Dear Editor,
Growth hormone-releasing hormone (GHRH) receptor (GHRHR) belongs to class B1 G protein-coupled receptors (GPCRs), and is mainly expressed in growth-stimulating somatotropic cells of the anterior pituitary gland (
Rivier et al., 1982;
Guillemin et al., 1982;
Mayo et al., 2003). GHRH secreted from the hypothalamus binds and activates GHRHR, evoking G
s-coupled cAMP signaling and stimulating growth hormone (GH) secretion, which governs systemic growth and lipid and protein metabolism (
Granata et al., 2025). Consequently, either insufficient or excessive GHRHR signaling is associated with diseases: isolated GH deficiency and dwarfism when the signal is too low; gigantism, acromegaly, lipodystrophy, and certain cancers upon abnormally high stimulation (
Corazzini and Salvatori, 2013). GHRHR is therefore a clinically validated target whose agonists are sought for GH deficiency syndromes, whereas antagonists may restrain hormone-dependent tumors (
Mayo et al., 2003;
Halmos et al., 2025). The representative antagonist, MIA-602, a synthetic 30‑amino acid peptide (
Zarandi et al., 2017), has demonstrated potent antitumor activity and exerts broad antioxidative benefits (
Perez et al., 2012;
Zhang et al., 2020). PCO371, an allosteric agonist originally developed for parathyroid hormone 1 receptor (PTH1R), also potently activates GHRHR (EC
50 within one order of magnitude of that for PTH1R) (
Zhao et al., 2023;
Kobayashi et al., 2023). We have previously reported the cryogenic electron microscopy (cryo-EM) structures of GHRHR–G
s and splice variant 1 (SV1)–G
s complexes bound by GHRH, revealing the molecular basis of peptide binding, receptor activation, and SV1-mediated biased signaling (
Cong et al., 2021;
Zhou et al., 2020).
To elucidate the structural basis for its conformational plasticity and ligand-dependent signaling, we studied three different functional states of GHRHR, the ligand-free and Gs-coupled (Apo), the allosteric agonist PCO371-bound and the antagonist MIA-602-bound, using cryo-EM and molecular dynamics (MD) simulations. The Apo GHRHR–Gs and PCO371–GHRHR–Gs cryo-EM structures were determined with global resolutions of 3.04 Å and 2.88 Å, respectively (Figs. 1A, 1B, S1 and S2). The high-resolution cryo-EM maps are sufficiently clear to allow for structural modelling of the receptor, the Gs heterotrimer, and the bound small-molecule agonist PCO371 (Fig. S3 and Table S1).
Structural superposition reveals that the transmembrane domain (TMD) of the
Apo GHRHR aligns well with that of the GHRH-bound GHRHR (PDB: 7CZ5), showing a low Cα root-mean-square deviation (RMSD) of 1.01 Å (Figs. 1C and S4A). This structural similarity suggests that the
Apo receptor can adopt a conformation compatible with G
s coupling, which is supported by functional studies showing low-level basal G
s activity in the absence of agonist using a sensitive NanoBiT assay (
Cong et al., 2021;
Zhou et al., 2020). While the intracellular region of GHRHR adopts a conserved architecture upon G
s coupling, the extracellular halves exhibit a pronounced structural adaptability, primarily dictated by the absence of bound peptide agonist and the integrity of extracellular domain (ECD; Figs. 1D–G and S4). This plasticity is most evident in extracellular loop 1 (ECL1) and the extracellular half of TM1, with subtle variations in ECL3 and TM7, whereas ECL2 and TMs 2–5 remain largely rigid. In GHRH-bound structures (both GHRHR and SV1), ECL1 folds inward toward the orthosteric pocket to engage the middle region (residues 10–18) of GHRH, which is stabilized by specific interactions, including hydrogen bonds between R11
GHRH and the backbone of H194
ECL1, and between S18
GHRH and D193
ECL1, as well as stacking interactions between Y10
GHRH and F187
2.72b [class B1 GPCR numbering in superscript (
Wootten et al., 2013)] (Fig. 1E). In contrast, in the
Apo GHRHR, ECL1 adopts a naturally extended conformation along TM2 and TM3, stabilized by an intramolecular polar network involving TM2–ECL1–TM3 segments (e.g., hydrogen bonds between sidechains of T192
ECL1 and H194
ECL1; backbone atoms between H194
ECL1 and F197
ECL1/S198
ECL1; and sidechain of S198
ECL1 and the backbone/sidechain atoms of S189
ECL1) (Fig. 1F). This results in a substantial outward displacement of ECL1 in
Apo GHRHR, with the Cα atoms of C195
ECL1 moving by 16.7 Å and 16.2 Å relative to its position in GHRH-bound GHRHR and SV1, respectively (Figs. 1D and S4A). A similar outward movement of ECL1 is observed in the
Apo SV1 structure, further validating the adaptability of this loop in response to ligand binding (Fig. S4B). Notably, the integrity of the ECD significantly influences ECL1 dynamics (Fig. S4E). The extracellular half of TM1 represents another key element of the peptide-binding pocket with marked adaptability (Fig. 1G). In GHRH-bound structures, this segment forms an extended α-helix (8 residues) that shapes the binding pocket through hydrophobic contacts (e.g., L119
1.29b and E123
1.33b) and stacking interactions (especially F126
1.36b). In the
Apo GHRHR, however, this region is disordered and could not be modeled, underscoring its role in ligand-induced structural reorganization (Fig. 1G). Similarly, ECL3 (e.g., D350
ECL3) and the extracellular ends of TM7 (e.g., L354
7.35b) slightly shift inward upon GHRH binding but adopt outward positions in the
Apo GHRHR and SV1 structures (Fig. S4F).
