Clear cell renal cell carcinoma (ccRCC) is the most common form of kidney cancer, responsible for millions of new diagnoses and approximately 2 million deaths per year worldwide. While surgery can effectively treat early-stage ccRCC, advanced and metastatic cases remain difficult to treat and carry poor survival rates.
In recent years, immunotherapy - treatments that mobilize the immune system against cancer - has become a major pillar of advanced kidney cancer treatment. However, not all patients respond to immunotherapy, and identifying who will benefit remains a critical challenge. New biological markers (biomarkers) that predict treatment response are urgently needed.
The gene MYBL1 has been studied in other cancers. It has been linked to sorafenib resistance in liver cancer, cancer stem cell behavior, salivary gland cancer, and glioma (a type of brain cancer). Despite this growing body of evidence pointing to MYBL1 as a cancer-promoting molecule, no previous study had investigated its role specifically in kidney cancer.
This study set out to comprehensively characterize MYBL1's role in ccRCC, combining large-scale computational analysis of public gene expression databases with laboratory experiments in kidney cancer cell lines and mouse models.
Gene expression and clinical data from 537 ccRCC patients were obtained from the Cancer Genome Atlas (TCGA) database and a separate European dataset (E-MTAB-1980) containing 240 patients. This two-cohort approach allowed findings to be discovered in one dataset and independently validated in another.
Multiple computational tools were used simultaneously to estimate the immune microenvironment - six different algorithms (XCELL, MCPCOUNTER, CIBERSORT, TIMER, EPIC, and QUANTISEQ) were applied to quantify the abundance of different immune cell types within tumors. Using multiple methods increases confidence in the results.
To predict immunotherapy response, two computational approaches were used: the Immunophenoscore (IPS) from The Cancer Immunome Database (a machine learning-based score indicating likelihood of responding to immune checkpoint inhibitors) and the TIDE algorithm (which measures T-cell dysfunction and exclusion as proxies for immunotherapy resistance).
Laboratory experiments were conducted to validate the computational findings. MYBL1 knockdown - using small hairpin RNA molecules to silence the gene in cancer cells - was performed in four kidney cancer cell lines. The effects on cell growth, colony formation, DNA replication, and cell death were measured. Mouse xenograft experiments confirmed the findings in living organisms.
Analysis of both RNA expression data and protein-level immunohistochemical images confirmed that MYBL1 is significantly overexpressed in ccRCC tumor tissue compared to normal kidney tissue. This elevation was consistent across both the TCGA database and the GTEx normal tissue database, as well as in human protein images from The Human Protein Atlas.
Survival analysis in both the TCGA ccRCC cohort and the E-MTAB-1980 validation cohort showed that patients with higher MYBL1 expression had worse overall survival, disease-free survival, and progression-free survival. These consistent findings across two independent patient populations strengthen confidence that MYBL1 expression is a meaningful prognostic signal.
Examining the relationship between MYBL1 and clinical tumor features, researchers found that MYBL1 was significantly higher in patients with more advanced lymph node (N-stage) and metastatic (M-stage) disease, suggesting it plays a role in cancer spread. Interestingly, MYBL1 levels did not differ significantly between different T-stages or tumor grades.
Multivariate statistical analysis confirmed that MYBL1 expression is an independent risk factor for ccRCC - meaning its negative effect on prognosis exists separately from other known risk factors like tumor stage or patient age, supporting its potential clinical value as a standalone biomarker.
Comparing the gene expression profiles of patients with high versus low MYBL1 expression, researchers identified 290 differentially expressed genes (154 reduced, 136 elevated). These genes were involved in diverse biological processes including transport of amino acids and ions, blood vessel regulation, and embryonic development patterns.
Gene Set Enrichment Analysis (GSEA) using the Hallmark gene set identified three pathways that were significantly activated in patients with high MYBL1 expression: the inflammatory response, G2M checkpoint (a cell cycle control point), and E2F targets (genes involved in driving cell division). Each of these pathways plays a known role in cancer progression.
Activation of the G2M checkpoint and E2F targets suggests that high MYBL1 is linked to cells that are actively proliferating and bypassing normal cell division controls. This is consistent with MYBL1's known function in other cancers as a promoter of uncontrolled cell growth.
The inflammatory response pathway being activated in high-MYBL1 tumors is particularly relevant in the context of immunotherapy. Chronic inflammation within tumors is known to create a complex, often immunosuppressive environment that can prevent the immune system from effectively killing cancer cells.
Applying six different immune quantification algorithms to TCGA ccRCC data, researchers found that high MYBL1 expression was consistently associated with a remodeled immune microenvironment. Patients with high MYBL1 had more regulatory T cells (Tregs), M2 macrophages, neutrophils, B cells, monocytes, and CD8+ T cells, but fewer endothelial cells, within their tumors.
