Radiation enteritis (RE) is a common complication in cancer patients receiving radiotherapy. Although studies have shown the changes of this disease at clinical, pathological and other levels, the dynamic characteristics of local microbiome and metabolomics are hitherto unknown. We aimed to examine the multi-omics features of the gut microecosystem, determining the functional correlation between microbiome and lipid metabolites during RE activity. By delivering single high-dose irradiation, a RE mouse model was established. High-throughput 16S rDNA sequencing and global lipidomics analysis were performed to examine microbial and lipidomic profile changes in the gut microecosystem. Spearman correlation analysis was used to determine the functional correlation between bacteria and metabolites. Clinical samples were collected to validate the above observations. During RE activity, the intestinal inflammation of the mice was confirmed by typical signs, symptoms, imaging findings and pathological evidences. 16S datasets revealed that localized irradiation dramatically altered the gut microbial composition, resulting in a decrease ratio of Bacteroidetes to Firmicutes. Lipidomics analysis indicated the remarkable lipidomic profile changes in enteric epithelial barrier, determining that glycerophospholipids metabolism was correlated to RE progression with the highest relevance. Spearman correlation analysis identified that five bacteria-metabolite pairs showed the most significant functional correlation in RE, including Alistipes-PC(36:0e), Bacteroides-DG(18:0/20:4), Dubosiella-PC(35:2), Eggerthellaceae-PC(35:6), and Escherichia-Shigella-TG(18:2/18:2/20:4). These observations were partly confirmed in human specimens. Our study provided a comprehensive description of microbiota dysbiosis and lipid metabolic disorders in RE, suggesting strategies to change local microecosystem to relieve radiation injury and maintain homeostasis.

Overload of palmitic acids is linked to the dysregulation of ceramide metabolism in nonalcoholic steatohepatitis (NASH), and ceramides are important bioactive lipids mediating the lipotoxicity of palmitic acid in NASH. However, much remains unclear about the role of ceramidases that catalyze the hydrolysis of ceramides in NASH. By analyzing the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO) database, we found that alkaline ceramidase 3 (ACER3) is upregulated in livers of patients with NASH. Consistently, we found that Acer3 mRNA levels and its enzymatic activity were also upregulated in mouse livers with NASH induced by a palmitate-enriched Western diet (PEWD). Moreover, we demonstrated that palmitate treatment also elevated Acer3 mRNA levels and its enzymatic activity in mouse primary hepatocytes. In order to investigate the function of Acer3 in NASH, Acer3 null mice and their wild-type littermates were fed a PEWD to induce NASH. Knocking out Acer3 was found to augment PEWD-induced elevation of C_18:1-ceramide and alleviate early inflammation and fibrosis but not steatosis in mouse livers with NASH. In addition, Acer3 deficiency attenuated hepatocyte apoptosis in livers with NASH. These protective effects of Acer3 deficiency were found to be associated with suppression of hepatocellular oxidative stress in NASH liver. In vitro studies further revealed that loss of ACER3/Acer3 increased C_18:1-ceramide and inhibited apoptosis and oxidative stress in mouse primary hepatocytes and immortalized human hepatocytes induced by palmitic-acid treatment. These results suggest that ACER3 plays an important pathological role in NASH by mediating palmitic-acid-induced oxidative stress.

Post-hepatectomy liver dysfunction is a life-threatening morbidity that lacks efficient therapy. Bioactive lipids involved in macrophage polarization crucially regulate tissue injury and regeneration. Herein, we investigate the key bioactive lipids that mediate the cytotherapeutic potential of polarized-macrophage for post-hepatectomy liver dysfunction. Untargeted lipidomics identified elevation of ceramide (CER) metabolites as signature lipid species relevant to M1/M2 polarization in mouse bone-marrow-derived-macrophages (BMDMs). M1 BMDMs expressed a CER-generation-metabolic pattern, leading to elevation of CER; M2 BMDMs expressed a CER-breakdown-metabolic pattern, resulting in upregulation of sphingosine-1-phosphate (S1P). After infusing M1- or M2-polarized BMDMs into the mouse liver after hepatectomy, we found that M1-BMDM infusion increased M1 polarization and CER accumulation, resulting in exaggeration of hepatocyte apoptosis and liver dysfunction. Conversely, M2-BMDM infusion enhanced M2 polarization and S1P generation, leading to alleviation of liver dysfunction with improved hepatocyte proliferation. Treatment of exogenous CER and S1P or inhibition CER and S1P synthesis by siRNA targeting relevant enzymes further revealed that CER induced apoptosis while S1P promoted proliferation in post-hepatectomy primary hepatocytes. In conclusion, CER and S1P are uncovered as critical lipid mediators for M1- and M2-polarized BMDMs to promote injury and regeneration in the liver after hepatectomy, respectively. Notably, the upregulation of hepatic S1P induced by M2-BMDM infusion may have therapeutic potential for post-hepatectomy liver dysfunction.

