Decoding the triglyceride-glucose index in metabolic dysfunction-associated steatotic liver disease: integrative insights from mendelian randomization, cross-tissue transcriptomics, and spatial multi-omics

Background: The triglyceride-glucose (TyG) index, an insulin resistance marker linked to the progression of metabolic dysfunction-associated steatotic liver disease (MASLD), underscores the redox imbalance-mediated crosstalk between MASLD and cardiovascular-liver-metabolic health (CLMH), although its causal mechanisms and molecular drivers remain unresolved. Methods: We employed a multi-omics framework to integrate Mendelian randomization (MR) and transcriptome-wide association studies (TWAS). MR leveraged 192 genome-wide significant SNPs for TyG from the UK Biobank, employing inverse-variance weighted (IVW) and generalized summary-data MR (GSMR). Transcriptomic integration utilized four approaches: Multi-marker Analysis of GenoMic Annotation (MAGMA) for gene-set enrichment; Joint-Tissue Imputation PrediXcan (JTI-PrediXcan) for tissue-specific expression; Sparse Multi-Tissue Imputation Xcan (SMulTiXcan) for cross-tissue meta-analysis; and Fine-mapping of Causal Gene Sets (FOCUS) for Bayesian fine-mapping. Co-morbid genes were validated using Functional Summary-based Imputation (FUSION) and prioritized based on the Polygenic Priority Score (PoPS). Single-cell spatial transcriptomics (sc-ST) in embryonic mice (E16.5) mapped tissue-specific expression via genetically informed spatial mapping (gsMap). Results: MR analysis demonstrated a causal effect of TyG on MASLD risk (IVW: odds ratio [OR] = 1.58, 95%CI = 1.04-2.38, P = 0.030; GSMR: OR = 1.43, 95% CI = 1.27-1.61, P = 5.20 × 10 −9 ). TWAS identified 12 co-morbid genes (C2orf16/SPATA31H1, FNDC4, GCKR, GMIP, HAPLN4, LPAR2, MAU2, MEF2B, NDUFA13, NRBP1, TM6SF2, ZNF513). Independent validation using the FUSION framework confirmed nine TyG-MASLD comorbid genes with genome-wide significant false discovery rate-adjusted associations. Notably, TM6SF2 (TyG-PoPS = 7.2491) and GCKR (TyG-PoPS = 6.7102) showed strong positive associations in TyG, while NDUFA13 exhibited negative scores in MASLD (PoPS = − 0.5028). Spatial mapping revealed conserved enrichment of APOA1, APOB, and APOC4 (sc-ST P<0.001) in murine liver and vascular tissues. Organ-specific analysis showed significant MASLD signals included the liver (sc-ST P = 6.43 × 10 −5 ), adrenal gland (Cauchy P = 0.0064), and connective tissue (sc-ST P = 3.29 × 10 −5 ). Conclusion: This study establishes TyG as a causal MASLD driver mediated by redox-sensitive hubs and evolutionarily conserved apolipoproteins, linking hepatic lipid peroxidation to systemic metabolic dysregulation. Targeting these pathways may mitigate dual hepatic-cardiovascular risks, advancing precision therapies for CLMH.