, which typically increase alongside mitochondrial DNA, resulting in an increase in nascent mitochondria[17]. SIRT1 can also directly deacetylate nuclear factor erythroid 2-related factor 2 (Nrf2) or indirectly activate it through the deacetylation of PGC-1α. Nrf2 is a key regulator of the cellular antioxidant defense system[18]. Once activated, it upregulates various antioxidant proteins in response to oxidative conditions, thereby reducing intracellular oxidative damage. Thus, activation of the SIRT1/PGC-1α axis has been identified as a potential therapeutic target for axonal diseases and neuropathy.

Electroacupuncture (EA) has been widely recognized for its efficacy in the clinical treatment of DPN. EA activates multiple bioactive substances to relieve pain, and its benefits have been demonstrated in both preclinical and clinical trials. It is now used globally to treat various pain conditions[19-22]. Additionally, EA has been shown to reduce neuroinflammation[23,24], lower blood sugar[25,26] and lipid levels[27,28], and increase nerve conduction velocity (NCV) in various neuropathies[29,30]. However, the underlying mechanisms of EA in treating DPN remain unclear. In our study, we used a high-fat diet (HFD) and low-dose streptozotocin (STZ)-induced rat model to investigate the neuroprotective effects of EA on experimental diabetes. Given the positive effects of the SIRT1/PGC-1α axis on mitochondrial biogenesis and antioxidant defense mechanisms, we aimed to explore the potential mechanisms by which EA treats DPN.

MATERIALS AND METHODS

Establishment and grouping of experimental animal models

All animal experiments were carried out in accordance with the Principles of Laboratory Animal Care and the Guide for the Care and Use of Laboratory Animals published by the National Science Council of China (grant No. 202205A011). Healthy adult male Sprague-Dawley rats (n = 38) weighing 200-220 g and aged 8 weeks were housed in the specific pathogen-free facility of the Animal Experimental Center at Nanjing University of Traditional Chinese Medicine. The rats had ad libitum access to food and water. Before the experiment, rats were fed a standard diet for 1 week and then randomly divided into four groups: Control group (n = 8); model group (n = 10); EA group (n = 10); and EA + EX527 group (n = 10). While the control group continued on a standard diet, the model, EA, and EA + EX527 groups were switched to HFD to simulate symptoms of insulin (INS) resistance in rats, which continued until the end of the experiment. After 2 weeks of HFD, the rats in the model group were injected with low dose STZ (Aladdin, China; 35 mg/kg in 0.1 M citric acid/sodium citrate buffer, pH 4.5), but the rats had to fast for 10 h before the injection to sufficiently destroy part of the pancreatic β cell function. A model of type 2 diabetes mellitus (T2DM) was deemed successfully established if the rats’ blood glucose levels were ≥ 16.7 mmol/L and remained stable for 1 week. Nine weeks after the development of diabetes, a pain threshold behavioral test was carried out. Those rats confirmed to have significant differences after the combined HFD-STZ intervention were considered to be successfully modeled as type 2 DPN (T2DPN) and included in the study. Rats that failed to exhibit symptoms of T2DPN were humanely euthanized with an intraperitoneal injection of 25% urethane (10 mL/kg).

EA treatment

Rats designated for treatment were anesthetized with 1%-2% isoflurane and positioned supine on a heating pad to maintain their body temperature. At the “Tianshu” (ST25) point on the abdomen (location: The junction between the upper one-third and lower two-thirds of the line connecting the xiphoid process to the pubic symphysis, 0.5 cm lateral to the midline, bilaterally), two non-insulated stainless steel acupuncture needles were obliquely inserted to a depth of 3 mm. These needles were connected to the output terminal of an EA device (LH402A, Beijing Huawei Technologies Co., Ltd.) to provide intermittent wave electrical stimulation at a frequency of 2/15 Hz and an intensity of 2 mA. The treatment was conducted at a fixed time every day for 20 min and lasted for a total of 6 weeks. Rats in the model group were anesthetized with 1%-2% isoflurane for the same duration to eliminate the possibility of blood glucose fluctuations caused by gas anesthesia.

Inhibitor administration

Following previous experimental protocols[31], EX527 (Abmole, United States) was prepared under dark conditions. The primary solution consisted of 10 mg of EX527 powder dissolved in 0.1 mL of dimethylsulfoxide (Abmole, United States). This solution was then added to a buffer containing sulfobutylether-β-cyclodextrin in physiological saline and diluted to a final concentration of 1 mg/mL.

Behavioral test

After successfully establishing the T2DM rat model with the combination of HFD-STZ over 9 weeks, behavioral tests were conducted at a fixed time every week to evaluate the mechanical and thermal sensitivity of rats. Before administering mechanical or thermal stimuli, we placed the rats on a metal mesh (or a glass plate) to allow full acclimation to the environment and to eliminate unnecessary errors. Experimental procedures for each group of rats are shown in Figure 1.

Figure 1 Experimental timeline of type 2 diabetic peripheral neuropathy rats modeling, behavior testing, and intervention. EA: Electroacupuncture; STZ: Streptozotocin.

