Male C57BL/6J mice (10–12-week-old with body weight of 24 ± 2 g) were purchased from the Animal Center of the Fourth Military Medical University. All mice were housed under specific pathogen-free conditions and fed standard animal chow and water. The animal experiments were approved by the Ethics Committee of the Fourth Military Medical University (license no. 20211016).

The mice were exposed to a controlled climate chamber (ambient temperature: 41.0 °C ± 0.5 °C, humidity: 60% ± 5%; XinRuan, Shanghai, China) without water and food. The rectal temperature was monitored as the core temperature (Tc) using a digital thermometer (BW-TH1101, Billion, Shanghai, China) every 15 min. When the Tc reached or exceeded 42.0 °C, the mice exhibited hind limb weakness and impaired locomotion, indicating the onset of HS. Then, the mice were maintained at room temperature (25.0 °C ± 0.5 °C) to recover until they were sacrificed (Leon et al., 2005; Luo et al., 2023a; Ji et al., 2024). Routine blood tests were performed using a DF-3000Vet automatic animal blood cell analyzer (Beijing, China).

The mice were randomly assigned to control (CTR) (n = 14), HS + normal saline (NS) (n = 15), and HS + APS (25, 50, 100, and 200 mg/kg BW; n = 14, 16, 16, and 12, respectively) groups. APS was administered intraperitoneally, and the concentrations were selected based on previous studies to ensure an appropriate balance between safety and therapeutic efficacy (Dong et al., 2019; Zhang et al., 2020; Ding et al., 2021; Ming et al., 2022; Li et al., 2023; Wei et al., 2023). After 1 week of treatment (once daily via intraperitoneal injection), the mice were exposed to heat stress, except for the CTR mice. Highly purified APS (purity exceeding 90%) was purchased from Tianjin Cinorch Pharmaceutical Co., Ltd. (Lot# 221205, China), with an average molecular weight ranging from 20,000 to 60,000 Da (patent no. ZL00811547.8). APS was purified using the water-extraction and alcohol-precipitation technique (Zhou et al., 2013). First, the root of AM was soaked and decocted in the stilled boiling water for extraction. After vacuum condensation, the liquid extract was precipitated and washed with various concentrations of alcohols to obtain the crude product (i.e., bound polysaccharides). This crude product then underwent filtration, elution, and ultrafiltration. The concentrated supernatant was further precipitated and washed with alcohols. Finally, the precipitates were lyophilized for analysis and experiments, yielding purified APS as a white powder suitable for human injection, which has been approved under the National Medical Products Administration of China. APS was detected using size-exclusion chromatography with refractive index detector (SEC-RID). The plant name (AM) had been confirmed through http://www.theplantlist.org.

Blood samples were collected using two methods: (1) medial canthus vein puncture before modeling and (2) terminal puncture via ocular enucleation at 3 h after HS that rectal temperature reached its lowest point followed by cervical dislocation (Li et al., 2021). All procedures were performed under isoflurane inhalational anesthesia and in compliance with the guidelines of the Institutional Animal Care and Use Committee. After cervical dislocation, the tissues were obtained and fixed with 4% paraformaldehyde. Subsequently, the tissues were embedded in paraffin and processed into 5-µm thick sections. Routine hematoxylin and eosin (H&E) staining was performed, and the specimens were imaged under a light microscope using a digital camera system (SlideView VS200, Olympus, Tokyo, Japan). Assessment of pathological changes in the intestines was blindly performed using the Olympus OlyVIA system (Tokyo, Japan). The villus height and crypt depth of the duodenal, jejunal, and ileal tissues were measured in six randomly selected fields of each slide. The crypt depth and muscle thickness in the colon tissue were measured in 10 random fields.

Intestinal lamina propria (LP) cells from mice were isolated using our previously described protocol (Wang et al., 2023). For FCM, the cells were incubated with anti-mouse CD16/CD32 Abs (BD Biosciences, San Jose, CA, United States) for 20 min at 4 °C, and then fluorochrome-labeled antibodies (eBioscience, Thermo Fisher Scientific) were added in a total volume of 100 μL, mixed thoroughly, and incubated for 30 min at 4 °C. FCM analysis was performed using an SA3800 spectral cell analyzer (Sony Biotechnology, Tokyo, Japan).

