TIE2 Receptor Circuit Drives Venous Malformation Growth Beyond PI3K Mutations

Venous malformations are chronic vascular anomalies affecting roughly 1 in 5,000–10,000 individuals, ranging from superficial lesions to life-threatening masses. They are predominantly caused by somatic gain-of-function mutations in PIK3CA (encoding PI3Kα) or TEK (encoding TIE2), both of which hyperactivate the PI3K–AKT–mTOR signaling axis in endothelial cells. Despite therapeutic use of rapamycin (sirolimus) and alpelisib, established VMs rarely resolve completely, motivating the search for additional disease-driving mechanisms.

Using two complementary Pik3ca-H1047R mouse models — one with mosaic endothelial expression driven by Tie2-Cre and another with inducible Cdh5-CreERT2 — researchers performed single-cell RNA sequencing on thousands of endothelial cells from VM lesions versus normal tissue. Lineage tracing confirmed that mutant cells clonally expand and adopt a post-capillary venous identity, expressing canonical venous markers while suppressing arterial gene programs. This transcriptional shift was accompanied by strong suppression of FOXO1, a transcription factor that AKT directly phosphorylates and inactivates. FOXO1 target genes — most critically ANGPT2 — were markedly downregulated in mutant endothelial cells compared to wild-type neighbors.

ANGPT2 normally functions as an endothelial-derived TIE2 antagonist that limits receptor activation. Its suppression by PI3K-driven AKT/FOXO1 signaling thus removes a key brake on TIE2. Simultaneously, aberrant smooth muscle cell (SMC) recruitment to VM lesions — verified both in mouse models and in resected human VM specimens — delivers excess ANGPT1, a potent paracrine TIE2 agonist. Phospho-TIE2 staining confirmed elevated receptor activation in both mouse and human VMs. This ligand imbalance (less ANGPT2, more ANGPT1) creates a feedforward loop: PI3K hyperactivity → FOXO1 suppression → ANGPT2 loss + SMC-derived ANGPT1 gain → TIE2 hyperactivation → further PI3K stimulation.

Critically, the study tested whether blocking downstream mTOR (via rapamycin) could resolve advanced lesions in mice with established VMs. Despite biochemical pathway inhibition, mTOR blockade had limited effects on lesion size or endothelial cell number at this stage. In contrast, pharmacological TIE2 inhibition (using rebastinib) or ANGPT neutralization substantially suppressed VM growth, reducing lesion area and endothelial cell expansion. These experiments were conducted in mice with established lesions treated over several weeks, providing a pre-clinical proof of concept for targeting the upstream receptor rather than downstream effectors.

Human tissue validation strengthened translational relevance: immunostaining of surgically resected VM specimens from patients showed the same pattern of reduced ANGPT2 expression, excessive SMC infiltration, and elevated pTIE2 compared to normal venous tissue. The convergence of mouse genetic models and human pathology data supports the feedforward PI3K–FOXO1–ANGPT–TIE2 circuit as a conserved disease mechanism. The authors propose that TIE2 inhibitors, already in clinical development for other indications, warrant evaluation in PIK3CA-driven VMs as single agents or in combination with PI3K-targeted therapies.