Anwei decoction alleviates chronic atrophic gastritis by modulating the gut microbiota-metabolite axis and NLRP3 inflammasome activity

Abstract

BACKGROUND

Chronic atrophic gastritis (CAG) is a clinically refractory gastric disease often characterized by high recurrence rates and adverse drug reactions. Anwei decoction (AWD), a traditional Chinese medicine formula, has been shown to significantly improve clinical symptoms in patients with CAG, as demonstrated by a multicenter cohort study (overall effective rate: 82.5%, P < 0.01). However, the unclear molecular mechanisms and therapeutic targets of AWD limit its international acceptance.

AIM

To investigate the therapeutic mechanisms of AWD against CAG from an integrated perspective.

METHODS

In this study, N-methyl-N’-nitro-N-nitrosoguanidine was used to establish a CAG rat model. Serum-derived constituents transferred from AWD were first identified using ultra-high-performance liquid chromatography coupled with tandem mass spectrometry. The concentrations of inflammatory cytokines in serum samples were determined by enzyme-linked immunosorbent assay. Moreover, gastric mucosal tissues were analyzed by quantitative real-time polymerase chain reaction to measure messenger RNA (mRNA) levels of the NLRP3 inflammasome. Western blotting was used to detect the protein expression of NLRP3, caspase-1, and interleukin (IL)-1β. To elucidate the regulatory mechanisms underlying AWD treatment, structural alterations of the gut microbiota (GM) and associated metabolites were analyzed using integrated high-throughput sequencing (16S rRNA) and liquid chromatography-mass spectrometry based untargeted metabolomics. This comprehensive approach systematically clarified AWD’s multi-target therapeutic mechanisms against CAG.

RESULTS

AWD notably reduced serum levels of pro-inflammatory cytokines, such as IL-1β, IL-18, tumor necrosis factor-α, and lipopolysaccharide, demonstrating significant statistical differences (all P < 0.01). Additionally, AWD substantially inhibited NLRP3 mRNA expression in gastric mucosal tissue (P < 0.01) and concurrently decreased the protein abundance of NLRP3, IL-1β, and caspase-1 (all P < 0.01), thereby suppressing inflammasome signaling activation. GM analysis indicated that AWD intervention significantly increased the relative abundance of beneficial bacteria. Associated microbial metabolites likely inhibited the NLRP3 inflammasome pathway by modulating immune cell function. Non-targeted metabolomics further indicated that AWD exerted anti-inflammatory effects by regulating critical metabolic pathways, including the Kaposi’s sarcoma-associated herpesvirus infection pathway, autophagy processes, and glycosylphosphatidylinositol-anchor biosynthesis.

CONCLUSION

AWD alleviates the pathological progression of CAG through multi-target synergistic mechanisms. On one hand, AWD directly suppresses gastric mucosal inflammation by inhibiting NLRP3 inflammasome activation. On the other hand, AWD remodels intestinal microbiota-metabolite homeostasis, enhances intestinal barrier function, and regulates mucosal immune responses.

Key Words: Anwei decoction; Chronic atrophic gastritis; Gut microbiota-metabolite axis; NLRP3 inflammasome; Traditional Chinese medicine

Core Tip: Chronic atrophic gastritis (CAG) is a precancerous lesion of the stomach, with a prevalence of 34.7% in China and a risk of gastric cancer that is 4 times that of ordinary people. Existing therapies have high side effects and cannot reverse pathological damage. Traditional Chinese medicine Anwei decoction inhibits NLRP3 inflammasomes, reshapes the balance of microflora, breaks through the limitations of single targets, and provides a new strategy for CAG treatment, which can delay cancer and reverse pathological damage through the regulation of the “microbiota-metabolism-immunity” network.

INTRODUCTION

Chronic atrophic gastritis (CAG) is a precancerous lesion characterized by chronic inflammatory infiltration of the gastric mucosa, glandular atrophy, and intestinal metaplasia[1]. Its global prevalence ranges from 20% to 40%, with East Asia showing notably high rates (34.7% in China)[2,3]. Studies indicate that patients with CAG have a four-fold greater risk of gastric cancer compared to healthy individuals[4]. Gastric cancer remains among the top three cancers globally in terms of mortality rates[5]. Therefore, developing effective interventions to control inflammation and delay carcinogenesis is of significant clinical importance. Current CAG management primarily involves lifestyle interventions and pharmacotherapy, such as proton pump inhibitors (PPIs), histamine H2 receptor antagonists, Helicobacter pylori (H. pylori) eradication therapy, and gastric mucosal protective agents[6]. However, prolonged PPI use may increase the risks of gastric cancer[7] and cardiovascular events[8]. Standard H. pylori eradication regimens face resistance issues that compromise efficacy and may exacerbate complications[9]. Additionally, existing treatments can alleviate symptoms and eliminate H. pylori but cannot reverse established gastric mucosal atrophy or intestinal metaplasia[10]. Some patients continue to experience lesion progression even after successful H. pylori eradication.

Traditional Chinese medicine (TCM) has demonstrated clear advantages in enhancing gastric mucosal defense, improving gastric microcirculation[11], regulating inflammation and immune responses, and modulating intestinal microbiota[12]. Jianwei Xiaoyan Granule has shown efficacy by inhibiting angiogenesis through the hypoxia inducible factor-1α-vascular endothelial growth factor pathway and reducing gastric mucosal inflammation[13]. Additionally, the incidence of adverse reactions associated with TCM is significantly lower compared to Western medicine. The therapeutic benefits of TCM are attributed to its multi-system regulatory effects, which not only enhance the overall functional status of patients but also strengthen immune function. This dual mechanism allows TCM to effectively relieve clinical symptoms and enhance patient tolerance, making it suitable for long-term chronic disease management.

