IDENTIFICATION AND FUNCTIONAL ANALYSIS OF OCCLUDIN PHOSPHORYLATION

Open Access
- Author:
- Sundstrom, Jeffrey
- Graduate Program:
- Cell and Molecular Biology
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- August 23, 2006
- Committee Members:
- David A Antonetti, Committee Chair/Co-Chair
John Michael Flanagan Jr., Committee Member
Robert A Gabbay, Committee Member
Bruce A Stanley, Committee Member
Christopher J Lynch, Committee Member - Keywords:
- AKT1
tight junction
phosphorylation
occludin
VEGF - Abstract:
- The critical importance of tight junctions in defining various tissue compartments is highlighted by the multiple disease states associated with its dysfunction. Our laboratory is specifically interested the molecular mechanisms underlying elevated retinal microvascular permeability associated with diabetic retinopathy. This breakdown of the blood-retinal barrier is known to be at least partially caused by an elevation in growth factors and cytokines. At the molecular level, these alterations are associated with a loss of junctional integrity that is, at least in part, related to an observed increase in occludin phosphorylation, a redistribution of occludin away from cell borders, and a decrease in occludin content. Critical to the experimental system used in this thesis, this pathology has been effectively modeled in vitro as VEGF stimulation of primary retinal endothelial cells results in increased occludin phosphorylation associated with increased permeability. The role of occludin in regulating the barrier properties of tight junctions has been extensively studied. In addition to changes in permeability associated with changes in occludin content, several groups reported an association between the state of occludin phosphorylation and the integrity of the paracellular barrier. Studies utilizing 2D-gel analysis suggest that occludin is phosphorylated on multiple residues in the basal state and three additional residues are phosphorylated in response to growth factors. Therefore, it is not surprising that studies relying on end points that assess global changes in occludin phosphorylation have often yielded conflicting results. Given that occludin is phosphorylated on multiple sites, a consensus regarding the state of occludin phosphorylation and tight junction dynamics will depend upon the identification of specific, in vivo phosphorylation sites. The goal of this thesis was to determine the mechanism by which phosphorylation of occludin leads to alterations in tight junction function. The hypothesis of this dissertation is that VEGF-induced phosphorylation of occludin results in attenuated binding to scaffolding proteins at tight junctions and/or the redistribution of occludin away from tight junctions and into early endosomes. The first aim of these studies was to map physiologically relevant occludin phosphorylation sites. To this end, we partially purified occludin from primary endothelial cells treated with VEGF and identified five putative phosphorylation sites using mass spectrometry and bioinformatics-based inclusion criteria. Phosphorylation of occludin on Ser490, a residue located within the C-terminal coiled-coil domain, was subsequently shown to occur in vivo with the development of a phosphospecific antibody. Growth factors known to increase permeability stimulated Ser490 phosphorylation in both epithelial and endothelial monolayers. Immunocytochemistry studies revealed that pSer490-occludin is localized in cytoplasmic puncta that, in response to growth factors, coalesced into vesicles that appear to be early endosomes. Protein interaction studies indicate that phosphorylation at Ser490 attenuates the interaction between occludin and ZO-1 while crystal structure analysis of S490D-occludin413-522 demonstrates a disruption in the positively charged ZO-1 binding surface. Taken together, these data suggest a critical role for the coiled-coil domain of occludin in the targeting occludin to tight junctions. Specifically, growth factor stimulation of Ser490 phosphorylation acts as a molecular switch to disassemble occludin from tight junctions. This is accomplished, at least in part, by an attenuated interaction with ZO-1 and may also involve the movement of occludin into endosomes. The second aim of this thesis was to facilitate the identification of in vivo phosphorylation sites. To this end, we developed MODICAS, software designed to identify PTMs. Peptide mapping was conducted by MODICAS and merged with bioinformatics-based inclusion criteria. This approach was tested on AKT1 partially purified in triplicate from a single retina. While the unambiguous identification of phosphorylation sites with tandem MS was not successful, our combined approach resulted in the identification of 9 phosphorylation sites. Importantly, 6 of these 9 sites have been previously identified by others as in vivo AKT1 phosphorylation sites. As such, this combination of empirically derived MS data with bioinformatic-based inclusion criteria was shown to be extremely efficient and accurate in the identification of phosphorylation sites on AKT1. This approach should prove useful in several contexts, such as clinical samples, in which limited sample material prevents the acquisition of quality tandem MS spectra. Collectively, these studies strongly suggest that Ser490 phosphorylation of occludin plays a central role in regulating the barrier properties of tight junctions. In addition, several other sites of occludin phosphorylation have yet to be further characterized and may further illuminate the role played by occludin in the regulation of tight junction function. Importantly, understanding the mechanisms involved in the inappropriate remodeling of tight junctions may prove to be pivotal in the design of novel therapeutics for treatment of diseases such as diabetic retinopathy. These studies illustrate the importance of monitoring regulatory PTMs, the identification of which we have greatly facilitated by coupling novel phosphopeptide mapping software with bioinformatics.