Blood Vessel Protein Epac1 Found to Fuel Melanoma Growth Through YAP/TAZ Signalling

Scientists at Heidelberg University have identified a molecular switch inside blood vessel cells that appears to be a critical enabler of melanoma growth, opening a potential new avenue for attacking one of the deadliest forms of skin cancer. The protein, called Epac1, is produced in unusually large amounts by endothelial cells — the cells that line blood vessels — within human melanoma tumors. When the researchers removed Epac1 from the endothelium of mice, melanomas grew significantly more slowly and built far fewer blood vessels, starving the tumors of the supply network they need to expand.

The study, published in the journal Angiogenesis, focuses on a long-standing puzzle in cancer biology: how tumors recruit and remodel blood vessels to sustain their growth. Tumors cannot enlarge beyond a few millimeters without recruiting new vasculature, a process known as tumor angiogenesis. The vessels that form inside tumors are typically abnormal — leaky, tortuously branched, and poorly covered by supportive pericyte cells — which creates hypoxic, acidic conditions that paradoxically stimulate even more angiogenic signalling, most notably through vascular endothelial growth factor, or VEGF, and its receptor VEGFR2. The new work reveals that Epac1 sits at a central junction of this process, coordinating both the chemical VEGF signals and the mechanical forces that shape tumor vessel formation.

Epac1, encoded by the gene RAPGEF3, is a guanine nucleotide exchange factor activated by the intracellular messenger cyclic AMP. It exists in two isoforms, but only Epac1 is expressed in endothelial cells, where it participates in regulating junctional dynamics and cell polarity. To determine whether Epac1 matters in tumor angiogenesis specifically, the team first analyzed publicly available single-cell RNA sequencing data from human primary melanomas and healthy human skin. They found that RAPGEF3 expression was markedly enriched in the endothelial cells of melanoma tissue compared with the endothelium of healthy skin, and that the gene was predominantly active in vascular rather than lymphatic endothelial cells. This human evidence suggested that the protein might be a genuine feature of the tumor vasculature rather than an artifact of experimental models.

The researchers then turned to mouse models. In mice lacking Epac1 throughout the body, subcutaneously implanted B16F10 melanoma cells still formed palpable tumors, but growth slowed dramatically after day fourteen — precisely the stage at which vascularization becomes the limiting factor for tumor expansion. By day eighteen, tumors in the knockout animals were roughly half the size of those in normal mice, and staining for the endothelial marker CD31 revealed a significantly smaller vascular area within them. Because global deletion could not distinguish effects in blood vessel cells from effects in other stromal cells, the team generated a second mouse line in which Epac1 could be deleted selectively in endothelial cells using a tamoxifen-inducible Cre system. The result was the same: reduced tumor volume and reduced vascularization, confirming that the endothelial supply of Epac1 itself drives melanoma growth.

To understand the molecular mechanism, the researchers isolated tumor endothelial cells from the melanomas and performed whole-transcriptome sequencing. In the Epac1-deficient vessels, gene ontology analysis showed broad downregulation of angiogenesis-associated pathways, including sprouting angiogenesis, endothelial cell migration, and proliferation. Key pro-angiogenic genes such as Kdr (which encodes VEGFR2), Angpt1, Hif1a, Dll1, and Dll4 were all reduced. Strikingly, the analysis also revealed suppressed expression of canonical target genes of the transcriptional co-activators YAP and TAZ — including Ccn1, Ccn2, and Ankrd1 — pointing to a previously unknown connection between Epac1 and the Hippo signalling pathway in tumor vessels.

YAP and TAZ are mechanosensitive transcriptional regulators that act as major effectors of the Hippo pathway. Their activity depends less on how much protein is present and more on where the protein resides: when they translocate into the nucleus, they partner with TEAD-family transcription factors to switch on genes that promote cell proliferation, migration, and survival. In the Epac1-deficient endothelial cells, the researchers found that total YAP and TAZ protein levels were unchanged, but nuclear localization was significantly reduced, and TEAD-dependent reporter activity was markedly diminished. A parallel experiment using a selective pharmacological inhibitor of Epac1’s enzymatic activity in human umbilical vein endothelial cells produced the same suppression of YAP/TAZ target gene expression, confirming that the catalytic function of the protein is required to maintain this transcriptional program.