While the presence of more CD8+ T cells might seem beneficial, the simultaneous increase in immunosuppressive Tregs and M2 macrophages creates a hostile environment that can neutralize the cancer-killing potential of the CD8+ cells. M2 macrophages are known to promote tumor growth and metastasis, and Tregs actively suppress other immune cells' activity.
MYBL1 expression was also positively correlated with higher immune scores and stromal scores - measures of the overall immune and connective tissue activity within the tumor microenvironment. Higher immune and stromal scores can reflect a more reactive but ultimately immunosuppressive tumor environment.
Crucially, patients with higher MYBL1 expression had significantly higher levels of four major immune checkpoint proteins: PD-1, CTLA4, PD-L1, and PD-L2. These checkpoints are the very targets of immunotherapy drugs. Their elevated presence in high-MYBL1 tumors means these patients might have tumors with more 'immune brakes' engaged - which could mean either better response to checkpoint inhibitor drugs, or paradoxically more resistance due to the deeper suppression.
To move beyond computational predictions to actual biological proof, researchers silenced MYBL1 in four different kidney cancer cell lines using RNA interference technology. The MYBL1 gene was confirmed to be overexpressed in all four cancer cell lines compared to normal kidney cells, consistent with the database analysis.
When MYBL1 was knocked down (silenced), kidney cancer cells showed significantly reduced proliferation in multiple assays - cells grew more slowly in CCK-8 viability tests, formed fewer colonies in colony formation assays, and showed reduced DNA replication in EdU incorporation assays. All three tests pointed to the same conclusion: MYBL1 drives cancer cell growth.
Cell death (apoptosis) increased dramatically when MYBL1 was silenced. Flow cytometry analysis showed a significantly higher percentage of cells undergoing programmed cell death in MYBL1-knockdown cells compared to control cells, confirming that MYBL1 normally helps cancer cells survive and resist natural death signals.
In a mouse xenograft model - where kidney cancer cells were implanted under the skin of mice - silencing MYBL1 significantly slowed tumor growth in living animals, validating that MYBL1 is genuinely required for ccRCC tumor growth beyond the artificial laboratory setting. Immunohistochemistry of patient tissue also confirmed higher MYBL1 in ccRCC tumors compared to normal kidney tissue from the same patients.
Building on MYBL1's influence over gene expression in ccRCC, researchers used a machine learning algorithm called LASSO logistic regression to identify the best subset of MYBL1-related genes that collectively predict patient prognosis. This approach selects only the most informative genes, avoiding overfitting and ensuring clinical practicality.
The final prognosis signature included 8 genes: CASR, F11, IGF2BP3, TAGLN3, PLPPR1, SIM2, RALYL, and RUFY4. Each was assigned a mathematical weight based on its relationship to survival, and their combined expression was used to calculate a risk score for each patient. High-risk patients had significantly worse overall survival than low-risk patients.
The signature showed strong predictive performance across three independent datasets: the TCGA training cohort, an internal validation cohort, and the external E-MTAB-1980 validation cohort. The 1-year, 3-year, and 5-year prediction accuracy (AUC) values were 0.77, 0.74, and 0.71 respectively - showing reliable performance across different time horizons.
Patients in the high-risk group had significantly lower rates of predicted immunotherapy response (26.7% vs 38.9% in the low-risk group). This means the MYBL1-based risk score is not just a survival predictor but also an immunotherapy response predictor, helping identify which patients are most likely to benefit from checkpoint inhibitor treatment.
This study establishes MYBL1 as a novel biomarker for clear cell kidney cancer - the first of its kind to examine this gene specifically in ccRCC. MYBL1 is elevated at both the RNA and protein levels in kidney cancer, is linked to worse survival and more advanced metastatic disease, and actively drives cancer cell growth and survival.
The finding that MYBL1 reshapes the immune microenvironment and is correlated with elevated immune checkpoint expression makes it directly relevant to immunotherapy decision-making. Higher MYBL1 may help identify patients whose tumors have a more complex immune environment that might influence whether they respond to or resist checkpoint inhibitors.
The 8-gene machine learning prognosis signature derived from MYBL1-related genes provides a practical clinical tool. With further validation, this signature could be applied to patient tumor samples to stratify risk and guide treatment decisions, including whether to prioritize immunotherapy or alternative approaches.
The authors note important limitations: the patient populations analyzed were predominantly Western, which may limit applicability across different ethnic groups with potentially different genetic backgrounds. Additionally, the exact mechanism by which MYBL1 promotes ccRCC at the molecular level remains to be fully elucidated in future research. These are priorities for follow-up studies.