Approximately 460 million individuals globally are affected by diabetes, with over half of these patients developing diabetic neuropathy. Among these, approximately one-third experience neuropathic pain [1,2]. Diabetic neuropathic pain (DNP) presents in various forms, including burning, cramping, and sinking sensations in the lower extremities, and may also be accompanied by depressive symptoms [3]. The etiology of pain associated with diabetic peripheral neuropathy is potentially linked to inflammation, oxidative stress, and mitochondrial dysfunction [3]. Currently, DNP is primarily managed using antidepressants, opioids, gamma-aminobutyric acid analogs, and topical agents such as capsaicin [4]. However, these treatment modalities often fall short of optimal efficacy despite the diversity of available strategies.

Exosomes, which are small vesicles approximately 100 nm in diameter containing lipids, proteins, and microRNAs, have emerged as a promising therapeutic avenue [5]. Notably, exosomes have demonstrated the ability to mitigate neuropathic pain by inhibiting microglial activation. Specifically, exosomes derived from M2 macrophages exhibit a high efficiency in reprogramming M1-type microglia into the M2 phenotype, thereby offering a potential novel approach for the treatment of DNP [6].

Organisms, ranging from bacteria to plants and animals, possess a 24-hour cycle known as the circadian rhythm [7]. A seminal study conducted in 1982 involving hamsters was the first to identify circadian rhythms in pain behavior, and since then, circadian oscillations have been observed in various pain-related clinical conditions [8]. For instance, neuropathic pain and temporomandibular joint pain reach a threshold of intolerance by 8 p.m., trigeminal neuralgia and fibromyalgia patients report heightened pain levels in the morning, individuals with biliary colic experience intensified pain at night, and post-surgical patients exhibit increased pain severity in the morning. Subsequent research revealed that pain intensity in diabetic patients fluctuates throughout the day, with heightened intensity observed during nighttime, which corresponds to the patient's resting period. Notably, this circadian rhythm of pain persisted even after opioid treatment [9,10]. Chronotherapy has demonstrated efficacy in clinical settings by administering treatments to patients at specific times according to their biological or disease-specific circadian rhythms, prioritizing optimal outcomes over patient or physician convenience. For instance, research has shown that glucocorticoids administered in the evening are more effective in alleviating pain and other symptoms of rheumatoid arthritis compared to those taken in the morning [11]. Additionally, there exists a circadian variation in the expression of opioid receptors, with studies in rodents indicating that opioids administered at night result in superior analgesic effects and quicker pain relief [12]. Despite the demonstrated efficacy of chronotherapy in treating various conditions, its application in patients with DNP remains understudied.

The most commonly used model of diabetic neuropathy is the use of streptozotocin (STZ)-induced diabetic rats [13]. In this study, an STZ-induced diabetic DNP rat model was employed to assess mechanical allodynia and thermal hyperalgesia. Intrathecal injections of M2 macrophage-derived exosomes were administered to the rats at varying time points corresponding to their active and resting phases. This approach aimed to investigate potential variations in therapeutic outcomes based on the timing of exosome administration and its subsequent impact on pain-related behaviors.

1. Induction of mouse bone marrow-derived M2 macrophages

Bone marrow-derived macrophages from C57BL/6 mice were prepared in much the same way as previously described in the literature [14]. In brief, bone marrow was extracted from the femur and tibia using a syringe following the dislocation of the mice's cervical vertebrae. The cells were then cultivated in Dulbecco's Modified Eagle Medium (DMEM) containing 10% fetal bovine serum, 1% penicillin and streptomycin, and 10 ng/mL M-CSF for a period of seven days. The supernatant was removed, and a complete medium containing 20 ng/mL interleukin 4 (IL-4) was added to induce macrophage differentiation towards the M2 phenotype. Two days later, the M2 macrophage surface marker CD206 was detected by flow cytometry.

2. Isolation and identification of exosomes

When M2 macrophages reached 80% fusion, the medium was replaced with exosome-removed fetal bovine serum and incubated for an additional 48 hours. The medium was collected and centrifuged for 10 minutes at 300 × g and 2,000 × g. After removal of cellular debris by centrifugation at 10,000 × g for 30 minutes, the supernatant was collected and the exosome pellet was collected by centrifugation at 15,000 × g for 90 minutes. The pellet was washed with phosphate-buffered saline (PBS) to remove contaminating proteins and centrifuged again at 150,000 × g for 90 minutes. Exosomes were either stored at –80°C or used immediately for downstream experiments. The vesicle diameter distribution of exosomes was analyzed using nanoparticle tracking analysis. Transmission electron microscopy (TEM) was used to observe the morphology of exosomes. Western blotting was used to detect specific exosome surface markers such as CD9, CD63, CD81, and TSG101.

3. Animals

Healthy and clean 6–8 weeks old male Sprague-Dawley rats, weighing 220–250 g, were purchased from Beijing Sprague-Ford Biotechnology Co. In mammals, the biological clock system controls circadian rhythms in the body to adapt to the alternation of day and night caused by the rotation of the Earth. There are pieces of external temporal information, called zeitgebers, that synchronize the biological clock cycle with the environmental cycle, and light is the most important zeitgeber. The time when the zeitgeber is involved in regulation is called zeitgeber time (ZT) [15]. In the light-dark cycle of this experiment, ZT0 is the beginning of darkness, and ZT12 is the beginning of light. The light and dark time was 12 hours each, ZT0-ZT12 for darkness (active period of rats), and ZT12-ZT24 for light time (the resting period of rats). Throughout the study, animals were used as little as possible to minimize their suffering. The animal ethics approval number is HLK-20240221-001. The experiments were done in the laboratory of Taiyuan Central Hospital.

