Beta-arrestins are well-known mediators involved in both activation and termination of G-protein-coupled receptor (GPCR) signaling pathways. Beta-arrestins transduce signals by binding to, and either activating or inhibiting, various downstream signaling proteins, leading to a variety of cellular responses, which can occur independent of the heterotrimeric G-protein coupling. Since about 40% of currently used therapeutics target GPCRs, understanding the functions of beta-arrestins is an important biological question. We established a model that shows beta-arrestins mediate the actin cytoskeletal reorganization downstream of protease-activated receptor-2 (PAR-2) by temporally and spatially controlling cofilin activation. Cofilin severs actin filaments and is controlled by opposing actions of LIM kinase (LIMK) (which inactivates it by phosphorylation on Ser3) and cofilin-specific phosphatase, chronophin (CIN) that activates it. My study elucidated how beta-arrestin-1 can interact and regulate LIMK and cofilin in PAR-2-enhanced actin dynamics. First, beta-arrestin-1 bound and inhibited LIMK activity directly. At least two specific binding sites on beta-arrestin-1 to LIMK1 or to cofilin were identified within residues 1 to 99 and 183-418 by using multiple recombinant beta-arrestin-1 truncations and a deletion mutant (residues 146 to 182 were deleted) in sandwich immunoassays. The deletion mutant showed high binding affinity and direct inhibition of LIMK1, which suggested that if beta-arretsin-1 is induced into the appropriate conformation for exposing the amino and carboxyl binding regions, beta-arrestin-1 can directly bind and inhibit LIMK1 activity. Beta-arrestin-1 may protect cofilin from phosphorylation and inactivation, too. Results from spot peptide arrays confirmed the amino region of beta-arrestin-1 was the specific binding site for LIMK1. Co-immunoprecipitation and bioluminescence resonance energy transfer assay further proved the interaction of beta-arrestin-1 and LIMK1 in cells. The absence of beta-arrestins disrupted the cofilin activation and caused abnormal development of growth cones and mature dendritic spines in mouse hippocampus. This dissertation demonstrated that the specific regions in beta-arrestin-1 scaffold and regulate the kinase activity of LIMK1. Thus, beta-arrestin-1 contributes to the spatial regulation of cofilin activation by sequestering cofilin with CIN while inhibiting LIMK at cell protrusions.

ABSTRACT OF THE DISSERTATION

The Pro-Inflammatory but not Protective Effects of Protease-Activated-Receptor-2 In the Airways are Abolished in Beta-Arrestin-2 Mice in OVA, Cockroach Frass and Alternaria-Induced Models of Allergic Asthma

by

Heddie Lynn Nichols

Doctor of Philosophy, Graduate Program in Biomedical Sciences

University of California, Riverside, December 2012

Dr. Kathryn A. DeFea

Proteinase Activated Receptor-2 (PAR₂), a G Protein Coupled Receptor (GPCR) activated by serine-like proteases, is reported to have both protective and pro-inflammatory effects in the airway. Given these dual and apparently opposing actions, both inhibitors and activators of PAR₂ have been proposed as therapeutics for asthma. PAR₂ can signal through two independent pathways: a G-protein-dependent and a beta--arrestin-2-dependent/G-protein-independent one. The beta-arrestin-dependent pathway promotes leukocyte migration, while bronchiolar smooth muscle relaxation requires G-protein signaling intermediates. These studies address the hypothesis that inflammatory responses to PAR₂ activation are mediated by beta-arrestins, while prostaglandin production and smooth muscle relaxation are not. Our initial studies focused on a mouse ovalbumin model for PAR₂-modulated airway inflammation to focus specifically on PAR₂ effects. During the course of our studies, two models of PAR₂ dependent airway inflammation were introduced. They are: Alternaria alternata, a fungus that commonly grows in homes and on plant; and Cockroach Frass from Blatella germanica a common household pest. These models induce a more robust inflammatory response and are more physiologically relevant as they used common household allergens. As determined by flow cytometry, cytospin and immunohistochemistry, PAR₂-induced overall lung inflammation, mucus production, airway responsiveness and recruitment of eosinophils and CD4+-lymphocytes to the Broncho Alveolar Lavage Fluid (BALF) were abolished in beta-arrestin-2-/-, compared with wild type mice. These results were exacerbated in our AltA and BG models. In contrast, PAR₂ promoted equivalent bronchial epithelium-dependent tracheal smooth muscle relaxation and production of PGE2 in both wild type and beta-arrestin-2-/- mice. Our data suggest that the PAR₂-enhanced inflammatory process is beta-arrestin-2-dependent, while the `protective' anti-constrictor effect of bronchial epithelial PAR₂ is beta-arrestin-independent.