To examine neutrophils, paraffin-embedded sections were prepared following the standard protocol and then incubated for 2 h at room temperature with mouse monoclonal antibody against myeloperoxidase (MPO; GB15224, Servicebio, China). Next, these specimens were incubated with a Cy3-labeled goat anti-mouse secondary antibody. The nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI) and then visualized under a fluorescence microscope (Olympus).

Total RNA was extracted using TRIGene reagent (GenStar, China) and converted to cDNA using the Hifair Ⅱ 1st Strand cDNA Synthesis Kit (Yeasen Biotechnology, Shanghai, China). qPCR analysis was performed using the SYBR Master Mix (GenStar), and the results were quantified via the 2^−ΔΔCT method. β-actin was used as the internal control. Transcript levels were normalized to transcript levels of β-actin.

To avoid the cage effect, the mice were randomized to treatment groups within a cage before the first injection. Freshly collected feces were obtained and stored immediately at −80 °C. The 16S rRNA sequencing of the gut microbiota was performed using high-throughput sequencing (Genedenovo Biotechnology Ltd., Guangzhou, China). There were six samples in each group.

To assess the serum vitamin B level, a chemiluminescence assay kit was used (Keming Biotechnology, Suzhou, China). Briefly, the sample, standard solution, and reagents were added following the manufacturer’s protocol, mixed thoroughly, and reacted at 25 °C for 20 min. The absorbance at 390 nm was measured using a microplate reader (Infinite200 PRO, Tecan, Männedorf, Switzerland), and the ΔA value and vitamin B6 content were calculated based on the standard curve.

The main components of APS are D-glucose, glucuronic acid, galacturonic acid, hexuronic acid, rhamnose, and arabinose (Xu et al., 2024b). The corresponding SMILES expression levels were obtained from PubChem (https://pubchem.ncbi.nlm.nih.gov/). The potential target genes of these six main components were predicted by three target prediction databases with corresponding screening conditions: SwissTarget Prediction Database (http://swisstargetprediction.ch/, screening parameter: probability of >0, species: Homo sapiens), SuperPred database (http://prediction.charite.de/, screening parameter: model accuracy >80%), and TargetNet database (http://targetnet.scbdd.com/, screening parameter: probability of >0). HS-related genes were obtained from GeneCards (https://www.genecards.org/) and OMIM (https://omim.org/) databases by entering “heat stroke” as a keyword. The potential target genes predicted by each APS component were intersected with all HS-related genes to obtain common genes, which are the putative therapeutic targets. The protein–protein interaction (PPI) network was constructed using the String (https://cn.string-db.org/) database (the organism was selected as human and the score was set to high confidence of 0.7). The DAVID functional annotation tool (https://david.ncifcrf.gov/) was used to perform Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) functional enrichment analyses.

Apoptotic cells in the small intestinal tissue were detected using the TUNEL assay, following the manufacturer’s protocol (Thermo Fisher Scientific). Apoptotic cells were imaged under an Olympus microscope. The TUNEL-positive cells were counted in 10 randomly selected high-power fields at 400× magnification.

Protein samples were prepared using ristocetin-induced platelet aggregation lysate and a protease inhibitor, separated through sodium dodecyl sulfate–polyacrylamide gel electrophoresis, and transferred to polyvinylidene fluoride membranes. Proteins were detected using the enhanced chemiluminescence reagent and Chemidoc imaging system (Bio-Rad) following incubation with primary and secondary antibodies. The following primary antibodies were obtained from Cell Signaling Technology (Massachusetts, United States) and Proteintech (Chicago, United States): caspase-3, cleaved caspase-3, caspase-9, cleaved caspase-9, Bax, Bak, ZO-1, and occludin. The expression level of the target protein was quantified as the ratio of the gray intensity to that of β-actin.