4. Establishment of the rat DNP model and experimental grouping

After the rats were reared for 5 days to adapt to the environment, they fasted without food and water for 12 hours, and the type 1 diabetes model was induced by a single intraperitoneal injection of STZ (Sigma-Aldrich) with a mass concentration of 1% STZ (dissolved in sodium citrate buffer, pH 4.5) at 65 mg/kg, and the normal group was injected with an equal volume of sodium citrate buffer. After 72 hours, blood samples were collected from the tail vein to measure the fasting blood glucose level, and a blood glucose level higher than 16.7 mmol/L included in this experiment was considered as the establishment of a diabetic animal model, and those with blood glucose < 16.7 mmol/L could be supplemented with a small dose of STZ (10–20 mg/kg). The mechanical withdrawal threshold (MWT) was measured on the 14th day after STZ administration, and rats with MWT ≤ 80% of the basal value were selected as the DNP rat model.

They were divided into four groups of 10 animals each using the random number table method: normal control group (control group), DNP group, DNP rats in the ZT0 exosome-injected group (DNP + exo-ZT0 group) and DNP rats in the ZT12 exosome-injected group (DNP + exo-ZT12 group). A PE-10 catheter (American Health Medical Instruments International, Model: 188700) was inserted into the subarachnoid space through the L4-L5 intervertebral space one day before STZ injection. Two weeks following the STZ injection, an intrathecal administration of 20 µL of exosomes at a concentration of 1 mg/mL was conducted for three consecutive days in the DNP + exo-ZT0 and DNP + exo-ZT12 groups at ZT 0 and 12, respectively. In contrast, the DNP group received an intrathecal injection of 20 µL of saline over the same three-day period. Five rats in each group were given sodium pentobarbital (50 mg/kg) to induce unconsciousness and euthanized twenty-one and twenty-eight days following the STZ injection. The lumbar portion of the spinal cord was then removed for further studies.

5. Detection of blood glucose and weight indicators

Blood was collected from the tail vein of rats, which fasted for 8 hours before each collection. The body weight and blood glucose level of rats were measured once before STZ administration using an electronic weighing meter (Shenzhen Bonso Electronics Co, Ltd) and a glucometer (Sanno), respectively, and the blood glucose level and body weight of the animals were measured once on each of the 3rd, 7th, 14th, 21st, and 28th days after STZ administration.

6. Determination of MWT and thermal withdrawal latency (TWL) in rats

The 50% MWT and TWL were sequentially measured before STZ administration and on days 14, 17, 19, 21, and 28 after STZ administration. The specific time of day for measurements was randomized for the control group, the DNP + exo-ZT0 group, and the DNP + exo-ZT12 group. Observing whether the pain of DNP rats had a circadian rhythm, the MWT and TWL of rats in the DNP group were measured at ZT8 (darkness, the active period of rats) and ZT20 (light, the resting period of rats), respectively, according to the methods described previously in the literature [16]. Briefly, rats were placed in a glass box with a grid bottom, and after acclimatization to the environment, starting at 2.0 g, the skin of the hind paws of the rats was stimulated with Von Frey fiber filaments (0.40 g, 0.60 g, 1.0 g, 1.4 g, 2.0 g, 4.0 g, 6.0 g, 8.0 g, 15.0 g). If the rats showed foot shrinkage or foot licking recorded as a positive response, the stimulation was continued with a weaker strength of 1.4 g. If the rats did not show foot shrinkage or foot licking recorded as a negative response, the stimulation was continued with a stronger strength of 4.0 g (up-and-down approach). The interval between two stimulations was at least 5 seconds. When a different response from the previous one appeared (e.g., no paw shrinkage became paw shrinkage or paw shrinkage became no paw shrinkage), the rat was stimulated 4 times, plus 2 previous stimulations, for a total of 6 stimulations, and then the 50% paw-shrinkage threshold of the rat could be determined. If the stimulus strength required to produce a paw retraction response was greater than 15 g or less than 0.4 g, the rat's paw retraction threshold was recorded as 15 g or 0.4 g.

The temperature of the hot plate (Shanghai Xinsoft Information Technology Co.) used for measuring TWL was set at 52°C ± 1°C, and the rats were placed on the hot plate to record the time from placing the rats on the hot plate to the appearance of foot licking, jumping, hissing, and other thermal stimulation responses. Each rat was measured 3 times, and the average of the three measurement times was taken as the latency period of the TWL. The experiment was stopped if the above reaction did not occur for more than 15 seconds.

7. RNA extraction and quantitative reverse transcription polymerase chain reaction (qRT-PCR)

Total RNA was extracted from the rat lumbar segment spinal cord samples or microglia using the Ultra Pure RNA Extraction Kit (Kangwei Century Co., Ltd.) and reverse transcribed to cDNA using the reverse transcription kit. PCR reactions were performed using the MagicSYBR Mixture Kit (Kangwei Century Co., Ltd.) and the Bio-Rad CFX 96 Real-Time PCR System (Bio-Rad Laboratories). The internal reference gene was β-actin, and the target gene was analyzed for relative quantification by the 2-ΔΔCt method. Primer sequences are listed in Table 1.