The extracellular plasticity in the
Apo state of GHRHR contrasts with observations in other class B1 GPCRs. For instance, in the
Apo structures of the glucagon receptor (GCGR) and glucose-dependent insulinotropic polypeptide receptor (GIPR), the orthosteric peptide-binding pocket is partially occupied by the inward-folded receptor segments, namely, ECL2 in GCGR and the TM6–ECL3 junction in GIPR (
Cong et al., 2024) (Fig. 1H). In GHRHR, however, the corresponding regions (ECL2 and ECL3) remain outward-facing, resulting in a more open extracellular vestibule in the absence of ligand. Collectively, these atomic-level observations demonstrate that the structural adaptability of the GHRHR extracellular domain, governed by agonist occupancy and ECD integrity, enables selective engagement of its cognate peptide.
The cryo-EM structure of the PCO371–GHRHR–Gs complex definitively resolves the binding mode of this synthetic agonist (Figs. 1B and S3B). PCO371 binds to a well-defined, predominantly hydrophobic pocket located deep within the intracellular half of the TMD, spatially distinct from the orthosteric pocket, with the closest atom of PCO371 positioned approximately 11.2 Å from the bound GHRH (Fig. 2A). Several key residues (F3386.49b on TM6 and Y3767.57b on TM7) act as gatekeepers, shaping and enclosing the pocket by forming a constricted region that sequesters the ligand from the lipid bilayer (Fig. 2A). This binding mode unequivocally classifies PCO371 as an allosteric agonist that modulates receptor activity from a unique intracellular site.
Of note, PCO371 binding elicits a distinct and localized conformational adjustment within the TM6 (Figs. 2B and S5). To avoid severe steric clash with PCO371, the TM6 kink region undergoes a marked outward displacement and partial unwinding. Quantitatively, the Cα atom of L3376.48b in PCO371-bound GHRHR shifts by 7.95 Å relative to its position in the Apo state and by 8.46 Å relative to the GHRH-bound state (Fig. 2B). Such a local perturbation propagates to the extracellular side of the receptor, leading to notable displacements in ECL3 and the extracellular tips of TM7 and TM1 (Fig. 2B). In contrast, TMs 2–5 exhibit highly aligned structural features across Apo, GHRH-bound, and PCO371-bound complexes (Fig. S5). This specific pattern indicates that the allosteric modulation of PCO371 is remarkably precise, primarily perturbing the helices (TM1, TM6, and TM7) that directly construct its binding site and associated extracellular loops, while leaving the structural framework of the active receptor largely unchanged.
PCO371 adopts a U-shaped conformation, stabilized by extensive interactions with both GHRHR and Gαs (Figs. 2C and S6). The dimethylhydantoin group, positioned deeply in the pocket, acts as a critical anchor to the G protein interface by forming a salt bridge with R1612.46b. Furthermore, it engages in polar interactions with the backbone atoms of Y391 (Gαs) and the sidechain atom of N3808.47b (GHRHR). The central dimethylphenyl group is stabilized by robust aromatic stacking interactions with H1652.50b, Y3767.57b, and Y391 (Gαs). The methyl substituents on this phenyl ring project into hydrophobic sub-pockets formed by L2273.54b, L3346.44b, and L394 (Gαs), effectively locking this group in place. The spiro-imidazolone moiety is advantageously lodged within the remodeled kink of TM6, where it forms multiple hydrogen bonds with Y3767.57b and the mainchain atoms of three TM6 residues (I3356.46b, L3376.48b, and F3386.49b). These polar interactions act as a molecular staple, rigidly anchoring the ligand to TM6, thereby accounting for the necessity of the local helical rearrangement (Fig. 2C). Finally, the trifluoromethoxyphenyl tail extends to the narrow cleft between TM6 and TM7, where its phenyl ring engages in stacking interactions with F3386.49b and is simultaneously enveloped by a hydrophobic sleeve formed by L3326.43b, I3356.46b, L3376.48b, I3717.52b, and L3757.56b, while its trifluoromethoxy group projects toward the lipid-facing exterior (Fig. S6A).
