Colorectal cancer remains one of the most common malignancies worldwide and a leading cause of cancer-related death, and patients with advanced disease still face poor outcomes because of metastasis and therapy resistance. A study published in the journal Cell Division on 25 March 2026 now reports a molecular axis that may help drive this aggressive behavior. The research, led by Song Yu, Ji Li and Shaolin Li, identifies the neuronal adhesion protein CNTN3 as a putative oncogene in colorectal cancer and shows that its expression is transcriptionally regulated by PRRX2, a transcription factor already known for promoting mesenchymal phenotypes in several other cancers. Together, the authors propose the PRRX2-CNTN3 axis as a mechanism linking epithelial-to-mesenchymal transition, or EMT, with metastatic potential in colorectal tumors.
The investigation began with bioinformatics screening of colorectal cancer gene expression datasets GSE9348, GSE20916 and GSE166427 from the Gene Expression Omnibus. Using a threshold of log fold change greater than 2.0 and an adjusted P value below 0.05, the analysis identified 1,413, 429 and 597 differentially expressed genes in the respective datasets. Intersection analysis narrowed these to 202 overlapping genes, among which five, including CNTN3, showed the most pronounced differential expression across all three datasets. Their elevated expression in tumors was further confirmed using the TCGA-COAD dataset. Notably, CNTN3 had previously been suggested as a potential biomarker for colorectal tumorigenesis, which prompted the team to examine its precise role in disease progression.
CNTN3 is not an obvious cancer suspect at first glance. It is a glycosylphosphatidylinositol-anchored neuronal membrane protein belonging to the immunoglobulin superfamily, best known for regulating axon guidance, synaptogenesis and neural network formation during development. Earlier research considered its expression largely restricted to specific neuron groups such as cerebellar Purkinje cells, dentate gyrus granule cells and neurons in the superficial layers of the cerebral cortex. More recent work, however, has connected CNTN3 to cancer: it has been described as a diagnostic and prognostic marker correlated with tumor cell adhesion, invasion and poor clinical outcomes in lung adenocarcinoma, and its overexpression is associated with aggressive tumor behavior in glioblastoma. Its role in colorectal cancer had remained poorly characterized.
The new study reports that CNTN3 expression was significantly elevated in colorectal cancer tissues and cell lines, and that this high expression correlated with tumor metastasis and advanced clinical stages in the TCGA-COAD dataset. The authors suggest that CNTN3 may serve as a potential biomarker of poor prognosis in the disease. To probe function, the researchers silenced CNTN3 in the Lovo and SW480 colorectal cancer cell lines using shRNA. Knockdown significantly reduced cell viability, decreased DNA replication activity in EdU assays and markedly reduced colony formation. Caspase-3/7 activity, a marker of apoptosis, was reduced following CNTN3 knockdown, which the authors interpret as evidence that the protein normally helps suppress apoptotic cell death in these cells.
The team then turned to EMT, the process by which tumor cells lose epithelial characteristics, such as E-cadherin expression, and gain mesenchymal markers like vimentin, thereby acquiring invasive and migratory capabilities. In CNTN3-silenced cells, epithelial markers ZO-1, CDH1 and OCLN1 were upregulated while mesenchymal markers VIM1, SNAI1 and SLUG were downregulated. Immunofluorescence staining showed increased CDH1 and reduced N-cadherin, vimentin and alpha-SMA expression. Wound healing assays demonstrated significantly reduced migration rates at 24 hours, and Transwell assays, with and without Matrigel, confirmed impaired migration and invasion. These results support the hypothesis that CNTN3 contributes to the metastatic phenotype of colorectal cancer cells in vitro.
Having established CNTN3 as a functional player, the researchers searched for what drives its expression. Analysis of the CNTN3 promoter with the UCSC browser and the JASPAR transcription factor database flagged PRRX2 as a candidate regulator. Chromatin immunoprecipitation followed by quantitative PCR confirmed significant enrichment of CNTN3 promoter segments in anti-PRRX2 precipitates, and dual-luciferase reporter assays in 293T cells showed that PRRX2 overexpression significantly enhanced luciferase activity driven by the CNTN3 promoter. PRRX2 itself was found at elevated levels in colorectal cancer cell lines and in the TCGA-COAD dataset, correlating with poor patient outcomes, lymph node metastasis and advanced clinical staging, and its expression positively correlated with CNTN3 expression in the same dataset.