In recent years, TCM’s advantages in treating chronic inflammatory diseases have gained attention. Anwei decoction (AWD), developed by Professor Lin PX, a renowned Chinese medicine physician, exemplifies TCM prescriptions used to treat chronic gastric diseases. Based on the principles of “strengthening the spleen, clearing heat, resolving blood stasis, and promoting tissue regeneration”, AWD comprises ten traditional herbal medicines, including Coptis chinensis, Pinellia ternata, dried ginger, white peony, lily, black medicine, salvia, coix seed, licorice, and woody fragrance. According to TCM theory, CAG primarily results from spleen and stomach qi deficiency combined with dampness, heat, and blood stasis. In this prescription, Astragalus and Coptis chinensis is the principal herb, effectively tonifying qi, strengthening the spleen, and clearing heat and dampness. Pinellia ternata strengthens the spleen and invigorates qi, while salvia promotes blood circulation and eliminates stasis. The balanced combination of these herbs aligns closely with the etiology and pathogenesis of CAG. Previous experiments by our research group demonstrated that AWD improves the intestinal epithelial cell microenvironment by regulating the expression of glucose regulated protein 78, phosphoenolpyruvate carboxy kinase, ATF6, CCAAT/enhancer binding protein homologous protein, and caspase 12 and alleviates gastric mucosal pathology through autophagy[14]. Clinically, AWD is widely employed for gastric diseases, significantly reducing gastric discomfort and promoting mucosal repair. However, its multi-target mechanisms still require systematic clarification.

Gut microbiota (GM) imbalance contributes to the progression of CAG through two primary pathways. Firstly, dysbiosis initiates NLRP3 inflammasome activation, resulting in elevated secretion of interleukin (IL)-1β and IL-18. Secondly, it decreases gastrin production, causing hypoacidity and subsequently creating a self-perpetuating cycle[15]. H. pylori infection activates the NLRP3 inflammasome through stimulation of the Toll-like receptor (TLR) 2/4-nuclear factor kappa-B (NF-κB) signaling axis. Additionally, H. pylori-derived virulence factors provoke the generation of reactive oxygen species (ROS), facilitating inflammasome formation, and exacerbating inflammation via caspase-1-dependent pyroptosis and the secretion of IL-1β and IL-18[16]. Importantly, the NLRP3/IL-1β signaling cascade triggers gastric epithelial cell proliferation through activation of NF-κB and signal transducer and activator of transcription 3 (STAT3) pathways, promoting intestinal metaplasia, which is recognized as a precancerous gastric stomach[17].

Recent studies indicate that TCM treats CAG by regulating immune responses, improving the gastric mucosal environment, and modulating GM[18]. Inflammation plays a crucial role in the initiation and progression of CAG[19]. Activation of the NLRP3 inflammasome is central to CAG pathogenesis, promoting inflammation and gastric epithelial dysplasia via the caspase-1/IL-1β/IL-18 axis[20,21]. H. pylori triggers NLRP3 inflammasome activation through the TLR2/4-NF-κB-ROS pathway, subsequently enhancing inflammation through cleavage of pro-inflammatory cytokines IL-1β and IL-18 and inducing pyroptotic cell death[22]. The NLRP3 inflammasome also activates downstream caspase-1, converting pro-IL-1β into mature IL-1β, which then stimulates the STAT3 and NF-κB pathways in an autocrine or paracrine manner, promoting abnormal gastric epithelial proliferation and subsequent intestinal metaplasia[23]. Moreover, GM is closely associated with host health, and its dysregulation contributes to CAG development[24]. Additionally, altered gut microbial metabolites can influence host physiological and pathological processes.

The integration of multi-omics approaches, including 16S rRNA sequencing, metabolomics, and bioinformatics, has become a powerful paradigm for elucidating the complexity of chronic diseases[25]. In this study, we applied such an integrated strategy to investigate the mechanism of AWD. An interactive regulatory network of “GM-metabolism-immunity” was constructed, enhancing the comprehensiveness and credibility of the findings. Furthermore, AWD significantly reduced gastric mucosal inflammation by modulating multiple metabolic pathways, including the Kaposi’s sarcoma-associated herpesvirus infection pathway, autophagy-related pathways, glycerophospholipid metabolism, and endocrine resistance. These discoveries expand the understanding of inflammatory regulatory pathways and provide novel therapeutic targets and intervention strategies for CAG. AWD also restores gastrointestinal microenvironment balance by targeting the GM-metabolite axis, offering a new perspective of “microecological regulation” for CAG treatment and overcoming limitations of existing therapies involving acid suppression and antibiotics. Additionally, AWD inhibits NLRP3 inflammasome activity, reducing the risk of CAG progression to gastric cancer and preventing malignant transformation. This study integrates the multi-dimensional “microbiota-metabolism-inflammation” regulatory mechanism for the first time, overcoming the limitations of single-target therapy and promoting modern interpretations of TCM theory.

MATERIALS AND METHODS

Laboratory animals

Fifty healthy specific pathogen free (SPF)-grade Wistar rats (25 males and 25 females, 6 weeks old, weight 150 ± 20 g) were provided by Hunan Tianqin Biotechnology Co., Ltd. [production license number: No. SCXK (Xiang) 2019-0013]. All experimental procedures complied rigorously with the Animal Ethics Code for Laboratory Welfare [animal use license No. SYXK (Gui) 2019-0001] and were officially authorized by the Animal Ethics Committee at Guangxi University of Traditional Chinese Medicine (approval No. DW20211117-178).

To minimize the confounding effects of diet and environment on the GM, all experimental rats were housed in the same SPF-grade facility under strictly controlled environmental conditions. They were allowed free access to the same standard feed, and rats from different experimental groups were randomly co-housed in different cages on the same rack to avoid cage effects.

Experimental drugs

AWD consists of the following medicinal herbs: Coptis chinensis (3 g), Pinellia ternata (9 g), ginger (3 g), white peony (10 g), lily (10 g), black medicine (6 g), Salvia miltiorrhiza (10 g), coix seed (15 g), licorice root (5 g), and muxiang (5 g). All herbal ingredients were obtained from the Chinese Medicine Pharmacy at Ruikang Hospital, which is affiliated with Guangxi University of Chinese Medicine.