β-arrestins, originally discovered in the context of G protein-coupled receptor (GPCR) desensitization and internalization, also function in signaling of these receptors independently of G protein coupling. These novel functions involve the roles for β-arrestins as scaffolds. It has been reported that β-arrestins interact with a number of binding partners including trafficking proteins, cytosolic kinases, cytoskeletal proteins, and non-receptor tyrosine kinase. Downstream of protease-activated receptor-2 (PAR-2), β-arrestin scaffolds the components of the ERK cascade, Raf-1, MEK1, and ERK1/2 with the receptor at the plasma membrane, leading to activation of cytoplasmic/membrane ERK1/2 signaling independent of G-protein coupling. Furthermore, we previously demonstrated that stimulation of PAR-2 resulted in prolonged activation of ERK1/2 in pseudopodia in a β-arrestin-dependent manner and β-arrestins were required in PAR-2 mediated ERK1/2 activation at the membrane and cell migration in metastatic tumor cell lines, suggesting β-arrestin-dependent ERK1/2 activation might play a role in cell motility. Although a number of recent studies reported that β-arrestins are required for ERK1/2 activation independently of G protein coupling, molecular mechanism of β-arrestin- mediated ERK1/2 activation via c-Raf has remained unclear. We hypothesized that the ability of β-arrestins to scaffold and prolong MAPK signaling at the membrane, is dependent upon precise molecular interactions that are facilitated by interaction of β-arrestins with PAR-2. To investigate the hypothesis, we determined the sites/domains in β-arrestin-1 that interact with components of the ERK module (c-Raf, MEK1, and ERK1/2) both in vitro and in cells using GST pull down assay, sandwich immunoassay, and co-immunoprecipitation and investigated the role of these identified regions of β-arrestin for interaction and activation of ERK1/2 using truncated mutants of β-arrestin or β-arrestin mutants lacking domains. In addition, a mysterious link in functional mechanism of β-arrestin-mediated ERK1/2 activation is how c-Raf relieves autoinhibition engendering a conformation change from a closed, inactive state to an open, active state, without small GTPase Ras, which is critical for c-Raf activation in the classical G protein dependent ERK activation. We hypothesized that β-arrestins behave similar to Ras for β-arrestin-dependent ERK activation to relieve autoinhibition of c-Raf. We have proved that β-arrestin-1 binds to the regulatory domain of c-Raf where Ras was indentified to bind. Therefore, binding of β-arrestin-1 to c-Raf might ensure formation of scaffolding complex containing β-arrestin-1 and the ERK cascade and play a critical role to activate c-Raf in β-arrestin-dependent ERK1/2 activation.

