A total of 48 specific pathogen-free male Sprague-Dawley rats (age, 6–8 weeks; weight, 250–320 g) were obtained from the Experimental Animal Center of Xinjiang Medical University (Ürümqi, China). All experimental protocols adhered to international guidelines for the care and use of laboratory animals (12) and were approved by the Animal Experiment Ethics and Welfare Committee of General Hospital of Xinjiang Military Command (ethics no. DWLL2023021802; Ürümqi, China).

The rats were housed in a small animal facility at the Desert Medicine Laboratory, General Hospital of Xinjiang Military Command, under controlled conditions (temperature: 22±2°C; relative humidity, 40±5%; 12 h light/12 h dark cycle). Before experimentation, rats underwent a 7-day adaptive period with unrestricted access to food and water.

Experiments were conducted in the Simulated Climate Cabin for Special Environment of Northwest China (Ürümqi, China). The environmental parameters for the experimental groups were set at 41±1°C temperature and 15±5% relative humidity, whereas the control group was maintained under standard conditions (22±2°C, 40±5% relative humidity). Before modeling, all rats underwent 7 days of incremental-load adaptive training on a horizontal treadmill. The training schedule was as follows: Treadmill speed was set at 10 m/min on day 1, at 12 m/min on day 2, was increased by 1 m/min daily between days 3 and 5 (up to 14 m/min), and was stabilized at 15 m/min for days 6 and 7. Each training session lasted 15 min. To encourage continuous exercise, a constant current stimulation electrode (1 mA) was placed at the end of the treadmill for negative reinforcement (13). After adaptive training, the rats were randomly allocated into six groups (n=8/group) using stratified randomization: Control, EHS, physiological saline, and low-, medium- and high-dose nanocurcumin groups. Based on reported safe intravenous curcumin dosages in humans (120 mg) (14,15), rat doses of 9.75, 19.5 and 29.25 mg/kg were calculated using standard body surface area normalization.

During EHS modeling undertaken by all rats, the rats exercised at 15 m/min in a desert environment (41±1°C temperature, 15±5% relative humidity). Rectal temperatures were measured and recorded every 5 min. The endpoint criteria for successful modeling included core temperatures of ≥42.5°C in three consecutive measurements at 5-min intervals, accompanied by exercise incapacity (loss of righting reflex). Upon reaching these criteria, the rats were immediately removed from the heat chamber.

After successful establishment of the model, nanocurcumin was administered intravenously in a nanocrystal formulation (particle size, ~70 nm; concentration, 30 mg/ml), provided by the Institute of Pharmacology and Toxicology, Academy of Military Medical Sciences (Beijing, China). The formulation was diluted in sterile normal saline to achieve the desired concentrations. All injections were administered via the tail vein immediately after EHS induction (post-treatment intervention). The saline group received an equivalent volume of sterile normal saline. The control group received no treatment and was kept under normal conditions without EHS induction or drug administration. Rats were euthanized 4 h after nanocurcumin administration for sample collection.

After the observation period, the rats were anesthetized intraperitoneally with pentobarbital sodium (30 mg/kg) and secured on an operating table. Blood samples (3–5 ml per rat) were collected from the abdominal aorta under deep anesthesia. After blood collection, the rats were humanely euthanized by exsanguination under continued anesthesia. Death was confirmed by the absence of heartbeat, respiratory movements, corneal reflex and response to paw pinch prior to tissue collection. All procedures adhered strictly to institutional guidelines for laboratory animal care and use, and humane endpoints were predefined. Humane endpoints included severe neurological dysfunction (for example, loss of righting reflex, seizure activity, or unresponsiveness), marked impairment of motor function (such as inability to stand or maintain posture), signs of severe physiological distress (for example, abnormal respiration patterns and profound weakness), and any condition judged by veterinary staff to indicate moribund state or irreversible distress. No animals died spontaneously or before the planned experimental endpoint. Blood samples were collected from the abdominal aorta into vacuum collection tubes containing appropriate anticoagulants. Brain tissues were collected on ice, immediately transferred to cryogenic tubes and placed in fixative solutions.

Serum samples were allowed to clot at room temperature for 30 min, after which they were centrifuged at 1,000 × g, at 4°C for 15 min, and the supernatants were collected for biochemical and ELISA analyses. Aliquots of serum samples were stored at −80°C until further analysis.

