Nanocomposite hydrogel acupoint therapy for sustained pregabalin delivery and long-term neuropathic pain relief

Abstract

Neuropathic pain (NP) presents significant management challenges due to the limited efficacy and adverse effects associated with current therapeutic options. In this study, we present an innovative nanocomposite hydrogel designed for acupoint administration (PG@PLGA-gels) that synergistically combines nanomedicine with traditional acupuncture principles. This system is engineered by incorporating pregabalin (PG)-loaded poly (lactic-co-glycolic acid) nanoparticles into a chitosan/β-glycerophosphate thermosensitive hydrogel, facilitating minimally invasive administration at the Huantiao (GB30) acupoint. Compared to PG solution, PG@PLGA-gels exhibited a sustained drug release profile, extending over 12 days, with an initial release of 46.44% within the first 24 h, and increased systemic exposure by a factor of 2.6-fold. Following acupoint injection, plasma PG concentrations reached their peaked at 16 h (7985.94 ± 96.19 ng/mL) and remained detectable for up to 288 h, whereas PG solution peaked at 4 h (4928.33 ± 124.71 ng/mL) and declined to near-baseline levels by 192 h. In rat models of paclitaxel-induced neuropathy and chronic constriction injury, acupoint administration of PG@PLGA-gels significantly mitigates mechanical and cold allodynia from day 8 to day 20, while no obvious sedation-related behaviors were observed. Transcriptomic analysis identified treatment-associated changes in spinal cord gene expression profiles, with enrichment in pathways related to neuronal structure, synaptic organization, excitatory signaling, and neuroendocrine regulation. Together, these results support PG@PLGA-gels as an acupoint nanomedicine platform that combines controlled drug delivery with acupoint-based modulation to achieve prolonged analgesia with minimized adverse effects.

Graphical abstract

1. Introduction

Neuropathic pain (NP) is a chronic and debilitating condition resulting from by lesions or diseases affecting the somatosensory nervous system. Epidemiological studies suggest a prevalence between 6.9% and 10%, with allodynia and hyperalgesia occurring in 15%–50% of affected patients [1,2]. Current first-line treatments, which include antidepressants, antiepileptics, and pregabalin (PG), often provide only partial relief for many patients. Their efficacy is frequently compromised by systemic adverse effects such as dizziness, somnolence, ataxia, and poor long-term adherence [[3], [4], [5], [6]]. PG, one of the most frequently prescribed medications, exerts its analgesic effects through α2δ-1 calcium channel modulation but is associated with dose-dependent sedation and cognitive impairment, raising concerns about its tolerability and potential for misuse [[7], [8], [9]]. These challenges underscore the necessity for localized drug-delivery strategies that can enhance analgesic efficacy while reducing systemic exposure.

Acupoint injection represents a clinically used approach that delivers pharmacological agents into specific anatomical sites, thereby integrating local stimulation with pharmacological activity [[10], [11], [12], [13]]. The Huantiao (GB30) acupoint, situated in proximity to the trajectory of the sciatic nerve, is extensively utilized in the treatment of sciatic neuropathy and is commonly employed in experimental NP models [[14], [15], [16]]. Localized administration at the GB30 acupoint facilitates drug deposition in the vicinity of affected neural tissues, potentially enhancing therapeutic efficacy while minimizing systemic distribution. Nonetheless, traditional acupoint injections are characterized by rapid drug diffusion and absorption, leading to a brief residence time and limited therapeutic duration.

Recent advancements in nanotechnology and injectable biomaterials present novel opportunities to address these challenges [17,18]. Biodegradable poly (lactic-co-glycolic acid) (PLGA) nanoparticles are capable of encapsulating small-molecule drugs, thereby enhancing their stability and enabling sustained release with high biocompatibility [[19], [20], [21], [22], [23], [24]]. Thermosensitive chitosan (CS)/β-glycerophosphate hydrogels remain injectable at low temperatures and undergo an in situ sol-gel transition under physiological conditions, forming a depot that permits prolonged local retention [[25], [26], [27], [28], [29], [30]]. The incorporation of nanoparticles into hydrogels facilitates a dual-stage controlled delivery system, wherein the hydrogel extends local residence time and the nanoparticles adjust the subsequent drug-release profile.

Despite the clinical significance of acupoint injection and the technical benefits of nanocomposite hydrogels, there is a lack of research investigating nanoparticle-loaded hydrogels for site-specific delivery at the GB30 acupoint or assessing their analgesic efficacy in NP management. Therefore, the aim of this study was to develop and systematically evaluate a PG-loaded PLGA nanoparticle-embedded chitosan thermogel (PG@PLGA-gels) for minimally invasive administration at GB30. This study specifically aimed to characterize the physicochemical properties, drug release profile, and biocompatibility of the PG@PLGA-gels; to assess their analgesic efficacy in rat models of paclitaxel-induced neuropathy and chronic constriction injury; and to evaluate whether acupoint administration of PG@PLGA-gels could provide sustained pain relief while mitigating the central side effects associated with PG.

2. Methods

2.1. Animals

Sprague–Dawley (SD) male rats (180–220 g, Shanghai Bicai Keyi Biotechnology Co., Ltd), were housed under 12 h light/dark cycle with free access to food and water. Experiments were approved by the Ethics Committee of Shanghai University of Traditional Chinese Medicine (PZSHUTCM2411250003) and complied with National Institutes of Health (NIH) guidelines.