Hind paw withdrawal threshold: The Von Frey test was used to evaluate the hind paw withdrawal threshold in response to mechanical stimulation in rats. The rats were placed on a metal mesh plate and confined under an inverted acrylic box. Von Frey fibers were then gently pressed against the soles of the rats’ hind paws and slowly increased in pressure. The threshold at which the rats lifted their hind paws in response to the mechanical stimulation was recorded to evaluate their sensitivity to mechanical pain. During the experiment, the rats’ foot soles were stimulated alternately at 5-min intervals for a total of three times, and the average value of the measurements was recorded.

Hind paw withdrawal latency: The Hargreaves test was used to quantify the thermal pain threshold in rats. Rats were placed on a glass plate, and their movement was restricted with an inverted acrylic box, while an infrared heat source was directed at the rats’ hind paws. The duration of each stimulation was limited to 20.1 s to avoid scalding of rats. The latency to paw withdrawal in response to thermal stimulation was recorded to assess the sensitivity of the rats to thermal pain. Throughout the experiment, the soles of the rats’ feet were stimulated alternately at 5-min intervals for a total of three times, with the average value of the responses recorded.

Microcirculatory blood perfusion

Rats anesthetized with 4% urethane were positioned prone on a heating pad to maintain their basal body temperature. The probe of the laser Doppler blood flow meter (PeriFlux 5000, Perimed, Sweden) was fixed to align with the medial skin of the lower limb of the rats. Once the readings stabilized, skin blood flow was recorded for 1 min to calculate the average blood perfusion unit. The same setup was used for measuring the blood flow of the sciatic nerve, but the probe was directed at the exposed middle segment of the sciatic nerve rather than the skin.

NCV

Anesthetized with 4% urethane, rats were placed in a prone position on a heating pad to maintain their basal body temperature. After depilatory cream was applied to remove hair from the lower limbs, incisions were carefully made to expose the sciatic nerve. Measurements were taken using the PowerLab 8/35 system (AD Instruments, Australia). The stimulation electrode was positioned at the sciatic notch of the rat, and the recording electrode was placed in the interdigital muscles. A single pulse of square-wave electrical stimulation was administered. The distance between the two electrodes and the latency from the onset of the electrical stimulation to the occurrence of the action potential were recorded. The calculation formula was as follows: NCV (m/s) = distance × latency.

Immunohistochemistry and quantification of intraepidermal nerve fiber density

Before tissue collection, the rats were euthanized by cervical dislocation. The footpads were carefully excised and thoroughly fixed in 4% paraformaldehyde. The fixed tissues were then dehydrated in a 30% sucrose solution, embedded in optimal cutting temperature compound, and stored at -80 °C until sectioning in a cryostat. Longitudinal frozen sections (50 µm in thickness) were cut using a freezing microtome (Leica, Germany) and collected into a 96-well plate. The subsequent processing steps refer to the methods described in previous literature[32]. The intraepidermal nerve fiber density (IENFD) was quantified by calculating the average number of nerve endings at the dermal-epidermal junction per millimeter.

Hematoxylin and eosin staining

After excised and transferred to 4% paraformaldehyde, the sciatic nerves of mice euthanized by cervical dislocation were fixed and embedded in paraffin. Transverse and longitudinal sections of the sciatic nerve were cut into 6 µm slices using a microtome (Leica, Germany) and mounted onto glass slides. The prepared samples were observed under an optical microscope (Olympus, Japan) after staining with hematoxylin and eosin (H&E).

Transmission electron microscopy

After careful excision of the rat sciatic nerve, a middle segment approximately 1 mm in length was preserved. This tissue section was immersed in precooled electron microscopy fixative for 1 h at 4 °C. Following fixation, the tissue was washed three times with 0.1 M PBS solution at 4 °C for 15 min each. Next, the tissue was immersed in 1% osmium tetroxide for 1 h, followed by another three washes with 0.1 M PBS solution for 15 min each. After gradient dehydration, the dehydrated tissues were infiltrated with increasing concentrations of embedding medium for 4 h, 12 h, and 6 h, respectively. Then, pure embedding medium was added to the molds and polymerized overnight at 37 °C. The embedded samples were cured in an oven at 60 °C for approximately 48 h. The blocks were trimmed and sectioned and ultimately prepared into ultra-thin sections of 60 nm. The sections were stained with lead citrate for 20 min, washed with double-distilled water, and left to dry overnight at room temperature. The final prepared tissue sections were examined using a 100 kV transmission electron microscope (HT7700, HITACHI, Japan).

Blood and tissue sample collection for ELISA and Western blot analysis

At the end of the experiment (week 15), rats were fasted for 12 h and anesthetized with isoflurane. The blood was collected from the aorta abdominalis and stored at 4 °C, followed by centrifugation (at 3000 rpm for 15 min) to obtain serum. The sciatic nerve on both sides of the rats were quickly removed, washed, and stored at -80 °C for western blot (WB) and ELISA.

ELISA

According to the manufacturer’s instructions, different types of ELISA kits (Nanjing Jiancheng Bioengineering Institute, Nanjing) were used to determine the levels of glucose and lipid metabolism, oxidative stress, and inflammation-related indicators in the serum and sciatic nerve tissue homogenates from different groups of mice. The final results were quantified using a microplate reader (BioTek, United States) based on colorimetric analysis.

