Identification of potential feature genes in non-alcoholic fatty liver disease using bioinformatics analysis and machine learning strategies

Inhibition of cysteine-serine-rich nuclear protein 1 ameliorates ischemia-reperfusion injury during liver transplantation in an MAPK-dependent manner

Hepatic ischemia-reperfusion injury (HIRI) is a critical pathophysiological process during liver transplantation (LT). Multiple genes and signal pathways are dysregulated during HIRI. This study aims to identify genes as potential therapeutic targets for ameliorating HIRI. Datasets containing samples from the human donor liver (GSE151648) and mouse HIRI model (GSE117066) were analyzed to determine differentially expressed genes (DEGs). The selected DEGs were confirmed by real-time PCR and western blot in the hepatocyte hypoxia-reoxygenation (HR) model, mouse HIRI model, and human liver samples after transplantation. Genetic inhibition was used to further clarify the underlying mechanism of the gene in vitro and in vivo. Among the DEGs, CSRNP1 was significantly upregulated (|log FC|= 2.08, P < 0.001), and was positively correlated with the MAPK signal pathway (R = 0.67, P < 0.001). CSRNP1 inhibition by siRNA significantly suppressed apoptosis in the AML-12 cell line after HR (mean Annexin+ ratio = 60.62% vs 42.47%, P = 0.0019), but the protective effect was eliminated with an additional MAPK activator. Knocking down CSRNP1 gene expression by intravenous injection of AAV-shRNA markedly reduced liver injury in mouse HIRI model (ALT: AAV-NC vs AAV-shCsrnp1 = 26,673.5 ± 2761.2 vs 3839.7 ± 1432.8, P < 0.001; AST: AAV-NC vs AAV-shCsrnp1 = 8640.5 ± 1450.3 vs 1786.8 ± 518.3, P < 0.001). Liver-targeted delivery of siRNA by nanoparticles effectively inhibited intra-hepatic genetic expression of Csrnp1 and alleviated IRI by reducing tissue inflammation and hepatocyte apoptosis. Furthermore, CSRNP1 inhibition was associated with reduced activation of the MAPK pathway both in vitro and in vivo. In conclusion, our results demonstrated that CSRNP1 could be a potential therapeutic target to ameliorate HIRI in an MAPK-dependent manner.

Bioinformatics-based analysis of the relationship between plasminogen regulatory genes and photoaging

BACKGROUND

Ultraviolet radiation causes skin photoaging by producing a variety of enzymes, which impact both skin health and hinder beauty. Currently, the early diagnosis and treatment of photoaging remain a challenge. Bioinformatics analysis has strong advantages in exploring core genes and the biological pathways of photoaging.

AIMS

To screen and validate key risk genes associated with plasminogen in photoaging and to identify potential target genes for photoaging.

METHODS

Two human transcriptome datasets were obtained by searching the Gene Expression Omnibus (GEO) database, and the mRNAs in the GSE131789 dataset were differentially analyzed, and then the weighted gene co-expression network analysis (WGCNA) was performed to find out the strongest correlations. Template genes, interaction analysis of differentially expressed genes (DEGs), modular genes with the most WGCNA correlations, and genecard database genes related to plasminogen were performed, and further Kyoto genes and Genome Encyclopedia (KEGG) pathway analysis. Two different algorithms, least absolute shrinkage and selection operator (LASSO) and support vector machines-recursive feature elimination (SVM-RFE), were used to find key genes. Then the data set (GSE206495) was validated and analyzed. Real-time PCR was performed to validate the expression of key genes through in vitro cellular experiments.

RESULTS

IFI6, IFI44L, HRSP12, and BMP4 were screened from datasets as key genes for photoaging and further analysis showed that these genes have significant diagnostic value for photoaging.

CONCLUSION

IFI6, IFI44L, HRSP12, and BMP4 play a key role in the pathogenesis of photoaging, and serve as promising potential predictive biomarkers for photoaging.

A data-driven approach to decode metabolic dysfunction-associated steatotic liver disease

Metabolic dysfunction-associated steatotic liver disease (MASLD), defined by the presence of liver steatosis together with at least one out of five cardiometabolic factors, is the most common cause of chronic liver disease worldwide, affecting around one in three people. Yet the clinical presentation of MASLD and the risk of progression to cirrhosis and adverse clinical outcomes is highly variable. It, therefore, represents both a global public health threat and a precision medicine challenge. Artificial intelligence (AI) is being investigated in MASLD to develop reproducible, quantitative, and automated methods to enhance patient stratification and to discover new biomarkers and therapeutic targets in MASLD. This review details the different applications of AI and machine learning algorithms in MASLD, particularly in analyzing electronic health record, digital pathology, and imaging data. Additionally, it also describes how specific MASLD consortia are leveraging multimodal data sources to spark research breakthroughs in the field. Using a new national-level 'data commons' (SteatoSITE) as an exemplar, the opportunities, as well as the technical challenges of large-scale databases in MASLD research, are highlighted.

Identification and validation of disulfidptosis-associated molecular clusters in non-alcoholic fatty liver disease

Objective: Non-alcoholic fatty liver disease (NAFLD) is the most prevalent liver disease in the world, and its pathogenesis is not fully understood. Disulfidptosis is the most recently reported form of cell death and may be associated with NAFLD progression. Our study aimed to explore the molecular clusters associated with disulfidptosis in NAFLD and to construct a predictive model. Methods: First, we analyzed the expression profile of the disulfidptosis regulators and immune characteristics in NAFLD. Using 104 NAFLD samples, we investigated molecular clusters based on differentially expressed disulfidptosis-related genes, along with the related immune cell infiltration. Cluster-specific differentially expressed genes were then identified by using the WGCNA method. We also evaluated the performance of four machine learning models before choosing the optimal machine model for diagnosis. Nomogram, calibration curves, decision curve analysis, and external datasets were used to confirm the prediction effectiveness. Finally, the expression levels of the biomarkers were assessed in a mouse model of a high-fat diet. Results: Two differentially expressed DRGs were identified between healthy and NAFLD patients. We revealed the expression profile of DRGs in NAFLD and the correlation with 22 immune cells. In NAFLD, two clusters of molecules connected to disulfidptosis were defined. Significant immunological heterogeneity was shown by immune infiltration analysis among the various clusters. A significant amount of immunological infiltration was seen in Cluster 1. Functional analysis revealed that Cluster 1 differentially expressed genes were strongly linked to energy metabolism and immune control. The highest discriminatory performance was demonstrated by the SVM model, which had a higher area under the curve, relatively small residual and root mean square errors. Nomograms, calibration curves, and decision curve analyses were used to show how accurate the prediction of NAFLD was. Further analysis revealed that the expression of three model-related genes was significantly associated with the level of multiple immune cells. In animal experiments, the expression trends of DDO, FRK and TMEM19 were consistent with the results of bioinformatics analysis. Conclusion: This study systematically elucidated the complex relationship between disulfidptosis and NAFLD and developed a promising predictive model to assess the risk of disease in patients with disulfidptosis subtypes and NAFLD.