The apelin system is an evolving transmitter system consisting of the G protein coupled apelin receptor and two endogenous peptide ligands, apelin and elabela. It is implicated as a potential therapeutic for a number of diseases; however, the endogenous peptides are limited by half-life and bioavailability. This study aims to identify and pharmacologically characterise apelin agonists in vitro and in vivo and to evaluate their therapeutic potential in pulmonary arterial hypertension as a model disease. CMF-019 was identified as the first G protein biased apelin agonist. To date, suitable small molecule apelin agonists as experimental tool compounds have been limited and CMF-019 represents an important advance. CMF-019 was active in vivo, producing an increase in cardiac contractility and vasodilatation, similar to apelin. These effects were achieved without receptor desensitisation, supporting the remarkable G protein bias observed in vitro. Furthermore, it was disease-modifying in vitro in an endothelial cell apoptosis assay but despite this, did not prevent pulmonary arterial hypertension in a monocrotaline rat model of the disease. An apelin mimetic peptide possessing an unnatural amino acid, MM202, conjugated chemically via a polyethylene glycol linker to an anti-serum domain antibody (AlbudAb) was also characterised. The product MM202-AlbudAb represents the first time an AlbudAb has been conjugated chemically to an unnatural peptide mimetic, providing protection from proteolysis and glomerular filtration. Importantly, it retained binding to albumin and demonstrated in vitro and in vivo activity at the apelin receptor. In conclusion, this thesis has identified and pharmacologically characterised two novel apelin agonists that possess significant advantages over the endogenous peptides. CMF-019 is suitable as an experimental tool compound and, as the first G protein biased small molecule, provides a starting point for more suitable therapeutics. In addition, MM202-AlbudAb proves that unnatural peptides can be conjugated to AlbudAb, supporting use of this technology in other small-peptide ligand transmitter systems.

<p>Chemokine receptors are members of the seven transmembrane receptor (7TMR) superfamily and are regulated by the G-protein coupled Receptor Kinase (GRK)/ b-arrestin system. CCL19 and CCL21 are endogenous agonists for the chemokine receptor CCR7. They are known to be equipotent in promoting Gi/o mediated calcium mobilization, chemotaxis and inhibition of adenylyl cyclase activity. Here we test the hypothesis that these ligands are biased agonists that differentially activate the G-protein coupled Receptor Kinase (GRK)/ b-arrestin system.</p><p>In order to test whether these ligands have distinct activity, murine T lymphocytes were used to compare the effects of CCL19 and CCL21 activation of CCR7 at endogenous expression levels. While each ligand stimulates similar chemotactic responses, we also find that CCR7 ligands lead to differential signaling. For instance, CCL19 is markedly more efficacious than CCL21 for the activation of ERK and JNK, but not AKT in these cells. Furthermore, ERK activation and chemotaxis are maintained as separate pathways, also distinguishable by their dependency upon PKC and PI3 kinase, respectively. Thus, CCL19 and CCL21 stimulate equal activation of PI3 kinase, AKT, and chemotaxis, but are in fact biased agonists leading to differential activation of MAP kinase in murine T lymphocytes. </p><p>To determine the mechanism of CCR7 ligand bias, we used HEK-293 cells expressing CCR7 to compare the proximate signaling events following CCL19 and CCL21 activation. We found striking differences in the activation of the GRK/ b-arrestin system. CCL19 leads to robust CCR7 phosphorylation and b-arrestin2 recruitment catalyzed by both GRK3 and GRK6 while CCL21 activates GRK6 alone. This differential GRK activation leads to distinct functional consequences. Only CCL19 leads to the recruitment of b-arrestin2-GFP into endocytic vesicles and classical receptor desensitization. In contrast, each agonist is fully capable of signaling to MAP kinase through b-arrestin2 in a GRK6 dependent fashion. </p><p>Therefore, CCR7 and its ligands represent a natural example of ligand bias whose mechanism involves differential GRK isoform utilization by CCL19 and CCL21 despite similar G-protein signaling. This study suggests that the GRK signatures of 7TMRs can determine the function of discrete pools of b-arrestin and thus guide its cellular effects.</p> / Dissertation