2.2. Synthesis and characterization of PG@PLGA-gels

The PG@PLGA nanoparticles were prepared by a double-emulsion solvent evaporation (W/O/W) method [31,32]. Briefly, 1 mg PG (Sinopharm Chemical Reagent Co., Ltd.) dissolved in 200 μL deionized water was emulsified with 2 mL of PLGA solution (50:50 lactide:glycolide ratio, 0.6 dL/g, 5 mg/mL in dichloromethane; Sigma-Aldrich) using an ultrasonic processor (Ton 5 s/Toff 6 s, 99 cycles, twice) in an ice bath. The primary emulsion was added to 6 mL of 1.5% polyvinyl alcohol (Mw 13,000–23,000, 87-89% hydrolyzed, Sigma-Aldrich) and sonicated for 5 min to form the secondary emulsion. After solvent evaporation, the suspension was centrifuged at 10,000 g at 4^oC for 5 min, and the supernatant was wash three times using 50 kDa ultrafiltration tubes to remove impurities. CS/β-GP hydrogels were prepared by slowly adding 0.5 mL β-GP (56%, Sigma-Aldrich) in NaOH into 2 mL CS (3% w/v, deacetylation ≥95%, viscosity 100-200 mPa s, Aladdin) in glacial acetic acid (β-GP:CS = 1:3), with stirring for 30 min in an ice bath. PG@PLGA-gels were obtained by dispersing freeze-dried PG@PLGA nanoparticles into CS solution, followed by hydrogel formation under the sam conditions.

The hydrodynamic size of PG@PLGA nanoparticles was measured by dynamic light scattering (Malvern Zetasizer ZEN3600, Malvern Panalytical Ltd, UK). Morphology was observed using scanning microscopy (JSM-7600F, JEOL) and transmission electron microscopy (TECNAI G2 F20, FEI). Gelation time was determined by the vial-inverting method at 37 °C. Degradation was evaluated by incubating blank gels and PG@PLGA-gels in PBS (pH 7.4) 500 μg/mL lysozyme, at 37 °C, 100 rpm, for 3, 6, 9, 12, and 15 days. Samples were weighed before (M₀) and after incubation (M_t), and degradation rate was calculated as M_r (%) = M_t/M₀ × 100%. Swelling was assessed by immersing dried gels (W₀) in PBS at 37 °C for 30, 60, 90, 120, 150, and 180 min. After blotting, swollen gels were weighed (W_t), and swelling rate was calculated as (W_t - W₀)/W₀ × 100%. PG content was quantified by HPLC (Agilent 1200, California, USA) with a Hypersil BDS C8 column. The mobile phase was 95:5 (v/v) potassium phosphate buffer (pH 6.9) acetonitrile, flow rate 1 mL/min, detection at 205 nm, column temperature 26 °C, and injection volume 10 μL [33]. Encapsulation efficiency (EE) and drug loading (DL) were calculated as: DL (%) = (PG encapsulated/PG@PLGA nanoparticles) × 100%; EE (%) = (PG encapsulated/PG added) × 100%.

2.3. In vitro cytotoxicity assay

The C2C12 cell line, hippocampal neuronal cell line (HT22), and human umbilical vein endothelial cell line (HUVEC) were cultured in DMEM with 10% fetal bovine serum (Gibco) and 1% penicillin/streptomycin (Gibco) at 37 °C an atmosphere containing 5% CO₂. Cytotoxicity of PG@PLGA-gels was assessed using CCK-8 assay (Beyotime, Shanghai, China). Cells (5 × 10³/well) were seeded in 96-well plates, incubated overnight, and subsequently exposed to hydrogel extracts for 24, 48, 72 h. After adding CCK-8 solution, absorbance was measured at 450 nm using a microplate reader. Cell viability was further evaluated by live/dead staining. Cells (1 × 10⁵/well, 24-well plates) were incubated with 72 h hydrogel extracts, stained with Calcein AM (2 μM) and PI (8 μM) for 30 min, and imaged under a fluorescence microscope (Leica, Germany).

2.4. In vivo histological and immunofluorescence analyses

At 1, 4, 7, and 14 days after injection of either blank gels or PG@PLGA-gels at the GB30 acupoint, the rats (n = 3 at each time point) were euthanized, and samples of the sciatic nerves, adjacent muscles, and major internal organs (heart, liver, spleen, lung and kidney) were harvested. The collected tissues were fixed in 10% formalin, embedded in paraffin, sectioned, and stained with hematoxylin-eosin (H&E) for histological examination under optical microscopy. For immunofluorescence, sections of the sciatic nerves and muscles were dewaxed, rehydrated, and blocked with goat serum. They were then inculbated overnight at 4 °C with primary antibodies targeting interleukin-1β (GB1111, Servicebio, China), interleukin-6 (GB11117, Servicebio, China), and tumor necrosis factor-α (GB11188, Servicebio, China). After PBST washes, the sections were incubated with secondary antibody (GB25303, Servicebio, China), counterstained with DAPI (G1012, Servicebio, China), and visualized using a fluorescence microscope (Leica, Germany). The fluorescence intensity was semi-quantitatively assessed using ImageJ software.

2.5. In vitro drug release assessment

To evaluate the release of PG, PG@PLGA-gels (n = 3 per group) were prepared in tubes at 37 °C and subsequently transferred into 10 mL of PBS (pH 7.4). The samples were incubated at 37 °C with agitation at 100 rpm. At predetermined intervals ranging from 1 to 360 h, supernatants were collected and replaced with fresh PBS. PG concentration was quantified by HPLC as described in Kasawar [33].

