Quantitative Systems Pharmacology and Biased Agonism at Opioid Receptors: A Potential Avenue for Improved Analgesics

Abstract

Chronic pain is debilitating and represents a significant burden in terms of personal and socio-economic costs. Although opioid analgesics are widely used in chronic pain treatment, many patients report inadequate pain relief or relevant adverse effects, highlighting the need to develop analgesics with improved efficacy/safety. Multiple evidence suggests that G protein-dependent signaling triggers opioid-induced antinociception, whereas arrestin-mediated pathways are credited with modulating different opioid adverse effects, thus spurring extensive research for G protein-biased opioid agonists as analgesic candidates with improved pharmacology. Despite the increasing expectations of functional selectivity, translating G protein-biased opioid agonists into improved therapeutics is far from being fully achieved, due to the complex, multidimensional pharmacology of opioid receptors. The multifaceted network of signaling events and molecular processes underlying therapeutic and adverse effects induced by opioids is more complex than the mere dichotomy between G protein and arrestin and requires more comprehensive, integrated, network-centric approaches to be fully dissected. Quantitative Systems Pharmacology (QSP) models employing multidimensional assays associated with computational tools able to analyze large datasets may provide an intriguing approach to go beyond the greater complexity of opioid receptor pharmacology and the current limitations entailing the development of biased opioid agonists as improved analgesics.

Keywords: Quantitative Systems Pharmacology; biased agonism; improved analgesics; multidimensional signaling network; opioid receptors.

Figure 1

Schematic representation of the conventional approach to developing G protein-biased opioid agonists. One of the main challenges in contemporary medicinal chemistry is the development of safer and more effective analgesics for pain treatment. In the conventional drug discovery approach, the first goal for the development of biased ligands is the identification of agonists exerting their effects through functionally selective mechanisms and the association of these mechanisms with a desired in vitro cellular response (1). While G proteins and arrestins are undoubtedly critical signaling effectors that regulate both normal and abnormal physiology, attributing the complex GPCR signaling to proximal transducers is too reductive. Frequently, the signaling pathways directly responsible for the therapeutic and detrimental side effects are largely considered a black box because they are not thoroughly predictable and understandable. This is due to a lack of knowledge at the whole system level, which integrates multiple effectors and interdependent networks. Furthermore, the drug behavior is influenced by its exposure at the site of action in tissue/organ (2), which may be difficult or impossible to measure experimentally. Due to the many differences in physiological systems upon measuring bias, the translation from in vitro profiles of biased signaling into in vivo animal models (3) represents the major complication in the search for safer bias. The complexity of this scenario is profoundly incremented by the variability in drug response at the healthy volunteer (4) and patient (5) levels, which arises from differences in the proteome, genome, disease states, lifestyle, and history.

Figure 2

An innovative QSP-based approach for G protein-biased opioid agonists development. In both conventional and Quantitative Systems Pharmacology (QSP)-driven approaches, the key milestones for developing G protein-biased opioid agonists with improved efficacy/safety profiles are similar. However, mathematical modeling and sophisticated computation add quantitative and integrative perspectives (summarized in the yellow boxes) to activities that are currently qualitative or isolated. First, opioid receptor signaling is multidimensional and involves peculiar transducers/effectors. The multiple G proteins and arrestin isoforms, the biphasic modulation of MAPK (e.g., ERK1/2-JNK) through G protein- and/or arrestin-mediated processes, and the opioid receptors homo/heteromerization need to be included in this multifaceted scenario (1). QSP models focus on interactions among these multiple elements and may help uncover key and novel GPCR events related to the desired phenotype of interest and the disease pathophysiology. The creation of a multi-scale model that incorporates data at several temporal and spatial scales (biomolecules, cells, tissue, organ, organism) can provide more accurate predictions of bias ligand PK/PD relationships in tissues and organs and the probable outcome due to genre signaling differences (2), which deeply impact on the biased agonists’ effects. QSP can effectively promote the scale-up from animals to humans while correctly accounting for physiological and genetic differences. In particular, the platform may help to identify the dose and doses regimens with the highest probability of success (3), optimizing the clinical trial design and minimizing time and costs necessary for R&D. Furthermore, it may facilitate the transition from healthy volunteer to patient studies and the identification of new clinical biomarkers related to a specific response/behavior (e.g., analgesia, sedation, respiratory depression, cognitive impairment, abuse liability) (4). Finally, adopting a QSP approach to biased agonism at opioid receptors, which also incorporates patient-to-patient variability (5), may easily increase the likelihood of developing a successful G protein-biased opioid analgesic with increased efficacy/safety and lower side effects.