Table 1 Primer sequence

Name	Sequence (5'-3')
iNOS	F: CTTCCGGGCAGCCTGTGAGACG R: ATCCCCAGGTGTTCCCCAGGTAGG
Arg-1	F: CGGCTTGCGAGATGTGG R: TAGCCGGGGTGAATACTGG
IL-1β	F: GGCAACTGTCCCTGAACTCAAC R: AAGCTCCACGGGCAAGACATA
IL-10	F: CACTGCTATGTTGCCTGCTCTT R: GTCTGGCTGACTGGGAAGTGG
CD86	F: GATTGCAGGTCCCAGTTCACTTC R: CCACTGTCCTGCTTGGACTCAC
CD206	F: TTCGGACACCCATC-GGAATTT R: CACAAGCGCTGCGTGGAT
β-action	F: CCCATCTATGAGGGTTACGC R: TTTAATGTCACGCACGATTTC

8. Western blot

Rat lumbar segment spinal cord samples or microglia were lysed in RIPA lysate (Shanghai Biyuntian Biotechnology Co., Ltd.), and the protein concentration was determined using a BCA kit (Beijing Lamblade Biotechnology Co., Ltd.). Approximately 30 ug of protein extracts were loaded and separated on sodium dodecyl sulfate-polyacrylamide gels and transferred to polyvinylidene difluoride membranes. The membrane was closed with 1% bovine serum albumin (BSA) for 2 hours. The membrane was incubated overnight at 4°C with primary antibody at a dilution of 1:1,000, then incubated with horseradish peroxidase-conjugated secondary antibody for 2 hours. Proteins were detected using an ECL reagent (Shanghai Biyuntian Biotechnology Co., Ltd.). Protein band density was semi-quantified using ImageJ. The primary antibodies were as follows: anti-β-actin, anti-iNOS, anti-ARG-1, anti-IKBα, anti-p-IKBα, and anti-p-NF-KB p65 (purchased from ImmunoWay Biotechnology Company).

9. Enzyme-linked immunosorbent assay (ELISA)

For rat lumbar spinal cord samples, tissue homogenates were prepared in a homogenization buffer. The homogenates were centrifuged at 15,000 × g for 30 minutes at 4°C, and the supernatant was collected and stored at –80°C for cytokine assays. For microglia, cell culture supernatants were collected, centrifuged at 250 × g for 10 minutes, and kept at –20°C for cytokine measurements. The levels of tumor necrosis factor-α (TNF-α) and IL-10 were measured according to the ELISA kit instructions (Wuhan Eliot Bioscience Co., Ltd.).

10. Immunofluorescent staining

Rat lumbar spinal cord sections were dewaxed, rehydrated, and then immersed in citrate buffer for antigen repair. Thereafter, the sections were closed with 1% BSA and then incubated overnight at 4°C with the following primary antibodies (purchased from ImmunoWay Biotechnology Company): anti-ionised calcium-binding articulating molecule-1 (Iba-1) (1:100), anti-iNOS (1:200), and anti-ARG-1 (1:200). Sections were then incubated with a mixture of fluorescein isothiocyanate or Cy3-conjugated secondary antibody (1:500, Jackson ImmunoResearch Laboratories Inc) for 1 hour. Nuclei were restained with 4,6-diamidino2-phenyl-indole dihydrochloride, and the sections were washed with PBS and then shaken-dried to be sealed with an anti-fluorescence-quenching sealer. The slices were observed under a fluorescence microscope and images were captured.

11. Microglia in vitro studies

To observe the effects of exosomes on microglia in vitro, the authors purchased BV2 microglia from the cell bank of the Chinese Academy of Medical Sciences and cultured them in a DMEM complete medium. Cells were inoculated at a density of 1 × 105 cells/mL into 6-well plates containing serum-free DMEM overnight before incubation with lipopolysaccharide (LPS) and exosomes. The next day, cells were stimulated with LPS (500 ng/mL) and LPS (500 ng/mL) + exosomes (200 ng/mL), respectively, for 24 hours. Cells and medium were then collected for analysis by Western blot, ELISA, and qRT-PCR.

12. Statistical analyses

All experiments were performed in at least three independent biological replicates. Data are expressed as mean ± standard error of the mean. Statistical analyses were performed using GraphPad software 10.0 and SPSS 27.0. The t-test was used to compare two groups, one-way analysis of variance was used to calculate the P value for comparisons between more than two groups, and Tukey's test was used for multiple comparisons between groups. Values of P < 0.05 were considered statistically significant.

1. Characterisation of M2 macrophage-derived exosomes

The detection of the surface marker CD206 in M2 macrophages via flow cytometry (Fig. 1A) suggests that the activation of the macrophage M2 phenotype can be induced by IL-4. Following the isolation of exosomes from the supernatants of activated M2 macrophages, the exosomes underwent characterization for morphology, size, and surface marker expression. TEM analysis revealed that M2 macrophage-derived exosomes exhibited a circular vesicle-shaped morphology with an average diameter of 139.5 nm (Fig. 1B, C). M2-Exo extracellular marker proteins CD9, CD63, CD81, and TSG101 were positive (Fig. 1D). These results support the successful preparation of M2 macrophage-derived exosomes with typical exosome morphology and molecular features.

Figure 1. (A) Detection of rat macrophage marker CD206 by flow cytometry. (B, C) Morphology and particle size analysis of exosomes observed by transmission electron microscopy. (D) Detection of marker proteins of exosomes.