WB analysis

The protein homogenate of the sciatic nerve was prepared using ristocetin-induced platelet aggregation lysis buffer supplemented with a protease inhibitor (Thermo Scientific). Protein concentration was determined by the bicinchoninic acid method (Thermo Scientific). An equal amount of protein sample was separated using sodium dodecyl sulfate polyacrylamide gel electrophoresis in 4%-12% gradient gel (L00686C, GenScript, United States) and then transferred to a polyvinylidene difluoride membrane. The membrane was immersed in 5% bovine serum albumin and incubated at 37 °C for 1 h to block non-specific binding. Subsequently, the polyvinylidene difluoride membrane was incubated overnight at 4 °C with specific primary antibodies targeting the following proteins: SIRT1 (1:1000, Abcam), PGC-1α (1:200, Abmart), Nrf1 (1:1000, Cell Signaling Technology), Nrf2 (1:500, SAB), TFAM (1:1000, Abcam), and interleukin (IL)-1β (1:1000, Abcam), followed by incubation with a horseradish peroxidase-conjugated secondary antibody at room temperature for 1 h. The final protein bands were visualized by chemiluminescence using enhanced chemiluminescence WB substrate.

Statistical analysis

The data were statistically analyzed using SPSS 24.0 (IBM Corp., Armonk, NY, United States) and graphically represented using GraphPad Prism 8.0.1 (GraphPad Inc., La Jolla, CA, United States). Results are expressed as mean ± standard error of the mean. The comparisons between before and after EA intervention were made with paired t-test, and the comparisons between the two different groups were made with an independent t-test. The comparisons between multiple groups were performed using one-way analysis of variance. Differences were considered statistically significant when P < 0.05.

RESULTS

Improvement of glucose and lipid metabolism in T2DPN rats by EA

To determine whether the low-dose STZ combined with HFD successfully established a rat model of T2DM and to evaluate the regulatory effects of EA on glycolipid metabolism in T2DPN, we first measured body weight, blood glucose, and related indexes of glucose and lipid metabolism in the serum of rats from each group. The weight gain of rats with HFD-STZ-induced diabetes decreased significantly (P < 0.05; Figure 2A). The blood glucose levels in HFD-STZ-induced rats were significantly higher than those in the control group (≥ 16.7 mmol/L; P < 0.01; Figure 2B). Glycated serum protein, which reflected average glucose levels over the prior 3 weeks and was independent of food intake, also increased in the model group (P < 0.01; Figure 2C). During the intraperitoneal glucose tolerance test, glucose metabolism was impaired, indicating reduced glucose tolerance (P < 0.05; Figure 2D and E). Increased serum INS levels suggest INS resistance (P < 0.05; Figure 2F). Abnormal levels of high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, total cholesterol, and triglycerides indicated a disruption in lipid metabolism (P < 0.01; Figures 2G-J). These results confirmed the successful establishment of the T2DM rat model.

Figure 2 Metabolism index of rats in different groups. A and B: Weekly changes of random blood glucose (A) and weight (B) in rats after streptozotocin injection; C-J: Evaluations in serum of rats in each group after intervention (n = 5). C: Glycosylated serum protein; D: Intraperitoneal glucose tolerance test (IPGTT); E: Area under the curve (AUC) for the IPGTT test; F: Fasting insulin (INS); G: Triglycerides (TG); H: Total cholesterol (T-CHO); I: High-density lipoprotein level (HDL-C); J: Low-density lipoprotein (LDL-C). ^aP < 0.05, ^bP < 0.01 model group vs control group; ^cP < 0.05, ^dP < 0.01 electroacupuncture (EA) group vs model group; ^eP < 0.05 EA + EX527 group vs EA group. GSP: Glycated serum protein.

Compared with the model group, rats in the EA group showed significant reductions in body weight (P < 0.05), blood glucose (P < 0.01), glycated serum protein (P < 0.01), intraperitoneal glucose tolerance test (P < 0.05), area under the curve (P < 0.05), and INS levels (P < 0.01) after treatment, suggesting that EA improved glucose metabolism, reduced symptoms of impaired glucose tolerance, and decreased INS resistance. EA also improved the abnormal levels of high-density lipoprotein cholesterol (P < 0.01), low-density lipoprotein cholesterol (P < 0.01), total cholesterol (P < 0.01), and triglycerides (P < 0.01), thereby alleviating lipid metabolic disorders. All these effects observed in the EA group were also present in the EA + EX527 group, indicating that the inhibition of SIRT1 did not significantly affect the role of EA in improving glucose and lipid metabolism.

Protection of peripheral nerve function of T2DPN rats by EA

After confirming the ameliorative effect of EA on glycolipid metabolism, we proceeded to investigate its impact on peripheral neuropathy. Peripheral nerve terminals, which lack myelin protection, are more susceptible to damage under long-term hyperglycemia. After approximately 9 weeks of diabetes development, there was a significant decrease in the latency of withdrawal response (P < 0.01) and lowered pain thresholds (P < 0.01) compared to the control group (Figure 3A and B). Significant changes were observed before and after the intervention. With the progression of diabetes, the withdrawal threshold (P < 0.01) and withdrawal latency (P < 0.05) of diabetic rats decreased (Figure 3C and D). EA treatment prevented the reduction in pain thresholds (P < 0.01) and restored the latency for paw withdrawal (P < 0.01), bringing them closer to normal levels. These changes in pain thresholds indicated that the pain sensitivity symptoms in untreated T2DPN rats continued to worsen as the disease progressed, which could be alleviated by EA treatment.

