**Background:** Postoperative cognitive dysfunction (POCD) is a common complication in elderly patients, affecting 10%–60% within the first week after surgery, with one-third experiencing long-term dysfunction. The gut-brain axis is increasingly recognized as a potential mediator, but the specific microbial species and metabolic pathways involved in POCD remain poorly understood. This study aimed to investigate the role of gut microbiota and their metabolites in POCD using an aged mouse model.
**Methods:** Eighteen-month-old male C57BL/6 mice were randomly assigned to a control group (n=6, air treatment) or a multiple anesthesia/surgery (AS) group (n=5). The AS group received 4% isoflurane induction, followed by 1.5% isoflurane maintenance for 2 hours with a laparotomy, repeated for 3 consecutive days. Cognitive function was assessed using the Morris Water Maze Test (MWMT) and depressive behavior via the Open Field Test (OFT) 72 hours after the last anesthesia. Synaptic damage was evaluated by Western blot for PSD95 in the hippocampus and prefrontal cortex. Neuroinflammation was measured by qPCR for IL-6 and IL-1β in the hippocampus. Fecal samples were collected for metagenomic sequencing (NovaSeq 6000) and non-targeted metabolomics (UPLC-MS/MS). Bioinformatic analyses included alpha/beta diversity, LEfSe, KEGG functional annotation, and Spearman correlation networks.
**Key Results:** AS-exposed aged mice showed significantly longer escape latency in the MWMT (Fig. 1C, D) and reduced time in the target quadrant (Fig. 1E), with no difference in swimming speed (Fig. 1F). In the OFT, AS mice had decreased central activity duration (Fig. 2C), reduced total distance (Fig. 2B), slower average speed (Fig. 2D), and increased quiescent time (Fig. 2E). Western blot revealed reduced PSD95 in the hippocampus and prefrontal cortex (Fig. 2F–H). Hippocampal IL-6 and IL-1β levels were elevated (Fig. 2I). Metagenomic analysis showed no significant difference in alpha diversity (Fig. 3A–D), but beta diversity (PCA, NMDS) was significantly different between groups (ANOSIM R=0.39, P=0.010; Fig. 3G, H). At the phylum level, AS mice had higher Bacteroidetes, Proteobacteria, Verrucomicrobia, and lower Firmicutes (Fig. 4A). LEfSe identified 53 differentially abundant taxa (LDA>3, P<0.05), with AS mice enriched in Bacteroidetes, Rikenellaceae, Akkermansiaceae, and Mucispirillum, while controls had more Lachnospiraceae (Fig. 5). At the species level, AS mice had higher s_Bacteroidales_unclassified, s_Muribaculaceae_bacterium_Isolate-104_HZI, s_Bacteroides_acidifaciens, and s_Mucispirillu_schaedleri, and lower s_Eubacterium_rectale, s_Lachnospiraceae_bacterium_A2, and s_Ruminococcus_sp_1xD21-23 (Fig. 4C). KEGG functional analysis showed enrichment in metabolic pathways including amino acid, purine, and carbon metabolism (Fig. 6). Metabolomics identified 2,723 features in positive ion mode and 724 in negative ion mode, with 112 distinct metabolites differentiating groups (Fig. 9). Key down-regulated metabolites included choline, spermidine, thiamine, and multiple long-chain polyunsaturated fatty acids (LCPUFAs) such as linoleic acid and 9,10-dihydroxy-12Z-octadecenoic acid. KEGG enrichment of differential metabolites highlighted purine metabolism, amino acid metabolism, linoleic acid metabolism, and glycerophospholipid metabolism (Fig. 9E, F). Spearman correlation analysis revealed that beneficial bacteria (e.g., Lachnospiraceae, Eubacterium, Butyrivibrio) were positively correlated with spermidine, phosphocholine, acylcarnitines, LCPUFAs, and thiamine, and negatively correlated with escape latency and quiescent time (Fig. 10). Conversely, Rikenellaceae, Akkermansiaceae, and Mucispirillum showed opposite correlations.
**Clinical Implications:** This study provides evidence that anesthesia/surgery in aged mice induces gut microbiota dysbiosis characterized by a reduction in anti-inflammatory and short-chain fatty acid-producing bacteria (e.g., Lachnospiraceae, Eubacterium) and an increase in potentially pro-inflammatory taxa (e.g., Bacteroidetes, Akkermansia). These microbial changes are associated with decreased levels of neuroprotective metabolites (spermidine, choline, thiamine, LCPUFAs), which may contribute to synaptic damage, neuroinflammation, and cognitive/behavioral deficits. The findings suggest that modulation of the gut microbiota—through probiotics, prebiotics, or dietary interventions—could be a potential therapeutic strategy to prevent or mitigate POCD in elderly patients. However, the study is limited by small sample sizes and the inherent differences between animal models and human disease, warranting further clinical validation.