2. Injection of exosomes during the resting period in rats relieves neuropathic pain in rats with DNP

Following STZ injection, rats developed hyperglycemia within one week and maintained elevated blood glucose levels throughout the experimental period, as shown in Fig. 2B. Compared to control rats, those treated with STZ exhibited a significant decrease in body weight, as illustrated in Fig. 2A. Furthermore, all STZ-treated rats displayed increased water intake and polyuria. Based on these characteristics, STZ induces persistent diabetes in rats. Treatment with exosomes did not affect the blood glucose concentration and body weight of the diabetic rats. Compared with the control rats, the diabetic rats with MWT ≤ 80% of the basal value after 2 weeks of STZ injection persisted in mechanically abnormal pain, suggesting that the DNP rat model was successfully induced. At 14, 17, 19, 21, and 28 days after STZ injection, rats in the DNP group had their MWT and TWL measured at ZT8 and ZT20, respectively, and it was found that the rats had higher MWT and TWL when tested at ZT8, suggesting that the neuropathic pain in the DNP rats had a higher MWT and TWL during the night time when the rats were active. Exosome treatment reduced MWT and TWL in the diabetic rats. Notably, compared with the DNP + exo-ZT0 group, the DNP + exo-ZT12 group more significantly alleviated mechanical allodynia and thermal hyperalgesia in the diabetic rats (e.g., Fig. 2C, D). These results suggest that there is a circadian rhythm of pain in DNP rats and that administration of the drug during the daytime when the rats are at rest can more effectively relieve neuropathic pain in rats.

Figure 2. Intrathecal injection of exosomes does not change body weight (A) or blood glucose levels (B), but reduces mechanical (C) and thermal pain in diabetic rats (D). DNP rats were given streptozotocin (STZ, intraperitoneal injection, 65 mg/kg), and blood glucose levels and body weights were measured before (baseline) and on days 3, 7, 14, 21, and 28 after STZ administration, respectively. Mechanical and thermal pain thresholds were measured before (baseline) and on days 14, 17, 19, 21, and 28 after STZ administration. Data are expressed as mean ± standard error of the mean (n = 10, day 28 n = 5). DNP: diabetic neuropathic pain, ZT: zeitgeber time, DNP + exo-ZT0: DNP rats in the ZT0 exosome-injected group, DNP + exo-ZT12: DNP rats in the ZT12 exosome-injected group, TWL: thermal withdrawal latency, MWT: mechanical withdrawal threshold. Comparison with control group, *P < 0.05, ***P < 0.001; with DNP (ZT8) group, #P < 0.05, ###P < 0.001; with DNP (ZT20) group, &&P < 0.01, &&&P < 0.001; with DNP+exo-ZT0 group, $P < 0.05, $$$P < 0.001.

3. Exosomal administration during the resting period in rats promotes polarization of rat spinal cord microglia from the M1 to the M2 phenotype

The concentrations of pro-inflammatory factor TNF-α and anti-inflammatory factor IL-10 in the spinal cord tissues of rats in each group were detected by ELISA on 7 and 14 days after exosome treatment. The results showed that the pro-inflammatory cytokine TNF-α was significantly higher and the anti-inflammatory cytokine IL-10 was significantly lower in the spinal cord tissues of rats in the DNP group compared with the control group. Both DNP + exo-ZT0 and DNP + exo-ZT12 groups significantly decreased the concentration of pro-inflammatory cytokines and increased the concentration of anti-inflammatory cytokines compared to the DNP group. However, DNP + exo-ZT12 treatment significantly promoted the secretion of anti-inflammatory cytokines and inhibited the secretion of pro-inflammatory cytokines compared to DNP + exo-ZT0 (Fig. 3A, B). Since microglia can have two different phenotypes, the authors suspected that exo-ZT12 administration promoted the polarisation of DNP rat microglia from the M1 phenotype to the M2 phenotype. qRT-PCR analyses were performed to analyse the expression of M1 (iNOS, IL-1β) and M2 (Arg1, IL-10) genes. As shown in Fig. 3C, D, M2 gene expression was significantly higher and M1 gene expression was lower in the DNP + exo-ZT0 and DNP + exo-ZT12 groups compared to the DNP group. Interestingly, M2 gene expression was higher and M1 gene expression was lower in the DNP + exo-ZT12 group compared to the DNP + exo-ZT0 group. qRT-PCR results were also confirmed by Western blot analysis (Fig. 3E, F). In addition, double immunofluorescence staining was performed with Iba1 using representative M1-associated iNOS and M2-associated Arg-1 markers to assess the characteristic polarisation of different groups of microglia in rat lumbar segment spinal cord tissues 14 days after exosome injection. As shown in Fig. 4, Iba-1 was considered a marker of microglia activation, and Iba1 expression was significantly increased in the DNP group compared with the control group. The DNP + exo-ZT0 and DNP + exo-ZT12 groups showed significantly fewer iNOS-positive microglia and increased levels of Arg-1-positive microglia in spinal cord tissues compared with the DNP group. Interestingly, there was a decrease in the number of iNOS-positive microglia and an increase in the number of Arg-1-positive microglia in the DNP + exo-ZT12 group compared to the DNP + exo-ZT0 group, which highlights the effect of exosome injection during the resting period in daytime rats on microglia polarisation in vivo. Thus, these results suggest that resting-period exosome injection has a significant effect on the ratio of anti-inflammatory to pro-inflammatory phenotypes of microglia in DNP rats and shifts microglial cell polarisation from an M1 to an M2 phenotype.