Figure 3 Relevant indexes of peripheral nerve function of rats in different groups. A and B: Weekly changes of withdrawal threshold (A) and latency (B); C and D: Changes before and after intervention of withdrawal threshold (C) and latency (D); E and F: Representative immunofluorescence images of nerve fiber in hind paw skin of rats and (E) measurement of intraepidermal nerve fiber density (IENFD) (F); G-I: The dotted line indicates the dermal-epidermal junction, and the arrow refers to the nerve fiber endings passing through the dermal-epidermal junction, which is used for nerve fiber counting. Levels of motor nerve conduction velocity (MNCV) (G), lower limb skin blood flow (H), and sciatic nerve blood flow (I) of rats in each group after intervention (n = 3-5). ^aP < 0.05, ^bP < 0.01 model group vs control group; ^cP < 0.05, ^dP < 0.01 electroacupuncture (EA) group vs model group; ^eP < 0.05, ^fP < 0.01 EA + EX527 group vs EA group; ^gP < 0.05, ^hP < 0.01 post-intervention vs pre-intervention.

Additionally, immunofluorescence staining of paw skin revealed that IENFD decreased in the model group but increased in the EA group. IENFD is considered the “gold standard” for assessing damage to small nerve fibers and is inversely correlated with mechanical and thermal detection thresholds to some extent (P < 0.05; Figure 3E and F). NCV and blood perfusion also showed an expected decrease (P < 0.05; Figure 3G-I), which was reversed by EA treatment (P < 0.05). EX527, a SIRT1 inhibitor, diminished the improvements in mechanical and thermal pain thresholds in T2DPN rats provided by EA and even led to worsening hyperalgesia and late-stage hypoalgesia, accompanied by lower IENFD (P < 0.01), slower NCV (P < 0.05), and reduced blood perfusion (P < 0.01). Based on these results, we speculated that EA exerts a neuroprotective effect on T2DPN, potentially mediated by the modulation of SIRT1.

Restoration of the morphology of sciatic nerve in T2DPN rats by EA

T2DPN is characterized by functional abnormalities and significant morphological lesions in large, myelinated nerves. H&E staining and transmission electron microscopy were used to observe the morphology of the sciatic nerve, primarily to evaluate the structure of the myelin sheaths and axons. As shown in Figure 4, H&E staining revealed the normal morphology of the sciatic nerve, including intact myelin sheaths, axons, and Schwann cells. Nerve fibers in the longitudinal section were orderly arranged, and in the cross-section, axons wrapped by uniform myelin sheaths were visible. Transmission electron microscopy demonstrated the normal ultrastructure of nerve fibers, including homogeneous fibers and a clear lamellar structure. In T2DPN rats induced by HFD-STZ, the sciatic nerve displayed disorganization, with lymphocyte infiltration, axonal swelling, vacuolation, and extensive demyelination, accompanied by abnormally folded myelin structures. Following EA intervention, the aberrant morphology of the sciatic nerve in T2DPN rats was significantly improved, presenting a complete and uniform circular structure, regular arrangement, and only slight demyelination. However, in the EA + EX527 group, where SIRT1 was inhibited, the effect of EA on nerve morphology improvement was blocked. Axons appeared atrophied or even lost, with extensive demyelination, and the morphology was highly irregular.

Figure 4 Representative images of hematoxylin and eosin staining and transmission electron microscopy. A and B: Representative images of hematoxylin and eosin staining. Left, low magnification (× 400). Right, high magnification of the boxed regions on the left; C: Eosin staining and transmission electron microscopy. Blue arrows show duplication of myelin. Yellow arrows show separation of the lamellae. A: Axis cylinder; AA: Axonal atrophy; AS: Axon swelling; D: Demyelination; EA: Electroacupuncture; MS: Medullary sheath; L: Lymphocytic infiltration; SC: Schwann cell.

EA improved the expression level of SIRT1/PGC-1α axis-related protein

To investigate the neuroprotective mechanism of EA in T2DPN, we measured the levels of proteins related to the SIRT1/PGC-1α axis in the rat sciatic nerve (Figure 5A). WB analysis showed a significant decrease in the expression of SIRT1 (P < 0.01) and PGC-1α (P < 0.01) in the sciatic nerve of T2DPN rats induced by HFD-STZ (Figure 5B and C). Similarly, the downstream proteins Nrf1 (P < 0.05), Nrf2 (P < 0.05), and TFAM (P < 0.01) exhibited consistent low expressions (Figure 5D-F). In contrast, after 6 weeks of EA treatment, the expression levels of these proteins associated with the SIRT1/PGC-1α axis were significantly improved (P < 0.05). When SIRT1 was inhibited by the injection of EX527, the expression of SIRT1 in the sciatic nerve was partially reduced, diminishing the positive effects of EA on downstream PGC-1α (P < 0.01). This inhibition also led to a reduction in the expression of Nrf1 (P < 0.01), Nrf2 (P < 0.05), and TFAM (P < 0.01), proteins related to mitochondrial biogenesis and oxidative stress regulation. These results suggest that EA may exert its neuroprotective effects by modulating the SIRT1/PGC-1α axis.

