Gene Silencing by Methylation Marks the Aggressive Sarcomatoid Shift in Kidney Cancer

Deep within kidney tumors, a dangerous transformation can occur that pathologists often miss and oncologists fear: clear cell renal cell carcinoma, the most common form of kidney cancer, can shed its epithelial identity and morph into a sarcomatoid variant that spreads aggressively and resists standard therapies. Now, a research team at the Portuguese Oncology Institute of Porto has uncovered a molecular signature of this transformation, showing that two genes are switched off by chemical tags on their DNA precisely when tumors turn sarcomatoid. The findings, published in Epigenetics Communications, could eventually allow doctors to detect this aggressive shift from a simple blood test rather than a surgical specimen.

The stakes are considerable. Renal cell carcinoma accounted for roughly 440,000 new diagnoses and 180,000 deaths worldwide in 2022, and clear cell renal cell carcinoma, abbreviated ccRCC, makes up 70 to 80 percent of those cases. While targeted therapies such as tyrosine kinase inhibitors and immune checkpoint inhibitors have improved outcomes for many patients, sarcomatoid differentiation remains a grim turning point. It appears in about 4 to 5 percent of renal cancers, more frequently in the clear cell subtype than in papillary or chromophobe tumors, and is thought to represent a dramatic activation of epithelial-to-mesenchymal transition, the biological program by which epithelial cells abandon their tight connections and acquire the motile, spindle-shaped character of mesenchymal cells.

Under the microscope, this shift is unmistakable once seen: an abrupt transition from orderly epithelial architecture to pleomorphic spindle cells, with loss of the classical features that define ccRCC. Yet catching it early is notoriously difficult. Kidney tumors sit deep in the retroperitoneum, so up to a quarter of patients already have metastatic disease at diagnosis, and among those, roughly 20 percent harbor sarcomatoid features. Needle biopsies, the main alternative to surgery for patients too frail for nephrectomy, frequently miss sarcomatoid regions entirely because of sampling limitations. The Porto team, led by Fernanda Fernandes-Pontes and senior authors Vera Miranda-Gonçalves and Carmen Jerónimo, set out to find a molecular fingerprint that could reveal sarcomatoid change even when the biopsy shows only conventional tumor.

Their strategy began with publicly available data from The Cancer Genome Atlas, accessed through the cBioPortal for Cancer Genomics. From the Kidney Renal Clear Cell Carcinoma Firehose Legacy dataset, they filtered out cases lacking clear histological annotation, leaving 33 tumors with documented sarcomatoid features and 50 conventional ccRCC tumors. Using the Shiny Methylation Analysis Resource Tool, or SMART App, they extracted CpG sites, the cytosine-guanine dinucleotides where methyl groups attach to DNA, from promoter regions within 1,500 base pairs upstream of each gene’s transcription start site, restricting the analysis to CpGs located within CpG islands as annotated by the UCSC genome browser.

From an extensive literature review anchored on a molecular characterization of sarcomatoid renal cancer by Motzer and colleagues, the researchers compiled 76 candidate genes spanning angiogenesis, cell cycle regulation, signaling pathways, cellular differentiation, metabolism, epigenetic enzymes, and immune function. In the TCGA cohort, 64 of these genes were significantly downregulated in sarcomatoid tumors and 12 were upregulated. The team then applied a crucial filter: they searched for genes whose promoter methylation levels anti-correlated with their transcript levels, using Spearman’s correlation coefficient between methylation beta-values from the Illumina HumanMethylation450 array and log2-transformed TPM expression values from RNA sequencing. This inverse relationship is the statistical signature of a gene being silenced by its own promoter methylation.

Two genes emerged with striking clarity. CDHR5, a mucin-like cadherin family member involved in cell adhesion, showed three promoter CpGs with significant inverse correlations ranging from Rho of −0.34 to −0.41, and all three sites were significantly more methylated in sarcomatoid tumors. UPB1, an enzyme catalyzing the final step of pyrimidine degradation, displayed six inversely correlated CpGs with correlations from −0.11 to −0.75, two of which were significantly hypermethylated in the sarcomatoid group. Correspondingly, both genes showed significantly reduced transcript levels in sarcomatoid ccRCC. A third gene, TLCD1, which regulates the incorporation of long-chain polyunsaturated fatty acids into cell membranes and is tied to metabolic reprogramming, was upregulated in silico and included for validation, though its role would prove more complicated.