Post-hepatectomy liver failure (PHLF), defined as acute liver failure following hepatectomy, remains a major complication for postoperative mortality lacking early detection approaches. This study aimed to leverage cutting-edge artificial intelligence (AI) techniques for extensive temporal feature analysis using perioperative data, to advance the detection of PHLF to the first 24 h after surgery. This nationwide multicenter retrospective study was conducted with a total of 1832 patients across six geographically diverse hospitals in China. This China cohort was divided into 681 cases for training the deep-learning model and 1151 cases for validation. Perioperative electronic health record (EHR) data were collected for each patient, including the basic characteristics and preoperative, intraoperative, and early postoperative factors within the first 24 h after surgery. 242 cases from the Medical Information Mart for Intensive Care (MIMIC)-IV database, predominantly comprising Caucasian patients with limited perioperative EHR data, were included to assess robustness in Western populations. PHLF was diagnosed by concurrent elevated prothrombin time/INR and hyperbilirubinemia on or after postoperative day 5 and graded according to the International Study Group of Liver Surgery criteria. The proposed algorithm employed a powerful foundation model (Bio-Clinical Bidirectional Encoder Representation from Transformers) and a context-aware transformer module to perform in-depth temporal feature investigation of perioperative data to enable early detection of PHLF. Our approach was compared with state-of-the-art machine learning and deep learning methods. T-distributed stochastic neighbor embedding and Shapley additive explanation analysis were conducted for model interpretability. The versatility of our model for routine clinical practice was systematically evaluated. This study is registered with ClinicalTrials.gov (NCT06532214). Our model demonstrated high accuracy in detecting PHLF within the first 24 h after surgery, achieving an AUC of 0.952 in internal validation and 0.884 in external validation of the China cohort, outperforming other competing algorithms. In the MIMIC-IV cohort with challenging incomplete EHR data, our model had an AUC of 0.654, which was still superior to alternative algorithms, indicating generalization potential for the Western population. Interpretability analysis showed that our model could effectively encode all perioperative factors into discriminative high-dimensional feature embeddings with temporal correlations into consideration. Importantly, our model holds clinically acceptable interpretability that the patients with model-detected PHLF are likely experiencing the onset of liver failure comprehensively driven by reduced liver volume, extensive surgical injury, and underlying liver disease. Further evaluation demonstrated that our model is of superior versatility to support perioperative phase-agnostic PHLF prediction, risk stratification for PHLF, and incomplete variable inputs. Importantly, our model demonstrates a high capacity for predicting clinically relevant PHLF. The clinicians' prediction assisted by our model was substantially improved over the clinician-only predictions (AUC = 0.778 vs. 0.637, P = 0.009). Our model achieved state-of-the-art performance in accuracy, generalizability, interpretability, and versatility for early detection of PHLF within the first 24 h after surgery. This approach holds promise for transforming perioperative management of hepatectomy and improving the rescue of life-threatening PHLF. The work was partially supported by a grant from the 10.13039/501100002920 Research Grants Council of the Hong Kong Special Administrative Region , China (Project Reference Number: T45-401/22-N), in part by grants from the 10.13039/100014717 National Natural Science Foundation of China (82170647, 82270661, and 62372441), in part by grants from the 10.13039/501100021171 Basic and Applied Basic Research Foundation of Guangdong Province (2023A1515010088, 2024A1515013204, and 2023A1515030268) and in part by grant from Shenzhen Science and Technology Program (Grant No. RCYX20231211090127030).