For AWD preparation, herbs were added to 1000 mL distilled water. The mixture was boiled over high heat and then simmered for 30 minutes. After the initial decoction, the solution was filtered through gauze. The herbs were decocted a second time using the same procedure, and both filtrates were combined. The final solution was concentrated to obtain AWD at 2 g crude drug per mL for subsequent experiments.

The positive control group received Weifuchun capsules (WFC) (Hangzhou Huqingyutang Pharmaceutical Co., Ltd., National Drug Approval No. Z20090697, 0.35 g per capsule). This medication has a well-established efficacy and strong evidence-based support. It was officially approved by the China Food and Drug Administration in 1982 and is currently the only Chinese patent medicine approved for the treatment of precancerous lesions of gastric cancer in China. It has been included in authoritative documents such as the consensus on integrated traditional Chinese and western medicine diagnosis and treatment of chronic gastritis in China and the expert consensus on clinical application of Weifuchun in treating precancerous lesions of gastric cancer, with a clearly defined clinical position.

The animal dosage of AWD in this study (10.7 g/kg) was converted from its clinical human dosage using the body surface area normalization method. The routine daily clinical dosage of AWD is one dose per day, with the total amount of medicinal materials per dose approximately 102 g. Based on an average adult weight of 60 kg, the clinical equivalent dosage is approximately 1.7 g/kg. According to the conversion factor between humans and rats for body surface area from a simple practice guide for dose conversion between animals and humans (typically 6.3), the equivalent dosage for rats was calculated as: Clinical equivalent dosage (g/kg) × 6.3 approximately 10.7 g/kg/day[26]. The animal dosage of WFC (0.43 g/kg) was also calculated using the same method.

Experimental design

The experimental design (Figure 1) included three groups: A normal control group (NC), a model group (MD), an AWD intervention group, and a WFC intervention group. The CAG rat model was established according to a previously described method[14]. After successful model establishment, rats in the AWD group received AWD (10.71 g/kg), WFC (0.43 g/kg) by gavage daily for 8 consecutive weeks. Gastric mucosal tissues from each group were collected the day after the drug intervention. Histopathological examination of gastric mucosal tissues was performed using hematoxylin and eosin (HE) staining. Serum concentrations of inflammatory markers [IL-1β, IL-18, tumor necrosis factor (TNF)-α, and lipopolysaccharide (LPS)] were quantified through enzyme-linked immunosorbent assay and an LPS detection kit. Expression levels of NLRP3 inflammasome messenger RNA (mRNA) were assessed using quantitative real-time polymerase chain reaction (qRT-PCR), while Western blotting (WB) was employed to determine protein expression levels of caspase-1, IL-1β, and NLRP3. Additionally, intestinal contents were collected for 16S rDNA amplicon sequencing and non-targeted metabolomics to comprehensively evaluate AWD’s effects on GM and metabolite profiles.

Figure 1 Experimental design: After 7 days of acclimatization, 10 rats were randomly selected as the normal control group. The remaining rats underwent chronic atrophic gastritis model induction. Upon successful modeling, rats were randomly divided into the model group receiving saline gavage, the Anwei decoction (AWD) intervention group receiving AWD (10.71 g/kg) and the Weifuchun capsules (WFC) intervention group (receiving WFC at 0.43 g/kg) daily for 8 weeks. The normal control group was provided regular feeding and drinking water free. CAG: Chronic atrophic gastritis; MNNG: N-methyl-N’-nitro-N-nitrosoguanidine; HE: Hematoxylin-eosin; IL: Interleukin; TNF: Tumor necrosis factor; LPS: Lipopolysaccharide; ELISA: Enzyme-linked immunosorbent assay; qRT-PCR: Quantitative real-time polymerase chain reaction; mRNA: Messenger RNA.

Animal model construction

After an initial adaptive feeding period of one week, fifty SPF-grade Wistar rats, evenly divided by sex, were randomly allocated into two distinct groups: NC group (n = 10) and MD group (n = 50). The CAG model was established by administering N-methyl-N’-nitro-N-nitrosoguanidine (MNNG) dissolved in drinking water at a concentration of 150 μg/mL for 24 weeks. To control the daily intake of MNNG, rats were individually housed and provided with a graduated drinking bottle. Every day, 40 mL of either MNNG solution (MD group) or plain water (NC group) was provided per rat, and the remaining volume was recorded to monitor consumption. Starting from the 8^th week after modeling initiation, two rats from the MD group were randomly selected every two weeks for gastric histopathological assessment via HE staining to monitor lesion progression. At the 24^th week, gastric mucosal specimens were collected for HE staining, and the pathological status was quantitatively evaluated according to the updated Sydney system. The results indicated successful construction of the CAG rat model. Rats in the MD group scoring ≥ 2 (moderate to severe) in either glandular atrophy or chronic inflammatory infiltration were considered successful models. Based on this criterion, 45 out of 50 rats were successfully modeled, yielding a modeling success rate of 90%. All subsequent interventions and analyses were performed on these 45 successfully modeled rats and the 10 NC rats.

Statistical analysis

All data were presented as mean ± SD from three independent repetitions. SPSS 22.0 (SPSS, Chicago, IL, United States) was used to analyze the data. Data between two groups were analyzed using the t test, and the quantitative data in multiple groups were tested by one-way analysis of variance. Differences were defined as statistically significant at P < 0.05.

RESULTS

Identification of drug components in serum

Ultra-high performance liquid chromatography with quadrupole time-of-flight mass spectrometry analysis identified AWD components in medicated serum. Nineteen serum-migrating compounds (Table 1) were successfully identified by acquiring total ion chromatograms (Figure 2) in positive and negative ion modes. Relevant studies have shown that these compounds, such as isoliquiritigenin and caffeic acid, exhibit significant anti-inflammatory and antioxidant effects[27-30].