Figure 5 Effect of electroacupuncture on the expression of silent information regulator type 2 homolog-1/peroxisome proliferator-activated receptor-gamma coactivator-1α in sciatic nerve. A: Protein bands of silent information regulator type 2 homolog-1 (SIRT1), mitochondrial transcription factor A (TFAM), peroxisome proliferator-activated receptor-gamma coactivator-1α (PGC-1α), nuclear respiratory factor 1 (Nrf1), and nuclear factor erythroid 2-related factor 2 (Nrf2); B-F: Relative expression of SIRT1 (B), PGC-1α(C), TFAM (D), Nrf1 (E) and Nrf2 (F). β-tubulin was used as a loading control of SIRT1 and TFAM. β-actin was used as a loading control for PGC-1α, Nrf1, and Nrf2. (n = 3). ^aP < 0.05, ^bP < 0.01 vs control group; ^cP < 0.05, ^dP < 0.01 vs model group; ^eP < 0.05, ^fP < 0.01 vs electroacupuncture (EA) group.

Reduction of oxidative stress and inflammatory reaction in sciatic nerve by EA

We assessed indicators related to oxidative stress to further validate the regulatory effect of EA on oxidative stress in the sciatic nerve of T2DPN rats through the SIRT1/PGC-1α pathway. Under hyperglycemia, the levels of pyruvate (P < 0.05) and lactic acid (P < 0.01) in the sciatic nerve of the model group rats were elevated, indicating an increased amount of glucose available for metabolism in the sciatic nerve (Figure 6A and B). However, the peripheral nervous system in T2DPN generally suffers from energy deficiency, and this abnormal metabolic phenotype reflects mitochondrial dysfunction. EA significantly reduced the levels of pyruvate (P < 0.05) and lactic acid (P < 0.01) in the sciatic nerve of rats, while the injection of EX527 hindered the ability of EA to decrease pyruvate (with no significant effect) and lactic acid levels (P < 0.01). High concentrations of pyruvate and lactic acid are believed to have detrimental effects on neurons, and excessive ROS are generated through a weakened ETC, leading to oxidative stress[33]. We measured the antioxidant enzymes superoxide dismutase (SOD; P < 0.05), catalase (CAT; P < 0.01), and glutathione peroxidase (GSH-PX; P < 0.01) in the sciatic nerve as well as the lipid peroxide malondialdehyde (MDA; P < 0.01; Figure 6C-F), which reflect the extent of oxidative damage, using ELISA. The results indicated that the levels of antioxidant enzymes SOD, CAT, and GSH-PX in the sciatic nerve of the model group were lower than those in the control group, while the concentration of MDA was higher. EA increased the levels of SOD (P < 0.05), CAT (P < 0.01), and GSH-PX (P < 0.05) in the sciatic nerve of T2DPN rats and decreased the MDA (P < 0.01) content. EX527 inhibited the antioxidant effects of EA. Oxidative stress can induce an inflammatory response. Results showed that the levels of the proinflammatory factor tumor necrosis factor-alpha (P < 0.01) and IL-β (P < 0.05) in the sciatic nerve of rats in the model group were increased (Figure 6G-I), while the anti-inflammatory factor IL-6 was decreased (P < 0.05; Figure 6J). EA treatment reduced the levels of proinflammatory factors tumor necrosis factor-alpha and IL-β in the sciatic nerve of rats (P < 0.01). Compared to the EA group, the levels of proinflammatory factors in the sciatic nerve of the EA + EX527 group were increased (P < 0.01). The content of IL-6 did not change significantly in the model group, EA group, and EA + EX527 group.

Figure 6 Comparison of pyruvate, lactic acid, oxidative stress, and inflammatory factor-related indicators in sciatic nerve of rats in each group. A-G: The sciatic nerve of rats in each group (n = 5); A: Pyruvate; B: Lactic acid; C: Superoxide dismutase (SOD); D: Catalase (CAT); E: Glutathione peroxidase (GSH-PX); F: Malondialdehyde (MDA); G: Tumor necrosis factor-alpha (TNF-α); H and I: Protein band (H) and relative expression of interleukin (IL)-1β (I; n = 3); J: IL-6 in the sciatic nerve of rats in each group (n = 5). ^aP < 0.05, ^bP < 0.01 vs control group; ^cP < 0.05, ^dP < 0.01 vs model group; ^eP < 0.05, ^fP < 0.01 vs electroacupuncture (EA) group.

DISCUSSION

T2DM, characterized by elevated blood sugar levels and INS resistance, can lead to various complications, including T2DPN, which significantly affects the quality of life and life expectancy of patients. Multiple downstream mechanisms, such as the polyol pathway and protein glycosylation, are simultaneously triggered by uncontrolled blood glucose levels and lipid abnormalities, indicating that the development of T2DPN cannot be halted by targeting a single pathway[6]. Consequently, more effective therapies are needed to alleviate abnormal sensory symptoms and protect damaged peripheral nerves. The use of EA for the treatment of T2DPN has been validated in both animal and clinical trials, demonstrating reductions in blood sugar levels and pain as well as improvements in NCV[34].

