Long non-coding RNA TMPO-AS1 flags poor outcomes in lung adenocarcinoma

Lung adenocarcinoma is the most common histological subtype of lung cancer, and it remains one of the hardest to treat. A new bioinformatics study, published in the journal Cell Division, points to a long non-coding RNA called TMPO-AS1 as a molecule that may help explain why some patients fare so poorly. By weaving together data from several public cancer databases, the researchers describe a regulatory circuit in which TMPO-AS1 sits at the top, a tumour-suppressive microRNA called hsa-let-7b-5p sits in the middle, and a panel of downstream genes links the axis to both tumour heterogeneity and the immune cells that infiltrate lung tumours.

The team, led by Sakshi Nirmal and corresponding author Rajeev Nema, began with a screening exercise. They pulled a list of sixteen long non-coding RNAs known to be upregulated in lung adenocarcinoma from the lnc2cancer 3.0 database, then interrogated the Kaplan-Meier Plotter resource to see which of them tracked with patient survival. Only one candidate survived the filter: TMPO-AS1, which showed a hazard ratio of 2.16 for overall survival, meaning that patients with higher expression had more than double the risk of death over the analysed period. The association was even stronger in smokers, with a hazard ratio of 2.34, and in female smokers, where it climbed to 3.13.

Validation came from independent platforms. A pan-cancer survey using the OncoMX database found TMPO-AS1 consistently elevated across The Cancer Genome Atlas tumour types, with a fold change of 2.11 in lung cancer. Differential expression checks in the UALCAN and ENCORI databases, and a separate analysis of TCGA-LUAD RNA-sequencing data using R-based statistical packages, all confirmed significant overexpression in tumour tissue compared with normal tissue, with p-values ranging from 6.3 × 10⁻⁵³ down to 1.05 × 10⁻¹³. The authors conclude that TMPO-AS1 is overexpressed in lung adenocarcinoma and may serve as a molecular classifier and a prognostic biomarker associated with poor clinical outcomes.

Long non-coding RNAs are transcripts longer than 200 nucleotides that do not code for proteins, yet many of them regulate gene expression through epigenetic, transcriptional and post-transcriptional routes. One influential framework is the competitive endogenous RNA, or ceRNA, hypothesis: a long non-coding RNA can act as a molecular sponge, binding microRNAs and thereby relieving those microRNAs’ normal repression of their messenger RNA targets. TMPO-AS1 has previously been implicated in proliferation, invasion and metastasis across several malignancies, and has been reported to sponge tumour-suppressive microRNAs including miR-200c, miR-320a and miR-329-3p.

To map the sponge in lung adenocarcinoma, the researchers used the miRNet database to identify eight candidate microRNAs that interact with TMPO-AS1. Correlation analysis in ENCORI singled out hsa-let-7b-5p as the strongest partner, showing a significant negative correlation with TMPO-AS1 (R = −0.277). In other words, when TMPO-AS1 was high, hsa-let-7b-5p tended to be low, a pattern consistent with the sponging model. Survival analysis in the Kaplan-Meier Plotter showed that low expression of hsa-let-7b-5p was associated with worse overall survival in lung adenocarcinoma (hazard ratio 0.71, 95% confidence interval 0.53–0.96, p = 0.023), with a median survival of 42.17 months in the low-expression cohort versus 54.4 months in the high-expression cohort. Expression analysis in matched normal and tumour samples confirmed that the microRNA was downregulated in tumours (p = 0.025).

With the middle of the axis in place, the team built what they call a heterogeneous gene model around hsa-let-7b-5p. Using the CancerMIRNome database, they identified the top ten genes positively correlated with the microRNA (including CPED1, EFCAB14, COLEC12, DLC1, SEPT4, TGFBR3, RNF144B, CD59, MAT2B and CHMP3) and the top ten negatively correlated genes (including ESPL1, AURKB, CCNF, EZH2, PLK1, E2F2, PUS1, AURKA, KIFC1 and HASPIN). Correlations were independently validated in ENCORI, and each gene’s prognostic value was tested across three survival endpoints: overall survival, first progression and post-progression survival. Stratification by smoking status and histological subtype narrowed the field to a six-gene panel: four positively associated genes (CPED1, TGFBR3, RNF144B and CD59) and two negatively associated genes (AURKB and KIFC1) that were the most robust prognostic markers in LUAD smokers.