Figure 2 Total ion chromatogram for natural product identification in samples. A: Total ion chromatogram (TIC) of blank serum (positive ion mode); B: TIC of blank serum (negative ion mode); C: TIC of serum containing Anwei decoction (AWD) (positive ion mode); D: TIC of serum containing AWD (negative ion mode); E: TIC of AWD (positive ion mode); F: TIC of AWD (negative ion mode).

Table 1 Results of serum composition identification of Anwei decoction.

No	Compound	Reference lon	Formula	RT/m	m/z (actual)	m/z (theoretical)	Annotation MW	Calculate MW	δ/ppm	Sort
1	Isoliquiritigenin[1]	[M+H]+1	C15H12O4	12.241	257.08105	257.08054	256.07356	256.07378	0.00022	Flavonoids
2	Methyl succinic acid[2]	[M-H]-1	C5H8O4	6.224	131.0338	131.03375	132.04226	132.04107	-0.00118	Dicarboxylic acid
3	4-oxoproline[3]	[M-H]-1	C5H7NO3	2.761	128.03412	128.03406	129.04259	129.0414	-0.0012	Amino acids
4	Caffeic acid[4]	[M+H]+1	C9H8O4	18.562	149.02344	179.03419	180.04226	148.01615	-32.02611	Phenolic acids
5	Citric acid[5]	[M-H]-1	C6H8O7	3.046	191.01913	191.01883	192.027	192.02641	-0.00059	Tricarboxylic acids
6	2-phenoxypropanoic acid	[M+FA-H]-1	C9H10O3	9.34	165.05478	165.05458	166.06299	120.05657	-46.00642	Carboxylic acids
7	Nor-9-carboxy-δ9-THC[6]	[M+H-H2O]+1	C21H28O4	14.998	345.20596	345.20358	344.19876	362.20915	18.01039	Cannabinoids
8	(2E)-4-hydroxy-3,7-dimethyl-2,6-octadien-1-ylbeta-D-glucopyranoside	[M-H]-1	C16H28O7	19.81	377.18655	377.18167	332.1835	378.19383	46.01033	Glycoside
9	1-[2-(1,3-Benzodioxol-5-yl)-3-methyl-1-benzofuran-5-yl]-1,2-propanediol	[M+H-H2O]+1	C19H18O5	16.836	309.11191	309.11212	368.10962	368.1121	0.00248	Alcohols
10	Ethyl palmitoleate	[M+H]+1	C18H34O2	20.671	283.26306	283.2634	282.25588	282.25578	-0.0001	Aliphatic
11	Gibberellin A7[7]	[M-H]-1	C19H22O5	15.936	329.13959	329.13947	330.14672	330.14686	0.00014	Diterpenoids
12	(2R)-2-[(2R,5S)-5-[(2S)-2-hydroxybutyl] oxolan-2-yl] propanoic acid	[M+H]+1	C11H20O4	13.463	217.14366	217.14362	216.13616	216.13638	0.00023	Carboxylic acids
13	Avocadyne 1-acetate[8]	[M+H]+1	C19H34O4	19.822	309.24216	277.21643	326.24571	308.23502	-18.01068	Aliphatic
14	4-{[(3S)-3-{5-[4-(dimethyl amino) phenyl]-1,3,4-oxadiazol-2-yl}-1-pyrrolidinyl] methyl} benzoic acid	[M-H]-1	C22H24N4O3	17.359	391.17963	391.17639	392.18484	392.1869	0.00206	Heteroaromatic
15	6-acetylmorphine[9]	[M+H]+1	C19H21NO4	8.764	328.15442	328.15405	327.14706	327.14714	0.00008	Morphinans
16	Methyl 3-(methylamino)-2-[(4-methyl-1,2,3-thiadiazol-5-yl)carbonyl] but-2-enoate	[M-H]-1	C10H13N3O3S	17.614	357.10123	357.10373	255.06776	358.1085	103.04074	Esters
17	4-(dimethylamino)-N-{[(2R,4S,5S)-5-(1-piperidinylmethyl)-1-azabicyclo[2.2.2] oct-2-yl]methyl} benzamide	[M+Na]+1	C23H36N4O	22.502	367.28183	367.28198	384.28891	344.29257	-39.99635	Amides
18	2-hydroxymyristic acid	[M-H]-1	C14H28O3	19.672	243.1964	243.19646	244.20384	244.20367	-0.00017	Long-chain fatty
19	4-(1-adamantyl)-2-methyl-1,3-thiazole	[M+H]+1	C14H19NS	8.375	234.13382	234.1333	233.12382	233.12635	0.00253	Thiazoles

The data in the third column of the table represents the ion mode of the compound. Determining the ion mode not only improves the accuracy and reliability of identification, but also provides adduct ion information, assists in molecular formula deduction, and obtains more comprehensive secondary mass spectrometry fragmentation information. MW: Molecular weight.

AWD reduces inflammatory infiltration in gastric mucosa

Significant differences in gastric mucosal pathology were observed among the groups through HE staining (Figure 3). Gastric mucosal histopathology was scored according to the updated Sydney system; results are presented in Figure 3 and Table 2. The MD group exhibited pronounced inflammatory cell infiltration, neutrophil infiltration, and glandular atrophy or reduction in the lamina propria. In contrast, pathological changes in gastric mucosa from the AWD and WFC groups were significantly alleviated (P < 0.01), with the AWD group demonstrating greater improvement.

Figure 3 Hematoxylin-eosin-stained histopathological images of gastric mucosa from the normal control group, model group, Anwei decoction group, and Weifuchun capsules intervention groups. Detachment of gastric mucosal epithelial cells (yellow arrow); dilation of gastric glands (blue arrow); atrophy or reduction of gastric glands (black arrow); hyperplasia of fibrous connective tissue (orange arrow); inflammatory cell infiltration (purple arrow). The black square indicates the location of the magnified field. NC: Normal control; MD: Model; AWD: Anwei decoction; WFC: Weifuchun capsules.

Table 2 Quantitative histopathological scoring of gastric mucosa in each group, mean ± SD.

