Homology modeling of a Class A GPCR in the inactive conformation: A quantitative analysis of the correlation between model/template sequence identity and model accuracy

GPR101: Modeling a constitutively active receptor linked to X-linked acrogigantism

GPR101 is a G protein-coupled receptor (GPCR) implicated in a rare form of genetic gigantism known as X-linked acrogigantism, or X-LAG. In particular, X-LAG patients harbor microduplications in the long arm of the X-chromosome that invariably include the GPR101 gene. Duplications of the GPR101 gene lead to the formation of a new chromatin domain that causes over-expression of the receptor in the pituitary tumors of the patients. Notably, GPR101 is a constitutively active receptor, which stimulates cells to produce the second messenger cyclic AMP (cAMP) in the absence of ligands. Moreover, GPR101 was recently reported to constitutively activate not only the cAMP pathway via Gs, but also other G protein subunits (Gq/11 and G12/13). Hence, chemicals that block the constitutive activity of GPR101, known as inverse agonists, have the potential to be useful for the development of pharmacological tools for the treatment of X-LAG. In this study, we provide structural insights into the putative structure of GPR101 based on in-house built homology models, as well as third party models based on the machine learning methods AlphaFold and AlphaFold-Multistate. Moreover, we report a molecular dynamics study, meant to further probe the constitutive activity of GPR101. Finally, we provide a structural comparison with the closest GPCRs, which suggests that GPR101 does not share their natural ligands. While this manuscript was under review, cryo-electron microscopy structures of GPR101 were reported. These structures are expected to enable computer-aided ligand discovery efforts targeting GPR101.

Homology Modeling of Class A G-Protein-Coupled Receptors in the Age of the Structure Boom

With 700 members, G protein-coupled receptors (GPCRs) of the rhodopsin family (class A) form the largest membrane receptor family in humans and are the target of about 30% of presently available pharmaceutical drugs. The recent boom in GPCR structures led to the structural resolution of 57 unique receptors in different states (39 receptors in inactive state only, 2 receptors in active state only and 16 receptors in different activation states). In spite of these tremendous advances, most computational studies on GPCRs, including molecular dynamics simulations, virtual screening and drug design, rely on GPCR models obtained by homology modeling. In this protocol, we detail the different steps of homology modeling with the MODELLER software, from template selection to model evaluation. The present structure boom provides closely related templates for most receptors. If, in these templates, some of the loops are not resolved, in most cases, the numerous available structures enable to find loop templates with similar length for equivalent loops. However, simultaneously, the large number of putative templates leads to model ambiguities that may require additional information based on multiple sequence alignments or molecular dynamics simulations to be resolved. Using the modeling of the human bradykinin receptor B1 as a case study, we show how several templates are managed by MODELLER, and how the choice of template(s) and of template fragments can improve the quality of the models. We also give examples of how additional information and tools help the user to resolve ambiguities in GPCR modeling.

New Insights into the Stereochemical Requirements of the Bombesin BB1 Receptor Antagonists Binding

Members of the family of bombesinlike peptides exert a wide range of biological activities both at the central nervous system and in peripheral tissues through at least three G-Protein Coupled Receptors: BB1, BB2 and BB3. Despite the number of peptide ligands already described, only a few small molecule binders have been disclosed so far, hampering a deeper understanding of their pharmacology. In order to have a deeper understanding of the stereochemical features characterizing binding to the BB1 receptor, we performed the molecular modeling study consisting of the construction of a 3D model of the receptor by homology modeling followed by a docking study of the peptoids PD168368 and PD176252 onto it. Analysis of the complexes permitted us to propose prospective bound conformations of the compounds, consistent with the experimental information available. Subsequently, we defined a pharmacophore describing minimal stereochemical requirements for binding to the BB1 receptor that was used in silico screening. This exercise yielded a set of small molecules that were purchased and tested, showing affinity to the BB1 but not to the BB2 receptor. These molecules exhibit scaffolds of diverse chemical families that can be used as a starting point for the development of novel BB1 antagonists.

Development of the first in vivo GPR17 ligand through an iterative drug discovery pipeline: A novel disease-modifying strategy for multiple sclerosis

The GPR17 receptor, expressed on oligodendroglial precursors (OPCs, the myelin producing cells), has emerged as an attractive target for a pro-myelinating strategy in multiple sclerosis (MS). However, the proof-of-concept that selective GPR17 ligands actually exert protective activity in vivo is still missing. Here, we exploited an iterative drug discovery pipeline to prioritize novel and selective GPR17 pro-myelinating agents out of more than 1,000,000 compounds. We first performed an in silico high-throughput screening on GPR17 structural model to identify three chemically-diverse ligand families that were then combinatorially exploded and refined. Top-scoring compounds were sequentially tested on reference pharmacological in vitro assays with increasing complexity, ending with myelinating OPC-neuron co-cultures. Successful ligands were filtered through in silico simulations of metabolism and pharmacokinetics, to select the most promising hits, whose dose and ability to target the central nervous system were then determined in vivo. Finally, we show that, when administered according to a preventive protocol, one of them (named by us as galinex) is able to significantly delay the onset of experimental autoimmune encephalomyelitis (EAE), a mouse model of MS. This outcome validates the predictivity of our pipeline to identify novel MS-modifying agents.