2.6. In vivo drug release anaysis

For fluorescence imaging, 10 μL of DiD dye (excitation at 644 nm/emission at 665 nm, Invitrogen) was incorporated into PLGA nanoparticles prior to gel preparation. To minimize potential adverse effects on animal welfare, repeated longitudinal imaging of the same animal was avoided due to the dense imaging schedule and the requirement for deep anesthesia and local hair removal. Therefore, in vivo fluorescence imaging was conducted using two independent and parallel experimental batches of rats. Each batch was scanned at three predefined time points, with different animals used at each time point to ensure that no animal was imaged more than once. The animals were divided into two groups: DiD group and DiD/PLGA-gels group (n = 3 per group at each time point). All animals were randomly assigned to groups prior to imaging, and imaging was performed under consistent settings. Rats received intramuscular injections at the GB30 acupoint with 1 mL of either the DiD solution or DiD-labeled PLGA-gels (DiD/PLGA-gels) under isoflurane anesthesia. Fluorescence images were acquired at 3, 24, 48, 96, 192, and 288 h post-injection using an IVIS Lumina III imaging system (PerkinElmer, Caliper Life Sciences, MA, USA) and were analyzed using the accompanying software.

2.7. Pharmacokinetic study

Male rats (n = 3 per group) were randomly assigned to receive acupoint injections of either PG solution, PG nanoparticles (NPs), or PG@PLGA-gels. Blood samples were collected from the orbital venous plexus at 0.5, 1, 2, 4, 8, 16, 24, 96, 192, and 288 h post-administration. The samples were subsequently centrifuged at 14,000 rpm to isolate plasma. Plasma PG concentrations were measured using LC-MS/MS (AB Sciex Q-trap 4500), following a modified protocol by Zhang [34]. Pharmacokinetic parameters were calculated using DAS2.0 software (Mathematical Pharmacology Professional Committee of China, Shanghai, China).

2.8. GB30 injection

Following the confirmation of adequate anesthesia, the rats were positioned prone on a flat surface. The GB30 acupoint was anatomically situated at the junction of the lateral one-third and medial two-thirds of the line connecting the greater trochanter and the sacral hiatus [15,35]. This anatomical location is depicted in Fig. S1. A 1 mL syringe was inserted perpendicularly to the skin at the GB30 acupoint to a depth of 11–12 mm, taking into consideration that the sciatic nerve trunk lies deep to this point [36]. Occasionally, unilateral hindlimb twitching was observed during needle insertion, indicative of proximity to the sciatic nerve. In such instances, the needle was retracted by 1–2 mm to prevent direct nerve stimulation prior to injection. A non-acupoint was selected at the root of the tail, characterized by an absence of significant vascular and neural structures, thus distinguishing it from the GB30 location (Fig. S1). To minimize the potential for injectate leakage, the needle was maintained in position for 30 s post-injection before being withdrawn slowly.

2.9. Neuropathic pain models and behavioral assessments

Two rat models were used: paclitaxel-induced NP (PINP) and chronic constriction injury (CCI). In the PINP model, rats were administered intraperitoneal paclitaxel (1 mg/kg, MedChemExpress) every other day for a total of four doses [37]. In the CCI model, under anesthesia, the right sciatic nerve was exposed and loosely ligated with four 4-0 chromic gut at 1 mm intervals [38]. Successful modeling was characterized by a reduction exceeding 50% in the mechanical withdrawal threshold (MWT) after one week. Two NP model animals cohorts were randomly allocated into six groups (n = 5 per group): the control group, the saline group (receiving acupoint injections of saline at GB30 post-modeling), the PG solution group (receiving acupoint injections of PG solution at GB30 post-modeling), the blank gels group (receiving acupoint injections of blank gels at GB30 post-modeling), the PG@PLGA-gels group (receiving injections of PG@PLGA-gels at the tail root post-modeling, serving as a non-acupoint control), and the PG@PLGA-gels/GB30 group (receiving acupoint injections of PG@PLGA-gels at GB30 post-modeling). In the PG@PLGA-gels/GB30 group, the total gel volume was 1 mL, with an effective drug loading of PG at 26 mg. In the other groups, PG was administered at a dosage of 10 mg/kg/day, amounting to total 26 mg per rat, with an injection volume of 1 mL [39]. A 1 mm syringe needle was inserted to a depth of 12-14 mm to reach the sciatic nerve trunk [40].

The assessment of mechanical allodynia was conducted using the von Frey test with the electronic von Frey Anesthesiometer (IITC Life Science Instruments, USA) [41]. Each rat was individually placed in an enclosure with a plastic chamber on a wire mesh floor and allowed to acclimate for at least 30 min. Subsequently, a polypropylene tip was applied to the plantar surface of the hind paw with a progressively increasing pressure until a functional response was elicited (e.g., lifting, licking, flicking, shaking or jumping). The maximum threshold pressure was recorded, and the measurements were repeated 3 times for each rat with a 3-min interval between each measurement. Cold allodynia was evaluated using the acetone drop test (ADT) [42]. Specifically, 50 μL of acetone (4^oC) was dropped through a micropipette to the right medial plantar region. The response within the first 20 s was quantified using the acetone test score (ATS): 0, no response; 1, one rapid hindpaw flick/stamp; 2, two or more hindpaw flicks/stamps; 3, periods of flicking/stamping with licking of plantar hindpaw [43]. Each measurement was repeated 3 times per rat, with a 5-min interval between repetitions. The sum of the three scores was used for data analysis.

2.10. Analysis of spinal RNA sequencing data

After behavioral testing, lumbar spinal cord segments (L2-L6) were harvested from the PG@PLGA-gels-PINP group, PG@PLGA-gels-CCI group, PINP group, CCI group, and blank control group (n = 3 per group). RNA extraction and sequencing were performed by Shanghai Biochip Co., Ltd. Differentially expressed genes (DEGs) were identified using the R package DeSeq2, with a fold change threshold of 1.5 and a P-value cutoff of 0.05. Heatmaps and hierarchical clustering analyses were performed using R package ComplexHeatmap. Functional enrichment analysis of DEGs, including Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway annotations, and Gene Set Enrichment Analysis (GSEA) was conducted using the Rpackage clusterProfiler, with annotation facilitated by the R package org.Rn.eg.db. All visualizations were created using the R package ggplot2.