IEC-6 cells (originated from rat small intestinal crypts) were purchased from ATCC and maintained in Dulbecco’s modified Eagle medium containing 10% fetal bovine serum. Cells were treated with different doses of APS (100, 200, 400 μg/mL) 1 h before challenging with lipopolysaccharide (LPS; 10 μg/mL), followed by a 24-h cultivation period. Each experiment was replicated in at least three independent parallel tests.

Cells were collected and washed with assay buffer, incubated with Calcein-AM and propidium iodide (PI) solutions following the manufacturer’s protocol (Solarbio, Beijing, China), and photographed under a fluorescence microscope (EVOS, Life Technology). The percentage of live (Calcein-AM⁺PI⁻) cells was calculated.

Cell viability was measured using the CCK-8 assay. Briefly, cells were seeded into 96-well plates and treated with LPS with or without APS for 24 h. Then, 10 μL of the CCK-8 reagent was added to each well. The plate was incubated at 37 °C for 4 h. The absorbance was recorded at 450 nm using a multiscan FC microplate reader (Thermo Scientific).

Statistical data were analyzed using GraphPad Prism 9.0 (GraphPad, La Jolla, CA, United States). Survival curves were plotted using the Kaplan-Meier method and compared using the log-rank test. Student’s t-test or one-way analysis of variance was used to compare different groups. P-values of <0.05 were considered to indicate statistical significance.

First, the mouse model of HS was established. Although different doses of APS pretreatment can improve the survival rate of HS mice compared with the HS + NS mice, only 50 mg/kg BW APS pretreatment group showed a significant difference (Figure 1A). Compared with the HS + NS group, the increasing rate of Tc in the HS + APS groups was slowed down, evidenced by the time to Tc max of the mice that was significantly longer in the HS + APS (>50 mg/kg BW) groups than in the HS + NS group (Figure 1B).

Regarding cell proportion, HS exposure can significantly reduce the proportion of lymphocytes and increase the proportion of granulocytes in the blood, whereas 50 and 100 mg/kg doses of APS pretreatment can reverse these effects (Figure 1C). Collectively, pretreatment with 50 mg/kg BW of APS had the best effects and was used for follow-up experiments.

APS improves HS-induced damage to organs. As shown in Figure 1D, multiple organ injuries developed during HS. The organs in the HS + APS group were more swollen, darker in color, and even had more bleeding points than were those in the HS + NS group. Further histopathological examination of various organs indicated that APS treatment reduced the risk of HS onset in mice. Specifically, the HS + NS group had damaged liver lobules and abnormally congested hepatic sinusoids and venules. The disruption of some glomerular structures was accompanied with congested interstitial blood vessels and lymphocyte infiltration; the alveolar wall capillaries showed signs of dilation and congestion, with hemorrhage in the alveolar cavities and lymphocyte infiltration. Moreover, alterations were noted in the spleen, where the structures of both white and red pulps were disturbed with unclear boundaries; all these injuries were ameliorated in the mice pretreated with APS (Figure 1E).

Next, we assessed the intestines histologically; H&E staining revealed that HS induced the loss of the intestinal epithelium and damage occurred in both the small and large intestines. In the HS + NS group, the villi in the small intestines (duodenum, jejunum, and ileum) were remarkably shortened, but the depth of the crypts was not altered. In most of the mice, the intestinal villi tips were detached. In addition, it is accompanied with enhanced inflammatory cell infiltration in the LP. However, APS attenuated the pathological damage to the small intestines in HS mice (Figures 2A,B). Meanwhile, increased depth of the crypts and a thickened mucus layer were observed in the colons. Similar to the small intestine, APS also attenuated the pathological damage in the large intestines of HS mice (Figures 2A,C).

Inflammatory cytokines were measured at the time of sample collection to test the efficacy of APS on the intestines. The mRNA levels of IL-6, TNF-α, and IL-1β were elevated following HS exposure, whereas APS pretreatment showed marked inhibition of inflammatory cytokines, indicating marked alterations in the intestinal inflammation of the mice (Figure 2D).

