Kappa opioid receptor (KOPR) is involved in many physiological functions and pharmacological responses such as analgesia, anti-pruritic effect, sedation, motor incoordination and aversion (Simonin et al., 1998; Liu-Chen, 2004). The cellular mechanisms following activation of KOPR involve in part Gi/o protein-dependent pathways (Law et al., 2000). Following KOPR activation, the receptor is phosphorylated and arrestins are recruited. Arrestins mediate agonist-dependent KOPR desensitization, internalization and down-regulation (Liu-Chen, 2004). In recent years, arrestins were found to initiate arrestin-dependent downstream signaling. Thus, agonist-promoted KOPR phosphorylation plays a pivotal role in KOPR regulation and signaling. Previous studies from our lab showed that in Chinese hamster ovary (CHO) cells stably transfected with the human KOPR (hKOPR), U50,488H induced phosphorylation (Li et al., 2002a); however, sites of phosphorylation were not determined. Using LC-MS/MS, our lab recently identified four residues (S356, T357, T363 and S369) to be the sites of U50,488H-promoted phosphorylation in the mouse KOPR (mKOPR) stably expressed in N2A cells (Chen et al., 2016). Antibodies were generated against phosphopeptides and purified and three antibodies were found to have high specificity for the mKOPR phosphorylated at S356/T357, T363 and S369, respectively (Chen et al., 2016). Our lab previously showed that while U50,488H promoted robust hKOPR phosphorylation and internalization, etorphine induced little phosphorylation and internalization, although both were potent full agonists in enhancing [35S]GTPγS (Li et al., 2002a; Zhang et al., 2002; Li et al., 2003). Etorphine caused lower levels of KOPR phosphorylation at all the four residues than U50,488H by immunoblotting with the phospho-specific antibodies (Chen et al., 2016). Using the SILAC (stable isotope labeling by amino acids in cell culture) approach, we have found that compared to etorphine, U50,488H promoted higher levels of single phosphorylation at T363 and S369 and double phosphorylation at T363+S369 and T357+S369 as well as triple phosphorylation at S356+T357+S369 (Chen et al., 2016). These results indicate that an above-threshold phosphorylation is required for KOPR internalization. It has been reported that KOPR is involved in neuronal differentiation and neurogenesis. In the first chapter, I focused on whether there are differences in the mechanisms underlying neurite outgrowth induced by U50,488H and etorphine. In the chapter 2, mechanisms of KOPR phosphorylation were characterized in detail using phospho-specific KOPR antibodies. Protein kinase C was found, for the first time, to be involved in agonist-promoted KOPR phosphorylation. The roles of PKC in behavioral effects induced by KOPR agonists in mice were examined. For the chapter 1, in Neuro2a mouse neuroblastoma cells stably transfected with the hKOPR (N2A-3HA-hKOPR), U50,488H robustly induced neurite outgrowth, but etorphine caused outgrowth to a much lower extent. G protein-dependent pathway was found to be involved in the actions of both agonists, but β-arrestin-dependent pathway was not. Inhibition of ERK1/2 phosphorylation decreased neurite outgrowth promoted by both agonists, indicating the roles of MAP kinase cascades in KOPR agonist-induced neuritogenesis. In contrast, β-arrestin2, 14-3-3ζ, GEC1 and Rap1 are not involved in U50,488H- or etorphine-promoted neurite outgrowth. Thus, the two agonists appear to share the same signaling pathways and the difference between two agonists is likely due to the lower efficacy of etorphine. For the chapter 2, U50,488H caused phosphorylation of the mKOPR at S356, T357, T363 and S369 in N2A cells stably transfected with FmK6H (FmK6H-N2A cells). NorBNI abolished U50,488H-induced KOPR phosphorylation at all four residues. GRKs (GRKs2, 3, 5 and 6) and PKCs were involved in U50,488H-mediated KOPR phosphorylation. In addition, PKC also participated in agonist-independent KOPR phosphorylation. This is the first time that PKC was shown to be involved in agonist-induced KOPR phosphorylation. We found that U50,488H caused KOPR phosphorylation at T363 and S369 in the mouse brain and PKC participated in phosphorylation of S369, but not T363, by using the PKC inhibitor chelerythrine (CHL). Thus, we further characterized effects of PKC inhibition on KOPR-mediated behaviors in CD1 mice. PKC was involved in KOPR-mediated sedation, motor incoordination and conditioned place aversion, but not analgesia and anti-scratching effect in mice. Studies in this thesis revealed the mechanisms of KOPR-mediated neurite outgrowth and KOPR-mediated phosphorylation and the involvement of PKC in KOPR-mediated pharmacological effects in vivo. These studies push the frontier of molecular pharmacology of the KOPR, which may be useful for development of KOPR agonists for therapeutic use. / Pharmacology