However, the specific mechanism behind these effects remains unclear. Consistent with previous research, our study found that EA intervention improved impaired glucose and lipid metabolism in T2DPN rats, reducing damage to peripheral nerve function and morphology. Nevertheless, the beneficial effects of EA on peripheral nerve function and morphology in T2DPN were eliminated by EX527, a specific SIRT1 inhibitor. Furthermore, our experiment showed that EA could increase the expression of the SIRT1/PGC-1α axis in the sciatic nerve and upregulate downstream Nrf1 and TFAM related to mitochondrial biogenesis and Nrf2 related to antioxidant defense, thereby reducing oxidative stress and inflammatory reactions in the nerve. The potential effects of EA were similarly obstructed by SIRT1 inhibition, which eliminated the enhanced effect of EA on the expression of SIRT1 and its downstream targets PGC-1α, Nrf1, Nrf2, and TFAM in the sciatic nerve as well as its beneficial effects on oxidative stress and the inflammatory response in the sciatic nerve. These findings suggest that the neuroprotective effect of EA on DPN involves promoting mitochondrial biogenesis and antioxidant defense mediated by the SIRT1/PGC-1α axis.

This study used HFD-STZ to induce a T2DPN rat model, which is considered one of the most commonly used methods to induce experimental diabetes and its complications in rodents[35]. As a preclinical research animal model, it is used to discover and evaluate new treatment methods for diabetic neuropathy and has gained wide acceptance[36].

Initially, Sprague-Dawley rats were fed HFD, followed by a low-dose STZ injection. The HFD can induce INS resistance, but when used alone, it does not cause hyperglycemia. The damage caused by STZ to pancreatic β cells is related to the dosage of STZ. A single high dose of STZ targets glucose transporter 2, which is abundant on the membranes of pancreatic islet cells, inducing inflammation and lymphocyte infiltration in the islets. This process further results in the necrosis of pancreatic β cells and a reduction in their volume[37], ultimately leading to INS deficiency, hyperglycemia, polydipsia, and polyuria[37,38]. Meanwhile, a low-dose STZ injection partially destroys β cells. In this scenario, compensatory mechanisms of β cells are activated to maintain normoglycemia by increasing β cell mass and augmenting β cell function, which also leads to hyperinsulinemia. The diabetic rats generated through this method maintain a stable hyperglycemic state and do not require additional INS injections, making them suitable models for emulating the pathological processes of human T2DM and its complications[39].

In this experiment, the model group rats exhibited consistent changes with previous literature, including increased blood sugar levels, impaired INS resistance, reduced glucose tolerance, and abnormal blood lipid profiles. These complex risk factors induce neuronal dysfunction. Under these circumstances, the inhibitory pathway may be impaired, or the harmful pathway may be excessively stimulated, leading to an imbalance between painful and non-painful stimuli, resulting in neuropathic pain[40]. Although there is no universally established timeline for the development of peripheral nerve abnormalities in T2DM rats induced by HFD-STZ, research indicates that rats can exhibit abnormal mechanical allodynia as early as 4 weeks after successful diabetes induction with HFD-STZ[39]. In this experiment, 9 weeks after the development of diabetes, the behavioral parameters of the model group rats showed significant changes, with the paw withdrawal threshold to mechanical and thermal stimulation decreasing. This was considered sufficient to meet the inclusion criteria for continued testing.

Glycolipid metabolism, neural function, and neural morphology were primarily observed to evaluate the metabolic regulation and neuroprotective effects of EA on HFD-STZ-induced T2DPN in rats. The neuroprotective mechanism of EA in T2DPN rats may be related to its activation of the SIRT1/PGC-1α axis in the peripheral nerves, promoting mitochondrial biogenesis and reducing oxidative stress and inflammatory responses. Therefore, during the EA intervention, rats were injected intraperitoneally with EX527, a specific SIRT1 inhibitor, and the same changes in the aforementioned indicators were observed.

The abnormal metabolism observed in diabetic rats includes hyperglycemia, INS resistance, and lipid metabolism disorders, which induce low-grade inflammation related to the susceptibility of peripheral nerve diseases[41,42]. Acupuncture treatment regulates glucose and lipid metabolism, and it has been reported that EA can stimulate the pancreatic intrinsic nervous system through the nerves innervating the acupuncture area, protecting pancreatic β cells and reducing INS resistance[43]. Consistent with previous studies, EA improved abnormal glucose and lipid metabolism in diabetes. Numerous studies have supported the role of SIRT1 in regulating glucose homeostasis in diabetes. Its mechanisms include regulating INS secretion and pancreatic β cell function, improving INS resistance by modulating INS signaling, inhibiting inflammation, and regulating hepatic glucose production[44]. However, the improvement of EA on blood glucose and lipid levels appears to be independent of SIRT1, and the regulatory effect of EA on these levels is not affected by SIRT1 inhibition. This suggests that the development of DPN may not be hindered solely by lowering blood sugar and blood lipids; therefore, the results of this experiment align with previous studies[45].