Group	Chronic inflammation	Active inflammation	Glandular atrophy	Intestinal epithelial metaplasia
NC	0 ± 0	0 ± 0	0 ± 0	0 ± 0
MD	2.17 ± 0.41b	0.33 ± 0.52b	1.17 ± 0.41b	0.33 ± 0.52b
AWD	0.17 ± 0.37b	0 ± 0b	0 ± 0b	0 ± 0b
WFC	0.33 ± 0.52a	0.33 ± 0.52	0 ± 0b	0 ± 0b

^aP < 0.05.

^bP < 0.01.

NC: Normal control; MD: Model; AWD: Anwei decoction; WFC: Weifuchun capsules.

AWD reduces serum levels of inflammatory factors IL-1β, IL-18, TNF-α, and LPS

The experimental results showed that, compared to the NC group, MD rats exhibited significantly increased inflammatory cytokine levels (P < 0.01). Following administration of AWD and WFC, the elevated cytokine levels in MD rats decreased significantly compared to untreated MD rats (P < 0.01). Notably, the AWD group demonstrated a more pronounced effect (Table 3).

Table 3 Levels of inflammatory factors in gastric mucosa across experimental groups, mean ± SD.

Group	IL-1β (pg/mg)	TNF-α (pg/mg)	IL-18 (pg/mg)	LPS (pg/mg)
NC	49.36 ± 7.36	86.46 ± 14.13	73.37 ± 10.74	348.38 ± 62.7
MD	122.7 ± 7.23b	185.97 ± 18.5b	143.86 ± 21.01b	719.75 ± 87.54b
AWD	62.69 ± 9.14b	111.4 ± 19.05b	90.99 ± 19.14b	459.82 ± 99.02b
WFC	80.08 ± 13.25b	140.29 ± 13.05b	120.98 ± 21.71a	579.85 ± 74.8a

^aP < 0.05.

^bP < 0.01.

NC: Normal control; MD: Model; AWD: Anwei decoction; WFC: Weifuchun capsules; IL: Interleukin; TNF: Tumor necrosis factor; LPS: Lipopolysaccharide.

AWD decreases NLRP3 mRNA expression in gastric mucosa

QRT-PCR analysis indicated significantly higher NLRP3 mRNA expression in the gastric mucosa of MD rats (7.89 ± 0.95) compared to the NC group (1.02 ± 0.12, P < 0.01, Figure 4). In contrast, NLRP3 mRNA expression was significantly reduced after AWD and WFC treatment (P < 0.01), with the AWD group showing a more pronounced effect. These results demonstrate that AWD markedly suppresses the gene expression of the NLRP3 inflammasome.

Figure 4 NLRP3 messenger RNA expression levels in gastric mucosa. Compared with the normal control group and compared with the model group. ^cP < 0.001. mRNA: Messenger RNA; NC: Normal control; MD: Model; AWD: Anwei decoction; WFC: Weifuchun capsules.

AWD reduces protein expression of caspase-1, IL-1β, and NLRP3

WB analysis was employed to measure protein expression of IL-1β, NLRP3, and caspase-1, revealing notably increased protein abundance in the MD group relative to NC animals (P < 0.01, Figure 5). Subsequent AWD and WFC treatments substantially reduced the expression of these proteins compared with untreated MD rats (P < 0.01). Notably, the AWD group showed a more pronounced effect.

Figure 5 Western blotting analysis. A: Western blot detection results of NLRP3 inflammasome-related protein expression levels in each group of cells; B: Protein expression levels of caspase-1; C: Protein expression levels of interleukin-1β; D: Protein expression levels of NLRP3. Compared with the normal control group and compared with the model group. ^cP < 0.001. IL: Interleukin; NC: Normal control; MD: Model; AWD: Anwei decoction; WFC: Weifuchun capsules.

Analysis of intestinal microbial diversity

Intestinal content samples collected from 18 rats across the NC, MD, and AWD groups were analyzed by sequencing the 16S rRNA gene amplicon. After noise reduction, amplicon sequence variants (ASVs) were identified, resulting in 1162 unique ASVs across the three groups. Specifically, the ASV numbers detected were 1245 in the NC group, 1123 in the MD group, and 916 in the AWD group. The Venn diagram (Figure 6A) clearly illustrated notable differences in microbial composition among the three groups.

Figure 6 Analysis of intestinal microbial diversity. A: Venn diagram; B: Heatmap integrated with Metastats significance markers, displaying differences in microbial composition determined by 16S rRNA amplicon sequencing; C: Non-metric multidimensional scaling analysis results. NC: Normal control; MD: Model; AWD: Anwei decoction; NMDS: Non-metric multidimensional scaling.

Species exhibiting significant differences among the groups were further analyzed. Using species abundance data at different taxonomic levels, the Metastats method was employed for hypothesis testing, and the resulting P values were adjusted to obtain Q values. Species significantly differing in abundance were identified based on P values and Q values, and integrated into a quantitative heatmap along with Metastats markers (Figure 6B). The results revealed considerable variation in the relative abundance and distribution patterns of microbial species among the three groups, suggesting that the structure and function of the microbial communities were altered.

Non-metric multidimensional scaling analysis (Figure 6C), yielded a stress value of less than 0.2, confirming that the analysis reliably depicted differences among samples.

Structure and changes in gut microbes

Composition of GM: Following sequencing data processing of intestinal contents from rats in the NC, MD, and AWD groups, species annotation based on ASV analysis was completed. The top 10 microbial taxa at the phylum and genus levels were selected according to relative abundance, and histograms depicting species abundance were constructed to visualize major bacterial community changes and the influence of AWD on microbiota composition.

The GM composition at the phylum level is illustrated in Figure 7A. Six dominant bacterial phyla were observed, including Spirochaetota, Bacteroidota, Firmicutes, Proteobacteria, Euryarchaeota, and Actinobacteria. The MD rats exhibited significantly increased levels of Spirochaetota relative to those in the NC group (P < 0.05). Conversely, AWD administration resulted in a notable decrease in the relative abundance of this phylum compared with untreated MD rats (P < 0.05).

