Structural and Functional Basis for Biased Agonism at the 5-hydroxytryptamine 5A Receptor

Dear Editor,

The 5-hydroxytryptamine 5A receptor (5-HT_5AR) belongs to the serotonin receptor family, which is classified as a group of G protein-coupled receptors (GPCRs) that are primarily associated with neuropsychiatric diseases. 5-HT_5AR is widely expressed in the central nervous system, including the cerebral cortex, hippocampus, cerebellum, olfactory bulb and dentate gyrus. It is involved in a wild spectrum of neuropsychiatric functions, including cognition, memory, circadian rhythms, and antinociception effects [1]. 5-HT_5AR predominantly signal through G_i/o, and also recruit β-arrestin upon activation. Recent studies indicate that β-arrestin recruitment influence the efficacy of numerous drugs targeting serotonin receptors. Although several structures of inactive and active 5-HT_5AR have been resolved, revealing interactions with several ligands as well as G_i/o proteins [2, 3], the role of 5-HT_5AR-targeting drugs in β-arrestin signaling remains poorly understood. The synthetic agonist 5-carboxamidotryptamine (5-CT) is commonly used as a tool ligand to evaluate the function of serotonin receptors, particularly the 5-HT_5AR, given its ten-fold higher affinity compared to 5-HT [4]. While 5-HT induces diverse signaling profiles across many other serotonin receptors, the biased agonism of 5-HT and 5-CT acting on 5-HT_5AR remain unclear.

To understand the molecular mechanism of 5-HT_5AR signaling through G_i and β-arrestin, we analyzed the structure of 5-CT stimulates 5-HT_5AR/G_i1 complex by cryo-electron microscopy (EM) with an overall resolution of 3.13 Å (Fig. S1a-b). The residues of 5-HT_5AR and ligand were well defined in the density map, except for the intracellular loop 3 (ICL3) due to the high flexibility (Fig. 1a). By constructing the receptor residue-ligand atom contact matrix, we found 5-CT form tight hydrophobic interactions with C125^3.36 and F301^6.51 and salt bridge with D121^3.32 (Fig. 1b). The amide group and the indole amine of 5-CT also interact with S204^5.42 and T126^3.36 by hydrogen bond (Fig. 1b). The binding pose of 5-CT exhibits slight conformational differences compared to that in a previous determined 5-CT/5-HT_5AR/G_i complex structure, indicating heterogeneity of 5-CT/5-HT_5AR interaction. This heterogeneity aligns with observations across multiple GPCR structures and likely captures near-native dynamics inherent to pharmacological mechanism of receptor-drugs interactions.

Fig. 1.

The structure and signaling characterization of 5-CT binding to 5-HT_5AR. a Cryo-EM density map and atomic model of the 5-CT/5-HT_5AR/G_i1/scFv16 complex. (5-HT_5AR: dark cyan; Gα_i1: orchid; Gβ₁: light coral; Gγ₂: medium aquamarine; 5-CT: yellow). b The binding mode of 5-CT in the binding pocket of 5-HT_5AR, the side chains of residues are shown as stick. c The effects of 5-CT and 5-HT on different mutants of the 5HT_5AR. The response curves for 5-CT induced G_i1 signaling (upper) and β-arrestin2 recruitment (lower) were measured by BRET and NanoBiT, respectively. d Bias factors (β value) of mutations relative to wild type (WT) calculated among three-agonist-induced receptor activation. Bias factors were derived from the curve fit parameters. Data are presented as mean ± SEM from four independent experiments, performed in triplicate. ∗ p < 0.033, ∗  ∗ p < 0.002 and ∗  ∗  ∗ p < 0.001 by one-way ANOVA followed by Dunnett’s multiple comparisons test.

In the contact matrix, S204^5.42 and E305^6.55 exhibit the highest contact frequencies with 5-CT (data not shown). To investigate the roles of these residues in 5-CT recognition, we introduced alanine substitutions at S204^5.42 and E305^6.55 into the receptor, following characterization of G_i1 activation by bioluminescence resonance energy transfer (BRET) and β-arrestin2 recruitment by NanoBiT. S204^5.42A abolished the β-arrestin2 recruitment and retained efficacy in G_i1 dissociation (Fig. 1c), indicating that the polarity of 5.42 was essential for β-arrestin2 recruitment. Additionally, E305^6.55A slightly enhanced the efficacy in G_i1 signaling (ΔE_max = 0.229 ± 0.07, p = 0.03) and potency in the β-arrestin2 recruitment (ΔpEC₅₀ = 0.452 ± 0.08, p = 0.006) of 5-CT. These results indicated that S204^5.42 are essential for signaling bias of 5-CT stimulating 5-HT_5AR, and interaction with E305^6.55 may not be essential for 5-CT to function. Although V194^45.52 did not form intense interaction with 5-CT, V194^45.52A reduced potency for both G_i dissociation (ΔpEC₅₀ = -0.841 ± 0.1, p < 0.001) and β-arrestin2 recruitment (ΔpEC₅₀ = -1.151 ± 0.08, p < 0.001), indicating its importance in ligand binding. Given its strategic location on ECL2, this residue is hypothesized to function as a putative gatekeeper for the binding pocket. The loss of its bulky side chain is expected to accommodate easier ligand dissociation, thereby preventing sustained receptor activation.