Functional experiments reinforced the regulatory relationship. Silencing PRRX2 in Lovo and SW480 cells concurrently reduced CNTN3 expression and significantly decreased cell viability, DNA replication and colony formation, while Caspase-3/7 activity increased. Migration and invasion assays showed reduced metastatic potential after PRRX2 knockdown. To test dependence in vivo, the team generated xenograft tumors in six-week-old female BALB/c nude mice, with six animals per group, using Lovo cells carrying PRRX2-targeting or control shRNA. In a rescue group, CNTN3 overexpression significantly reversed the suppression of tumor growth caused by PRRX2 silencing, with higher tumor volumes and weights, and immunohistochemistry confirmed restored CNTN3 expression in those tumors.
The reciprocal experiments completed the picture. When PRRX2 was overexpressed in Lovo and SW480 cells, mRNA levels of both PRRX2 and CNTN3 rose substantially, cell proliferation increased, and epithelial markers ZO-1, CDH1 and OCLN1 fell while mesenchymal markers VIM1, SNAI1 and SLUG rose. Additional CNTN3 knockdown counteracted these effects. Conversely, CNTN3 overexpression in PRRX2-knockdown cells restored viability, DNA replication, colony formation and migratory and invasive capacity, and reduced Caspase-3/7 activity. The authors conclude that the PRRX2-CNTN3 axis is essential for colorectal cancer progression in their models, modulating proliferation, apoptosis and metastatic behavior.
The study has acknowledged limitations that temper its clinical implications. The assessment of EMT relied primarily on molecular marker expression rather than high-resolution morphological tracking or single-cell imaging of partial EMT states. The downstream signaling cascades through which CNTN3 acts, potentially including AKT or MAPK/ERK pathways suggested by structural similarity to CNTN1, were not experimentally verified. Clinical relevance rests heavily on public databases such as TCGA and GEO, and independent cohort validation with patient tissue samples is ongoing. Moreover, the xenograft model confirmed tumor growth effects but did not employ a metastasis model, such as tail vein or orthotopic injection, to directly evaluate distant metastatic capacity in vivo. The findings therefore remain preclinical, derived from cell lines, database analyses and xenografts, and do not yet establish diagnostic or therapeutic utility in patients.
Even with those caveats, the work offers a mechanistic insight previously unexplored in other malignancies: a direct transcriptional relationship in which PRRX2 binds to and activates the CNTN3 promoter, integrating transcriptional control with EMT and aggressive cancer phenotypes. Unlike earlier studies that treated EMT as a downstream effect of either factor individually, this research unifies them into a single regulatory axis. The authors suggest that the axis may emerge as a potential therapeutic target for mitigating colorectal cancer progression, and they call for future studies on disrupting PRRX2-CNTN3 signaling, its possible influence on immune modulation and therapy resistance, and the precise signaling pathways downstream of CNTN3. For now, the study adds a candidate mechanism to the molecular map of colorectal cancer metastasis, one that will require independent clinical and in vivo validation before it can inform patient care.
Subject of Research: PRRX2 transcriptional regulation of CNTN3 in colorectal cancer epithelial-to-mesenchymal transition and metastasis
Article Title: PRRX2-CNTN3 axis promotes epithelial-to-mesenchymal transition and metastasis in colorectal cancer
Article References: Yu, S., Li, J., & Li, S. (2026). PRRX2-CNTN3 axis promotes epithelial-to-mesenchymal transition and metastasis in colorectal cancer. Cell Division, 21(1), Article 8. https://doi.org/10.1186/s13008-026-00179-z
Image Credits: AI Generated
DOI: 10.1186/s13008-026-00179-z
Keywords: colorectal cancer, CNTN3, PRRX2, epithelial-to-mesenchymal transition, metastasis, transcriptional regulation, EMT markers, xenograft model, TCGA-COAD, oncogene, cell proliferation, apoptosis
News Source: Nathaniel Bowman. (October 30, 2026). PRRX2-CNTN3 axis drives EMT and metastasis in colorectal cancer. Scienmag.