2.11. Biosafety analysis

For barbiturate-induced sleep assessment, naïve rats were randomly assigned to 4 groups (n = 5 per group): saline group, PG solution group, blank gels group, PG@PLGA-gels group. One hour after GB30 injection, sodium thiopental (40 mg/kg) was administered intraperitoneally. The duration of sleep was determined by measuring the interval between the loss and recovery of the righting reflex [44]. After 21 days, rats were euthanized, and major organs were harvested for H&E staining and histological examination. Muscle strength of rats (n = 5 per group) posti-GB30 injection was assessed by extensor postural thrust test [45]. Briefly, rats were positioned upright with their hind limbs extended, and the thrust force exerted on a weighing scale was recorded. These assessments were performed on days 1, 4, 8 and 12.

2.12. Statistical analysis

All data are presented as the mean ± SD and analyzed by GraphPad Prism 8.0 software. Statistical significance was evaluated in accordance with one-way analysis of variance (ANOVA) (more than two groups). Statistical significance was set at p < 0.05.

3. Results

3.1. Fabrication and characterization of PG@PLGA-gels

The preparation process of PG@PLGA is illustrated in Fig. 1A. PG@PLGA nanoparticles were synthesized using the double-emulsion solvent evaporation (W/O/W) technique. Following freeze-drying, the nanoparticles were dispersed in the CS/β-GP mixed solution and stirred in an ice bath to produce a PG@PLGA-gel liquid. Upon observing the phase transition at 37 °C, it was determined that the PG@PLGA-gel liquid could completely solidify within 210 s. Dynamic light scattering (DLS) analysis indicated that PG@PLGA nanoparticles possess a hydrodynamic diameter of 106.2 ± 2.3 nm, compared to the uncoated PLGA cores, which measure 91.2 ± 3.9 nm (Fig. 1B). These results were corroborated by transmission electron microscopy images (Fig. 1C). The optimal drug loading (DL) and encapsulation efficiency (EE) of PG within the PG@PLGA nanoparticles were found to be 65.8% and 67.2%, respectively (Table S1). At room temperature (25 °C), the PG@PLGA-gels appeared as a cloudy liquid, indicating superior injectability and fluidity (Fig. 1D(b)). Upon exposure to physiological temperature (37 °C), these hydrogels underwent a rapid phase transition to a gel state within 3.5 min, thereby establishing a stable reservoir at the site of injection (Fig. 1D(d and c)). Scanning electron microscopy (SEM) analysis confirmed that spherical nanoparticles were incorporated within the hydrogels, exhibiting a network structure that is denser than that of blank gels (Fig. 1D(g and h)). Higher magnification images demonstrated that the blank gel contained a significant number of dispersed voids, whereas the PG@PLGA-gels showed relatively fewer voids (Fig. 1D(i and j)). Notably, PG@PLGA nanoparticles were clearly visible on the surface of the gel (Fig. 1D(i and j)). Subsequently, PBS containing lysozyme was employed to replicate the physiological conditions for in vitro hydrogel degradation. The degradation rate of PG@PLGA-gels reached 33.24 ± 3.17% on day 6 and 85.92 ± 2.7% on day 15, with no statistically significant difference between the PG@PLGA-gels and the blank gels (Fig. 1E). To evaluate the in-gel compression within the acupoint tissue, a swelling test was performed. The swelling ratios of both groups increased progressively during the initial 60 min and subsequently stabilized (Fig. 1F). The maximum swelling rate was recorded at 330 ± 3% for the control gels and 235 ± 14% for the PG@PLGA hydrogels (Fig. 1F), with the latter being reduced due to the space occupied by the PG@PLGA nanoparticles.

Fig. 1.

Fabrication and physicochemical characterizations of PG@PLGA-gels. A. Schematic illustration of fabrication of PG@PLGA-gels. B. Hydrodynamic diameter of PLGA cores and PG@PLGA nanoparticles. C. Representative TEM images of PG@PLGA. Scale bar, 200 nm. D. (a, b): Macroscopic images of PG@PLGA-gels at room temperature; (c, d) photographs showing gel formation at 37 °C; (e, f) images demonstrating injectability through a syringe needle; (g, j) representative SEM images highlighting nanoparticle incorporation within the hydrogel network. Scale bar: 20 μm, Scale bar in (g, h) and 1 μm in (i, j). Red arrows indicate PG@PLGA nanoparticles. E. Degradation ratio of blank gels and PG@PLGA-gels in vitro (n = 3 per group). F. Swelling ratio of blank gels and PG@PLGA-gels in vitro (n = 3 per group). Data are presented as mean ± SD. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

3.2. Evaluation of drug release and pharmacokinetics of PG@PLGA-gels

The in vitro drug release experiment of PG@PLGA-gels was performed under conditions replicating the ionic environment. The PG standard curve is shown in Fig. S2. The release profiles demonstrated an initial burst release of 46.44 ± 1.4% within the first 24 h, followed by a sustained release phase (Fig. 2A). For in vivo drug release visualization, DiD-labeled PLGA hydrogels were injected at GB30. The fluorescence of DiD was consistently maintained and gradually attenuated over a period exceeding 12 days, whereas the fluorescence intensity in the DiD solution group decreased rapidly (Fig. 2B–C). The effect of PG@PLGA-gel injection on PG drug metabolism was evaluated through plasma pharmacokinetics testing. In the PG solution group, the plasma drug concentration reached its peak at 4 h (4928.33 ± 124.71 ng/mL) and became nearly undetectable by 192 h (Fig. 2D). Conversely, in the PG@PLGA-gels group, the plasma drug concentration peaked at 16 h (7985.94 ± 96.19 ng/mL) and persisted up to 288 h (Fig. 2D). Analysis indicated that the AUC₀→t for the PG@PLGA-gels group was 2.6 times greater than that of the PG solution group (Fig. S3).