We also characterized signaling of 5-HT_5AR mutants activated by endogenous ligand 5-HT as reference. S204^5.42A decreased potency of 5-CT (ΔpEC₅₀ = -0.952 ± 0.12, p < 0.001), but enhanced potency of 5-HT in G_i dissociation (ΔpEC₅₀ = 0.694 ± 0.05, p < 0.001). V194^45.52A exhibited similar impacts on 5-CT and 5-HT, reducing potency of 5-CT (G_i: ΔpEC₅₀ = -0.841 ± 0.1, p < 0.001; β-arrestin2: ΔpEC₅₀ = -0.952 ± 0.12, p < 0.001) and 5-HT (G_i: ΔpEC₅₀ = -1.151 ± 0.08, p < 0.001; β-arrestin2: ΔpEC₅₀ = -1.206 ± 0.03, p < 0.001) in both pathways, whereas E305^6.55A differently altered signaling of the two ligands (Fig. 1c). Specifically, E305^6.55A slightly enhanced potency of 5-CT in β-arrestin recruitment (ΔpEC₅₀ = 0.452 ± 0.08, p = 0.006), while it reduced potency of 5-HT in β-arrestin recruitment (ΔpEC₅₀ = -0.262 ± 0.04, p = 0.01) thereby biasing signaling of 5-HT towards G_i dissociation (Fig. 1d). These findings confirmed that S204^5.42 plays a crucial role in β-arrestin2 signaling by 5-HT_5AR. In contrast, E305^6.55 exhibits differential interaction modes with hydroxyl group of 5-HT compared to amide group of 5-CT. Mutagenesis and molecular dynamics simulations reveal distinct functional roles: D121^3.32 (Energy = -41.5 ± 9.5 kcal/mol, 55.4% of total) and E305^6.55 (Energy = -17.7 ± 10.8 kcal/mol, 23.6% of total) dominate 5-CT binding energetics, whereas S204^5.42 facilitates β-arrestin recruitment through conformational rearrangement without significant binding energy contributions (Fig. S1c). Most serotonin receptors do not possess a negatively charged residue at position 6.55. Notably, the exceptions, 5-HT_1ER and 5-HT_1FR, exhibit low affinity for 5-CT. This correlation suggests that the presence of an acidic acid at 6.55 (E^6.55) may negatively regulate 5-CT binding. Interestingly, in 5-HT_5AR simulation, E305^6.55 contributes a substantial favorable portion to the binding energy of 5-CT. We noticed that residues at 6.55 and 45.52 occupy roughly opposing positions. In 5-HT_1ER and 5-HT_1FR the amide-binding cavity is already constricted by the bulky I^45.52, and the presence of E^6.55 introduces steric clash. In contrast, V194^45.52 in 5-HT_5AR widens this cavity, allowing E305^6.55 to form a productive hydrogen bond with the 5-carboxamide group of 5-CT, thereby playing a positive, selectivity-determining role.

Together, our results demonstrated that S204^5.42A differentially affected the activation of downstream signaling by 5-CT and 5-HT, showing a clear β-arrestin2 signaling bias. The residue at 5.42 has been discovered contributing into the β-arrestin biased pharmacologic action of ligands in the serotonin receptors, as literatures reported [5]. The diversity of the molecular mechanism between serotonin receptors raised the possibility of designing selectively biased ligand for specific receptor in serotonin receptor family.

Authors’ contribution

X.Z., Z.S. and Z.X. initiated the structural study of 5-HT_5AR and its ligand; J.Y. and X.T. designed the expression constructs, and purified and prepared the 5-CT/5-HT_5AR/G_i/scFv16 complex under the supervision of Z.S. and Z.X.; J.Y. and X.T. performed cryo-EM screening, data collection, model building, refinement and MD simulations in the study; X.Z., P.C. and L.X. performed cell function assays on 5-HT and 5-CT; L.W. performed molecular dynamics simulation of 5-CT/5-HT_5AR/G_i1/scFv16 complex with the assistance from L.X.; L.X. analyzed data and prepared the figures with the assistance from X.Z.; X.Z. and L.X. wrote the manuscript; J.W., Z.S., W.Y. and Z.X. revised the manuscript. Z.X. planned and coordinated the entire project. All authors have read and approved the final manuscript.

Funding

This work was supported by the National Key R&D Program of China (2024YFA1107500 to Z.S.); the National Natural Science Foundation of China (32371288 to W.Y. and 82404716 to Z.X.); the Sichuan Science and Technology Program (2024ZDZX0055 to Z.S.); 1.3.5 Project for Disciplines of Excellence, West China Hospital, Sichuan University (ZYGD25003 to Z.S.).

Data availability

Cryo-EM coordinates and density maps of 5-CT/5-HT_5AR/G_i1/scFv16 complex have been deposited at the PDB and the Electron Microscopy Data Bank under accession codes 9W6W, EMD-65707.

Any additional data reported in this paper is available from the lead contact upon request.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interest.