Neutrophil infiltration was assessed through FCM analysis and immunofluorescence staining of MPO⁺ cells, as an overexuberant neutrophil response is known to cause epithelial damage in the inflammatory intestine (Williams and Parkos, 2007). As shown in Figure 3A, compared with the CTR mice, HS + NS mice had a higher proportion of neutrophils, and APS showed marked inhibition of neutrophils; however, the proportion of macrophages and T lymphocytes remained unchanged. These results indicated that in the early/acute phase of HS, neutrophils are first-line defenders, and their dysfunction inducing the innate immune inflammatory response leads to mortality and poor outcomes. Additionally, the LP of the ileal region was rich in immune cells, and the HS + NS mice had a higher number of MPO⁺ cells than the CTR mice, with the HS + APS group showing significantly less neutrophil infiltration than the HS + NS group (Figure 3B).

Modern pharmacological studies have shown that APS exerts favorable effects on organ damage. However, the role of APS in gastrointestinal tract and their underlying protective mechanisms remain unclear. Therefore, we aimed to determine the role of APS in the gut from the perspective of microbial transformation.

After read-quality filtering, denoising, and clustering, high-quality bacterial 16S rRNA gene sequences were successfully obtained from 12 samples. The gut microbial abundance curve revealed that the sample in each group had sufficient microbial abundance (data not shown). Gut microbiota was assessed using α-diversity indexes, with the Sob and Chao1 indexes reflecting the community richness and the Shannon and Simpson indexes reflecting the community diversity. As shown in Figure 4A, the Sob and Chao1 indexes of the APS group were significantly higher than those of the NS control group, whereas the Shannon and Simpson indexes showed no differences between the two groups, indicating that APS had a significant effect on microbiota richness but no effect on microbiota evenness.

Next, we compared the microbial community similarity using two β-diversity measures: Partial Least Squares Discrimination Analysis (PLS-DA) and Anosim. PLS-DA scatterplots are depicted in Figure 4B, which revealed that APS was significantly different from NC at the operational taxonomic unit level, indicating that the microbial community was distinctly altered by APS administration. Anosim analysis confirmed a significant difference between the two groups (P = 0.019).

As shown in Figure 4C, the Venn plot demonstrated that 105 genera coexisted in the two groups and nine genera were unique to the APS group, including Frisingicoccus, Gram-negative bacterium cTPY-13, Methylobacterium–Methylorubrum, Lactococcus, Microbulbifer, Nosocomiicoccus, Facklamia, Candidatus, Paracaedibacter, and Fusobacterium. The detailed results of cluster analysis of the intestinal flora are shown in Figure 4D. Firmicutes, Bacteroidetes, Verrucomicrobia, Proteobacteria, Desulfobacterota, Campilobacterota, and Actinobacteriota were the most dominant phyla in each group. In terms of abundance, the top three flora at the genus level were Akkermansia, Lactobacillus, and Lachnospiraceae_NK4A136_group.

Linear discriminant analysis (LDA) effect size (LEfSe) was used to screen specific species to further determine the effects of APS on intestinal flora. As shown in Figure 5A, compared with the NS group, the relative abundances of some genera, including eubacterium_fissicatena_group, Erysipelatoclostridiaceae, Bacteroides_caecimuris, Blautia, and Lachnoclostridium, significantly reduced, whereas the abundances of Gram_negative_bacterium_cTPY_13, Burkholderiales_bacterium_YL45, Burkholderiales, Parasutterella, and Sutterellaceae evidently increased in the APS group (P < 0.05, LDA score >3).

Bacterial functions were predicted using PICRUSt2 and Tax4Fun algorithms. Tax4Fun analysis showed that the functional categories of fatty acid metabolism, geraniol degradation, steroid hormone biosynthesis, and chemical carcinogenesis were significantly different between NS and APS groups (Figure 5B). PICRUSt2 analysis revealed that compared with the NC group, APS-induced bacterial functional changes in KEGG level 3 were mainly involved in vitamin B6 metabolism, riboflavin metabolism, toluene degradation, and protein processing in the endoplasmic reticulum (Figure 5C). Systemic vitamin B6 deficiency has previously been associated with increased oxidative stress and inflammation (Lysne et al., 2016). Riboflavin is a natural antioxidant, and the depletion of riboflavin causes endoplasmic reticulum stress (Zhang et al., 2022a; Zhang et al., 2022b). Moreover, vitamin B6 and riboflavin may play important protective roles against heat stress at the cellular level (Sammad et al., 2022; Zhang et al., 2022b). These results indicated that the impact of APS on the microbiota structure was also accompanied with changes in microbial function.