Figure 7 Structure and changes in gut microbes. A: Relative abundance of microbial taxa at the phylum level; B: Relative abundance of microbial taxa at the genus level; C: Linear discriminant analysis score distribution indicating microbial taxa significantly differing among groups; D: Taxonomic cladogram illustrating taxa with significant differences among groups. NC: Normal control; MD: Model; AWD: Anwei decoction; LDA: Linear discriminant analysis.

Figure 7B illustrates the bacterial composition at the genus level, predominantly consisting of Ligilactobacillus, Romboutsia, Bacteroides, Lactobacillus, Ruminococcus, and Treponema. The abundance of Bacteroides significantly declined in MD rats relative to NC animals (P < 0.05). However, AWD intervention markedly restored the abundance of this genus compared to MD rats (P < 0.05). This analytical approach successfully underscored the distinctions in gut microbial communities among various treatment conditions, offering compelling evidence that AWD effectively modulates intestinal microbiota composition.

Differential microbial species among groups: The linear discriminant analysis effect size (LEfSe) method was subsequently applied to investigate differences in GM composition among the CN, MD, and AWD groups, as illustrated in Figure 7C and D. LEfSe integrates linear discriminant analysis with nonparametric statistical testing to effectively determine microbial taxa that exhibit significant variations among the studied groups.

Results indicated that at the family level, the abundance of Bacteroidaceae and unidentified Christensenellaceae was significantly higher in the NC group than in other groups (P < 0.05). In the MD group, the abundance of Escherichia coli (species and genus level unidentified bacteria) significantly increased compared with other groups (P < 0.05). In the AWD group, the abundance of Lactobacillus murinus (species level) and Treponema (genus level) was significantly higher compared to other groups (P < 0.05).

These results suggest notable differences in microbial abundance among treatment groups at specific taxonomic levels, indicating potential biomarkers for differentiating microbial community structures among groups.

Differential analysis of metabolic profiles

To clearly elucidate metabolic changes among the NC, MD, and AWD groups, primary and secondary metabolites in samples were comprehensively identified using ultra-high-performance liquid chromatography coupled with tandem mass spectrometry combined with broadly targeted metabolomics. A total of 6429 metabolites were detected, comprising 4385 positive-ion metabolites and 2044 negative-ion metabolites.

Among positive-ion metabolites, the top ten metabolite categories were amino acids and derivatives (20.64%), benzene and substituted derivatives (16.47%), heterocyclic compounds (10.73%), aldehydes, ketones, and esters (8.43%), fatty acids (4.68%), hormones and related compounds (4.09%), organic acids and derivatives (7.51%), glycerophospholipids (3.56%), nucleotides and derivatives (2.15%), and terpenoids (1.91%) (Figure 8A).

Figure 8 Circular diagrams illustrating metabolite category composition. A: Positive-ion metabolites; B: Negative-ion metabolites. FA: Fatty acids; GL: Glycerolipids; GP: Glycerophospholipids; SP: Sphingolipids; ST: Steroid.

Among negative-ion metabolites, the top ten categories were benzene and substituted derivatives (16.18%), amino acids and derivatives (12.92%), heterocyclic compounds (12.55%), organic acids and derivatives (11.44%), aldehydes, ketones, and esters (8.87%), hormones and related compounds (6.4%), fatty acids (6.07%), alcohols and amines (4.42%), nucleotides and derivatives (2.85%), and carbohydrates and derivatives (1.82%) (Figure 8B).

Differential metabolite analysis showed significant changes in 132 metabolites between the MD and NC groups, with 34 upregulated and 98 downregulated (Figure 9A). Between the AWD and MD groups, 542 metabolites differed significantly, including 410 upregulated and 132 downregulated metabolites (Figure 9B). These results indicated distinct metabolic profiles across groups, highlighting unique metabolic signatures within each group.

Figure 9 Volcano plots of differential metabolites. A: Model group vs normal control group; B: Anwei decoction group vs model group. Each dot represents a metabolite, with green dots indicating downregulated metabolites, red dots indicating upregulated metabolites, and gray dots indicating metabolites with no significant difference. VIP: Variable importance in projection.

Pathway analysis of differential metabolites

To elucidate biological functions and pathways associated with significantly different metabolites (MD vs NC groups and AWD vs MD groups), differential metabolites were annotated and enriched using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database.

The differential metabolites identified between the MD and NC groups were annotated into various KEGG pathways, including 30 pathways under metabolism, 5 related to organismal systems, 6 under human diseases, 3 classified within environmental information processing, and 3 under cellular processes. Among these, 22 differential metabolites were annotated in the metabolism pathways, accounting for 91.67% of total annotated metabolites (Figure 10A). These metabolites were primarily enriched in pathways involving amino sugar and nucleotide sugar metabolism, nucleotide sugar biosynthesis, nucleotide metabolism, pyrimidine metabolism, and general metabolic pathways (Figure 10B).

Figure 10  Kyoto Encyclopedia of Genes and Genomes analysis. A: Kyoto Encyclopedia of Genes and Genomes (KEGG) classification of differential metabolites (model group vs normal control group); B: KEGG enrichment analysis (model group vs normal control group); C: KEGG classification of differential metabolites (Anwei decoction group vs model group); D: KEGG enrichment analysis (Anwei decoction group vs model group). KEGG: Kyoto Encyclopedia of Genes and Genomes.

Differential metabolites distinguishing the AWD group from the MD group were annotated into 44 metabolism pathways, 14 organismal systems pathways, 11 human diseases pathways, 3 environmental information processing pathways, and 6 cellular processes pathways. Notably, 66 differential metabolites (82.5% of the total annotated metabolites, Figure 10C) were classified into metabolism pathways. These metabolites predominantly enriched pathways including glycerophospholipid metabolism, retrograde endocannabinoid signaling, autophagy pathways (animal-specific and other types), endocrine resistance, glycosylphosphatidylinositol-anchor biosynthesis, and Kaposi sarcoma-associated herpesvirus infection pathways (Figure 10D).