Fig. 2.

Drug release from PG@PLGA-gels both in vivo and in vitro. A. The in vitro release curve of PG@PLGA-gel. B. In vivo fluorescence imaging of rats. C. The in vivo fluorescence intensity of DiD/PLGA gels group and DiD solution group, DiD/PLGA gels group was consistently stronger than that of the DiD solution, indicating the ability of the nanocomposite hydrogel to prolong drug release and prolong local drug retention. D. The changes of plasma drug concentration in PG solution group and PG@PLGA-gels group within 288 h after injection. Data are presented as mean ± SD (n = 3 at each time point/group).

3.3. Biocompatibility assessment of PG@PLGA-gels

In vitro assays demonstrated that PG@PLGA-gels exhibited negligible cytotoxicity towards HT22, C2C12, and HUVECs, with cell viability exceeding 95% after 72 h of incubation, corroborated by the live/dead assay results (Fig. 3A–B). Further in vivo evaluation of the local toxicity of PG@PLGA-gels on the sciatic nerve and adjacent muscular tissues revealed a progressive decrease in residual hydrogels, which nearly fully degraded by day 14 (Fig. 3C). H&E staining showed that PG@PLGA-gels did not induce significant pathological or morphological changes and were associated with minimal inflammatory cell infiltration (Fig. 3D). Biomaterials can elicit inflammatory responses upon in vivo implantation [46]. To evaluate the temporal variations in local inflammatory markers, immunofluorescence staining for IL-1β, IL-6, and TNF-α was conducted on neuromuscular tissues surrounding the GB30 acupoint at 1, 4, 7, and 14 days post-injection. Fig. 3E shows the composite images of the three cytokines at the specified time points, while Supplementary Figs. S4–S6 provide the respective individual fluorescence channels and a quantitative analysis of intensity. The PG@PLGA-gels elicited an elevated expression of IL-1β, IL-6, and TNF-α during the acute phase (days 1-4) in comparison to the saline group. Subsequently, cytokine levels gradually diminished and aligned with those of the saline group by day 14.

Fig. 3.

In vitro cytotoxicity and in vivo biocompatibility of PG@PLGA-gels. A. Cell viability of C2C12, HT22 and HUVEC cells co-cultured with PG@PLGA-gels solution extracted after 24 h, 48 h and 72 h. Data are presented as mean ± SD (n = 3 per group). B. Fluorescence images of C2C12, HT22 and HUVEC cells after co-culture with PG@PLGA-gels solution extracted after 72 h by staining with calcein-AM/PI. The living cells are labeled with green and the dead cells are labeled with red. Scale bar, 200 μm. C. Macroscopic examination of blank gels or PG@PLGA-gels at days 1, 4, 7 and 14 post-injection. Orange arrows point to hydrogels. Green arrows point to sciatic nerves. Yellow arrows point to the head of femur. D. Representative histological images of tissues with H&E staining at days 1, 4, 7 and 14 post-injection (M: muscle; N: sciatic nerve; S: stromal cells). Scale bar, 100 μm. E. Time-resolved immunofluorescence analysis (IL-1β, IL-6 and TNF-α) of sciatic nerve and surrounding muscles at days 1, 4, 7 and 14 post-injections. Green fluorescence represents the inflammatory cytokines and blue fluorescence represents the nuclei. Scale bar = 60 μm. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

3.4. Analgesic effect of PG@PLGA-gels at GB30 acupoint in NP model

The PINP model (n = 5 per group) was employed to simulate chemotherapy-induced NP in humans and assess the analgesic efficacy of PG@PLGA-gels (Fig. 4A). Within this model, the cumulative administration of paclitaxel significantly reduced MWT and increased ATS compared to the control group (all p < 0.001, Fig. 4B–D). Injection of PG solution at the GB30 produced transient analgesia at day 8 compared to control group (MWT and ATS, p < 0.001), whereas PG@PLGA-gels administrated at the same acupoint induced progressive and sustained improvement in both MWT and ATS from days 8 to 20 (Fig. 4B–D). At day 20, this effect surpassed that of the PG solution, blank gels (MWT and ATS, p < 0.001, Fig. 4B–D), but had no significant difference compared to non-acuppoint PG@PLGA-gels in ATS (p = 0.195). Notably, blank gels exhibited partial time-limited analgesic effects before day 16, but diminished with degradation, indicating potential additional neuromodulatory effects due to local gel deposition (p < 0.001, Fig. 4B–D). Normalized area under the curve (AUC) analyses visually presents the total analgesic effect of each group and further substantiated the significant analgesic effect of PG@PLGA-gels when injected at GB30 (Fig. 4C–E). The CCI model was used to replicate symptoms of chronic nerve compression in humans. The experimental workflow is illustrated in Fig. 5A (n = 5 per group). In CCI model, rats exhibited persistent mechanical and cold allodynia compared to control group (all p < 0.001, Fig. 5B–D), which were significantly alleviated by the administration of PG@PLGA-gels at GB30 (all p < 0.001, Fig. 5B–D). The total analgesic effiectiveness of PG@PLGA-gels injection at GB30 was demonstrated to achieve optimal outcomes through AUC analyses (Fig. 5C–E). All original AUC values are displayed in Fig. S7.

Fig. 4.