Figure 3. Exosome injection during the resting period of DNP rats can more effectively promote the polarisation of microglia from M1 to M2 phenotype. (A, B) Levels of pro-inflammatory factor TNF-α and anti-inflammatory factor IL-10 in the lumbar segment of the rat spinal cord were determined by ELISA after 7 days and 14 days of exosome injection (n = 3). (C, D) qRT-PCR to detect relative mRNA levels of M1-related iNOS, IL-1β, and M2-related Arg-1 and IL-10 in the lumbar segment of the rat spinal cord after 7 and 14 days of exosome injection (n = 3). Relative mRNA levels of M1-associated iNOS, IL-1β, and M2-associated Arg-1, IL-10 in the spinal cord (n = 3). (E, F) Western blot detection of iNOS, ARG-1 protein levels in the lumbar segment of rat spinal cords 7 days after exosome injection and 14 days after exosome injection (n = 4). Data are expressed as mean ± standard error of the mean. DNP: diabetic neuropathic pain, ZT: zeitgeber time, DNP + exo-ZT0: DNP rats in the ZT0 exosome-injected group, DNP + exo-ZT12: DNP rats in the ZT12 exosome-injected group, TNF-α: tumor necrosis factor-α, IL: interleukin, ELISA: enzyme-linked immunosorbent assay, qRT-PCR: quantitative reverse transcription polymerase chain reaction. Compared with control group, ns represents P > 0.05, **P < 0.01, ***P < 0.001; compared with DNP group, #P < 0.05, ##P < 0.01, ###P < 0.001; compared between the exosomal intervention groups, $P < 0.05, $$P < 0.01, $$$P < 0.001.

Figure 4. Significant activation of microglia in the dorsal horn of the spinal cord of diabetic rats, exosome administration enhances the expression of the M2 microglia marker ARG-1 and attenuates the expression of the M1 microglia marker iNOS, and nighttime administration is more effective than daytime administration. (A) Representative images of immunofluorescence assay for protein expression of Iba1 and ARG-1 in the lumbar spinal cord of the rat. (B) Representative images of Iba1, iNOS protein expression immunofluorescence assay (DAPI is blue fluorescence, Iba1 is red fluorescence, ARG-1 is green fluorescence), Bar = 100 um, quantitative analysis of immunofluorescence intensity (n = 3). Data are expressed as mean ± standard error of the mean. DNP: diabetic neuropathic pain, ZT: zeitgeber time, DNP + exo-ZT0: DNP rats in the ZT0 exosome-injected group, DNP + exo-ZT12: DNP rats in the ZT12 exosome-injected group, DAPI: 4,6-diamidino2-phenyl-indole dihydrochloride. *P < 0.05, **P < 0.01, ***P < 0.001 compared with control group; #P < 0.05, ###P < 0.001 compared with DNP group; $P < 0.05, $$P < 0.01 for comparison between exosome intervention groups.

4. Exosomes affect rat spinal cord microglia polarisation via NF-kB p65 and IKBα inflammatory signaling pathway

The NF-kB signaling pathway is a classical signaling pathway for inflammation and it plays an important role in the activation of M1 microglia. Inhibition of phosphorylation of NF-kB p65 and IKBα, which plays an important role in the transition of microglia to an anti-inflammatory M2 phenotype. Therefore, the authors explored the alteration of phosphorylation levels of NF-kB p65, and IKBα after exosome injection in DNP rats (Fig. 5). Western blot analysis showed that the levels of p-NF-kB p65 and p-IKBα were significantly down-regulated, and IKBα expression level was significantly up-regulated after 7 and 14 days of exosome treatment, while the levels of p-NF-kB p65 and p-IKBα were significantly up-regulated by exosome injection in ZT12 than ZT0. Injection of exosomes inhibited the phosphorylation of NF-kB p65 and IKBα more significantly than ZT0. These studies suggest that exosomes likely promote the transformation of microglia to the M2 phenotype by inhibiting the phosphorylation of NF-kB p65 and IKBα, and the better analgesic effect of ZT12-injected exosomes in rats may be related to their ability to inhibit the phosphorylation of NF-kB p65 and IKBα more significantly, thus affecting the polarisation of microglia.

Figure 5. Exosomes regulate microglia M1/M2 polarisation through NF-kB p65, IKBα signaling pathway. (A, B) Western blot strip plots of IKBα, p-IKBα, p-NF-kB p65 proteins in rat spinal cord on the seventh and 14th day after exosome injection, respectively, and quantification of grey values of protein expression levels (n = 4). Data are expressed as mean ± standard error of the mean. DNP: diabetic neuropathic pain, ZT: zeitgeber time, DNP + exo-ZT0: DNP rats in the ZT0 exosome-injected group, DNP + exo-ZT12: DNP rats in the ZT12 exosome-injected group. ***P < 0.001 compared with control group; ##P < 0.01, ###P < 0.001 compared with DNP group; $$P < 0.01, $$$P < 0.001 compared between exosome intervention groups.

5. Exosomes shift polarisation of BV2 microglia from M1 to M2 via NF-kB p65, IKBα inflammatory signalling pathway

To further explore the mechanisms by which exosomes regulate microglia phenotype, a series of experiments were performed in BV2 microglia in vitro. In these experiments, microglia activation was induced with LPS by co-culturing exosomes and LPS-activated microglia. Using ELISA, the authors found an increase in pro-inflammatory cytokines (IL-10) and a decrease in anti-inflammatory cytokines (TNF-α) in the LPS group; the opposite result was observed in the LPS + exosome group (Fig. 6A, B). qRT-PCR (Fig. 6C, D), and protein blotting analyses (Fig. 6E) confirmed the ability of exosome administration to convert the M1 phenotype to the M2 phenotype in BV2 microglial cells. Meanwhile, protein blotting analysis demonstrated that exosomes affected microglia polarisation via NF-kB p65, IKBα inflammatory signaling pathway in vitro experiments, which was consistent with the in vivo results.