This thesis is comprised of three parts. In the first part, we investigated zyklophin, a novel selective short-acting kappa opioid receptor (KOPR) antagonist, and its effects on scratching behaviors in Swiss-Webster mice. We investigated whether zyklophin was able to induce scratching in a dose-dependent fashion, and whether this scratching behavior could be blocked by pretreatment with nor-binaltorphimine (norBNI). We also used KOPR -/- mice to further clarify the role of the KOPR in this behavior. In the second part, we examined the role of the divergent amino acid at position 6.58 in the mu opioid receptor (MOPR) and the KOPR on the binding of beta-funaltrexamine ß-FNA). ß-FNA is an irreversible antagonist at the MOPR and a reversible agonist at the KOPR. Utilizing the recently published crystal structures of the MOPR and KOPR, we collaborated with Dr. Lei Shi, who employed molecular modeling to choose a residue in transmembrane helix 6 (TM6) to mutate at the same position in MOPR and KOPR. We then characterized the mutants by performing [3H]diprenorphine binding, competition binding by unlabeled &beta;-FNA, irreversible ß-FNA binding and [35S]GTPγS binding. In the third part, we investigated the concept of functional selectivity, or ligand bias, at the KOPR. We studied 23 different KOPR agonists in vitro using [35S]GTPγS binding as a measure of G protein activation and the on-cell Western (OCW) as a measure of ß-arrestin-mediated receptor internalization at the human KOPR (hKOPR), and from the results, chose 13 ligands to study at the mouse KOPR (mKOPR). We then selected biased ligands from the in vitro mKOPR results and studied their effects on scratching behavior, inhibition of pain behaviors and dysphoria as measured by the conditioned place aversion (CPA) test. We predicted that the G biased ligand would produce analgesia and anti-scratching effects at lower doses than would produce aversion in the CPA test, since analgesia has been shown to be G protein mediated and CPA has been shown to be arrestin mediated. Our first set of studies revealed that zyklophin (0.1, 0.3 and 1 mg/kg, s.c., behind the neck), induced vigorous scratching in a dose-dependent manner. 0.3 mg/kg zyklophin induced 150 scratches over a 30 minute period. The scratching was not blocked by pretreatment with 20 mg/kg norBNI (i.p.) 18-20 hours before injection of 0.3 mg/kg zyklophin s.c. in the nape of the neck. The scratching also persisted in KOPR -/- mice, in which the absence of the KOPR was confirmed by [3H]U69,593 binding (2 nM). In our second set of studies, we mutated the lysine at position 303 in the MOPR to glutamic acid (K303E), and the glutamic acid at the equivalent position in the KOPR to lysine (E297K). We transfected these mutant receptors into mouse neuroblastoma (N2A) cells. We found that the mutations had no effect on [3H]diprenorphine binding affinity or competition binding with [3H]diprenorphine and &beta;-FNA indicating a functional intact opioid receptor. The mutations also did not affect [35S]GTPγS binding EC50 or Emax values. The mutation K303E in the MOPR reduced irreversible binding by 2/3 compared to the wildtype MOPR. Finally, we found that there were several ligands that displayed bias at the hKOPR and the mKOPR. At the hKOPR, using dynorphin A as the reference ligand to calculate bias, ICI-199441 was the only G biased ligand, while enadoline, nalbuphine, pentazocine, salvinorin A, tifluadom and butorphanol were arrestin-biased. At the mKOPR, only salvinorin B methoxymethyl ether (MOM-SalB) was G-biased, and salvinorin B ethoxymethyl ether (EOM-SalB), ICI-199441, U50,488H, nalfurafine and 12-epi-salvinorin A (12epiSalA) were ß-arrestin-biased. Enadoline and salvinorin A were slightly arrestin biased with respect to dynorphin A. From the in vitro data at the mKOPR, we selected MOM-SalB as our G biased ligand, U50,488H as our arrestin biased ligand and additionally chose to investigate nalfurafine due to its use in clinical studies. We hypothesized that U50,488H and nalfurafine would produce aversion at lower doses than analgesia or anti-pruritic effects. We found that nalfurafine was the only ligand studied to have a separation between doses that produced analgesia and anti-scratching effects, with A50 values of 5.8 and 8 μg/kg, respectively, and only produced significant dysphoria at a dose of 20 μg/kg. U50,488H and MOM-SalB produced dysphoria at all doses tested (0.25-10 mg/kg and 0.01-0.3 mg/kg, respectively). U50,488H produced a dose-dependent analgesia and anti-scratching with A50 values of 0.58 mg/kg and 2.07 mg/kg, respectively. MOM-SalB was more potent than U50,488H in producing dose-dependent analgesia and anti-scratching, with A50 values of 0.017 mg/kg and 0.070 mg/kg, respectively. Therefore, we concluded that the in vitro bias is not able to accurately predict in vivo behaviors, and nalfurafine is the first selective full agonist at the KOPR to show ligand bias in vivo. / Pharmacology