Analgesic effect of acupoint injection of compound gel on PINP model. A. Illustration of experiments timeline in PINP model. The rat were subjected to six days of modeling. On day 7 behavioral tests were performed to confirm the success of modeling. Injection treatment was performed on the day 8, in each group, treatment was applied on the left leg, and VFT were performed on day 8 (1 h after the injection treatments), day 12, day 16 and day 20, and ATS were performed on day 8, day 14 and day 20. B. The ability of the rats to withstand mechanical stimulation (mechanical withdraw threshold, g), including blank control, saline control, PG solution, blank gel, PG@PLGA-gel + GB30 injection and PG@PLGA-gel + non-acupoint injection. C. Comparison of the AUC (baseline-normalized) of VFT values. *: P < 0.05. **: P < 0.01, ***: P < 0.001. D. The ability of the rats to withstand thermal stimulation (cold score). ANOVA followed by a post hoc test. *: P < 0.05. **: P < 0.01, ***: P < 0.001. E. Comparison of the AUC (baseline-normalized) of cold score values. *: P < 0.05. **: P < 0.01, ***: P < 0.001. Data are presented as mean ± SD (n = 5 per group).

Fig. 5.

Analgesic effect of acupoint injection of compound gel on CCI model. A. Illustration of experiments timeline in CCI model. The rats were operated on day 0. On day 7 behavioral tests were performed to confirm the success of modeling. Injection treatment was performed on the day 8, in each group, treatment was applied on the left leg, and VFT were performed on day 8 (1 h after the injection treatments), day 12, day 16 and day 20, and ATS were performed on day 8, day 14 and day 20. B. The ability of the rats to withstand mechanical stimulation (mechanical withdraw threshold, g), including blank control, saline control, PG solution, blank gel, PG@PLGA-gel + GB30 injection and PG@PLGA-gel + non-acupoint injection. C. Comparison of the AUC (baseline-normalized) over a prespecified window of VFT values. *: P < 0.05. **: P < 0.01, ***: P < 0.001. D. The ability of the rats to withstand thermal stimulation (cold score). E. Comparison of the AUC (baseline-normalized) over a prespecified window of cold score values. *: P < 0.05. **: P < 0.01, ***: P < 0.001. Data are presented as mean ± SD (n = 5 per group).

3.5. Transcriptomic profiling of PG@PLGA-gels at GB30 in NP models

To explore treatment-associated spinal molecular programs in an unbiased manner, transcriptomic profiling was performed in lumbar spinal cord tissues (L2–L6) from PINP and CCI rats two weeks after PG@PLGA-gels administration to identify treatment-associated transcriptional changes and enriched biological pathways. The PCA results of the two models are shown in Fig. 6A and F. In the PINP model, 1485 differential expression gene (DEGs) were identified, with 719 genes upregulated and 766 genes downregulated in the PG@PLGA-gel-PINP group compared to the untreated PINP group (Fig. 6B). Hierachical clustering of the top 30 upregulated and downregulated genes highlighted representative upregulated genes involved in neurol survival and vesicular transport, and downregulated genes associated with axonal remodeling (Fig. 6C). GO and KEGG enrichment analysis revealed that upregulated DEGs were mainly enriched in neurogenesis, synaptic transmission, and neuroendocrine regulation, whereas downregulated DEGs enriched in signal transduction (e.g., MAPK, Wnt, and calcium signaling), cytoskeletal organization, axon guidance, and metabolic processes (Fig. 6D and E). GSEA substantiated these findings, emphasizing pathways linked to cytoskeletal remodeling, inflammatory signaling, metabolic regulation, and hormone-mediated responses (Fig. S8).

Fig. 6.

Transcriptome sequencing of spinal cord tissues in PINP and CCI models. A. PCA of PINP group and PG@PLGA-gels PINP group. B. differential expression gene (DEGs) analysis between PG@PLGA-gel-PINP group and PINP group. C. The cluster heatmap analysis of the top 30 upregulated and downregulated genes between PG@PLGA-gel-PINP group and PINP group. D. Gene ontology (GO) enrichment analysis of the DEGs in the PG@PLGA-gels-PINP group. E. Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis of the DEGs in the PG@PLGA-gels-PINP group. F. PCA of CCI group and PG@PLGA-gels CCI group. G. DEGs analysis between PG@PLGA-gel-CCI group and CCI group. H. The cluster heatmap analysis of the top 30 upregulated and downregulated genes between PG@PLGA-gel-CCI group and CCI group. I. Gene ontology (GO) enrichment analysis of the DEGs in the PG@PLGA-gels-CCI group. J. Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis of the DEGs in the PG@PLGA-gels-CCI group.

In the CCI model, a total of 1614 DEGs were identified between the PG@PLGA-gel-CCI group and the untreated CCI group, with 750 genes upregulated and 864 genes downregulated in the PG@PLGA-gel-CCI group (Fig. 6G). Hierarchical clustering analysis revealed that the upregulated genes involved in energy metabolism, neuron differentiation and apoptosis, and response to calcium ion, whereas the downreg ulated genes were primarily associated with the nervous system and ion transport (Fig. 6H). GO and KEGG enrichment analysis revealed that upregulated DEGs were enriched in vesicle-mediated transport, synapse organization, neuronal differentiation, while downregulated DEGs were primarily enriched in excitiatory neurotransmission and ion transport (Fig. 6I and J). GSEA corroborated these findings, consistently showing enrichment in pathways such as circadian entrainment, glutamatergic synapse, calcium signaling, and PI3K-Akt signaling (Fig. S9). Despite the distinct etiologies of PINP and CCI, transcriptomic analysis revealed overlapping patterns of pathway enrichment across the two models, including pathways related to neuronal structure, synaptic organization, excitatory neurotransmission, ion transport, and metabolic regulation.

