Vascular endothelial growth factor, or VEGF-A, has a complicated reputation in lung biology. The factor was originally discovered as vascular permeability factor, a molecule secreted by tumor cells that made blood vessels leak fluid at an alarming rate. In the lung, that leakiness is precisely the problem in acute lung injury: when the endothelial lining of pulmonary microvessels becomes permeable, protein-rich fluid floods the airspaces, oxygen exchange collapses, and patients may require mechanical ventilation to survive. Earlier clinical studies had found elevated VEGF in the airspaces of patients with respiratory distress syndrome, and researchers have long debated whether the growth factor is a culprit driving vascular leak or, paradoxically, a repair signal that helps the damaged endothelium recover. The new findings suggest that part of the confusion may have arisen because investigators were measuring a family of molecules with opposite effects as if it were a single entity.

The key to that complexity lies in alternative splicing, the process by which a single gene yields multiple protein versions. In the final stretch of the VEGF-A gene, two different sets of terminal exons can be stitched onto the coding sequence. One choice produces the classical pro-angiogenic isoforms, exemplified by VEGF165a, which bind the receptor VEGFR2 on endothelial cells and trigger the signaling cascades that make vessels grow and leak. The alternative choice produces the so-called b-variants, including VEGF165b, which differ in just six amino acids at the carboxyl terminus. That subtle change, first characterized by researchers studying kidney and cancer biology in the early 2000s, is enough to render the variant weakly angiogenic and poorly permeability-inducing, while still allowing it to occupy the same receptor. The b-isoforms therefore behave as natural antagonists of their pro-angiogenic siblings, competing for receptor binding without delivering the same intracellular message.

The Guangzhou-led team began by asking whether this splice balance is disturbed in human lung injury. Analyzing blood samples from patients, they found that the circulating ratio of VEGF165b to total VEGF-A was significantly reduced in acute lung injury, and that the lower the proportion of the protective variant, the more severe the disease. That inverse correlation, reported alongside ethical approval from the First Affiliated Hospital of Guangzhou Medical University and Zhujiang Hospital of Southern Medical University, provided the first hint that a relative deficiency of VEGF165b might accompany, and possibly contribute to, the vascular catastrophe of respiratory distress syndrome. Correlation alone cannot establish causation, so the investigators turned to experimental models in which the relationship could be tested directly.

In mice, the researchers induced acute lung injury with lipopolysaccharide, the inflammatory toxin derived from bacterial cell walls that mimics Gram-negative sepsis, one of the most common triggers of the human syndrome. When the animals received exogenous VEGF165b, the consequences of the insult were markedly blunted: pulmonary edema diminished, inflammatory infiltration of the lung tissue decreased, and overall measures of injury improved. The converse experiment proved equally telling. When the team neutralized endogenous VEGF165b with antibodies, leaving the pro-inflammatory isoforms to act unopposed, lung damage became worse. Together, these gain-of-function and loss-of-function results established the splice variant as a genuine endogenous protector rather than a passive bystander, and they identified the VEGF165b-to-total-VEGF ratio as a candidate biomarker for tracking disease severity.

The mechanistic core of the study explains how a single splice decision can translate into barrier protection at the molecular level. Using human microvascular endothelial cells, human pulmonary artery endothelial cells, and primary pulmonary microvascular endothelial cells, the researchers showed that VEGF165b acts as a competitive antagonist at VEGFR2, selectively blocking the phosphorylation of the receptor at tyrosine 1175, a docking site whose activation is known to be essential for the permeability response. With that phosphorylation event suppressed, the downstream PI3K/AKT/eNOS pathway, which normally drives nitric oxide production and vessel relaxation during vascular leak, was quieted. The team traced the consequences further downstream to the Src kinase and the adherens junction protein VE-cadherin, the molecular glue that holds endothelial cells together. When Src phosphorylates VE-cadherin, the junctions loosen and fluid escapes; by preventing the upstream signal, VEGF165b reduced VE-cadherin phosphorylation and thereby kept the endothelial barrier intact.

Intriguingly, the protective variant does not stop at junctional reinforcement. Through RNA sequencing and network analysis of the signaling hubs affected by VEGF165b, the investigators found that the variant also suppresses the chemokine CCL2, also known as monocyte chemoattractant protein 1. CCL2 is a potent recruiter of circulating monocytes, and its reduction translated into a second, distinct anti-inflammatory mechanism: fewer monocytes adhered to the endothelial surface, limiting the wave of inflammatory cells that normally amplifies tissue damage in acute lung injury. The authors describe this as a dual mode of action, in which VEGF165b simultaneously strengthens the physical wall of the microvasculature and dampens the cellular inflammatory assault against it. Such a combination is rare among candidate therapies, which typically address either leakiness or inflammation but rarely both.

The findings also reframe a long-running therapeutic debate. Several anti-VEGF strategies, including the monoclonal antibody bevacizumab, have been explored for severe COVID-19 and other causes of pulmonary edema on the logic that blocking VEGF should reduce vascular leak. But blanket VEGF blockade carries well-documented risks, including hypertension, proteinuria, and thromboembolism, precisely because the growth factor performs essential maintenance roles in healthy vessels. The splice-variant view offers a more surgical alternative: rather than eliminating VEGF signaling altogether, one could in principle restore the balance between pro-permeability and anti-permeability isoforms, either by administering recombinant VEGF165b, which has already been tested as an anti-cancer agent in animal models, or by shifting the splicing machinery toward the protective variant. The study’s demonstration that neutralizing VEGF165b worsens injury adds a note of caution for any therapy that removes this natural brake indiscriminately.

As with any preclinical advance, the road from mouse models and cell cultures to the intensive care unit is long. The patient data establish an association between a low VEGF165b fraction and disease severity, but clinical trials will be needed to determine whether replenishing the variant improves outcomes, at what dose, and at what stage of illness. Questions also remain about how the splicing balance is regulated in the acutely inflamed lung, and whether splicing factors such as the SRSF family, which other studies have implicated in acute lung injury, could serve as drug targets. Nevertheless, the conceptual payoff is substantial. The work reveals that vascular homeostasis in the lung depends on the critical equilibrium of VEGF splice variants, positions VEGF165b as both a biomarker and a potential therapeutic agent for acute lung injury and respiratory distress syndrome, and offers a new strategy for treating vascular barrier dysfunction in critical lung disease, one of the most stubborn unsolved problems in modern critical care medicine.

Subject of Research: The role of the VEGF-A splice variant VEGF165b in protecting against acute lung injury by antagonizing VEGF165a/VEGFR2 signaling

Article Title: VEGF165b attenuates LPS-induced acute lung injury via antagonizing the VEGF165a/VEGFR2 signaling pathway

Article References: Deng, J., Wang, X., Wu, Z., Bai, Q., Luo, Q., Zhou, L., Yao, Y., Wang, H., Zhong, K., Liu, H., Chen, X., & Wang, T. (2026). VEGF165b attenuates LPS-induced acute lung injury via antagonizing the VEGF165a/VEGFR2 signaling pathway. Angiogenesis, 29(4), Article 75. https://doi.org/10.1007/s10456-026-10090-4

Image Credits: AI Generated

DOI: 10.1007/s10456-026-10090-4

Keywords: acute lung injury, ARDS, VEGF165b, VEGFR2, vascular permeability, endothelial barrier, alternative splicing, VE-cadherin, CCL2, lipopolysaccharide, angiogenesis, pulmonary edema