Figure 6. Exosomes convert BV2 microglia from M1 phenotype to M2 phenotype through NF-kB p65, IKBα signaling pathway. (A, B) The concentration of pro- and anti-inflammatory cytokines in BV2 microglia in different groups. (C, D) mRNA expression levels of M1-related genes (CD86) and M2-related genes (CD206) detected by qRT-PCR. (E) The protein expression levels of M1- and M2-related genes and the expression of NF-kB p65 and IKBα signaling pathway in BV2 microglia in different groups were detected by Western blot. Data were expressed as mean ± standard error of the mean (n = 3). LPS: lipopolysaccharide, TNF-α: tumor necrosis factor-α, IL: interleukin, qRT-PCR: quantitative reverse transcription polymerase chain reaction, ns: no significance. Compared with control group, **P < 0.01, ***P < 0.001, ****P < 0.001; compared with LPS group, #P < 0.05, ###P < 0.001.

In this study, the authors present novel findings demonstrating that intrathecal administration of M2 macrophage-derived exosomes during the resting phase in STZ-induced DNP rats results in significant analgesic effects. Additionally, a circadian rhythm in pain perception was observed among diabetic rats. The data indicate an elevation in the pro-inflammatory cytokine TNF-α and a reduction in the anti-inflammatory cytokine IL-10 within the spinal cord. Treatment with exosomes led to a decrease in inflammatory cytokine levels and an increase in anti-inflammatory cytokine levels. Furthermore, the analgesic effects of exosomes are facilitated through the induction of microglial polarization towards the M2 phenotype and the inhibition of NF-kB p65 and IKBα phosphorylation.

Following the administration of STZ, rats exhibited hyperglycemia within 72 hours and developed sensory dysfunction, including mechanical nociception and thermal nociceptive sensitization, after two weeks, corroborating previous findings [17]. It has been demonstrated that exosomes derived from mesenchymal stem cells can suppress the activation of astrocytes and microglia [18]. Alterations in the morphology of microglia, such as morphological hypertrophy, increased thickness, and retraction of processes, can be observed, were observed in STZ-induced DNP rats, where microglia serve as immune cells specific to the central nervous system [19]. Microglia exhibit two phenotypes: the M1 phenotype, which disrupts synaptic communication between neurons and contributes to neuropathic pain, and the M2 phenotype, which secretes anti-inflammatory cytokines and exerts a palliative effect on neuropathic pain [20]. Activation of microglia with elevated M1/M2 ratios has been documented in various neuropathic pain models, such as the chronic constriction injury (CCI) model and the spinal cord injury model [21,22]. Studies have demonstrated that restoring the M1/M2 microglia ratio is advantageous in treating neuropathic pain [23]. In the current investigation, exosomes derived from M2 macrophages significantly facilitated the conversion of M1-type microglia to the M2 phenotype and elevated the mechanical allodynia and thermal hyperalgesia in diabetic rats. Additionally, numerous studies have examined astrocytes in DNP models. One study demonstrated that activation of spinal cord astrocytes facilitated pain relief in DNP mice. Conversely, another study did not observe astrocyte activation in STZ-induced diabetic rats. Additionally, other studies have reported a reduction in immune-responsive astrocytes in the spinal cord associated with diabetes. These findings warrant further investigation to elucidate the underlying mechanisms [19,24,25].

The regulation of circadian rhythms is governed by the supraoptic nucleus of the hypothalamus [26]. A survey of individuals with DNP revealed that over half experienced heightened pain during nighttime hours, with approximately 30% reporting consistent pain levels throughout the day and night [27]. This study represents the first investigation into the measurement of MWT and TWL in rats at various time intervals. The findings revealed a novel observation that rats with DNP exhibited higher MWT and TWL during their active phase compared to their resting phase. Additionally, an investigation into neuropathic pain in CCI rats demonstrated that administering analgesics during the rats' resting period resulted in decreased nociceptive sensitivity and prolonged drug efficacy, while no significant analgesic effect was observed when the drug was administered during the active period [28]. In investigations of molecular mechanisms within the rat hippocampal microglia, it was observed that certain genes displayed rhythmic expression patterns, with a notable decrease in expression coinciding with or occurring during periods of activity in rats [29]. Additionally, it was found that microglia in rats exhibited an increased secretion of inflammatory factors during active phases compared to resting phases [29]. This observation may provide insight into the enhanced efficacy of exosome administration during periods of rest in rats with diabetic neuropathy. Notably, pain itself can affect circadian rhythms, with fibromyalgia and cancer-induced pain disrupting the patient's circadian rhythms; weak light in a dark environment can also have an effect on circadian rhythms; and women with DNP have greater nociceptive sensitivity compared to men, with nociceptive sensations varying more during the day than at night [30–32].

Previous research has indicated the involvement of the NF-κB signaling pathway in neuroinflammation [33,34]. This finding was further substantiated by the authors’ research, in which in vivo and in vitro experiments were conducted demonstrating that exosomes derived from M2 macrophages inhibited the phosphorylation of NF-kB p65 and IKBα. This inhibition facilitated the polarization of microglia towards the M2 phenotype, potentially contributing to the enhanced efficacy of ZT12-injected exosomes. Additionally, ion channels, including Nav channels, potassium channels, Cav channels, and hyperpolarization-activated cyclic nucleotide-gated ion channels, are closely associated with painful diabetic neuropathy. Developing targeted drugs against these ion channels represents a promising direction for future research [35,36].