Serum levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), creatinine (Cr), blood urea nitrogen (BUN), creatine kinase (CK) and lactate dehydrogenase (LDH) were measured using an automated biochemical analyzer (Shenzhen Mindray Bio-Medical Electronics Co., Ltd.). Concentrations of IL-6, TNF-α, IL-1β, thyrotropin-releasing hormone (TRH), thyroid-stimulating hormone (TSH), corticotropin-releasing hormone (CRH), adrenocorticotropic hormone (ACTH), total triiodothyronine (T3), total thyroxine (T4), free triiodothyronine (FT3), free thyroxine (FT4) and corticosterone (CORT) were determined using rat ELISA kits following the manufacturer's protocols. IL-6 (cat. no. F3066-A), TNF-α (cat. no. F3056-A), IL-1β (cat. no. F2923-A), TRH (cat. no. F3393-A), ACTH (cat. no. F3440-A), T3 (cat. no. F0487-OA), FT3 (cat. no. F0004-ZA), T4 (cat. no. F0107-ZA), FT4 (cat. no. F0010-ZA) and CORT (cat. no. F0096-OA) kits were purchased from Fankewei Biotechnology Co., Ltd. TSH was measured using a kit from Beijing Bioss Biotechnology Co., Ltd. (cat. no. BSKR62969), and CRH was measured using a kit from Wuhan Saipei Biotechnology Co., Ltd. (cat. no. SP13191). Each sample was tested in duplicate. Absorbance was measured at 450 nm using a microplate reader, and concentrations were calculated from standard curves.

Brain tissues were fixed in 10% neutral buffered formalin at room temperature (20–25°C) for 7 days, followed by dehydration and paraffin embedding according to standard protocols. Paraffin-embedded tissues were sectioned at 4-µm thickness. For H&E staining, paraffin sections were deparaffinized in xylene and rehydrated through graded ethanol. Sections were stained with hematoxylin solution for 3–5 min at room temperature, followed by eosin staining for 1–2 min at room temperature using a Servicebio H&E staining kit (Wuhan Servicebio Technology Co., Ltd.), according to the manufacturer's instructions. For Nissl staining, sections were incubated in cresyl violet staining solution (Wuhan Servicebio Technology Co., Ltd.) for 2–5 min at room temperature, followed by differentiation in 0.1% acetic acid and dehydration, according to the manufacturer's instructions. After staining, sections were dehydrated, cleared and mounted with neutral resin. Images were acquired using a light microscope.

The 3D structure of curcumin was retrieved from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/). Potential targets of curcumin were predicted using SwissTargetPrediction (http://www.swisstargetprediction.ch), similarity ensemble approach (SEA; http://sea.bkslab.org) and PharmMapper (16). For SEA analysis, target screening was limited to ‘Homo sapiens’ with a Max-Tc score >0.8. For PharmMapper, targets with a Fit Score ≥3.0 and P<0.05 were selected. Results from these databases were merged, duplicates removed and a comprehensive curcumin-target dataset was established. EHS-related targets were collected from GeneCards (https://www.genecards.org), OMIM (https://www.omim.org) and National Center for Biotechnology Information (NCBI) Gene (https://www.ncbi.nlm.nih.gov/gene) using the key word ‘exertional heat stroke’. After duplicates were removed, an EHS-target dataset was created. Potential therapeutic targets of curcumin for EHS were identified using the Jvenn online tool (https://www.bioinformatics.com.cn/static/others/jvenn/) by intersecting the curcumin-target and EHS-target datasets.

Protein-protein interaction (PPI) analysis was conducted using the STRING database (version 11.5; http://string-db.org), restricting species to ‘Homo sapiens’ with a confidence score ≥0.4. The resulting PPI network was imported into Cytoscape software (version 3.10.4; http://cytoscape.org/) (17) for visualization and topological analysis. Hub genes were identified using the CytoHubba plugin based on degree ranking, reflecting the number of direct interactions with other nodes. Significant functional modules within the network were further classified using the MCODE plugin in Cytoscape with the parameters set as follows: Degree Cutoff, 2; Node Score Cutoff, 0.2; K-Core, 2; and Max Depth, 100. Modules demonstrating high connectivity and biological relevance were selected for subsequent analyses.

Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were conducted using the clusterProfiler package in R (version 4.5.2; http://www.r-project.org/) (18). The Benjamini-Hochberg method was used to correct for multiple testing, and adjusted P-values (P.adjust <0.05) were considered statistically significant.