Introduction Adhesion GPCRs (aGPCRs) constitute the second largest family of the GPCR superfamily, and yet their properties are also the least understood. Growing research on the biological functions of aGPCRs suggest their implications in various (patho)physiological processes, such as cell migration, organ development and cancers. Moreover, due to the unique architecture of a large extracellular region (ECR) containing a plethora of adhesion motifs, aGPCRs are vital as a mechanosensor which transduces extracellular mechanical stimuli into intracellular signal transduction. One distinct feature of aGPCRs among the GPCR superfamily is the possession of a conserved extracellular fold termed GPCR autoproteolysis-inducing (GAIN) domain in perhaps all members within the class. The cleavage at the last loop of the GAIN domain leads to the formation of two non-covalently associated N- and C-terminal fragments (NTF/CTF). A peptide stretch in the start of the CTF acts as a tethered agonist (TA) which is responsible for at least part of the signaling volumes of an activated receptor. Despite the strict conservation of the GAIN domain and its importance in the activation mechanism of aGPCRs, some other fundamental properties of the receptors, with reference to GAIN domain cleavage, have not been rigorously analysed in a biological context. Thus, this study aims to: 1. Explore the structural and molecular determinants that affects GAIN domain cleavage; 2. Investigate the consequences of GAIN domain cleavage towards (i) surface trafficking, and (ii) phosphorylation of receptors. Results Abolishment of GAIN domain cleavage in Polycystin-1, the only other protein family possessing the GAIN domain, was found to eliminate its surface expression, which is a cause of polycystic kidney/liver disease. However, whether such relationship is also true for aGPCRs has not been systematically analysed. Therefore, the study started with profiling the kinetics of surface delivery of several members of aGPCRs. Mutations on the -2 or +1 residues of the GPCR proteolytic site (GPS) (thereby abolishing GAIN domain cleavage) affected the steady-state surface and total expressions of the receptors differently, and had variable effect towards different receptor members. However, the observations from steady-state kinetics are also a resultant output from numerous processes involved in proteostasis. To further dissect whether GPS mutations affect the surface trafficking of the receptors, a pulse-chase assay called the ‘Retention Upon Selective Hook’ (RUSH) assay was employed, wherein the synthesised receptor molecules conjugated to a streptavidin-binding peptide are withheld in the ER by the co-expressed, ER-resident streptavidin, and are only released upon the addition of biotin that outcompete the receptor-streptavidin binding, creating a synchronised transport. By adapting the RUSH assay on some aGPCR members, the attenuation of surface trafficking by GPS mutations has become more apparent. The tested receptors were found to have a deficit in the quantity of surface population, rather than a change in rate of trafficking, upon the introduction of GPS mutations. This implies that the cells may utilise GAIN domain cleavage as a quality checkpoint for ER exit of aGPCRs. As the GAIN domains of at least some aGPCRs were found to be cleaved before ER exit, and as the rate of surface delivery was generally not affected by GAIN domain cleavage, the influence of GAIN domain cleavage may arise earlier during the receptor maturation in the ER. However, while the mechanisms of GAIN domain cleavage have been elucidated previously, they rely heavily on purified domains. The fundamental questions of when exactly the GAIN domain is cleaved and what additional determinants apart from the GPS sequence contribute to GAIN domain cleavage during receptor biogenesis have still not been answered. In combination with molecular dynamics (MD) simulation studies on the GAIN domain of rat isoform of ADGRL1, F803 was found to be crucial in the proteolysis by forming an edge-π interaction with H836 (-2 position of the cleavage site), such that H836 is in close proximity to the hydroxyl group