3.6. PG@PLGA-gels reduced toxicity of PG

PG solution prolonged thiopental-induced anesthesia compared to saline (p < 0.001), indicating CNS depressant effects, whereas PG@PLGA-gels showed no difference from saline (Fig. 7A). There was a reduction in extensor muscle strength after GB30 injection (p < 0.001) but recovered to baseline levels within three days (Fig. 7B). Body weight remained stable across groups without obesity tendency at the given dose (10 mg/kg/d) (Fig. S10). Histological analysis of major organs (heart, liver, spleen, lung and kidney) revealed no evidence of treatment-related pathological changes in any group 14 days post-injection (Fig. S11).

Fig. 7.

Systemic safety and local motor function evaluation of PG@PLGA-gel. A. Effect of PG, PLGA and PG@PLGA-gel on sleep duration (sleeping time, min), ANOVA followed by a post hoc test, **: P < 0.01, ***: P < 0.001. B. Effect of GB30 injection on hindlimb motor function, ANOVA followed by a post hoc test, **: P < 0.01, ***: P < 0.001. Data are presented as mean ± SD (n = 5 per group).

4. Discussion

In this study, PG@PLGA-gels were formulated for acupoint injection at GB30, exhibiting superior and prolonged analgesic efficacy compared to PG solution or blank gels in two established NeuP models. The therapeutic advantages observed are likely attributable to the synergistic effects of extended local drug availability and acupoint-based modulation. Corresponding with the behavioral outcomes, transcriptomic analyses indicated treatment-related alterations at the pathway level in the spinal cord, suggesting potential involvement of neuronal structure, synaptic organization, excitatory signaling, and neuroendocrine-related processes. These findings offer system-level insights into the biological programs that may accompany the analgesic effects of PG@PLGA-gels.

GB30 was chosen as the injection site for several compelling reasons. Firstly, GB30 is one of the most frequently utilized and well-recognized acupoints for managing NP, demonstrating significant analgesic and antinociceptive effects across various pain types [15,47,48]. Secondly, stimulation of GB30 has been shown to produce analgesic effects in rat models of paclitaxel-induced NP and chronic constriction injury [14,16]. Thirdly, from an anatomical perspective, GB30 in rats is located directly above the sciatic nerve trunk, allowing the drug-loaded hydrogel to exert its effects locally on the target nerve. Lastly, this region is characterized by thick muscle tissue and is situated away from vital organs, enhancing procedural safety and making it suitable for repeated injections. Additionally, the dense network of small blood vessels at this site is anticipated to facilitate the degradation and clearance of the gel within the body. In our study, the swelling of the hydrogel occurs within the dense musculature surrounding the GB30 region, an area sufficiently compliant to accommodate volume expansion without causing tissue damage or ischemia. Given that the acupoints are three-dimensional spatial structures composed of various tissues, including nerves, blood vessels, muscles, and connective tissues [49,50], targeted stimulation induced by hydrogel swelling in the acupoint area can achieve specific therapeutic effects [51]. In our study, the PG@PLGA hydrogels exhibited favorable injectability, moderate degradation, and sustained stability, supporting their potential application in acupoint drug delivery systems.

The hydrogel-based acupoint injection offered distinct advantages in drug delivery [52]. A recent study has elucidated the therapeutic synergy of this acupoint gel, particularly in the context of magnetic-responsive hydrogels for pain modulation through acupoint stimulation [53]. This synergy arises from the enhancement of drug efficacy through both physical and chemical stimuli. Application of functional gel acupoint injection for the treatment of rheumatoid arthritis demonstrated a reduction in the toxic effects of conventional anti-inflammatory drugs while enhancing the analgesic efficacy of acupoints [54]. Our study builds upon this foundation by examining the role of drug release, distinguishing itself from prior research by employing a first-line drug as the active agent. PLGA is noted for its ability to improve the stability and bioavailability of drugs by protecting them from degradation and facilitating their transport across biological barriers, which holds significant implications for neurological diseases [55]. In our study, the experimental results regarding drug release and pharmacokinetics demonstrated that PG@PLGA-gels facilitate controlled release, enhance local retention, and prolong systemic circulation. Additionally, our findings confirmed that PG@PLGA-gels exhibit minimal cytotoxicity to nerve and muscle cells in vitro, underscoring their potential as a safe biomaterial for localized administration. Furthermore, PG@PLGA-gels demonstrated favorable biocompatibility and biodegradability for localized drug delivery. Neuropsychiatric disorders, including somnolence and dizziness, are frequently observed side effects of pregabalin PG solution has been shown to extend sleep duration and induce drowsiness, as sleepiness, dizziness, and loss of consciousness are common adverse effects associated with PG [56,57]. In our work, safety evaluations confirmed that PG@PLGA-gels reduced pregabalin-associated CNS depression and did not induce systemic toxicity, underscoring their translational potential. Overall, this combination presents a promising strategy for the treatment of chronic pain, as the mechanical and pharmacological effects synergistically target pain at its source.