Correlation analysis

To systematically clarify possible interactions between host metabolites and GM, correlation analyses of the top 20 significantly differential metabolites and microorganisms were conducted. Heatmap analysis revealed the association patterns between key metabolites and GM.

As shown in Figure 11A, calcium carbonate in the NC and MD groups positively correlated with Bacteroides, Alloprevotella, Butyricimonas, Turicibacter, Rothia, Paraprevotella, and Rikenella (P < 0.05), and negatively correlated with unidentified Enterobacteriaceae, unidentified Rhodospirillales, and Phascolarctobacterium (P < 0.05). Thymidine 5’-diphosphate (dTDP)-galactose negatively correlated with Monoglobus, unidentified Enterobacteriaceae, unidentified Rhodospirillales, and Phascolarctobacterium (P < 0.05).

Figure 11  Correlation analysis. A: Correlation heatmap of top 20 differential metabolites and microorganisms (model group vs normal control group); B: Correlation heatmap of top 20 differential metabolites and microorganisms (Anwei decoction group vs model group). In the correlation heatmap, microorganisms are represented by rows, while metabolites correspond to columns. Positive correlations are indicated by red ellipses, whereas negative correlations are shown with blue ellipses. Ellipse narrowness reflects correlation strength, with narrower ellipses denoting stronger associations. Cells lacking ellipses represent correlations that were not statistically significant (P > 0.05).

As illustrated in Figure 11B, Alloprevotella in the AWD group (compared to MD group) positively correlated with cucurbitacin D, dTDP-galactose, histidine (His)-serine (Ser)-leucine (Leu)-Ser-glutamate (Glu), 17,23-epoxy-29-hydroxy-27-norlanost-8-ene-3,15,24-trione, tomenin, thyroxine sulfate, and 5alpha-ergosta-7,22-diene-3beta,5-diol (P < 0.05). Conversely, Alloprevotella negatively correlated with halcinonide, PE-NMe2 [18:0/18:1(9Z)], isoleucine-aspartate-Leu-arginine, and (+)-penbutolol (P < 0.05). Anaerovorax positively correlated with 8-iso-16-cyclohexyl-tetranor-prostaglandin E2 (P < 0.05) and negatively correlated with (+)-penbutolol (P < 0.05).

These positive and negative correlations indicate that certain microorganisms may promote or inhibit specific metabolite formation or accumulation. They also illustrate the complexity of the gut microbial community, potentially linked to microbial interactions, host physiological states, and environmental factors.

DISCUSSION

This study is the first to elucidate how AWD improves CAG by inhibiting the NLRP3 inflammasome and regulating GM. Experimental results confirmed that AWD reduces gastric mucosal inflammation by modulating the NLRP3/caspase-1/IL-1β signaling pathway. Additionally, AWD regulated GM composition and metabolic function, restored intestinal microecological balance, and effectively treated CAG. This finding offers a novel strategy for treating CAG via the “gut-stomach axis” and serves as a paradigm for investigating multi-target mechanisms of TCM formulations, holding significant clinical translational potential.

In this study, high-resolution mass spectrometry was innovatively employed to systematically analyze AWD and drug-containing serum. Prototype components, such as isoliquiritigenin, caffeic acid, citric acid, and (2E)-4-hydroxy-3,7-dimethyl-2,6-octadien-1-yl-β-D-glucopyranoside, along with their metabolic transformations, were identified in vivo. Metabolites, including calcium carbonate, dTDP-galactose, cucurbitacin D, His-Ser-Leu-Ser-Glu, 17,23-epoxy-29-hydroxy-27-norlanost-8-ene-3,15,24-trione, tomenin, thyroxine sulfate, and 5alpha-ergosta-7,22-diene-3beta,5-diol, may exert anti-inflammatory effects through the “microbiota-metabolite-immunity” axis. This provides a biochemical basis for elucidating AWD’s multi-target mechanisms.

From a TCM perspective, the compatibility of isoliquiritigenin, caffeic acid, and citric acid in AWD embodies the principles of “simultaneous treatment of spleen and stomach” and “harmonization of liver and stomach”[31,32].

Simultaneous treatment of spleen and stomach: Isoliquiritigenin repairs gastric mucosa by strengthening the spleen and invigorating qi, thus enhancing digestive function. Caffeic acid clears heat and dampness, and citric acid facilitates digestion by promoting fluid circulation. Their combined action strengthens the spleen-stomach axis and restores the physiological function of “spleen rising and stomach descending”.

Harmonization of liver and stomach: Isoliquiritigenin harmonizes liver and spleen, inhibiting mucosal damage caused by liver stagnation. Caffeic acid soothes the liver and promotes blood circulation, blocking pathological interactions between liver qi and stomach dysfunction. Citric acid regulates digestive juice secretion, relieving liver-stomach disharmony. Their synergistic effect aligns with the TCM principle “treating the stomach without neglecting the liver”. This compatibility reflects classical teachings such as “understanding spleen function first when encountering liver disorders”, as stated in “treatise on typhoid fever”, and “internal damage to the spleen-stomach leading to various diseases”, from Li DY’s “treatise on the spleen and stomach”.

From a modern medical viewpoint, isoliquiritigenin, caffeic acid, and citric acid exert anti-inflammatory effects through multiple mechanisms, influencing GM and inflammatory responses. Isoliquiritigenin increases probiotics (Akkermansia), enhances short-chain fatty acid (SCFA) production, and activates G protein-coupled receptors, reducing inflammation and lipid accumulation[33]. Isoliquiritigenin also inhibits NLRP3 inflammasome activation by strengthening the intestinal barrier, reducing microbial metabolites (e.g., LPS) entering circulation[33,34]. Additionally, isoliquiritigenin’s metabolites formed through oxidation and reduction possess anti-inflammatory activity[35].

