Guidance cues and pattern prediction in the developing retinal vasculature: a combined experimental and theoretical modelling approach

Summary The aim of this project is to use the latest mathematical modelling (MM) techniques coupled with state-of-the-art 3-D imaging to discover how the final patterning of the mouse retinal vasculature plexus (RVP) is regulated in both normally developing mice and mice with severe vascular defects (VEGF-transgenics); similar to those observed to cause a lifetime of blindness in human babies . As the retina grows, its metabolic needs are supported by a 3-D network of blood vessels that form a characteristic pattern of capillary networks linking the arterioles and veins in its different layers. The final structure of the RVP is determined by molecular pathways in the tissue, on which the endothelial cells (the main cell forming the blood vessels) and pericytes (cells which give structural/functional stability to the vessels) migrate. The direction of migration of these cells is dependent on the concentration gradients of both soluble and matrix-associated factors which chemically (chemotaxis) attract the cells towards the rim of the optic cup. Using confocal and 2-photon microscopy (allowing the collection and assimilation of 3-D images of cells, pathways and chemotactic agents) we will examine the retina from different developmental stages of normal and neonatal mice with vascular malformations to discover which molecular pathways and chemotactic agents are critical in determining the patterning and final maturation of the RVP. The images are generated by labelling cells with specific dyes (for instance endothelium with an isolectin called BSI-B4; perictes, with a probe against smooth muscle actin and pathways/chemotactic agents with specific antibodies) exciting the tissue with a laser and capturing the light from the fluorescing dyes with a confocal microscope. As all the images collected in these studies are essentially snapshots of what happens at one particular time during the growth of the RVP, it is essential to be able to overlay these results and discover how the cells respond to the underlying expression patterns of the pathways and chemotactic agents being produced over time (temporally). This is where the powerful tool of MM can lead to new discoveries about how the combination of events (at the tissue, cell and molecular level) are regulated. The first proposed MM will initially rely on data generated from studies performed in normal and VEGF-transgenic neonatal mice. After collecting, digitising and quantitating a series of parameters (i.e. vessel lengths, branch-points, fractal dimension [how often a basic pattern is repeated at different scales], cell-type, location and concentration of molecules) this information is used to inform the MM, so that a virtual model of the RVP can be generated. This MM is then verified and improved, by testing its ability to predict the RVP patterning at later stages during development. Gaps in the biological data (in both the pathways and the chemotactic gradients) can be anticipated by the MM that will then be used to inform biological experiments by proposing new studies which will further elucidate the cellular and molecular mechanisms underlying the development of the 3-D structure of the RVP. This is a unique collaboration between biologists and mathematicians, in which both disciplines are instructive in discovering how a complex 3-D tissue, the RVP, grows and contributes to the final structure of the eye in both normal animals and animals with a serious ocular pathology. The final portion of the project will employ the MM to predict which therapeutic approaches to treat the VEGF-transgenic mice, will prevent progression of ocular pathology. This research will benefit basic biological understanding of how blood vessels grow in 3-D (which determines the growth of all organs), how the eye grows normally and more specifically in ocular conditions characterised by inapprpriate blood vessel formation (all major diseases that cause blindness in neonates and adults).

The retinal vascular plexus (RVP) is a complex 3-D structure which supports the metabolic demands of a rapidly developing and differentiating retina during neonatal life in mammals. The architecture is tightly regulated by coupling of angiogenic sprout growth along an underlying structural matrix, in response to chemotactic cues which are spatio-temporally resticted during the matruation of the RVP. Since the direct visualization of tissue, cellular, matrix and chemotactic cues in a temporal fashion is currently unattainable, we will combine the latest confocal/2-photon microscopic imaging techniques in 3-D (along with parameterization of these images) with mathematical modelling (MM) in order to generate a temporally-resolved 3-D architecture of the RVP, which recapitulates experimental observations. This should result in a realistic, temporally-defined 3-D vascular growth pattern, which grows along the pathways and responds to chemotactic gradients as defined from the in vivo studies. Initially the MM will entirely rely on data generated from in vivo studies, with verification and improvement of the model being performed. Later iterations of the MM, will be used to inform biological studies (such as focussing on specific chemotactic cues, time-points or indicators of cellular behaviour), thus integrating the biological and MM approaches. This is a unique collaboration between biologists and mathematicians, in which both disciplines are instructive in discovering how a complex 3-D tissue, the RVP, grows and contributes to the final structure of the eye in both normal animals and animals with a serious ocular pathology. Finally, the MM will be used to predict how agents can modify RVP growth in both normal and VEGF-transgenic mice, with one aim being to prevent progression of ocular pathology. This research will benefit basic biological understanding of how blood vessels grow in 3-D (which determines the growth of all organs) and more specifically in the eye.