To comprehensively evaluate the analgesic effects of GB30 injection of PG@PLGA-gels on NP induced by physical compression and drug damage, two well-established NP models were utilized in this study. The PINP model represents drug-induced toxic damage to sensory neurons, resulting in mitochondrial dysfunction, disruption of calcium homeostasis in neurons, and subsequent neuroinflammation [58]. The CCI model replicates axonal injury due to nerve compression through physical ligation, initiating a cascade of neuroinflammatory and central sensitization processes [59]. The molecular analysis of spinal cord tissues from CCI and PINP rats elucidated the mechanisms underlying the action of PG@PLGA-gels. The observed upregulation of neurotrophin signaling suggests that PG@PLGA-gels enhance neurogenesis and synaptic plasticity. Neurotrophins, such as brain-derived neurotrophic factor (BDNF), are integral to neuronal survival and repair, thereby supporting the restoration of pain-regulating mechanisms [60]. In our study, the neurotrophin signaling pathway was upregulated in the PINP group, which are essential for repairing damaged spinal cord circuits and enhancing pain control [61]. Moreover, the upregulation of genes such as Ptpn4 by PG@PLGA-gels at GB30, observed in both PINP and CCI, which are linked to neuronal differentiation and the inhibition of apoptosis, highlights the critical role of nerve regeneration during treatment [62]. Additionally, the administration of GB30 via PG@PLGA-gels demonstrated a modulation of pain signals by influencing synaptic activities, including serotonergic and dopaminergic synapses. Serotonin exerts its effects primarily through 5-HT₁ receptors, and activation of 5-HT₁A receptors in areas such as the spinal cord and brainstem results in the suppression of nociceptive transmission [63]. Furthermore, the dopaminergic synapse pathway is pivotal in modulating neuropathic pain through descending pathways [64]. Additionally, DEGs had revealed significant enrichment in endocrine-related pathways involved in the regulation of pain perception. and its receptors within the central nervous system, where they play recognized roles in neuroprotection, synaptic plasticity, and anti-inflammatory processes. In spinal cord injury models, gonadotropin-releasing hormone (GnRH) had been demonstrated to attenuate neuroinflammatory responses and promote functional recovery [65]. The observed upregulation of GnRH signaling suggested a treatment-induced activation of an intrinsic neuroadaptive program, which facilitated neuronal repair, suppresses glial activation, and strengthened downstream analgesic and trophic pathways [66]. The significant enrichment of endocrine-related pathways further suggested that PG@PLGA-gels at GB30 help restore neuroendocrine balance to influence neuroplasticity and inflammation. Overall, these findings demonstrate that PG@PLGA-gels provide a dual therapeutic action by promoting neuronal repair while suppressing maladaptive neuroplasticity, offering a promising strategy for the management of NP.

Our methodology integrates acupoint stimulation with pharmacological intervention to offer an innovative approach to chronic pain management, specifically addressing NP. This approach seeks to provide sustained analgesic effects while minimizing dependence on excessive medication. Unlike previous sustained-release hydrogel systems primarily designed for postoperative hyperalgesia [67], our strategy specifically targets NP phenotypes by combining acupoint-specific neural modulation with long-acting analgesic delivery. Furthermore, this method is adaptable to personalized medicine, allowing for the customization of drug types and injection sites based on the nature, severity, and individual characteristics of a patient's pain, thereby facilitating precise medical treatment. Overall, this approach presents a promising new strategy for the management of NP.

5. Strengths and limitations

Our research presents a targeted delivery method using PG-loaded PLGA nanoparticles within a thermosensitive chitosan hydrogel (PG@PLGA-gels) for injection at the GB30 site. This method potentially enhances current treatments in several ways. Firstly, it allows for sustained local drug release, minimizing systemic exposure and reducing the risk of pregabalin-induced sedation. Secondly, the in situ gelation at GB30 extends local retention and may work in harmony with acupoint-related neuromodulatory effects. Thirdly, a single administration providing prolonged analgesia could enhance patient compliance compared to repeated systemic dosing or topical treatments. The study was subject to several limitations. Firstly, the absence of a PLGA-gel/GB30 control group (lacking pharmacological agents) limited a more isolated assessment of the nanoparticle carrier's contribution to analgesic efficacy. Secondly, the long-term tissue responses to repeated local injections, including potential PLGA-induced acidification or inflammation, necessitate further investigation to establish a more thorough safety profile using mouse models. Thirdly, the precise placement of the needle at the GB30 site presents technical challenges that may impact reproducibility; thus, future research should consider exploring micro-volume multi-point injection strategies.

6. Conclusions

In conclusion, the present study illustrates that a CS/β-GP thermosensitive hydrogel is an effective vehicle for the sustained delivery of pregabalin at the GB30 acupoint. The PG@PLGA-gel system demonstrates prolonged analgesic efficacy in NeuP models while mitigating central side effects typically associated with systemic administration. By combining controlled drug release with acupoint-based administration, this study establishes a proof-of-concept framework for acupoint nanomedicine, underscoring its potential as a complementary strategy for the management of NeuP.

Ethics approval

The ethical approval number for this experiment is PZSHUTCM2411250003. All animal experiments were carried out following the National Institutes of Health (NIH, USA) guidelines for the care and use of laboratory animals in research.

Funding declaration

This work was supported by the grants from Natural Science Foun dation of China (NSFC) to K.W. (82474639, 81973940), Shanghai Acupuncture Clinical Research Center of Medicine Project to J.Z. (No. 20MC1920500), National Administration of Traditional Chinese Medicine high-Level Disciplines Construction Project to J.Z. (No.ZYYZDXK-2023068), and Three Year Action Plan for Shanghai to Further Accelerate the Inheritance, Innovation and Development of Traditional Chinese Medicine (2025-2027) to J.Z. (No. 1-1-2).

CRediT authorship contribution statement

Xiping Duan: Data curation, Investigation, Visualization, Writing – original draft. Manjing Li: Formal analysis, Investigation, Methodology. Zifan Liu: Investigation, Methodology. Wenyu Chen: Investigation, Validation. Tianlong Chan: Data curation, Formal analysis, Software. Ruoxi Zhang: Methodology. Jia Zhou: Funding acquisition, Resources, Supervision. Yichun Xu: Conceptualization, Methodology, Project administration, Resources, Supervision. Ke Wang: Conceptualization, Funding acquisition, Project administration, Resources, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Data availability

Data will be made available on request.