Yasmin Hashambhoy, a postdoctorate fellow in the Feilim Mac Gabhann lab at the Hopkins Department of Biomedical Engineering, is one step closer to fully understanding the mechanisms that govern angiogenesis, the growth of new blood vessels from an existing vascular network. Growth of the new blood vessels is directed by concentration gradients of vascular endothelial growth factor (VEGF), and the team has developed a computational model to simulate how the concentration gradients are regulated by a secreted inhibitor of VEGF. Their results were published this past October in Frontiers in Physiology.
New blood vessels are needed to supply tissues with blood and oxygen during development, tissue repair and cancer. "Ultimately, we'd like to understand how to regulate tumour angiogenesis effectively," Hashambhoy wrote in an email to The News-Letter.
The VEGF family of extracellular ligands is secreted by developing vessels in mouse embryonic stem cells, the model system that Mac Gabhann lab collaborators at the University of North Carolina use to study angiogenesis. Blood vessels in developing mouse embryos express membrane-bound tyrosine kinase receptors for VEGF. When VEGF binds two of these receptors simultaneously, the receptors activate cell proliferation and migration. Because the cells of the developing vasculature communicate with each other in order to prevent each cell from forming a new blood vessel sprout, sensitivity to VEGF gradients is crucial. During sprouting of the blood vessel, the cell with the highest local concentration of VEGF is most likely to become the tip cell of the growing sprout that will then determine the direction of growth by migrating towards increasing VEGF concentrations.
The concentration of VEGF is controlled by a number of mechanisms. Soluble VEGF becomes bound to receptors on the extracellular surface of existing blood vessels, therefore, reducing the local concentration of soluble VEGF. In order to further reduce the local concentration of VEGF, existing blood vessels secrete a soluble version of one of the receptors.
This soluble protein, sFlt-1, competes for VEGF binding with the membrane-bound signaling version. In an environment that already has adequate vasculature, large amounts of sFlt-1 bind VEGF and, therefore, prevent it from binding the membrane receptors that activate cell growth and proliferation. The existing blood vessels "sense" each other in this way and further growth is inhibited.
"Computational modeling is a very useful tool — we can use it to help us understand systems at a higher spatial and temporal resolution than would be available through experiments alone. A really useful model can make predictions that drive the next set of experiments. But models need to be constrained by and tested with experiments," Hashambhoy wrote in regards to her computational work with VEGF gradients. "This gives us a deeper understanding of biological systems and also helps to improve the model."
Computational modeling is an attempt to reproduce, predict and understand molecular phenomena using basic physical principles and parameters derived from experiments. Agent-based models, like those developed by the collaborators of the Mac Gabhann lab at the University of Virginia, "[are] a class of models where there are individual agents that act autonomously in response to their environment. Of course, their responses then affect the environment and are able to affect agents around them," she wrote. "For example, you might model a group of cells that each respond to receptor activation on their membranes by moving in a particular direction or by dividing. What's really interesting is seeing how the whole system responds over time."
Differential mathematical equations were written to describe the specific model system or problem, and then incorporated in computer codes. For example, Hashambhoy and colleagues derived diffusion equations for VEGF and sFlt-1 that include interactions with the extracellular matrix and degradation rates of the soluble proteins in the interstitial space. When the differential equations are solved and run on the modeling program Fortran, changes in VEGF and sFlt-1 concentration gradients are tracked over time from a set initial conditions. Other considerations that add to the complexity to the model include rates of secretion of VEGF and sFlt-1, insertion and internalization of membrane receptors, and relevant binding affinities and kinetics. The ultimate goal of the model was to determine how the concentration gradients of unbound VEGF vary around different parts of a blood vessel sprout, with and without sFlt-1 to inhibit VEGF.
The results of the model reproduce experimental results well; after 10 hours of simulated sFlt-1 secretion, the model predicts a very low concentration of VEGF and hence reduced activation of membrane receptors near the stalks of blood vessel sprouts. "When sFlt1 signaling is removed in the model, the simulation results predict that sprouts will emerge closer to each other. These predictions match what is seen in experiments by our collaborators, who observe that in the WT case, the distance between sprouts is greater than in the case where Flt1 is knocked out," Hashambhoy wrote.
Understanding how this secreted inhibitor directs blood vessel growth in healthy tissue has valuable consequences for drug development strategies targeting growing cancerous tumors. "This model is a useful tool for helping to understand how sFlt1 affects sprout growth. This can be applied to disease models such as cancer by incorporating tumour cells, which have different receptor concentrations and VEGF secretion rates.," Hashambhoy wrote. "This would affect the endothelial cell environment, and affect cell response. Ultimately, we'd like to understand how to regulate tumour angiogenesis effectively."
