Promoters play a central role in transcription initiation. In the human genome, most protein-coding genes are regulated by multiple alternative promoters, and differential promoter usage gives rise to distinct transcript isoforms. Dysregulated promoter selection has been implicated in aberrant tissue development and disease pathogenesis. However, how genetic variants influence promoter choice, particularly in a genome-wide and mechanistic manner, remains poorly understood. In this study, we integrate large-scale GTEx and TCGA datasets to systematically identify promoter usage quantitative trait loci (puQTLs) across multiple normal tissues and cancer types, thereby interrogating the functional impact of genetic variation from a pre-transcriptional regulatory perspective. Focusing on puQTLs, we investigate whether and how genetic variants modulate alternative promoter selection rather than overall gene expression.
By incorporating high-resolution 3D genomics data, we reveal that more than 50% of puQTLs engage in long-range chromatin interactions with distal promoters, indicating that promoter usage is frequently regulated through spatial genome organization rather than linear proximity alone. These findings establish a direct mechanistic link between puQTLs and three-dimensional chromatin architecture, distinguishing puQTLs from conventional QTLs that primarily act through local regulatory effects. Building on the biological characteristics of puQTLs—including their distinct regulatory modes, differences from canonical eQTLs, and tissue- and cancer-type specificity—we further integrate multi-omics data to delineate the transcriptional regulatory activity and molecular mechanisms of puQTLs. Overall, by connecting genetic variation, three-dimensional genome organization, and alternative promoter usage, this study uncovers fundamental regulatory principles underlying tissue development and tumorigenesis, and substantially advances our understanding of the functional consequences of noncoding genetic variants.