Thymidylate synthetase is the rate-limiting enzyme in the de novo thymidine synthesis pathway, catalyzing the conversion of deoxyuridine monophosphate to deoxythymidine monophosphate, a critical step required for DNA synthesis and repair. Because TYMS activity peaks during the S phase of the cell cycle when DNA replication is active and is extremely low during G0 and G1, it is fundamentally linked to cell proliferation capacity.
TYMS overexpression has been documented in multiple malignancies including colorectal, gastric, hepatocellular, pancreatic, and breast cancers, where it consistently correlates with poor clinical outcomes. In lung adenocarcinoma specifically, elevated TYMS expression has been associated with enhanced invasive capacity and shorter survival, suggesting it operates as both a driver of tumor progression and a potential therapeutic vulnerability.
Despite this evidence, no systematic multi-omics study had comprehensively characterized TYMS in LUAD across transcriptomic, proteomic, and genomic dimensions simultaneously. The interactions between TYMS expression and the tumor immune microenvironment, copy number variation, and drug sensitivity profiles remained poorly understood in LUAD specifically.
TYMS is also directly implicated in folate metabolism through its protein interaction network, connecting it to the same one-carbon metabolism pathway disrupted by MTHFR variants and targeted by folate antagonist chemotherapy drugs such as methotrexate, 5-fluorouracil, and pemetrexed. This places TYMS at the intersection of DNA synthesis, folate metabolism, and chemotherapy resistance mechanisms.
This study integrated transcriptomic, proteomic, and genomic data from TCGA, GTEx, CPTAC, and GEPIA2 to conduct a comprehensive bioinformatics analysis of TYMS across multiple analytical dimensions in LUAD. The target gene selection process began by intersecting the top 500 survival-associated genes with 4,235 differentially expressed genes between LUAD and normal tissue, yielding 148 candidate prognostic genes from which TYMS was selected based on functional enrichment in nucleotide metabolism and thymidine synthesis pathways.
mRNA expression was assessed using GEPIA2 and UALCAN, while protein expression was analyzed via UALCAN in conjunction with the CPTAC LUAD dataset of 109 tumor cases and 111 adjacent normal tissue samples. Immunohistochemistry images from the Human Protein Atlas provided visual confirmation of TYMS protein distribution in tumor versus normal lung tissue.
Immune cell infiltration was quantified using the TIMER platform with Spearman correlation analysis against B cells, CD8 plus T cells, CD4 plus T cells, neutrophils, and dendritic cells, as well as immune checkpoint molecules including CTLA4, PD-1, and IL-6. Somatic copy number variation analysis was performed using LinkedOmics to identify genes whose expression correlated with TYMS amplification, followed by GSEA pathway enrichment.
Drug sensitivity was assessed using the GDSC database, correlating TYMS mRNA expression levels against IC50 values for multiple anticancer agents to identify drugs whose efficacy is enhanced or reduced in high-TYMS tumors. Protein-protein interaction networks were constructed using STRING with an interaction score threshold above 0.40.
TYMS mRNA was significantly overexpressed in LUAD tissue compared to normal lung across both GEPIA2 and UALCAN analyses, and TYMS protein was confirmed as significantly upregulated in LUAD tissue by CPTAC proteomics with immunohistochemical validation from the Human Protein Atlas. The concordance between transcriptomic and proteomic upregulation provides stronger evidence of functional relevance than either level alone.
High TYMS expression was significantly associated with early tumor staging, specifically showing elevated protein expression in stage 1 tumors, suggesting that TYMS overexpression occurs early in LUAD development rather than emerging only in advanced disease. TYMS expression was also elevated in high-grade tumors (grade 3) and in patients in the 41 to 60 age cohort.
Survival analysis demonstrated that high TYMS expression correlated with reduced overall survival in LUAD patients, establishing its value as an independent prognostic indicator. This prognostic association in LUAD aligns with previously reported findings in colorectal, hepatocellular, and pancreatic cancers, confirming that TYMS follows a consistent oncogenic expression pattern across tumor types.