Tissue damage and inflammatory reactions caused by abnormal metabolism can harm the nociceptors in diabetic patients, including unmyelinated C fibers that mediate heat sensation and thinly myelinated Aδ fibers that mediate mechanical and cold sensations. This damage results in a decreased pain threshold and an increased response to harmful stimuli and spontaneous activities[46]. According to the results of the von Frey and Hargreaves tests, EA improved STZ-induced hyperalgesia. The interruption of peripheral and central nerve connections, due to the death of distal small fibers or sustained[47] abnormal input signals from damaged peripheral nerves, may lead to changes in peripheral and central neuroplasticity, potentially resulting in dysesthesia or even loss of sensation[48]. Emerging studies have indicated a relationship between SIRT1 and DPN; the expression of SIRT1 in neural tissue of the DPN experimental model has been shown to decrease, and activating SIRT1 can alleviate thermal hyperalgesia and mechanical allodynia in DPN[49]. In this experiment, rats with SIRT1 inhibition showed a transition from pain sensitivity to pain dullness, and the improvement effect of EA on abnormal behavioral parameters was significantly inhibited.

The impact of diabetes on peripheral nerve fibers was also reflected in IENFD. The effect of EA on the reduction of IENFD in T2DPN rats was demonstrated by staining and counting the IENFD at the terminal points of the rats. NCV primarily evaluates the function of large myelinated fibers[50]. The experiment showed that EA prevented the decline of NCV in T2DPN rats. The common myelin sheath injury during DPN damages the nodes of Ranvier, which are essential for the saltatory conduction of action potentials. Simultaneously, hyperglycemia activates the polyol pathway, consuming inositol, which is critical for nerve conduction. These pathological changes affect depolarization during stimulation, leading to a decline in NCV. EA has been proven to play a myelin-protective role in various nervous system diseases and regulates blood sugar, reducing the glucose substrate entering the polyol pathway and decreasing inositol consumption, which may explain why EA improves decreased NCV[51-53]. Additionally, we measured the blood flow of the lower limb skin and sciatic nerve using laser Doppler. The decline in blood flow is often related to peripheral small vessel abnormalities caused by diabetes, while EA increased the blood flow in the lower limb skin and sciatic nerve, thereby ensuring the nutritional support of peripheral nerves.

Histopathological examination revealed significant morphological changes in the peripheral nerves of rats with T2DPN, particularly in the myelin sheath, which serves as a physical barrier between axons and the external environment. Under normal physiological conditions, the myelin sheath envelops the axons to form a dense laminar structure. However, in T2DPN rats, the myelin sheath and axons become separated, and the myelin sheath may even be completely dissolved. An abnormal myelin sheath not only leads to direct contact between axons and harmful substances, such as inflammatory mediators, but also promotes the accumulation of neurotoxic substances and the transfer of lipotoxic substances to axons[54]. EA can improve the morphology of sciatic nerve axons and the myelin sheath, thereby ensuring normal physiological activity in nerves. The injection of a SIRT1 inhibitor weakened the protective effects of EA on peripheral small nerve fibers, myelinated nerve fibers, and small blood vessels in T2DPN rats. Specifically, after SIRT1 inhibition, severe damage to small nerve fibers resulted in a significant decrease in IENFD. Although SIRT1 inhibition did not significantly influence the regulatory effect of EA on glucose and lipid metabolism, it did lead to notable disruptions in neural morphology and function. The NCV results indicated that SIRT1 inhibition weakened the enhancement effect of EA on NCV. Research has shown that targeted SIRT1 therapy can restore endothelial cell function, improve microcirculation disorders, and enhance blood flow[55,56]. The experimental results demonstrated that SIRT1 inhibition affected the protective effect of EA on the function of small blood vessels in the lower limb skin and sciatic nerve blood flow. SIRT1 inhibitors not only eliminated the protective effects of EA on nerve morphology but also led to morphological deterioration.

NAD+-dependent SIRT1 is highly sensitive to changes in the energy demand of the body and is considered a sensor for fluctuations in nutritional and energy status, regulating multiple metabolic pathways in cells to maintain energy homeostasis[57,58]. SIRT1 plays a crucial role in enhancing mitochondrial function and alleviating inflammation and oxidative stress. It stimulates mitochondrial biogenesis and activates the antioxidant system by deacetylating downstream proteins[59,60]. Research has indicated that downregulation of SIRT1 is sufficient to induce behavioral abnormalities related to sensitization and pain in spinal cord neurons of healthy rats[44], while overexpression of SIRT1 in neurons can prevent or even reverse HFD induced DPN in mice[61]. Therefore, activation of SIRT1 is increasingly regarded as a therapeutic target for DPN.

SIRT1 activates the downstream protein PGC-1α, which acts as a transcriptional coactivator and a major regulatory factor[62,63] that stimulates a series of nuclear transcription factors such as Nrf1 and Nrf2[64]. This activation leads to an increase in the expression of TFAM[65], which has been shown to have a direct relationship with mitochondrial DNA levels in many mitochondrial and genetic models[17]. Current studies have confirmed that mitochondrial biogenesis is accompanied by an increase in oxidative phosphorylation capacity, a reduction in pathological oxidative stress, and the repair of mitochondrial dysfunction[65]. Mitochondrial dysfunction and the accompanying oxidative stress play a significant role in the pathophysiology of DPN[66-68]. Certain compounds have been shown in rodent models of diabetes to actively preserve mitochondrial function in nerves and regulate oxidative stress, providing neuroprotection[69,70].