Vitamin B6 deficiency has been reported to contribute to inflammatory disease, and vitamin B6 supplementation can reverse these effects in deficiency states. To verify the effect of APS on vitamin B6, we detected its level in the serum. In the resting state, APS did not significantly affect vitamin B6 levels. Within 3-h post-HS, vitamin B6 levels showed a compensable elevated trend. We then assessed the serum levels of vitamin B6 in the HS group pretreated with APS. However, compared with the CTR and APS groups, APS notably elevated serum vitamin B6 levels after HS (Figure 5D).

To further explore the underlying mechanism of APS treatment for HS, we used three target prediction databases to identify and analyze the potential target genes of the six major components of APS (D-glucose, glucuronic acid, galacturonic acid, hexuronic acid, rhamnose, and arabinose). A total of 33 potential targets were obtained in the SwissTarget database, 212 in the SuperPred database, and 248 in the TargetNet (Figure 6A). After deleting duplicate targets, a final list of 427 potential APS-related targets was obtained. In addition, 6456 HS-related genes were retrieved from the GeneCards database and 843 from the OMIM database, yielding a total of 6945 HS-related genes after removing duplicates (Figure 6B). The intersection of APS potential target genes and HS-related genes was then calculated, revealing 283 common genes, which were considered potential therapeutic targets of APS for HS treatment (Figure 6C). These target genes were next imported into the String database to construct a PPI network (Figure 6D) and underwent GO and KEGG functional enrichment analyses. The results demonstrated that these target genes were enriched in pathways such as phosphorylation, inflammatory response, positive regulation of ERK1 and ERK2 cascade, response to LPS, and response to hypoxia. Notably, the apoptosis-related pathway was enriched in both GO and KEGG analyses (Figure 6E).

Dysfunctions in apoptosis are implicated in HS pathogenesis. We used the TUNEL assay to detect apoptotic DNA fragmentation in the intestine. HS mice showed more cells positive for apoptotic DNA fragmentation than the NC mice; however, there were significantly fewer positive cells in the HS + APS mice compared with those in the HS + NS mice (Figure 7A).

The effects of APS on apoptotic pathways were further assessed via Western blot analysis. The results demonstrated a marked increase in the protein levels of Bax, Bak, and cleaved caspase-3 following HS exposure. Notably, pretreatment with APS significantly attenuated the HS-induced upregulation of intrinsic apoptotic proteins, as illustrated in Figure 7B. Concurrently, we observed a reduction in the expression of the tight junction proteins zonula occludens (ZO)-1 and occludin in the HS + NS group, indicating a compromised intestinal barrier integrity. However, this impairment was mitigated in the HS + APS group (Figure 7C). These findings suggest that APS has therapeutic potential in mitigating the intestinal damage caused by HS by modulating the apoptotic process.

Similar results were observed in IEC-6 intestinal epithelial cells. We first performed CCK-8 assay to evaluate the viability of cells treated with different doses of APS, and found that treatment with 200 μg/mL APS exhibited the best cell viability (Figure 8A). We then employed this dose of APS to conduct further Calcein-AM/PI staining. Results revealed a significant decrease in live cells after LPS treatment, and treatment with APS significantly mitigated the LPS-induced upregulation of cell death (Figure 8B). Western blot analysis demonstrated a substantial increase in the protein levels of Bax, Bak, cleaved caspase-3, and cleaved caspase-9 following LPS stimulation; however, APS treatment significantly attenuated the LPS-induced upregulation of these apoptotic proteins, as shown in Figure 8C.