The protein-protein interaction network centered on TYMS identified ten core interacting proteins, including key folate and nucleotide metabolism enzymes (DHFR, MTHFR, MTHFD1, SHMT1, DCTD) and DNA replication and cell cycle proteins (TK1, DTYMK, DUT, MCM2), placing TYMS at a hub connecting two critical biological processes relevant to tumor progression.
Pathway enrichment analysis using CPTAC proteomics data revealed that TYMS expression was significantly associated with alterations in the p53/Rb and mTOR signaling pathways. The p53 pathway is central to cell cycle regulation and apoptosis, and its dysregulation promotes uncontrolled proliferation and immune evasion. TYMS involvement in mTOR signaling suggests it may additionally promote tumor growth by regulating cellular metabolism and immune microenvironment composition.
Immune cell infiltration analysis using TIMER showed that high TYMS expression negatively correlated with B cell activity while positively correlating with neutrophil infiltration. TYMS expression also showed significant positive associations with CTLA4, PD-1, and IL-6, molecules that promote immune suppression and checkpoint-mediated escape. This immune signature suggests that TYMS overexpression contributes to an immunosuppressive tumor microenvironment unfavorable to anti-tumor immune responses.
Somatic copy number amplification of TYMS was significantly associated with poor LUAD prognosis. GSEA on TYMS-amplified tumors identified folate metabolism as the most enriched pathway, directly linking TYMS genomic amplification to the same metabolic pathway targeted by multiple chemotherapy regimens including methotrexate, 5-fluorouracil, and pemetrexed.
Genes whose expression significantly correlated with TYMS copy number variation were enriched in nucleotide metabolism and biosynthesis pathways, suggesting that TYMS amplification creates a broader metabolic reprogramming state rather than affecting thymidine synthesis in isolation.
TYMS expression showed a positive correlation with sensitivity to afatinib and gefitinib, both EGFR tyrosine kinase inhibitors, suggesting that high-TYMS tumors may be more responsive to EGFR-targeted therapy through interactions between TYMS-mediated metabolic regulation and EGFR signaling activity. A positive correlation with trametinib, a MEK inhibitor, was also identified.
Conversely, high TYMS expression negatively correlated with sensitivity to methotrexate, reflecting the mechanistic antagonism between TYMS activity and folate antagonist therapy. Methotrexate blocks folate metabolism while TYMS drives thymidine synthesis through the same pathway, so elevated TYMS may compensate for the methotrexate-induced folate depletion and reduce drug efficacy.
Negative correlations were also observed between TYMS expression and vorinostat, navitoclax, BX-912, FK866, and GSK1070916, suggesting that elevated TYMS levels broadly interfere with the mechanisms of these agents. For vorinostat, an HDAC inhibitor, the antagonism may reflect TYMS-mediated alteration of chromatin-related metabolic states that reduce histone deacetylase inhibitor efficacy.
This drug sensitivity profile has direct therapeutic implications: patients with high-TYMS LUAD may respond better to EGFR inhibitor-based regimens than to folate antagonist chemotherapy, and combining TYMS inhibition with existing chemotherapy agents could potentially overcome the resistance mechanism driven by TYMS amplification.
This multi-omics study establishes TYMS as an independent prognostic biomarker in lung adenocarcinoma, with overexpression at both mRNA and protein levels associated with poor survival, early-stage high-grade disease, immunosuppressive microenvironment characteristics, and folate metabolism pathway enrichment through copy number amplification.
The convergence of evidence across transcriptomics, proteomics, genomics, immune profiling, and drug sensitivity analysis provides a coherent functional map of how TYMS transitions from gene-level overexpression to clinical prognostic impact, a multi-level characterization that single-platform studies cannot achieve.
Limitations include the exclusively bioinformatics-based approach without validation in independent clinical specimens, and the absence of experimental mechanistic studies directly confirming how TYMS modulates the immune microenvironment or drug sensitivity in LUAD cell models. These gaps represent the most important priorities for follow-up research.
Future directions should include clinical sample validation of TYMS protein expression as a prognostic marker, experimental dissection of TYMS-CTLA4 and TYMS-PD1 interactions in the tumor immune microenvironment, and clinical investigation of whether TYMS-targeted combination regimens can overcome the folate metabolism resistance mechanism associated with TYMS copy number amplification.