Although sequence alignment reveals that the PCO371-binding pocket is highly conserved across class B1 GPCRs (
Kobayashi et al., 2023;
Zhao et al., 2023), a notable distinction lies within TM6 (Fig. 2D). In GHRHR, two consecutive positions (6.44 b and 6.49 b) are occupied by large aromatic phenylalanine residues (F333
6.44b and F338
6.49b), creating a unique, continuous aromatic surface on one side of the binding pocket. In contrast, PTH1R features a valine at position 6.44 b and a phenylalanine only at 6.49 b, and other class B1 receptors possess smaller aliphatic residues (e.g., leucine, valine, or threonine) at one or both of these positions. This dual-phenylalanine motif in GHRHR provides a more extensive and complementary aromatic environment for the central dimethylphenyl and tail phenyl groups of PCO371 (Figs. 2E and S6D). The enhanced van der Waals contacts and optimized π-π stacking interactions afforded by this configuration are likely to enhance binding affinity and more effectively stabilize the active conformation of the receptor–G protein complex (Fig. S7), thus constituting a key structural basis for the observed superior agonist potency of PCO371 at GHRHR compared to PTH1R (
Kobayashi et al., 2023). This analysis highlights how subtle variations within a conserved allosteric pocket, i.e., single-residue differences, can be exploited to achieve receptor subtype selectivity, offering a precise template for structure-based design of next-generation, selective allosteric modulators for class B1 GPCRs.
To fully explore the conformational spectrum of GHRHR, we undertook a parallel effort to resolve its inactive, antagonist-bound structure in the absence of intracellular effectors (Fig. S8). Negative-staining and initial cryo-EM analyses demonstrated sample homogeneity and the presence of the complex (Fig. S8D and S8E). However, 2D class averages indicated incomplete Fab occupancy and potential flexibility at fusion junctions, thereby precluding high-resolution 3D reconstruction from the current dataset (Fig. S8F). Given the challenges in obtaining a high-resolution structure of the antagonist-bound GHRHR via cryo-EM, we turned to MD simulations (Figs. S9 and S10). MIA-602 functions as an antagonist by enforcing a specific inactive conformational ensemble of GHRHR through a multi-tiered mechanism. Its focused binding mode restricts receptor dynamics and prevents the compaction of the central polar network, thereby aborting signal initiation (Fig. S10A-D). Consequently, the conformational coupling between TM6 and TM7 via the H165
2.50b–E223
3.50b–T331
6.42b–Y376
7.57b (HETY) motif is halted, and the intracellular domain is maintained in a dilated, G protein-incompatible state due to the disassembly of the TM2–TM6–TM7–H8 polar network (
Zhang et al., 2017) (Fig. S10E and S10F). This hierarchical suppression of the activation steps explains the potent and sustained antagonism exerted by MIA-602.
By integrating cryo-EM and MD simulations, this study delineates the structural landscape of the human GHRHR across key conformational states: the ligand-free (Apo) Gs-coupled state, the allosteric small-molecule agonist (PCO371)-bound Gs-coupled state, and the peptidic antagonist (MIA-602)-bound inactive state (without G protein coupling). The endogenous ligand GHRH triggers large-scale inward folding of ECL1, which acts as an extracellular clamp essential for full activation. The allosteric agonist PCO371 remodels a conserved intracellular pocket, functioning as an intracellular wedge that stabilizes the active Gs interface without inducing extracellular reorganization, thereby establishing a structural basis for biased signaling. Conversely, the antagonist MIA-602 enforces an inactive state by locking the conserved intracellular HETY motif, preventing the characteristic outward swing of TM6 required for G protein engagement. Our findings collectively underscore that GHRHR signaling preference, namely, its propensity to engage specific transducers in response to different ligand stimulation, is governed by distinct, ligand-specific conformational adaptations at both extracellular and intracellular interfaces (Supplementary Materials). By linking atomic-level structural adaptations to functional signaling preferences, this work offers valuable insights into the development of next-generation GHRHR therapeutics.
The Author(s) 2026. Published by Oxford University Press on behalf of Higher Education Press.