Recent research has demonstrated the potential of exosomes in alleviating pain resulting from nerve damage by modulating neuroinflammation and neuropathic pain through the delivery of proteins, mature miRNA, and translatable transcripts [37,38]. In the present study, mouse bone marrow-derived M2 macrophage-derived exosomes were used for the treatment of DNP, which to the authors’ knowledge is the first of its kind, and satisfactory results were achieved. Recent research has demonstrated the potential of exosomes as effective and safe nanocarriers for targeted gene therapy for pain through intranasal administration and tail vein injection, leveraging their small molecular weight and ability to traverse the blood-brain barrier [39,40]. Despite their promise for clinical use, exosomes encounter obstacles such as low yield, impurities, and limited loading efficiency, necessitating additional research efforts [41].

The management of DNP poses a significant challenge for healthcare providers and patients alike, underscoring the pressing need for further investigation and intervention. While some patients have exhibited a circadian rhythm in pain, the efficacy of associated treatment modalities remains suboptimal. This study demonstrates a notable reduction in pain in male rats with DNP through the administration of exosomes at varying time points. Future investigations should explore the potential application of exosome therapy in clinical settings, taking into consideration the gender-specific differences in neuropathic pain pathways between male and female rats. Additionally, the prevalence of female DNP patients in clinical populations warrants further examination by researchers.

The authors express their gratitude to the Taiyuan Central Hospital's laboratory professors at Peking University First Hospital for their assistance to the writers.

The datasets supporting the finding of this study are available from the corresponding author upon reasonable request.

No potential conflict of interest relevant to this article was reported.

This work was supported by grants from the Bureau of Science and Technology of Taiyuan City (202232 and 202268).

Wei Wei: Writing/manuscript preparation; Jun Fang: Writing/manuscript preparation; Baozhong Yang: Writing/manuscript preparation; Chenlong Cui: Resources; Jiacheng Wei: Investigation; Yating Xue: Investigation.

1. Saeedi P, Petersohn I, Salpea P, Malanda B, Karuranga S, Unwin N, et al; IDF Diabetes Atlas Committee. Global and regional diabetes prevalence estimates for 2019 and projections for 2030 and 2045: results from the International Diabetes Federation Diabetes Atlas, 9th edition. Diabetes Res Clin Pract 2019; 157: 107843.
   
2. Abbott CA, Malik RA, van Ross ER, Kulkarni J, Boulton AJ. Prevalence and characteristics of painful diabetic neuropathy in a large community-based diabetic population in the U.K. Diabetes Care 2011; 34: 2220-4.
   
3. Shillo P, Sloan G, Greig M, Hunt L, Selvarajah D, Elliott J, et al. Painful and painless diabetic neuropathies: what is the difference? Curr Diab Rep 2019; 19: 32.
   
4. Bhandari R, Sharma A, Kuhad A. Novel nanotechnological approaches for targeting dorsal root ganglion (DRG) in mitigating diabetic neuropathic pain (DNP). Front Endocrinol (Lausanne) 2022; 12: 790747.
   
5. Kalluri R, LeBleu VS. The biology, function, and biomedical applications of exosomes. Science 2020; 367: eaau6977.
   
6. Kim H, Wang SY, Kwak G, Yang Y, Kwon IC, Kim SH. Exosome-guided phenotypic switch of M1 to M2 macrophages for cutaneous wound healing. Adv Sci (Weinh) 2019; 6: 1900513.
   
7. Dibner C, Schibler U, Albrecht U. The mammalian circadian timing system: organization and coordination of central and peripheral clocks. Annu Rev Physiol 2010; 72: 517-49.
   
8. Oliverio A, Castellano C, Puglisi-Allegra S. Opiate analgesia: evidence for circadian rhythms in mice. Brain Res 1982; 249: 265-70.
   
9. Gilron I, Ghasemlou N. Chronobiology of chronic pain: focus on diurnal rhythmicity of neuropathic pain. Curr Opin Support Palliat Care 2014; 8: 429-36.
   
10. Gilron I, Bailey JM, Tu D, Holden RR, Weaver DF, Houlden RL. Morphine, gabapentin, or their combination for neuropathic pain. N Engl J Med 2005; 352: 1324-34.
    
11. Walton JC, Walker WH 2nd, Bumgarner JR, Meléndez-Fernández OH, Liu JA, Hughes HL, et al. Circadian variation in efficacy of medications. Clin Pharmacol Ther 2021; 109: 1457-88.
    
12. Palada V, Gilron I, Canlon B, Svensson CI, Kalso E. The circadian clock at the intercept of sleep and pain. Pain 2020; 161: 894-900.
    
13. Pathak R, Sachan N, Chandra P. Mechanistic approach towards diabetic neuropathy screening techniques and future challenges: a review. Biomed Pharmacother 2022; 150: 113025.
    
14. Ying W, Cheruku PS, Bazer FW, Safe SH, Zhou B. Investigation of macrophage polarization using bone marrow derived macrophages. J Vis Exp 2013; 76: 50323.
    
15. Tahara Y, Aoyama S, Shibata S. The mammalian circadian clock and its entrainment by stress and exercise. J Physiol Sci 2017; 67: 1-10.
    
16. Chaplan SR, Bach FW, Pogrel JW, Chung JM, Yaksh TL. Quantitative assessment of tactile allodynia in the rat paw. J Neurosci Methods 1994; 53: 55-63.
    
