Research / Clinical Summary

Brian Eliceiri, PhD Associate Professor, Surgery Tumor Growth, Invasion & Metastasis Program Contact by Email

Dr. Eliceiri’s research has been most influenced by two fundamental observations relating to endothelial barrier function: 1) cell-matrix-cell interactions modulate growth factor signaling and drug responsiveness in the vascular endothelium, and 2) cell-cell signaling is profoundly different in the endothelium of in vivo animal models and in vitro cell culture systems of endothelial dysfunction.

1) Cell-matrix-cell signaling
He and his colleagues have demonstrated that growth factors such as basic fibroblast growth factor (FGF2) and vascular endothelial growth factor (VEGF) signal by cooperative mechanisms through specific vintegrins established a new paradigm of intercellular signaling. Both growth factors and components of the extracellular matrix components combine to create unique signaling microenvironments. With this in mind, genetic models of disease can be used to study the scope of signaling diversity that can theoretically be achieved with combinations of molecules in various cell types or organs.

Building upon their observation that in vivo FGF2 induces integrin-dependent sustained kinase activity in the vascular endothelium that is absent in vitro, they are examining the role of additional components of matrix-dependent endothelial signaling. For example, although the non-receptor tyrosine kinase, Src, is expressed many cell types, Src selectively regulates VEGF signaling in the vascular endothelium in vivo. This observation could not have been predicted in vitro. They further analyzed the capacity for Src to modulate matrix-dependent endothelial signaling by demonstrating that Src-mediated phosphorylation of focal adhesion kinase (FAK) regulates the formation of a transient FAK-integrin v5 complex in vivo. While Src phosphorylates multiple sites in FAK in vitro, a site- and tissue-specific distribution of FAK phosphorylation was observed in response to VEGF, suggesting that the regulation of widely expressed molecules is selectively regulated depending on the cell-matrix context in vivo.

2) In vivo modeling of the endothelial barrier function
Cell-matrix-cell signaling is of paramount importance in the regulation of barrier integrity in the brain and on drug biodistribution. They exploit animal models with specific genetic modifications to test our central hypothesis that blood brain barrier (BBB) function in the brain is regulated by specific cell-cell communication within a neurovascular unit (i.e. endothelia, basal lamina, and astrocytes). Specifically, they employ advances in transgenic reporter mice and non-invasive imaging to: a) analyze BBB signaling and cell biology in vivo; b) measure consequences of BBB breakdown and neuroprotection; and c) apply non-invasive imaging to study changes in BBB integrity and monitor drug delivery to the central nervous system.

Their studies of VEGF-Src endothelial signaling in the regulation of cell-matrix-cell interactions have focused on the use of transgenic, knockout and conditional knockout mouse models for the functional analysis of specific BBB components. For example, they have shown that Src knockout mice have a leakage-resistant phenotype in which VEGF-induced vascular permeability is blocked in brain capillaries. Downstream of Src, they use a conditional knockout model of FAK to target the inducible and endothelial specific knockdown of endogenous FAK in the brain. To examine the capacity for agrin, a proteoglycan expressed in the basal lamina, to modulate growth factor bioavailability, transgenic and knockout agrin mice are used to assess signaling in endothelial cells and astrocytes. In this BBB model, the signaling consequences of growth factor bioavailability provides (1) the basis for analyzing disease-induced BBB breakdown and (2) genetic and pharmacological strategies for neuroprotection.

This versatility is demonstrated using the Src model to dissect the role of BBB breakdown on tumor infiltration. Eliceiri and his colleagues have shown that glioma invasion is blocked in Src-knockout mice, and is associated with decreased fibrin, a provisional perivascular matrix protein. This observation establishes a functional link between the BBB breakdown that is induced by tumor cells and the nascent accumulation of a matrix protein, which glioma cells exploit to infiltrate the parenchyma. Similarly, their studies of BBB breakdown following cerebral ischemia reveal that Src knockout mice are protected from ischemia-induced edema and brain damage. Furthermore, administration of Src inhibitors following stroke in mice leads to reduced edema and brain damage thereby recapitulating the neuroprotection observed in Src knockout mice. This first generation small molecule Src inhibitor has led to a second generation of Src inhibitors that are currently entering Phase I human clinical trials.

While a combination of cell biology and genetic techniques can focus on a specific molecule for further drug development, they also employ advances in imaging to examine changes in BBB integrity in intravital and non-invasive animal models. For example, the dual channel fluorescent analysis of the biodistribution of drug candidates (labeled with near-infrared dyes) in parallel with the fluorescence of a BBB component (e.g. green endothelium of a TIE2-GFP mouse) localizes a drug in the specific context of the BBB. Although several fluorescence strategies can be employed to track BBB breakdown and cell migration in vivo, we have observed that intravital imaging is best complemented by non-invasive imaging techniques. For this purpose, luminescence is well-suited to the imaging from deep with the brain where cell migration and astrocyte activation are monitored with firefly luciferase. The labeling of candidate drugs with renilla luciferase, a luciferase with distinct luminescent properties, enables biodistribution analysis in the context of animal models of BBB disruption.

They believe that the combination of a cell-matrix-cell model of BBB function addresses a critical challenge in targeted drug design. Advances in genetic models and imaging will facilitate the translation of candidate drugs optimized for personalized genetics and designed for specific BBB dysfunctions.

Update Summary via ONcLINE (password required)
Click here to request a new or forgotten password