Binding Interactions of Ergotamine and Dihydroergotamine to 5-Hydroxytryptamine Receptor 1B (5-HT1b) Using Molecular Dynamics Simulations and Dynamic Network Analysis

Ergotamine (ERG) and dihydroergotamine (DHE), common migraine drugs, have small structural differences but lead to clinically important distinctions in their pharmacological profiles. For example, DHE is less potent than ERG by about 10-fold at the 5-hydroxytrptamine receptor 1B (5-HT_1B). Although the high-resolution crystal structures of the 5-HT_1B receptor with both ligands have been solved, the high similarity between these two complex structures does not sufficiently explain their activity differences and the activation mechanism of the receptor. Hence, an examination of the dynamic motion of both drugs with the receptor is required. In this study, we ran a total of 6.0 μs molecular dynamics simulations on each system. Our simulation data show the subtle variations between the two systems in terms of the ligand-receptor interactions and receptor secondary structures. More importantly, the ligand and protein root-mean-square fluctuations (RMSFs) for the two systems were distinct, with ERG having a trend of lower RMSF values, indicating it to be bound tighter to 5-HT_1B with less fluctuations. The molecular mechanism-general born surface area (MM-GBSA) binding energies illustrate this further, proving ERG has an overall stronger MM-GBSA binding energy. Analysis of several different microswitches has shown that the 5-HT_1B-ERG complex is in a more active conformation state than 5-HT_1B-DHE, which is further supported by the dynamic network model, with reference to mutagenesis data with the critical nodes and the first three low-energy modes from the normal mode analysis. We also identify Trp327^6.48 and Phe331^6.52 as key residues involved in the active state 5-HT_1B for both ligands. Using the detailed dynamic information from our analysis, we made predictions for possible modifications to DHE and ERG that yielded five derivatives that might have more favorable binding energies and reduced structural fluctuations.

Influence of the Structural Accuracy of Homology Models on Their Applicability to Docking-Based Virtual Screening: The β2 Adrenergic Receptor as a Case Study

How accurate do structures of the β₂ adrenergic receptor (β₂AR) need to be to effectively serve as platforms for docking-based virtual screening campaigns? To answer this research question, here, we targeted through controlled virtual screening experiments 23 homology models of the β₂AR endowed with different levels of structural accuracy. Subsequently, we studied the correlation between virtual screening performance and structural accuracy of the targeted models. Moreover, we studied the correlation between virtual screening performance and template/target receptor sequence identity. Our study demonstrates that docking-based virtual screening campaigns targeting homology models of the β₂AR, in the majority of the cases, yielded results that exceeded random expectations in terms of area under the receiver operating characteristic curve (ROC AUC). Moreover, with the most effective scoring method, over one-third and one-quarter of the models yielded results that exceeded random expectation also in terms of enrichment factors (EF1, EF5, and EF10) and BEDROC (α = 160.9), respectively. Not surprisingly, we found a detectable linear correlation between virtual screening performance and structural accuracy of the ligand-binding cavity. We also found a detectable linear correlation between virtual screening performance and structural accuracy of the second extracellular loop (EL2). Finally, our data indicate that, although there is no detectable linear correlation between virtual screening performance and template/β₂AR sequence identity, models built on the basis of templates that show high sequence identity with the β₂AR, especially within the ligand-biding cavity, performed consistently well. Conversely, models with lower sequence identity displayed performance levels that ranged from very good to random, with no apparent correlation with the sequence identity itself.

Chiral analogues of (+)-cyclazosin as potent α1B-adrenoceptor selective antagonist

(+)-Cyclazosin [(+)-1] is one of most selective antagonists of the α_1B-adrenoceptor subtype (selectivity ratios, α_1B/α_1A = 13, α_1B/α_1D = 38-39). To improve the selectivity, we synthesized and pharmacologically studied the blocking activity against α₁-adrenoceptors of several homochiral analogues of (+)-cyclazosin featuring different substituents on the carbonyl or amine groups, namely (-)-2, (+)-3, (-)-4-(-)-8, (+)-9. Moreover, we studied the activity of some their opposite enantiomers, namely (-)-1, (-)-3, (+)-6, and (-)-9, to evaluate the influence of stereochemistry on selectivity. The benzyloxycarbonyl and methyl (4aS,8aR) analogues (+)-3 and (-)-6 improved in a significant way the α_1B selectivity of the progenitor compound: 4 and 14 time vs. the α_1D subtype and 35 and 77 times vs. the α_1A subtype, respectively. The study confirmed the importance of the hydrophobic cis-octahydroquinoxaline moiety of these molecules for the establishment of interactions with the α₁-adrenoceptors as well that of their (4aS,8aR) stereochemistry to grant selectivity for the α_1B subtype. Hypotheses on the mode of interaction of these compounds were advanced on the basis of molecular modeling studies performed on compound (+)-3.