of T838 for the initiation of the nucleophilic attack. Reconstruction of the edge-π interaction into ADGRB3, a naturally uncleavable receptor, partially reinstates its GAIN domain cleavage; but similar reintroduction on ADGRB2 has no effect on restoring the proteolysis. Nonetheless, this observation highlights the vitality of a proper folding of the GAIN domain, specifically the microenvironment of the cleavage site, in assisting in cleavage. The study continued with a systematic series of experiments that ultimately discover the roles of the CTF towards GAIN domain cleavage of aGPCRs. Firstly, to mimic the biogenesis of the receptor, the seven transmembrane (7TM) region of ADGRE2 (E2) was stepwisely truncated and then analysed for GAIN domain cleavage. It was observed that the extent of GAIN domain cleavage increases when the ECR of E2 precedes with more number of TMs. The proteolysis occurs, although less efficiently, as early as the first TM is synthesised. Interestingly, GAIN domain cleavage is unaffected when the TM region of the E2-1TM mutant was replaced by other single-pass TM, and whether it is trafficked to the surface or held in the ER, while the proteolysis of TM-less ECR mutants is largely impeded. Based on this observation, the ECR and the TM region was spaced either by a fluorophore moiety or a variable number of helical turns. Remarkably, the extent of GAIN domain cleavage of all tested receptors declined upon the increase in displacement with the lumenal side of the ER membrane, defining the importance of membrane proximity in the completion of proteolysis during the maturation of GAIN domain. In that, a new model of GAIN domain cleavage during biogenesis has been proposed, with appreciation of the GAIN domain as part of a higher-order stuctural organisation rather than an independent domain. A physiological extent of GAIN domain cleavage does not only require the folding of the GAIN domain, but also the membrane tethering property of the CTF, allowing a partial cleavage as little as one TM is generated, and a dynamic stability provided by the full CTF. In some aGPCRs, the contributions from CTF are more significant than the autonomous GAIN domain folding. The findings implicate more complex requirements for GAIN domain cleavage in a biological context, and hence supporting a possibility that GAIN domain cleavage is the rate-determining step for ER exit of the receptor, leading to the observations obtained in the kinetic study. Phosphorylation of L3 by PKC activated by distant signaling cascade(s) The last part of the study focused on characterising the mechanism of phosphorylation of ADGRL3 (L3) at Thr1140 (pT1140), which is a poorly explored field of aGPCRs. It was made possible by exploiting a phosphospecific antibody developed in collaboration. Coincidently, pT1140 was not dependent on the examined GPCR properties of the receptor, such as G protein coupling, dependence of the TA, and GRK-mediated phosphorylation. Instead, by series of pharmacological inhibitions, it was discovered that pT1140 originates from the action of novel PKCs (nPKCs). Co-expression of L3 and dominant-negative mutants or the catalytic domains of individual members of nPKCs reveals that PKC acts as a master regulator of the phosphorylation event, by directly phosphorylating the receptor and priming other members of the nPKCs for pT1140. Finally, possible origins of the PKC activation were explored. It was found that the stimulation of PKC occurs via actin disassembly, which can act downstreams of VEGF-A/VEGFR2 signaling, although the physiological relevance is still yet to be deciphered. Nonetheless, the observations opened up new directions of research in the aspect of crosstalks between different signaling cascades and the possible modulations of the signaling fidelity of aGPCRs. Additionally, the complexity of aGPCR signaling has been clearly demonstrated. Conclusion This study has further defined the importance of GAIN domain cleavage for surface trafficking of aGPCRs, a process crucial for extracellular interactions. Moreover, a novel mechanistic model of GAIN domain cleavage in relevance to biogenesis and maturation of the receptors has been postulated. Characterisation of a site-specific phosphorylation mechanism of L3 has illustrated the potential of complex interactions of aGPCRs with other signaling pathways in cells. The results collectively shed light on the structure-function relationship of aGPCRs, and pave ways for numerous potential areas for explorations in the future.