The identity of those genes is telling. The positively associated group includes established tumour suppressors, which were downregulated in parallel with hsa-let-7b-5p suppression under cancerous conditions. The negatively associated group includes genes with oncogenic potential, such as the Aurora kinases A and B and the kinesin KIFC1, which were upregulated when the microRNA was low. Together, the authors argue, this gene panel illustrates how TMPO-AS1-mediated sponging of a single microRNA could ripple outward into a coordinated transcriptional programme, contributing to the molecular heterogeneity that makes lung adenocarcinoma so difficult to treat.

The final layer of the analysis concerned the tumour immune microenvironment. Using the Gene Set Cancer Analysis (GSCA) database, which estimates the relative abundance of immune cell subsets in tumour samples from bulk gene-expression data, the researchers examined how their gene panel correlated with immune infiltration. Among the positively associated genes, TGFBR3, RNF144B and CD59 showed strong negative correlations with natural regulatory T (nTreg) cells (R = −0.53, −0.56 and −0.37, respectively) and positive correlations with natural killer T (NKT) cells (R = 0.52, 0.54 and 0.38). The negatively associated genes AURKB and KIFC1 showed the opposite pattern, correlating positively with nTreg cells (R = 0.56 and 0.55) and negatively with CD4-positive T cells (R = −0.39 and −0.40). TMPO-AS1 itself correlated positively with nTreg cells (R = 0.45) and negatively with CD4-positive T cells (R = −0.24).

The authors interpret these reciprocal patterns as evidence that the axis may help shape the immune landscape of lung tumours. High expression of the hsa-let-7b-5p-positive gene panel appears to accompany an antitumour immune profile, with more NKT cell infiltration and less nTreg activity, whereas suppression of that axis may accompany a pro-tumour environment dominated by nTregs, an immunosuppressive subset of CD4-positive T cells, and diminished broader CD4-positive T cell responses. The authors caution that nTreg enrichment need not mirror total CD4-positive T cell abundance, since the CD4 compartment is functionally heterogeneous, and that the divergence likely reflects a shift toward an immunosuppressive microenvironment rather than a simple loss of T cells.

It is important to keep the scope of this work in view. This is an in-silico study: every finding derives from publicly available transcriptomic, survival and immune-infiltration databases, and the authors themselves note that the results come from bulk tumour data without single-cell resolution. They also acknowledge that only selected immune populations were analysed in depth; CD8-positive T cells, macrophage subtypes, dendritic cells and B cells were not comprehensively assessed, and future work incorporating single-cell RNA sequencing and experimental validation will be needed to confirm causality and cell-type-specific mechanisms. What the study does deliver is a systems-level map: a candidate biomarker, TMPO-AS1, that is significantly overexpressed in lung adenocarcinoma and associated with unfavourable outcomes; a microRNA partner whose loss tracks with worse survival; a six-gene panel that links the axis to prognosis in smokers; and immune correlations that suggest, but do not yet prove, a role in immune evasion. The authors present the framework as a foundation for future experimental validation of the axis as a candidate target for RNA-based and immunomodulatory therapeutic strategies in lung adenocarcinoma, a step that would be needed before any of these correlations could be translated into clinical tools.

Subject of Research: Bioinformatic analysis of the TMPO-AS1/hsa-let-7b-5p regulatory axis and its association with prognosis and immune infiltration in lung adenocarcinoma

Article Title: The TMPO-AS1/hsa-let-7b-5p regulatory axis shapes transcriptional heterogeneity and immune landscape in lung adenocarcinoma

Article References: Nirmal, S., Saini, C., Baweja, B., Vats, P., Patidar, P., Jangir, K., & Nema, R. (2026). The TMPO-AS1/hsa-let-7b-5p regulatory axis shapes transcriptional heterogeneity and immune landscape in lung adenocarcinoma. Cell Division, 21(1), Article 10. https://doi.org/10.1186/s13008-026-00185-1

Image Credits: AI Generated

DOI: 10.1186/s13008-026-00185-1

Keywords: lung adenocarcinoma, TMPO-AS1, hsa-let-7b-5p, long non-coding RNA, ceRNA network, prognostic biomarker, immune infiltration, TCGA, bioinformatics, tumour microenvironment, microRNA sponging, transcriptional heterogeneity