Three-dimensional structures of hub proteins were retrieved from the Protein Data Bank (https://www.rcsb.org/). Protein structures were processed using PyMOL (version 2.3.0; http://pymol.org/) by removing water molecules and ligands, hydrogen atoms were added using AutoDock Tools (version 1.5.7; http://autodock.scripps.edu/) and structures were saved in pdbqt format. The 3D structure of curcumin was downloaded from PubChem and optimized using the MMFF94 force field in OpenBabel (version 3.1.1; http://openbabel.org/) to obtain its lowest-energy conformation. Molecular docking was performed using AutoDock Vina (version 1.2.5; http://vina.scripps.edu/) with an exhaustiveness setting of 25 to ensure docking accuracy. Binding affinities were evaluated based on calculated binding free energies (kcal/mol), where lower values indicate stronger binding.

Approximately 50 mg brain tissue was ground in liquid nitrogen and transferred into a 1.5-ml centrifuge tube. The tissue samples were then lysed in 300 µl RIPA buffer (Wuhan Servicebio Technology Co., Ltd.) containing 1 mM PMSF, 1X phosphatase inhibitor cocktail and 1X protease inhibitor cocktail. After incubation on ice for 30 min, the lysates were centrifuged at 12,000 × g for 20 min at 4°C, and supernatants were collected.

Protein concentrations were measured using a BCA protein assay kit (Wuhan Servicebio Technology Co., Ltd.) according to the manufacturer's instructions. Subsequently, equal amounts of protein (30 µg/lane) were separated by SDS-PAGE on 10% gels and transferred onto PVDF membranes. The membranes were blocked in 5% skimmed milk prepared in TBST (Tris-buffered saline containing 0.1% Tween-20) at room temperature for 1 h. The membranes were incubated overnight at 4°C with primary antibodies against TLR4 (cat. no. BS-1021R; 1:1,000; BIOSS), MyD88 (cat. no. BS-1047R; 1:1,000; BIOSS), NF-κB (cat. no. BS-0465R; 1:1,000; BIOSS) and β-actin (cat. no. 4970S; 1:2,000; Cell Signaling Technology, Inc.). After three washes in TBS-Tween (10 min each), the membranes were incubated with HRP-conjugated secondary antibodies (goat anti-rabbit IgG; cat. no. A0208; 1:10,000; Wuhan Servicebio Technology Co., Ltd.) for 1 h at room temperature. The protein bands were visualized using enhanced chemiluminescence (ECL) detection kit (Wuhan Servicebio Technology Co., Ltd.) reagent and captured using a gel imaging system. Band intensities were semi-quantified using ImageJ software (version 1.53t; National Institutes of Health) and normalized to β-actin.

All statistical analyses were performed using SPSS 25.0 (IBM Corp.), GraphPad Prism 10 (Dotmatics) and R 4.5.2. Data normality was assessed using the Shapiro-Wilk test and homogeneity of variance was evaluated using Levene's test. Data are presented as the mean ± SD. Differences among groups were analyzed using one-way ANOVA, followed by Tukey's post hoc test for multiple comparisons. All tests were two-sided and P<0.05 was considered to indicate a statistically significant difference.

A total of 188 potential curcumin targets were identified from the SwissTargetPrediction and SEA databases. Additionally, 5,314 EHS-related target genes were retrieved from the GeneCards and NCBI databases. Venn analysis identified 138 overlapping targets, which were considered potential therapeutic targets of curcumin in EHS (Fig. 1A). These intersecting targets were imported into the STRING database to construct a PPI network (Fig. 1B). Topological analysis using Cytoscape 3.10.4 showed that the PPI network comprised 134 nodes and 2,392 edges (Fig. 1C). Based on node degree, the CytoHubba plugin was used to identify the top seven hub genes: AKT1, TNF, EGFR, BCL2, STAT3, SRC and NFKB1 (Fig. 1D). These hub genes are primarily involved in inflammatory regulation, cell survival, apoptosis and signal transduction (19–23), suggesting that curcumin may exert therapeutic effects against EHS through multiple key signaling targets.

To further characterize the PPI network and improve visualization of the results, MCODE analysis was performed to identify highly interconnected functional modules. Six significant clusters were identified (Table I). Among these, clusters 1 and 3 were selected for further analysis because of their high connectivity and biological relevance.

Cluster 1 had the highest MCODE score (19.143) and contained 22 nodes and 402 edges, indicating that it represented the core functional module. This cluster included several key genes, including BRAF, FOS, STAT3, TNF, SRC, EGFR, AKT1, JAK2, BCL2, CHUK and NFKB1 (Fig. 1E). These genes were mainly associated with inflammation-related signaling pathways, including the NF-κB, PI3K-AKT, MAPK and JAK-STAT pathways (24,25).