Binding, Thermodynamics, and Selectivity of a Non-peptide Antagonist to the Melanocortin-4 Receptor

The melanocortin-4 receptor (MC4R) is a potential drug target for treatment of obesity, anxiety, depression, and sexual dysfunction. Crystal structures for MC4R are not yet available, which has hindered successful structure-based drug design. Using microsecond-scale molecular-dynamics simulations, we have investigated selective binding of the non-peptide antagonist MCL0129 to a homology model of human MC4R (hMC4R). This approach revealed that, at the end of a multi-step binding process, MCL0129 spontaneously adopts a binding mode in which it blocks the agonistic-binding site. This binding mode was confirmed in subsequent metadynamics simulations, which gave an affinity for human hMC4R that matches the experimentally determined value. Extending our simulations of MCL0129 binding to hMC1R and hMC3R, we find that receptor subtype selectivity for hMC4R depends on few amino acids located in various structural elements of the receptor. These insights may support rational drug design targeting the melanocortin systems.

GPCR Homology Model Generation for Lead Optimization

The vast increase of recently solved GPCR X-ray structures forms the basis for GPCR homology modeling to atomistic accuracy. Nowadays, homology models can be employed for GPCR-ligand optimization and have been reported as invaluable tools for drug design in the last few years. Elucidation of the complex GPCR pharmacology and the associated GPCR conformations made clear that different homology models have to be constructed for different activation states of the GPCRs. Therefore, templates have to be chosen accordingly to their sequence homology as well as to their activation state. The subsequent ligand placement is nontrivial, as some recent X-ray structures show very unusual ligand binding sites and solvent involvement, expanding the space of the putative ligand binding site from the generic retinal binding pocket to the whole receptor. In the present study, a workflow is presented starting from the selection of the target sequence, guiding through the GPCR modeling process, and finishing with ligand placement and pose validation.

Assessing GPCR homology models constructed from templates of various transmembrane sequence identities: Binding mode prediction and docking enrichment

GPCR crystal structures have become more readily accessible in recent years. However, homology models of GPCRs continue to play an important role as many GPCR structures remain unsolved. The new crystal structures now available provide not only additional templates for homology modelling but also the opportunity to assess the performance of homology models against their respective crystal structures and gain insight into the performance of such models. In this study we have constructed homology models from templates of various transmembrane sequence identities for eight GPCR targets to better understand the relationship between transmembrane sequence identity and model quality. Model quality was assessed relative to the crystal structure in terms of structural accuracy as well as performance in two typical structure-based drug design applications: ligand binding pose prediction and docking enrichment in virtual screening. Crystal structures significantly outperformed homology models in both assessments. Accurate ligand binding pose prediction was possible but difficult to achieve using homology models, even with the use of induced fit docking. In virtual screening using homology models still conferred significant enrichment compared to random selection, with a clear benefit also observed in using models optimized through induced fit docking. Our results indicate that while homology models that are reasonably accurate structurally can be constructed, without significant refinement homology models will be outperformed by crystal structures in ligand binding pose prediction and docking enrichment regardless of the template used, primarily due to the extremely high level of structural accuracy needed for such applications.

Using molecular dynamics for the refinement of atomistic models of GPCRs by homology modeling

Despite GPCRs sharing a common seven helix bundle, analysis of the diverse crystallographic structures available reveal specific features that might be relevant for ligand design. Despite the number of crystallographic structures of GPCRs steadily increasing, there are still challenges that hamper the availability of new structures. In the absence of a crystallographic structure, homology modeling remains one of the important techniques for constructing 3D models of proteins. In the present study we investigated the use of molecular dynamics simulations for the refinement of GPCRs models constructed by homology modeling. Specifically, we investigated the relevance of template selection, ligand inclusion as well as the length of the simulation on the quality of the GPCRs models constructed. For this purpose we chose the crystallographic structure of the rat muscarinic M3 receptor as reference and constructed diverse atomistic models by homology modeling, using different templates. Specifically, templates used in the present work include the human muscarinic M2; the more distant human histamine H1 and the even more distant bovine rhodopsin as shown in the GPCRs phylogenetic tree. We also investigated the use or not of a ligand in the refinement process. Hence, we conducted the refinement process of the M3 model using the M2 muscarinic as template with tiotropium or NMS docked in the orthosteric site and compared with the results obtained with a model refined without any ligand bound.