Protease activated receptor 2 (PAR2) and Neurokinin 1 receptor (NK1R) are 7 transmembrane receptors (7TMRs), which signal by G alpha q leading to Ca2+ release and protein kinase C (PKC) activation. Both receptors are desensitized by beta-arrestin binding to their C termini. They can also activate ERK1/2 through beta-arrestin scaffolding complexes. They differ in ERK1/2 mediated physiological outcomes: cell migration versus proliferation. Using beta-arrestins, PAR2 activates cofilin, to promote chemotaxis, which was not observed in NK1R. We hypothesized that the differences in beta-arrestin dependent signaling by these 7TMRs can be attributed to how the molecular scaffolds bind the C-tail of the receptors. Using wild type and C-tail chimeras, we showed that the rate of desensitization, internalization, subcellular localization post endocytosis as well as ERK1/2 dependent physiological responses depend on the nature of interaction of beta-arrestins with the C-termini. Bioluminescence resonance energy transfer (BRET) assays showed that PAR2 recruits both beta;-arrestin 1/2 faster and with greater affinity than NK1R. We further show that initial G alpha q signaling is necessary for beta-arrestin 1/2 recruitment to NK1R. PAR2 recruits beta-arrestin 1/2 even when G alpha q pathway is blocked. Assays with C terminal phosphorylation mutants of PAR2 indicated that, phosphorylation of certain residues are necessary for beta-arrestin 1/2 recruitment. Phosphorylation at putative PKC sites (S363 and T366) determine stability of PAR2-beta-arrestin 1/2 complex. This in turn ensures downstream cofilin activation and cell migration. BRET assays with PAR2 and G-protein coupled receptor kinase-2 (GRK-2), revealed that GRK2 is recruited to PAR2 in a dose dependent fashion. It is possible that GRK-2 maybe another kinase which regulates PAR2 signaling. Finally, using allergic proteases from Alternaria alternata and Blatella germanica we demonstrated that PAR2 activation by these proteases leads to beta-arrestin 1/2 recruitment and subsequent beta-arrestin 1 dependent cofilin activation and cell migration. Cell migration brought about by the beta-arrestin signaling arm of PAR2 might be an important molecular mechanism for migration of immune cells to the airways, which is an important symptom of allergic asthma.

Research of G-protein coupled receptor signaling (GPCR) for therapeutic applications have historically been a hot topic of research. In the past, GPCR studies have been primarily focused on the molecular mechanism of its signaling pathway debunking the components of either traditional G-protein signaling or by other novel and more complex signaling pathways. However, recent GPCR studies have transitioned to examining distinct behaviors exhibited by and physiological responses that are triggered by cell- and localization-specific GPCRs. These responses include inflammatory, anti-inflammatory, cell migration and metastasis. Further, research of GPCR signaling have extended to the utilization of "biased agonists" where these modified agonists activates a specific GPCR and have the ability to trigger the activation of a specific pathway and inhibit other pathways that otherwise is propagated via activation by the non-modified agonist.

In the first study, we focus on the two distinct pools of protease-activated receptor-2 (PAR-2) present in intestinal epithelial cells: an apical pool facing the lumen, and a basolateral pool facing the bloodstream. Studies have demonstrated that introduction of PAR-2 agonists such as 2-furoyl-LIGRL-O-NH2 (2fAP) to the lumen can activate apical PAR-2 while it has been predicted mast cells and recruited leukocytes release proteases that can activate basolateral PAR-2. However, whether both pools of PAR-2 are equally capable of signaling, and the possible distinction between apical and basolateral PAR-2 induced responses have yet been addressed. Here we identified the expression of both apical and basolateral PAR-2 in cultured CACO2-BBe monolayers, and in mouse proximal jejunum. We reveal apical and basolateral PAR-2 in the intestinal epithelium are accessible by PAR-2 agonist 2fAP, and have both temporally and mechanistically signaling distinctions. Further, we demonstrate the apical and basolateral pools of PAR-2 can trigger distinct physiological responses as well.

Secondly, we illustrate that PAR-2 activation enhanced bacteria uptake in cultured colonocytes via a clathrin and phosphotidyl-inositol-3-kinase dependent manner. Moreover, we show that PAR-2 activation stimulated bacteria uptake in mice intestines. These results suggested that activation of the receptor plays a crucial role in gastrointestinal barrier dysfunction and is indeed a contributing factor to bacterial infection, further allowing for better understanding of PAR-2 and its physiological role in the gastrointestinal system.

On the whole these studies attempts to provide clarification for the characterization of PAR-2 in the gastrointestinal system that can be used further in the research for the usage of PAR-2 as a therapeutic target.

