BIOPHYSICAL CHARACTERIZATION OF TIGHT JUNCTION PROTEINS
Open Access
- Author:
- Tash, Brian Richard
- Graduate Program:
- Cell and Molecular Biology
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- December 16, 2010
- Committee Members:
- John Michael Flanagan Jr., Dissertation Advisor/Co-Advisor
John Michael Flanagan Jr., Committee Chair/Co-Chair
David Antonetti, Committee Member
Ira Joseph Ropson, Committee Member
Thomas E Spratt, Committee Member
Anthony Edward Pegg, Committee Member - Keywords:
- ZO-1
tight junctions
occludin - Abstract:
- The tight junction (TJ) is a complex of transmembrane, peripheral membrane and signaling proteins that together form a barrier to the flow of solute and molecules through the paracellular space in polarized epithelial cells. The major transmembrane proteins of the TJ are occludin (bicellular-TJ) and tricellulin (tricellular-TJ), which regulate TJ barrier properties, and the claudins, which form the TJ barrier. These proteins are anchored to the actin cytoskeleton through the membrane peripheral scaffolding protein, ZO1. Specifically, the coiled-coil domain of occludin (occCC) located at the distal end of their intracellular C-terminal tail, interacts with the SH3-GuK domains of ZO1. The role of the occludin-ZO1 interaction in the regulation of TJ barrier properties has been studied extensively. Loss of this interaction correlates with reduced barrier properties of the TJ. Studies have supported a role for multi-site phosphorylation of occludin in regulating complex formation. It has been shown that phosphorylation on Ser490, within occCC, is involved in regulation of tight junction barrier properties by controlling binding to ZO1, occludin ubiquitination, and trafficking into early endosomes. However, the molecular mechanisms underlying the occludin-ZO1 interaction are unknown, and how Ser490 phosphorylation regulates TJ properties is obscure. A second site, Ser471, located in the negatively charged head of occCC, was also found to be phosphorylated in vivo, but its role is also unclear. The related, but distinct, tricellulin-ZO1 interaction is less well characterized and its phosphorylation state unclear. The hypothesis of this thesis is that phosphorylation of specific sites on occCC regulates its interaction with ZO1, thereby potentially regulating bicellular TJ complex formation and by extension, the equivalent region in tricellulin and tricellular TJs. To test this, experiments were designed to structurally characterize the interactions between the coiled-coil domains of either occludin or tricellulin, respectively, with the PDZ3-SH3-GuK (PSG) domains of ZO1. A range of biophysical techniques were used to characterize the PSG domains of ZO1 and occCC individually and in a complex, including circular dichroism (CD), hydrogen-deuterium exchange mass spectroscopy (HDEX) , multi-angle laser light scattering (MALLS) and nuclear magnetic resonance (NMR). The functional state of the protein was assessed using nuclear magnetic resonance (NMR) chemical shift perturbation assays. These studies revealed that binding of occCC to the GuK domain of ZO1 altered the structure of its PDZ3 domain, suggesting a mode of communication between the various protein binding sites within ZO1. To gain an understanding of the molecular mechanisms underlying the occludin-ZO1 protein complex and the location of Ser490 and Ser471 in this complex, various biophysical tools were used to characterize structurally occCC in a complex with the PSG domains of ZO1. Initially, small angle X-ray scattering (SAXS) was used to directly visualize protein complex formation in solution. These data revealed that occCC binds to the GuK domain of ZO1. However, due to the pseudo-symmetrical shape of the occCC molecule, it was unclear which of the two potential regions on occCC mediated the interaction. To resolve this ambiguity, NMR experiments were utilized to identify which residues in occCC are at, or near, ZO1 in the protein complex: Residues in the negatively charged head of occCC are adjacent to the ZO1 GuK domain in the complex. Ser490 is ~35 Å from the interface and unlikely to directly affect complex formation, whereas Ser471, in the negatively charged head of occCC, is uniquely positioned at the interface between occludin and ZO1. Together, this led to a model in which phosphorylation of Ser471 would strengthen complex formation, facilitated by ionic interactions between oppositely charged regions in each protein. To determine whether the distal C-terminal tail of tricellulin folds into a coiled-coil similar to occCC, and if so, whether it interacts with the PSG domain of ZO1 in a similar manner to occCC, CD, SEC-MALLS, NMR and SAXS were used. This region of tricellulin does adopt a coiled-coil structure (tricCC) similar to occCC, consistent with the high level of primary sequence similarity between the two. However, SAXS and NMR studies with tricCC and ZO1 PSG demonstrated that these proteins do not interact under the conditions tested. Although Ser471 is conserved in tricCC, the surface charge distribution of the head of tricCC is less negatively charged than the analogous region in occCC consistent with its inability to interact with ZO1 PSG under the conditions tested. Together, the data presented in this thesis identifies which residues in occCC interact with the scaffolding protein ZO1 and implicates Ser471 in complex formation. Further, although tricCC has a similar tertiary structure and a serine at the equivalent position of residue 471 in occludin, the charge distribution is sufficiently different that an analogous complex between tricCC and ZO1 is not formed. This leads to a model in which the negatively charged residues within the head of occCC are critical to mediate its interaction with ZO1. This charge density would be further strengthened by phosphorylation of Ser471.