17. Jiang Y, Wang J, Li H, Xia L. IL-35 promotes microglial M2 polarization in a rat model of diabetic neuropathic pain. Arch Biochem Biophys 2020; 685: 108330.
    
18. Kang J, Guo Y. Human umbilical cord mesenchymal stem cells derived exosomes promote neurological function recovery in a rat spinal cord injury model. Neurochem Res 2022; 47: 1532-40.
    
19. Tsuda M, Ueno H, Kataoka A, Tozaki-Saitoh H, Inoue K. Activation of dorsal horn microglia contributes to diabetes-induced tactile allodynia via extracellular signal-regulated protein kinase signaling. Glia 2008; 56: 378-86.
    
20. Wu J, Ding DH, Li QQ, Wang XY, Sun YY, Li LJ. Lipoxin A4 regulates lipopolysaccharide-induced BV2 microglial activation and differentiation via the notch signaling pathway. Front Cell Neurosci 2019; 13: 19.
    
21. Shu B, He SQ, Guan Y. Spinal cord stimulation enhances microglial activation in the spinal cord of nerve-injured rats. Neurosci Bull 2020; 36: 1441-53.
    
22. Yuan J, Fei Y. Lidocaine ameliorates chronic constriction injury-induced neuropathic pain through regulating M1/M2 microglia polarization. Open Med (Wars) 2022; 17: 897-906.
    
23. Zhang Y, Chen Q, Nai Y, Cao C. Suppression of miR-155 attenuates neuropathic pain by inducing an M1 to M2 switch in microglia. Folia Neuropathol 2020; 58: 70-82.
    
24. Afsari ZH, Renno WM, Abd-El-Basset E. Alteration of glial fibrillary acidic proteins immunoreactivity in astrocytes of the spinal cord diabetic rats. Anat Rec (Hoboken) 2008; 291: 390-9.
    
25. Liao YH, Zhang GH, Jia D, Wang P, Qian NS, He F, et al. Spinal astrocytic activation contributes to mechanical allodynia in a mouse model of type 2 diabetes. Brain Res 2011; 1368: 324-35.
    
26. Mohawk JA, Green CB, Takahashi JS. Central and peripheral circadian clocks in mammals. Annu Rev Neurosci 2012; 35: 445-62.
    
27. Turk DC, Dworkin RH, Allen RR, Bellamy N, Brandenburg N, Carr DB, et al. Core outcome domains for chronic pain clinical trials: IMMPACT recommendations. Pain 2003; 106: 337-45.
    
28. Xia T, Cui Y, Qian Y, Chu S, Song J, Gu X, et al. Regulation of the NR2B-CREB-CRTC1 signaling pathway contributes to circadian pain in murine model of chronic constriction injury. Anesth Analg 2016; 122: 542-52.
    
29. Fonken LK, Frank MG, Kitt MM, Barrientos RM, Watkins LR, Maier SF. Microglia inflammatory responses are controlled by an intrinsic circadian clock. Brain Behav Immun 2015; 45: 171-9.
    
30. Korszun A. Sleep and circadian rhythm disorders in fibromyalgia. Curr Rheumatol Rep 2000; 2: 124-30.
    
31. Gilron I, Bailey JM, Vandenkerkhof EG. Chronobiological characteristics of neuropathic pain: clinical predictors of diurnal pain rhythmicity. Clin J Pain 2013; 29: 755-9.
    
32. Vetter C. Circadian disruption: what do we actually mean? Eur J Neurosci 2020; 51: 531-50.
    
33. Li H, Yang J, Wang Y, Liu Q, Cheng J, Wang F. Neuroprotective effects of increasing levels of HSP70 against neuroinflammation in Parkinson's disease model by inhibition of NF-κB and STAT3. Life Sci 2019; 234: 116747.
    
34. Shabab T, Khanabdali R, Moghadamtousi SZ, Kadir HA, Mohan G. Neuroinflammation pathways: a general review. Int J Neurosci 2017; 127: 624-33.
    
35. Zenker J, Ziegler D, Chrast R. Novel pathogenic pathways in diabetic neuropathy. Trends Neurosci 2013; 36: 439-49.
    
36. Wang Q, Ye Y, Yang L, Xiao L, Liu J, Zhang W, et al. Painful diabetic neuropathy: the role of ion channels. Biomed Pharmacother 2024; 173: 116417.
    
37. Shiue SJ, Rau RH, Shiue HS, Hung YW, Li ZX, Yang KD, et al. Mesenchymal stem cell exosomes as a cell-free therapy for nerve injury-induced pain in rats. Pain 2019; 160: 210-23.
    
38. Ramanathan S, Ajit SK. MicroRNA-based biomarkers in pain. Adv Pharmacol 2016; 75: 35-62.
    
39. Sun G, Li G, Li D, Huang W, Zhang R, Zhang H, et al. hucMSC derived exosomes promote functional recovery in spinal cord injury mice via attenuating inflammation. Mater Sci Eng C Mater Biol Appl 2018; 89: 194-204.
    
40. Guo S, Perets N, Betzer O, Ben-Shaul S, Sheinin A, Michaelevski I, et al. Intranasal delivery of mesenchymal stem cell derived exosomes loaded with phosphatase and tensin homolog siRNA repairs complete spinal cord injury. ACS Nano 2019; 13: 10015-28.
    
41. Tran PH, Xiang D, Nguyen TN, Tran TT, Chen Q, Yin W, et al. Aptamer-guided extracellular vesicle theranostics in oncology. Theranostics 2020; 10: 3849-66.