Investigating the Role of G-protein Bias in the Anti-Nociceptive and Side Effect Profiles of Two Novel Mu Opioid Receptor Agonists Kurkinol and Kurkinorin

Abstract

<b>Chronic pain is a major problem worldwide, affecting 1 in 5 New Zealanders resulting in a decreased quality of life for the patient and a large socioeconomic problem costing an estimated $13-$14.5 billion a year. Current therapeutics target the mu opioid receptor (μ receptor) and include drugs such as morphine and fentanyl. While these drugs are highly effective in the treatment of strong acute pain, their long-term use is associated with tolerance to the analgesic effects and increasing rates of side effects such as respiratory depression and constipation. Due to their high abuse liability, they are also known to cause dependence and addiction when prescribed for extended periods. This is believed to have played a role in the opioid crisis in the United States and highlights the need for improved therapeutics for the treatment of chronic pain. </b><p><br></p><p>One mechanism that has been proposed to generate μ receptor agonists for the treatment of chronic pain with reduced analgesic tolerance and safer side effects is the development of G-protein biased agonists. Such agonists selectively activate the canonical G-protein signalling to a greater extent than the non-canonical β-arrestin2 pathway. This is based on previous work in β-arrestin2 knockout mice where the antinociceptive effects were increased, while side effects, including respiratory depression, tolerance, and constipation are reduced, increasing the therapeutic window. In this thesis, we aimed to assess the anti-nociceptive and side effect behavioural profiles of two novel μ receptor agonists, kurkinol (bias = 0.14) and kurkinorin (bias = 0.57), with a varying bias for the G-protein pathway to assess the role of this paradigm. </p><p><br></p> Evaluation of the behavioural profile of kurkinol and kurkinorin revealed that G-protein bias was correlated to increased anti-nociceptive potency and reduced tolerance in wildtype C57BL/6J mice. Furthermore, the anti-nociceptive potency of morphine was increased, and tolerance decreased in in β-arrestin2 knockout mice. While the level of tolerance was reduced for kurkinorin. However, in the chemotherapy-induced model of neuropathic pain, tolerance to kurkinol and kurkinorin developed at the same rate as morphine. Overall this work showed a poor correlation between G-protein bias and therapeutic window. With the G-protein selective kurkinol inducing worse respiratory depression, constipation, and motor coordination impairment compared to kurkinorin. Interestingly, respiratory depressive and constipation effects of kurkinol were not prevented in the β-arrestin2 knockout mice indicating that they are induced through the G-protein pathway. <p><br></p><p>These results highlight the change that has occurred in the biased agonism field over the last 4 years, with the lack of reproducibility of key experiments and poor translation of G-protein biased μ receptor agonists resulting in improved therapeutic windows both clinically and pre-clinically. Moreover, recent research has shown that pathway efficacy (i.e. partial agonism) and not G-protein bias is responsible for the behavioural profiles of compounds previously identified as G-protein biased. We, therefore, decided to further investigate the cell signalling profiles of our two novel agonists to assess them for partial agonism and to assess downstream signalling molecules activated by G-protein and β-arrestin2. </p><div><br></div> This revealed cell-specific inhibition of membrane hyperpolarisation in Hek293 and CHO cells stably expressing the human μ receptor, with kurkinol found to be the most potent in both cell lines, followed by kurkinorin, morphine, and [D-Ala2, N-MePhe4, Gly-ol]-enkephalin (DAMGO). However, no differences were identified between the μ receptor agonists in the activation of the inwardly rectifying channels in the CHO cell line. The assessment of pCREB (phosphorylated cAMP response-element binding protein) as a β-arrestin2 dependent pathway revealed poor activation by kurkinorin while kurkinol was a potent activator. Bias factors generated from this data showed poor correlations to therapeutic windows. While the differences om CREB phosphorylation was shown to have a stronger correlation to therapeutic windows generated from the behavioural data. <p><br></p><p>Overall this thesis has identified kurkinorin as a μ receptor agonist that induces potent anti-nociception with reduced side effects, without strong-G-protein bias. We also show that the highly selective μ receptor agonist kurkinol has improved anti-nociception with a worse side effect profile adding to the growing body of literature showing bias is not a good predictor in its current state. Furthermore, the discrepancies between cell lines, differential activation of subcellular pathways, and lack of reproducibility between bias equations indicate that the field has massively oversimplified a complex system. Which has, most likely, resulted in the poor translation of in vitro bias factors to clinically available μ receptor agonists for chronic pain. </p>