EA is known to upregulate the expression of SIRT1 and activate related pathways in various animal models[31,71-73], thereby exerting regulatory effects. Our study evaluated the regulation of SIRT1 expression and related downstream proteins in the peripheral nerves of rats with T2DPN, aiming to elucidate the potential mechanism underlying the neuroprotective effect of EA. We found that the expression of SIRT1 and downstream PGC-1α decreased in the sciatic nerves of T2DPN rats, whereas EA treatment increased the expression levels of SIRT1 and PGC-1α. Additionally, EA increased the expression of NRF1 and NRF2 in T2DPN, which are involved in initiating mitochondrial biogenesis through the activation of TFAM. Nrf2, a principal regulatory factor for cellular antioxidant defense, plays a crucial role in enhancing resistance to oxidative stress damage[74]. Inhibition of SIRT1 led to the downregulation of its downstream proteins.

The proper function and integrity of the peripheral nervous system require metabolites, particularly glucose, to ensure continuous energy support. Within the neuron-glia metabolic coupling mechanism, glucose is converted into lactate, which is transported to axons. This lactate is then oxidized to pyruvate and enters the tricarboxylic acid cycle within mitochondria for further metabolism, providing a source of reduction for electron transfer in the ETC. Due to the incomplete coupling of the mitochondrial oxidative phosphorylation process, proton leakage occurs during electron transfer in the ETC, resulting in the conversion of up to 2% of oxygen into superoxide under normal physiological conditions. In a hyperglycemic environment, excessive metabolic substrates create a high proton gradient in the ETC, causing uncoupling of oxidative phosphorylation and producing superoxide free radicals that exceed the antioxidative capacity of the body, ultimately leading to oxidative stress damage in cells.

Rats with T2DPN often experience oxidative stress and inflammatory damage in their neural tissues. Oxidative stress extensively damages large molecules such as proteins, lipids, and DNA within cells, initially affecting mitochondria[75]. This phenomenon explains the abnormal metabolic phenotype characterized by increased blood sugar and insufficient energy in nerve cells. The activation of NRF2 by SIRT1 can produce various antioxidants and cell-protective enzymes by binding to and activating the antioxidant response element in the genome, thereby alleviating oxidative stress induced by metabolic disorders in the body.

In our experiment, we measured metabolic substrates and oxidative stress indicators in the sciatic nerves of the rats. As discussed earlier, superoxide radicals generated during electron transfer can attack various biological molecules. Endogenous antioxidant enzymes, such as SOD, can dismutate these superoxide radicals into hydrogen peroxide, which can then be reduced to water via CAT and GSH-PX[76]. Upregulation of endogenous antioxidants plays a positive role in neuroprotective processes. Our results demonstrated that EA reduced pyruvate and lactate levels in the sciatic nerve of T2DPN rats while decreasing MDA, a lipid peroxidation-related indicator, and increasing the activity of antioxidant stress kinases SOD, CAT, and GSH-PX (Figure 7).

Figure 7 Schematic illustration of the possible mechanisms underlying neuroprotective effects of electroacupuncture in experimental diabetes. In a metabolically disturbed organism, metabolic flux induces reactive oxygen species (ROS) production and damages mitochondria, eventually leading to an energy homeostasis imbalance in nerve cells. Electroacupuncture may attenuate oxidative stress, break the vicious cycle, and promote mitochondrial biogenesis by activating the silent information regulator type 2 homolog-1/peroxisome proliferator-activated receptor-gamma coactivator-1α axis. CAT: Catalase; GSH-PX: Glutathione peroxidase; Nrf1: Nuclear respiratory factor 1; Nrf2: Nuclear factor erythroid 2-related factor 2; PDC-1α: Peroxisome proliferator-activated receptor-gamma coactivator-1α; SIRT1: Silent information regulator type 2 homolog-1; SOD: Superoxide dismutase; TCA: Electron transport chain; TFAM: Mitochondrial transcription factor A.

Oxidative stress is closely related to inflammation. EA not only enhances antioxidant capacity but also reduces the inflammatory response. Inhibiting SIRT1 expression diminished the effect of EA on oxidative stress and inflammatory responses in the sciatic nerve. Therefore, we concluded that the protective effect of EA on peripheral nerve injury in diabetes may be mediated by the SIRT1/PGC-1α axis, promoting mitochondrial biogenesis and reducing oxidative stress.

However, our research had limitations. Analyzing extracts from the nucleus rather than the entire protein fractions of the sciatic nerve using WB would provide more convincing results. Therefore, a study with higher precision would be needed to reconfirm our findings. Furthermore, corresponding immunofluorescence staining could more intuitively and persuasively present the upregulation of biogenesis and distribution of mitochondria in axons, confirming their positional relationship in nerve fibers. These aspects will be addressed in our future research.

CONCLUSION

Our experiment indicated that the improvement effect of EA on DPN rats might be related to the regulation of the SIRT1/PGC-1α axis. This axis not only activated mitochondrial biogenesis but also reduced oxidative stress and protected intact mitochondria.

ACKNOWLEDGEMENTS

We appreciate the staff of the Key Laboratory of Acupuncture and Medicine Research of the Ministry of Education in the Nanjing University of Chinese Medicine for their support during the preparation of this manuscript.