The team then traced how Epac1 links VEGF signalling to YAP/TAZ activation. In control endothelial cells, VEGF stimulation robustly induced expression of Vegfr2 and Ccn2, activated the small GTPase RhoA, and promoted phosphorylation of myosin light chain 2, a key downstream target of the RhoA effector kinase ROCK. In Epac1-deficient cells, all of these responses were blunted or abolished. Pharmacological inhibition of RhoA with C3 transferase reduced TEAD-driven transcription and suppressed VEGF-induced Ccn2 expression, while VEGF-driven nuclear translocation of YAP and TAZ in human endothelial cells was fully prevented by RhoA inhibition. Together, these experiments establish a signalling chain running from Epac1 through VEGFR2, RhoA, and the actin cytoskeleton to the nucleus, where YAP and TAZ execute their pro-angiogenic gene program.

Perhaps the most striking findings concerned mechanosensing — the ability of endothelial cells to respond to physical forces such as blood flow. Tumor vessels experience disturbed, oscillatory flow patterns that promote pro-angiogenic signalling and metastasis. Using an in vitro system that applies oscillatory shear stress to endothelial monolayers, the researchers showed that control cells aligned themselves with the direction of flow, whereas Epac1-deficient cells completely failed to do so. Under the same conditions, oscillatory shear stress normally increases the interaction between VE-cadherin, the junctional adhesion protein, and VEGFR2 — a partnership thought to form part of the endothelial mechanosensory complex. Proximity ligation assays revealed that this interaction was abolished in Epac1-deficient cells, as was shear-induced phosphorylation of VE-cadherin at tyrosine 658, an essential modification for mechanosensing, and the shear-induced nuclear accumulation of YAP and TAZ. When the researchers reintroduced Epac1 into the knockout cells using lentiviral vectors, VEGFR2 and Ccn2 expression, VEGF responses, and flow-induced alignment were all rescued, confirming the specificity of the effect.

The study also carries an important nuance about Epac1’s role in vascular biology. Previous work by the same group and others has shown that Epac1 is dispensable during physiological angiogenesis — the retinal vascular area of Epac1-deficient mice at postnatal day five is comparable to that of wild-type animals — yet pathological neovascularization in oxygen-induced retinopathy is significantly reduced without the protein. The new melanoma data fit this pattern: endogenous endothelial Epac1 appears to be largely irrelevant for normal vascular maintenance but becomes functionally indispensable in pathological settings characterized by excessive VEGF signalling and altered mechanical cues. This context-dependence is encouraging from a therapeutic standpoint, because it suggests that drugs targeting Epac1 might impair tumor vessel growth while sparing the normal vasculature, potentially reducing the side effects that complicate current anti-angiogenic therapies.

The authors caution that their study used melanoma as the tumor model, and that the prominent upregulation of Epac1 in tumor endothelium and its consequences for tumor growth may not generalize to all cancer types. Nevertheless, the identification of Epac1 as a central modulator that integrates VEGF signalling, RhoA-dependent cytoskeletal dynamics, YAP/TAZ transcription, and flow mechanosensing provides the most complete mechanistic picture to date of how a single endothelial protein can orchestrate pathological tumor vascularization. With Epac inhibitors already under investigation for other cancers, the prospect of starving melanomas by dismantling the molecular machinery of their blood vessels has moved a tangible step closer to the clinic.

Subject of Research: The role of endothelial Epac1 in regulating VEGFR2 and YAP/TAZ signalling during melanoma angiogenesis

Article Title: Endothelial Epac1 facilitates YAP/TAZ controlled melanoma growth and angiogenesis

Article References: Wibowo, Y. C., Ma, N., Ren, Y., Cordero, J., Gahn, J., Chen, Z., Levay, M., Ola, R., Feng, Y., Dobreva, G., Langer, H., Wieland, T., Vettel, C., & Jansen, S. (2026). Endothelial Epac1 facilitates YAP/TAZ controlled melanoma growth and angiogenesis. Angiogenesis, 29(4), Article 72. https://doi.org/10.1007/s10456-026-10080-6

Image Credits: AI Generated

DOI: 10.1007/s10456-026-10080-6

Keywords: Epac1, melanoma, angiogenesis, tumor endothelial cells, VEGFR2, YAP/TAZ, Hippo signalling, RhoA, mechanotransduction, VE-cadherin, VEGF, skin cancer