To confirm these computational findings in real tissue, the team assembled a cohort of 88 formalin-fixed, paraffin-embedded tumor samples from nephrectomies performed between 2001 and 2024 at their tertiary cancer center, including 34 tumors with sarcomatoid differentiation. A dedicated uropathologist annotated both the truly sarcomatoid regions and conventional ccRCC areas on stained slides for precise macrodissection. RNA was extracted and reverse-transcribed for quantitative PCR analysis of transcript levels, while DNA was bisulfite-converted, a chemical treatment that converts unmethylated cytosines to uracil while leaving methylated cytosines intact, allowing methylation to be quantified by multiplex quantitative methylation-specific PCR with primers and TaqMan probes designed to cover the exact CpGs identified in the TCGA analysis.

The validation was unambiguous. Sarcomatoid tumors showed significantly higher promoter methylation of both CDHR5 and UPB1, along with significantly lower transcript levels of both genes, mirroring the TCGA data. Immunohistochemistry on 64 cases, 32 of each type, revealed that UPB1 protein, which localizes to the cytoplasm, was significantly reduced in sarcomatoid tumors with a p-value of 0.0022, while CDHR5 membrane staining showed the same downward trend without reaching statistical significance. Notably, UPB1 immunostaining scores anti-correlated with the proliferative index measured by Ki67, and UPB1 protein was significantly lower in stage T2 through T4 sarcomatoid tumors compared to stage T1, suggesting a putative tumor-suppressor role. TLCD1, by contrast, failed to show significant expression differences in the tissue cohort, prompting the authors to hypothesize that other epithelial-to-mesenchymal transition pathways drive sarcomatoid change in ccRCC.

The clinical implications sharpened when the team examined biomarker performance. Individually, CDHR5 methylation discriminated sarcomatoid from conventional tumors with 70.4 percent sensitivity, while UPB1 methylation achieved 74.1 percent sensitivity with similar specificity. Combined into a two-gene panel, the markers reached 78.7 percent specificity and 63.0 percent sensitivity, with positive and negative predictive values in the same range. More strikingly, elevated CDHR5 methylation was significantly associated with metastatic disease at diagnosis, with a p-value of 0.0159, and with advanced stage III/IV disease at p equal to 0.0488. The authors propose a plausible mechanism: as a mucin-like adhesion protein, reduced CDHR5 expression, driven by promoter hypermethylation, may loosen tumor cell cohesion, enhance invasion and migration, and facilitate access to blood vessels, thereby promoting metastatic dissemination.

The researchers caution that this is a retrospective and exploratory study, and that functional validation of the genes’ roles and regulatory mechanisms remains an important next step. Yet the translational potential is clear. Because aberrant DNA methylation occurs early in tumorigenesis and can be detected in cell-free DNA circulating in blood, the CDHR5 and UPB1 methylation signature could, if validated in larger studies and liquid biopsies, provide a non-invasive molecular fingerprint of sarcomatoid differentiation. Such a test would complement histopathological work-up of core biopsies, flag patients whose conventional-appearing tumors harbor hidden aggressive biology, and inform therapeutic decisions, since sarcomatoid tumors are generally more resistant to tyrosine kinase inhibitors but may respond better to anti-PD-L1 immunotherapy. For a cancer that often hides its worst features until it is too late, a pair of silenced genes may prove to be the warning sign clinicians have been waiting for.

Subject of Research: Epigenetic regulation of CDHR5 and UPB1 promoter methylation in sarcomatoid clear cell renal cell carcinoma

Article Title: CDHR5 and UPB1 expression associated with promoter methylation in sarcomatoid differentiated clear cell renal cell carcinoma

Article References: CDHR5 and UPB1 expression associated with promoter methylation in sarcomatoid differentiated clear cell renal cell carcinoma. (n.d.). https://doi.org/10.1186/s43682-026-00053-7

Image Credits: AI Generated

DOI: 10.1186/s43682-026-00053-7

Keywords: clear cell renal cell carcinoma, sarcomatoid differentiation, DNA methylation, CDHR5, UPB1, epigenetics, biomarkers, liquid biopsy, The Cancer Genome Atlas, kidney cancer, epithelial-to-mesenchymal transition, quantitative methylation-specific PCR