A comprehensive analysis of the tryptophan metabolism-related gene signature to predict the prognosis of esophageal squamous cell carcinoma based on multi-omics.

BACKGROUND

Tryptophan (Trp) metabolism plays a vital role in tumor development and outcomes. However, Trp in esophageal squamous cell carcinoma (ESCC) remains poorly understood. We aimed to explore the role and mechanism of Trp metabolism in ESCC.

METHODS

We integrated single-cell RNA (scRNA) sequencing, bulk transcriptome, proteomics, and microbiome data from public databases. Tryptophan-related cell populations and their interactions were explored using the "seurat" R package at the single-cell level. Least absolute shrinkage and selection operator (LASSO) and univariate Cox regression were used to select prognostic TrpGs and construct a risk model. The overall survival, immune infiltration, checkpoint expression, drug sensitivity, and microbiota composition between high- and low-risk groups were evaluated. Independent prognostic factors were identified via multivariate Cox analysis and validated by qPCR analysis, and a nomogram was constructed for survival prediction.

RESULTS

We identified 28 differentially expressed tryptophan-related genes (DE-TrpGs), and fibroblasts emerged as the cell type with the highest TrpG score, although reduced in ESCC. Eighteen DE-TrpGs showed downregulation in tumor fibroblasts at the single-cell level. Fibroblast-epithelial communication involved the LAMININ, HSPG, and AGRN pathways. Five prognostic TrpGs (MAOA, AKR1A1, ALDH9A1, HAAO, and ALDH2) were selected to construct the risk model. The expression of MAOA, AKR1A1, ALDH9A1, HAAO, and ALDH2 was significantly downregulated in ESCC tumor tissues compared to non-tumor tissues. High-risk patients showed poorer overall survival (OS), distinct immune cell infiltration, elevated expression of KIR2DL1, LGALS9, TNFRSF18, and TNFRSF4, increased sensitivity to imatinib and axitinib, resistance to multiple chemotherapeutics, and reduced and abundance. HAAO, ALDH2, and lymph node stage were identified as independent prognostic factors and were used to develop a predictive nomogram.

CONCLUSION

We identified a Trp metabolism-associated fibroblast population in the ESCC tumor microenvironment (TME) and developed a five-gene TrpG signature for prognostic prediction in ESCC patients.