Caffeic acid suppresses inflammatory cytokines (TNF-α, IL-6) and enzymes (cyclooxygenase-2) by inhibiting the NF-κB signaling pathway[36]. Its antioxidant activity prevents lipid peroxidation and inflammation from oxidative stress. Indirectly, caffeic acid inhibits NLRP3 inflammasome activation by downregulating NF-κB signaling[37]. In vivo, caffeic acid converts into ferulic acid or other bioactive metabolites via methylation or oxidation[38].

Citric acid reduces NLRP3 inflammasome activation by regulating intracellular metabolism, reducing energy supply to inflammatory cells[39]. It primarily metabolizes through the tricarboxylic acid cycle, yielding carbon dioxide and water while providing energy[40]. In the gut, citric acid promotes Bifidobacterium growth, enhancing SCFA production. Thus, citric acid indirectly influences GM and metabolites, supporting intestinal health[41].

In recent years, considerable attention has been focused on elucidating how the NLRP3 inflammasome becomes activated in numerous chronic inflammatory conditions. The activation of this inflammasome can trigger inflammatory cascades, subsequently damaging epithelial cells of the gastric mucosa, disrupting cellular homeostasis, and ultimately contributing to CAG[42,43]. In this study, multiple inflammatory cytokines and NLRP3 inflammasome-related proteins were quantified using enzyme-linked immunosorbent assay and WB techniques. Results indicated that AWD significantly inhibited these inflammatory markers. This finding is consistent with previous studies on NLRP3 inflammasome activation in various chronic inflammatory diseases[43-45]. These results further support targeting NLRP3 inflammasome activation as a potential therapeutic strategy for CAG.

Additionally, to explore interactions between GM and host health, this study employed both 16S rDNA amplicon sequencing and non-targeted metabolomic profiling. The regulatory effects of AWD on intestinal microbiota composition and metabolic activity in rats with CAG were comprehensively evaluated from diverse analytical dimensions. Our findings that AWD modulates the GM and associated metabolites suggest a potential crosstalk along the gut-stomach axis. The human microbiota is a complex ecosystem whose role in biomedical sciences and disease therapy is increasingly recognized[46]. However, further studies are needed to fully elucidate the specific microbial alterations induced by AWD. It could induce GM remodeling by increasing beneficial bacteria (Bacteroides, Firmicutes, Ligilactobacillus, and Lactobacillus) and decreasing harmful bacteria (Spirochaetota and Proteobacteria). Moreover, metabolomic analysis indicated significant enrichment of differential metabolites between the AWD and MD groups in metabolic pathways closely associated with various physiological and pathological processes. For example, the Kaposi sarcoma-associated herpesvirus infection pathway is implicated in multiple human cancers[47]. Autophagy-related pathways are essential for maintaining cellular homeostasis, stress responses, and disease progression[48]. Enrichment of the glycerophospholipid metabolism pathway may relate to cell membrane metabolism and signaling[49], while the endocrine resistance pathway is involved in mechanisms underlying resistance to endocrine therapies[50]. Changes in metabolites between groups are closely associated with disease onset, progression, and therapeutic responses. These findings suggest AWD may alleviate inflammation by modulating multiple metabolic pathways.

Compared with omeprazole, a PPI commonly prescribed in Western medicine for gastritis, AWD has demonstrated unique advantages in treating CAG, particularly in inhibiting NLRP3 inflammasome activation and regulating GM. Omeprazole alleviates symptoms such as gastric pain and acid reflux by suppressing gastric acid secretion but has limited effects on gastric mucosal repair and inflammation. AWD can not only relieve symptoms but also promote mucosal repair and reduce inflammation through multi-target regulatory mechanisms. Some patients treated with omeprazole remain symptomatic and experience high recurrence rates[51]. Due to its multi-target regulation, AWD fundamentally improves pathological states and reduces recurrence risk. Long-term omeprazole use may lead to drug resistance and gastric wall thickening[52], while AWD follows the holistic principles of TCM and does not induce resistance. These characteristics suggest broad therapeutic potential for AWD in treating CAG.

While this study successfully delineates a novel GM metabolite NLRP3 axis through which AWD alleviates CAG, it has several limitations. First, the study was conducted in animal models, whose physiological and pathological mechanisms differ from those in humans; thus, clinical trials are essential to validate these findings. Second, although we described a novel GM metabolite NLRP3 axis for AWD, the precise molecular targets within gastric mucosa and the specific contributions of individual AWD components remain incompletely understood. The mechanisms discussed for constituents like isoliquiritigenin primarily rely on previous pharmacological reports, and direct causal relationships within the AWD CAG context require further confirmation. Additionally, other unmeasured confounding factors may have influenced the animal experiments. However, this study rigorously standardized diet, housing conditions, and the genetic background of animals, supporting the conclusion that observed microbial changes primarily resulted from AWD intervention.

To address these gaps, future research should employ a multifaceted strategy. Initially, transcriptomic and proteomic analyses of gastric tissues should identify key differentially expressed genes and proteins. Subsequent validation using gene knockout models or specific inhibitor interventions would conclusively establish causal molecular pathways and direct targets, shifting analysis from correlation to causation. Furthermore, constructing a network pharmacology model integrating AWD’s bioactive components, predicted targets, and associated signaling pathways can provide a more comprehensive, system-level perspective of its mechanism of action. Finally, long-term studies evaluating the durability and safety of AWD treatment are crucial to fully understand its effects on gastric mucosal pathology, inflammatory factor regulation, and GM stability, thereby providing a robust foundation for clinical application. Future research should also utilize germ-free mice colonized with specific microbiota or perform fecal microbiota transplantation experiments to establish direct causal evidence while minimizing confounding factors.

CONCLUSION

AWD alleviates the pathological progression of CAG through multi-target synergistic mechanisms. On one hand, AWD directly suppresses gastric mucosal inflammation by inhibiting NLRP3 inflammasome activation. On the other hand, AWD remodels intestinal microbiota-metabolite homeostasis, enhances intestinal barrier function, and regulates mucosal immune responses.