Cluster 3 contained 22 nodes and 126 edges. Representative genes in this cluster included MMP3, MMP7, MMP9, MMP13, ADAM17, NOS2, NOX4, TLR9, IKBKB and SERPINE1 (Fig. 1F). These genes were primarily associated with extracellular matrix remodeling, inflammatory responses and oxidative stress-related processes (26). The remaining clusters were not analyzed further because they exhibited lower connectivity or were less relevant to the primary biological processes investigated in the current study.

The 138 overlapping targets between curcumin and EHS were subjected to GO and KEGG enrichment analyses. GO analysis (Fig. 2A and B) showed that these targets were predominantly enriched in biological processes related to ‘response to peptide hormone’, ‘regulation of inflammatory response’, ‘cellular response to peptide hormone stimulus’, ‘response to stress’, and ‘cellular response to stimulus’. In the cellular component category, significant enrichment was observed in ‘membrane raft’ and ‘membrane microdomain’, as well as ‘focal adhesion’ and ‘cell-substrate junction’, suggesting involvement in receptor-mediated signaling complex assembly. Molecular function analysis revealed significant enrichment in ‘signal transducer activity’ and ‘cytokine receptor binding’.

KEGG pathway analysis (Fig. 2C and D) demonstrated that the overlapping targets were significantly enriched in multiple signaling pathways, including ‘PI3K-Akt signaling pathway’, ‘MAPK signaling pathway’, ‘Ras signaling pathway’, ‘AGE-RAGE signaling pathway in diabetic complications’, ‘TNF signaling pathway’, ‘FoxO signaling pathway’, ‘T cell receptor signaling pathway’, ‘Chemokine signaling pathway’, ‘IL-17 signaling pathway’, ‘HIF-1 signaling pathway’, ‘mTOR signaling pathway’, ‘Insulin signaling pathway’, ‘NF-kappa B signaling pathway’, ‘Toll-like receptor signaling pathway’, ‘VEGF signaling pathway’, ‘Notch signaling pathway’, and ‘B cell receptor signaling pathway’. These pathways are mainly involved in inflammatory regulation, immune response and cell survival signaling. These pathways are known to serve central roles in the systemic inflammatory response induced by EHS (1,7,27,28). Given the pivotal role of the TLR4/MyD88/NF-κB axis in amplifying inflammatory cascades, this pathway was selected for subsequent experimental validation. Collectively, these enrichment analyses suggested that curcumin may protect against EHS-induced brain injury through modulation of key inflammatory signaling networks.

To further investigate interactions between key target proteins and curcumin in the context of EHS treatment, molecular docking analysis was conducted. The docking results indicated that curcumin exhibited strong binding affinities with the active pockets of the identified target proteins. Interaction network analysis highlighted AKT1, SRC, NFKB1, BCL2, EGFR, STAT3 and TNF as highly interconnected nodes, potentially serving critical roles in the pathogenesis of EHS. Curcumin demonstrated favorable binding to these proteins, with calculated binding free energies of −6.85 kcal/mol (AKT1), −6.24 kcal/mol (SRC), −5.83 kcal/mol (NFKB1), −5.53 kcal/mol (BCL2), −5.56 kcal/mol (EGFR), −3.89 kcal/mol (STAT3) and −4.06 kcal/mol (TNF-α) (Fig. 3A-G). These data suggested that curcumin may modulate the activity of these target proteins by binding directly to their receptor sites, thereby potentially attenuating pathological processes associated with EHS. Additionally, based on KEGG pathway enrichment analysis highlighting the TLR and NF-κB signaling pathways, docking analyses were conducted for MyD88 and TLR4, two key upstream regulators of NF-κB signaling. Curcumin exhibited strong binding affinities with MyD88 (−7.9 kcal/mol) and TLR4 (−5.96 kcal/mol) (Fig. 3H and I). Collectively, these findings indicated that curcumin may exert therapeutic effects against EHS through interactions with these critical molecular targets, including key upstream regulators of the TLR4 signaling pathway.

Intravenous administration of nanocurcumin partially reversed dysfunction in the hypothalamic-pituitary-thyroid (HPT) and hypothalamic-pituitary-adrenal (HPA) axes induced by EHS in rats. Compared with in the control group, rats in the EHS group exhibited HPT axis activation accompanied by peripheral hormone conversion abnormalities. Specifically, serum levels of TRH, TSH, T4 and FT4 were significantly increased, whereas T3 and FT3 levels were significantly decreased, consistent with low T3 syndrome (Fig. 4A-F). Physiological saline partially attenuated the alterations in TRH, TSH, T3, T4, and FT4 levels induced by EHS, while FT3 remained unchanged. Nanocurcumin treatment partially corrected these abnormalities, Compared with the EHS group, low-dose nanocurcumin significantly reduced TRH and TSH levels, low- and medium-dose nanocurcumin reduced FT4 levels, medium- and high-dose nanocurcumin reduced T4 levels and increased T3 levels, and high-dose nanocurcumin increased FT3 levels.