Lastly, findings for which I was a supporting author in a recent publication, we examined the differential effects of β-arrestins on internalization, desensitization and signaling of PAR-2. We demonstrate here that β-arrestin-1 and β-arrestins-2 have distinctive roles in PAR-2 signaling, including mechanistic and temporal distinctions in signaling, signal termination, cellular localization of signaling components and degradation.

Beta-arrestins are adaptor proteins that can scaffold a number of signaling proteins to promote localized activity within the cell. Downstream of some GPCRs, β-arrestins can promote activation of the actin filament severing protein, cofilin, through two mechanisms: one involving inhibition of LIM Kinase (LIMK) which negatively regulates cofilin activity through phosphorylation on serine 3. The mechanism by which β-arrestin-1 regulates LIMK activity has not been elucidated; however, it has been shown to be important for cell migration downstream of protease-activate-receptor-2 (PAR-2), dendritic spine formation and opioid receptor function. Here my work demonstrate that β-arrestin-1 directly binds both cofilin and LIMK, and inhibits LIMK activity directly and investigate the mechanism by which inhibition of kinase activity occurs. Using serial truncations and site-directed mutagenesis, I identify crucial residues for cofilin and LIMK interaction within amino acids 1-99 of β-arrestin-1 and show that charged residues at 50 and 51 are crucial for binding to LIMK and R51 is required for LIMK inhibition, PAR2 stimulated cofilin dephosphorylation and cell migration. Additionally, our work reveals that amino acids 1-99 aminos of β-arrestin-1 bind both cofilin and LIMK with a higher apparent affinity than the full length and blocks PAR2-stimulated cofilin dephosphorylaton in HEK293 cells, suggesting it functions as a selective dominant negative β-arrestin-1, inhibiting specifically the cofilin pathway. Thus, residues in the N-terminus of β-arrestin-1 are involved in LIMK inhibition and cofilin activation and this, in turn, is important for cell migration downstream of PAR-2.

The goal of this dissertation project was to understand how activation of Protease-activated receptor-2 (PAR2)-beta-arrestin-2-dependent signaling leads to promotion of inflammation in the allergic response to the airborne allergen Alternaria alternata (A. alternata) in the lung. Exposure to A. alternata spores is a major problem for people susceptible to the allergens, and is a leading cause of asthmatic symptoms. The current standard of care for treatment of acute asthma exacerbations is not sufficient for many patients, highlighting the need for alternative therapies. PAR2 is of great interest as a therapeutic target because it promotes both pro- and anti-inflammatory responses in the airway, through opposing G-protein and β-arrestin signaling pathways, and it is activated by proteases released by household pathogens such as Alternaria. These studies examine: 1) the molecular mechanisms underlying Alternaria-induced asthma, delineating the signaling pathways and identifying a new PAR2 activating protease (AASP) present in Alternaria extracts and 2) characterizing a novel PAR2 antagonist (C391) for potential β-arrestin bias, and investigating its ability to inhibit Alternaria –induced inflammation. In cultured cells, I examined key readouts of G-protein and β-arrestin-dependent PAR2 signaling and demonstrate that Alternaria and AASP promote both pathways. In vivo, Alternaria and AASP promote pro-inflammatory responses in the lung that are dependent on PAR2 and β-arrestin-2. However, the proinflammatory components of Alternaria also promote the protective PAR2 effects in isolated bronchioles. My studies indicate that the PAR2 antagonist, C391, inhibits Alternaria-induced airway inflammation in vivo. In cultured cells, C391 inhibits PAR2-β-arrestin signaling in cells at an IC50 100-fold lower than that for inhibition of G-protein signaling, suggesting it may represent a β-arrestin biased antagonist. Altogether, I have found that PAR2 activation via β-arrestin-2 plays a major role in mediating pro-inflammatory responses induced by several allergens in the airway, and that the use of the novel antagonist using C391 may be a new avenue for therapeutic intervention. Thus, PAR2-βarrestin-2 activation acts as a central switch that allergens turn on but can be turned off therapeutically.