High-Content Screening in GPCR Drug Discovery: Challenges and Opportunities
Andrea Weston, Jonathan O'Connell, Martyn Banks, and Andrew Alt
Department of Applied Biotechnology, Bristol-Myers Squibb
Abstract
G protein-coupled receptors (GPCRs), also known as seven-transmembrane receptors, constitute the largest, most versatile, and most ubiquitous family of integral plasma membrane receptors. Accordingly, GPCRs have come to represent the most popular target class for therapeutic intervention. Classically, GPCRs are best known for their ability to initiate a signaling cascade via heterotrimeric G proteins. However, this prevailing model has expanded considerably over the past decade, and it is now well-accepted that GPCRs can signal independently of G proteins, through -arrestin. This and other additional layers of complexity have major implications in efforts to pharmacologically modify GPCRs, leading to the development of new approaches to study GPCR function and to screen for small molecule modulators. High-content screening (HCS), the automation of cellular microscopy coupled with image analysis, is a platform that is particularly well suited for GPCR drug discovery efforts; owing largely to the combined ability to multiplex, monitor kinetic events, and examine subcellular structures. This review focuses on some emerging themes in GPCR biology, and describes how HCS can be leveraged to better interrogate GPCR function.
Introduction: High-Content Screening to Support GPCR Drug Discovery
High-content screening (HCS) refers to the automation of cellular microscopy coupled to image analysis. This approach enables cell-based multiparameter assays, as well as subpopulation analysis. The ability to multiplex cellular parameters and to study them within a physiologically relevant environment has made HCS very attractive for drug discovery efforts, with the expectation that assays carried out using this more context-appropriate approach will produce hits with improved translation of efficacy to in vivo animal models and, ultimately, humans. G-protein-coupled receptor (GPCR) drug discovery efforts occupy a significant portion of the portfolios among several pharmaceutical companies and GPCR-targeted drugs comprise over 30% of currently marketed therapeutics. Of particular interest, the past 10 years have seen a dramatic paradigm shift in the basic concepts of GPCR-mediated cell function, including the identification of G-protein-independent/-arrestin-dependent signaling [1, 2], GPCR oligomerization [3], ligand-controlled GPCR trafficking [4], and allosteric modulation [5]. Exploiting these phenomena, in an effort to design drugs with very precise pharmacological profiles, has become highly desirable. High-content screening is particularly well suited to this effort, given that with this approach, GPCR function can be examined both kinetically and spatially in the cell, and multiple endpoints can be monitored within a single assay.
In practice, adoption of HCS as a high-throughput primary screening platform has been slow in coming, as the technology has not yet matured to be as robust and efficient as other existing approaches (herein HTS applies to compound collections on the order of 1x106 compounds). Although HCS inarguably provides more information-rich data than most established HTS approaches (which often provide a single readout from an average of all cells within a well of a microtiter plate) extracting this rich information is not trivial. In high-throughput mode, high-content assays could generate up to a terabyte worth of imaging data in a single screen - orders of magnitude greater than the volume produced by traditional HTS assays. The need to store this data and to make the images easily available for additional analysis requires a sophisticated informatics infrastructure and extensive support. In addition, for high-throughput screening, image acquisition is time consuming, the assays often incur higher per-well costs, and they frequently involve more challenging assay design. As a result of these challenges, high-content analysis has been largely limited to lower-throughput applications, such as target validation and lead optimization.
Despite the challenges of HCS, there are many applications for which this approach offers tremendous potential over more traditional cell-based assays. Going forward, HCS can be expected to have an integral role in the study of GPCR biology, as the inherent complexity in this area unfolds. Here we discuss the utility of high-content screening to understand GPCR-mediated cell function and to identify and characterize small molecule modulators for this target class. In addition, the benefits of HCS are weighed against the challenges of implementing high-throughput, high-content screening to support GPCR drug discovery.
A Changing Paradigm for GPCR-Mediated Cell Signaling
The classical view of GPCR-mediated signaling entails the coupling and activation of heterotrimeric G proteins in response to agonist stimulation of a GPCR. When a GPCR binds to its cognate G protein upon agonist stimulation, the resulting conformational change triggers the release of guanosine 5'-diphosphate (GDP) and the subsequent binding of guanosine 5'-triphoshpate (GTP) by the G subunit, followed by dissociation of the G subunit from the G complex [6]. The G protein is now in its active state, and the dissociated G and Gsubunits then relay signals to various downstream effectors, such as second-messenger-generating enzymes and ion channels, ultimately causing key changes in cell function [6]. An important event in GPCR-mediated cell modulation is the termination of GPCR signaling and receptor desensitization that follow agonist stimulation. This highly conserved desensitization restricts the magnitude and duration of GPCR signaling and results from receptor phosphorylation by G- protein-coupled receptor kinases (GRKs) followed by the recruitment and binding of -arrestin proteins to the cytoplasmic surface of the receptor [7]. Binding of -arrestin to an activated GPCR sterically disrupts coupling of the receptor to its G protein, thereby terminating or "arresting" the receptor response. Arrestins subsequently target the GPCR for endocytosis in clathrin-coated pits, by acting as a scaffold to link the GPCR with clathrin and AP-2 [8]. Once internalized, the GPCR can either be recycled back to the cell membrane for further stimulation (resensitization), or can be targeted for degradation [4].
In addition to the well-recognized role of -arrestin in terminating G protein coupling and mediating GPCR endocytosis, approximately a decade ago, -arrestins were also found to act as scaffolds, linking activation of GPCRs to numerous signaling networks (for review, see [1] ). Since these initial studies, interest in G-protein-independent signaling has grown considerably, and there has been a shift in the classical paradigm of GPCR signaling and in the way in which these receptors are interrogated. The first indication that -arrestins are mediators of cell signaling came from studies in which transfection of dominant negative mutants of -arrestin blocked 2-adrenergic receptor (2AR)-induced activation of mitogen-activated protein kinases (MAPKs) [9-11]. Following these studies, 2AR activation was found to induce the formation of src/-arrestin complexes [12-14]. -arrestins were subsequently found to scaffold specific components of signaling cascades at various other GPCRs including PAR-2, neurokinin-1 receptor, vasopressin 2 receptor (V2R), parathyroid hormone receptor (PTH1R), CXCR4 and multiple others (reviewed in [1]). The signaling networks that are scaffolded by -arrestins include MAPKs (the Raf-Mek-Erk cascade, JNK and P38) as well as other kinases such as PI3K, AKT and RhoA [1].
To date, -arrestin regulation of ERK1/2 is the best-characterized of all known -arrestin-mediated signaling events. Interestingly, at some GPCRs, ERK activation has been found to be downstream of both G protein- and -arrestin-dependent pathways and this dual activation has been shown to be temporally and spatially distinct (Figure 1) (reviewed in [15]). Specifically, G-protein-mediated Erk activation is quick and transient, with activated Erk levels peaking within the first five minutes of agonist stimulation, but undetectable after 10 minutes. This activated Erk localizes to the nucleus where it regulates gene transcription. Conversely, -arrestin-mediated activity is minimal within the first 10 minutes of stimulation, but accounts for the majority of detectable phosphorylated ERK after 30 minutes. -arrestin-activated Erk is retained in endocytic vesicles of cytosol [1, 16, 17]. The different patterns of ERK activation lead to distinct biological outcomes from the same GPCR, adding to the diversity of these receptors in mediating cellular physiology, and highlighting the potential for the identification and development of functionally selective GPCR agonists.
...To view the complete article, please login or subscribe if you don't already have an account.