In addition, serum CRH and CORT levels were significantly elevated, whereas ACTH was markedly reduced in rats in the EHS group, reflecting impaired feedback regulation of the HPA axis (Fig. 4G-I). Treatment with nanocurcumin partially ameliorated these EHS-induced changes, Among the three doses, high-dose nanocurcumin showed the most comprehensive improvement in HPA axis-related hormone profiles, particularly by restoring ACTH levels and reducing CRH and CORT levels, although low-dose nanocurcumin appeared to exert a stronger effect on CORT reduction. These findings suggested that nanocurcumin has protective and regulatory effects on neuroendocrine dysfunction associated with EHS.

To further evaluate the systemic effects of nanocurcumin and to determine whether acute toxicity occurred during the short-term observation period, serum inflammatory cytokines and biochemical markers related to peripheral organ and tissue damage were examined.

As shown in Fig. 5A-C, compared with in the control group, rats subjected to EHS exhibited significantly elevated serum levels of TNF-α, IL-6 and IL-1β, indicating a marked systemic inflammatory response. Serum cytokine levels in the NS group showed a partial reduction compared with the EHS group, but did not return to control levels. Conversely, nanocurcumin treatment effectively reduced serum TNF-α, IL-6 and IL-1β levels to varying degrees, with the most pronounced effects observed in the medium- and high-dose groups. These findings suggested that nanocurcumin may attenuate the systemic inflammatory response triggered by EHS.

Additionally, biochemical markers associated with liver, kidney, skeletal muscle and general tissue injury were analyzed. Compared with in the control group, the EHS group demonstrated significant increases in serum ALT, AST, Cr, BUN, CK and LDH levels (Fig. 5D-I). These elevated markers indicated biochemical evidence of liver injury, renal dysfunction, muscle damage and generalized tissue damage following EHS induction. The normal saline group did not exhibit a full reversal of these abnormalities. By contrast, nanocurcumin treatment reduced ALT, AST, BUN, CK and LDH levels in a dose-dependent or partially dose-dependent manner. Serum Cr levels also showed a downward trend following nanocurcumin treatment, especially at medium and high doses. Notably, medium- and high-dose nanocurcumin groups did not exhibit further increases in biochemical markers compared with in the EHS or NS groups. These results indicated that nanocurcumin does not exacerbate acute liver, kidney or tissue injury within the 4-h therapeutic window used in the current study; instead, it partially reverses biochemical abnormalities caused by EHS, demonstrating no apparent acute toxicity.

Histopathological analysis results (H&E staining) are shown in Fig. 6A and B. In the control group, neuronal morphology appeared normal, with intact cellular structures observed in the cerebral cortex and hypothalamus. By contrast, rats in the EHS group exhibited marked neuronal shrinkage, nuclear pyknosis, cytoplasmic eosinophilia and disrupted cellular arrangement, indicating substantial brain damage. Rats in the normal saline group displayed similar pathological changes, indicating no spontaneous recovery. Nanocurcumin treatment attenuated these neuronal injuries in a dose-dependent manner, with the medium- and high-dose groups showing less neuronal shrinkage, improved cellular arrangement and reduced nuclear condensation compared with in the EHS group.

Nissl staining further confirmed these findings (Fig. 6C and D). The EHS group exhibited reduced numbers of Nissl bodies and prominent neuronal atrophy in the hypothalamus. By contrast, nanocurcumin administration maintained neuronal integrity and partially restored Nissl body distribution, especially in the medium- and high-dose groups. These results suggested that nanocurcumin may effectively alleviate EHS-induced neuronal injury in the cerebral cortex and hypothalamus.

To elucidate the therapeutic mechanism of nanocurcumin in EHS, western blot analysis was conducted to measure the expression of key inflammatory signaling proteins in EHS-model rats (Fig. 7). Compared with in the control group, the protein expression levels of TLR4, MyD88 and NF-κB were significantly increased in EHS-model tissues, and normal saline treatment did not restore these parameters to baseline levels. Following nanocurcumin intervention, the expression levels of these proteins decreased in a dose-dependent manner. MyD88, TLR4 and NF-κB expression was significantly decreased in response to all nanocurcumin doses (low, medium and high), with the most pronounced effect observed at the highest dose, nearly restoring expression levels to those of the control group.