Cryo-EM structures of GnRHR: Foundations for next-generation therapeutics

Spodoptera frugiperda ( Sf9 , Expression systems) cells and insect Hi5 cells (Expression 391 systems) were grown in SIM HF medium (Sino biological ink.) at 27 °C and 120 rpm. HEK-293T cells were grown in a humidified 37 °C incubator with 5% CO 2 using Dulbecco’s Modified Eagle Medium containing 10% fetal bovine serum (VWR). The pGnRHR (full length) and fGnRHR (full length) were cloned into the pFastBac Vector, which contains a N-terminal FLAG tag (DYKDDDDV) and a tobacco etch virus protease site before the receptor using homologous recombination (CloneExpress One Step Cloning Kit, Vazyme). The native signal peptide was replaced with prolactin precursor sequence which is modified from the N-terminal of bovine prolactin sequence to promote protein expression, consistent with our previously established construct ( 28 ). The Gα q was designed based on a mini-Gα s skeleton with an N terminus replaced by Gα i for the binding of scFv16. All the three G-protein components, including rat Gβ1 and bovine Gγ2, were cloned into a pFastBac vector, respectively. The single-chain antibody fragment scFv16, which was also cloned into a pFastBac vector with a with a GP67 signal peptide and C-terminal His8 tag, was added to stabilize the nucleotide-free GnRHR–miniGα q complexes. GnRHR, miniGα q , Gβ1, Gγ2, and scFv16 were coexpressed in Hi5 insect cells (Expression system) using the Bac-to-Bac baculovirus expression system (invitrogen). Cell pellets were thawed and lysed in 20 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), PH 7.4, 100 mM NaCl, 10% glycerol, 5 mM CaCl 2 , 5 mM MgCl 2 , supplemented with 25 mU mL −1 apyrase and ethylenediaminetetraacetic acid (EDTA)-Free Protease Inhibitor Cocktail. The GnRHR–G q complex was formed in membranes by the addition of 10 μM GnRH. The suspension was incubated for 1.5 h at room temperature. The membrane was then solubilized using 0.5% Lauryl maltose neopentyglycol (LMNG) and 0.1% Cholesterol hemisuccinate for 2 h at 4 ˚C. The supernatant was collected by centrifugation at 80,000 g for 40 min and then incubated with FLAG beads for 2 h at 4 ˚C. After batch binding, the resin was loaded into a plastic gravity flow column and washed with 20-column volumes of the same buffer (20 mM HEPEs, pH 7.4, 100 mM NaCl, 10% glycerol) plus 0.01% LMNG, 0.001% glyco-diosgenin (GDN), 2 mM CaCl 2 , 2 mM MgCl 2 , and 5 μM GnRH, and finally eluted using 0.2 mg/mL FLAG peptide. The complex was then concentrated using an Amicon Ultra Centrifugal Filter (molecular mass cut-off, 100 kDa) and injected onto a Superose 6 Increase 10/300 GL column (GE Healthcare) equilibrated in the buffer containing 20 mM HEPES, pH 7.4,100 mM NaCl, 2 mM MgCl 2 , 0.00075% LMNG, 0.00025% GDN, and 5 μM GnRH used above, respectively. The complex fractions were collected and concentrated for electron microscopy experiments. For the preparation of cryo-EM grids, 3 μL of the purified protein at 6.8 mg/mL for the GnRH-pGnRHR–G q complex and 8.5 mg/mL for the GnRH–fGnRHR–G q complex were applied onto a glow-discharged holey Nitinol grid (M01 Au300-r1.2/1.3, Nanodim). Grids were plunge-frozen in liquid ethane using Vitrobot Mark IV (Thermo Fischer Scientific). Frozen grids were transferred to liquid nitrogen and stored for data acquisition. Cryo-EM imaging of the complex was performed on a Titan Krios at 300 kV in the Advanced Center for Electron Microscopy at Shanghai Institute of Materia Medica, Chinese Academy of Sciences (Shanghai China). A total of 9,698 movies for the GnRH–pGnRHR–G q complex and 8,454 movies for the GnRH–fGnRHR–G q complex were collected on a Titan Krios equipped with a Falcon 4 direct electron detection device at 300 kV. Images were taken with a pixel size of 0.73 Å, a defocus ranging from −1.0 to −2.0 μm, and a total dose of 50 e/Å 2 over 2.5 s exposure on each EER format movie, using the EPU software (FEI Eindhoven, Netherlands). Each movie was divided into 36 frames during motion correction. The single particle analysis of GnRHR–G q complexes was performed with cryoSPARC v4.3.1 ( 46 ). Dose-fractionated image stacks were subjected to motion correction by MotionCor2 ( 47 ). Contrast transfer function (CTF) parameters for the micrograph were estimated by patch CTF estimation ( 48 ). For the GnRH–pGnRHR–G q complex, the autopicked particles were extracted on a binned dataset with a pixel size of 1.46 Å and were subjected to reference-free two-dimensional (2D) classification to discard poorly defined particles. Rounds of two-dimensional classification gained a total of 494,682 particles, which were retained for three-dimensional (3D) classifications. After multiple rounds of heterorefinement, a total of 118,050 particles were subsequently subjected to nonuniform refinement and local refinement. The final density map was obtained with an overall resolution of 3.18 Å at a Fourier shell correlation of 0.143. For the GnRH–fGnRHR–G q complex, a total of 1,196,735 particles were picked from template-picked particles through several rounds of 2D classification. These selected particles were subjected to 3D classifications. A total of 257,378 particles were selected and subjected to nonuniform refinement, local refinement. The final density map was obtained with a nominal resolution of 2.67 Å at a Fourier shell correlation of 0.143. For the GnRH–GnRHR–G q complexes, the AlphaFold2 structure of African clawed fGnRHR and pGnRHR and the structures of miniG q , rat Gβ1, bovine Gγ2, and scFv16 (PDB: 8HCQ) were used as the initial model for model rebuilding and refinement against the electron microscopy map ( 29 , 49 ). The model was docked into the electron microscopy density map using ChimeraX ( 50 ), followed by iterative manual adjustment and rebuilding in COOT and ISOLDE ( 51 , 52 ). Real-space refinements were performed using Phenix programs ( 53 ). The model statistics were validated using MolProbity ( 54 ). The final refinement statistics were validated using the module “Comprehensive validation (cryo-EM)” in Phenix ( 55 ). The final refinement statistics are provided in SI Appendix , Table S1 . Structural figures were prepared in Chimera ( 56 ), ChimeraX, and PyMOL ( https://pymol.org ). The G q signaling of hGnRHR was assessed using the inositol phosphate IP1 accumulation assay with the IP-One HTRF kit (Cisbio). HEK293T cells were cultured in a 12-well plate at a density of 400,000 to 500,000 cells/mL, then transfected with 1.0 μg plasmid for expression of the receptors. After 24 h of incubation at 37 °C with 5% CO 2 , the cells were harvested by centrifugation and resuspended in stimulation buffer. A 7 μL volume of the cell suspension was transferred to a white 384-well plate, followed by the addition of 7 μL of ligands. The reaction was incubated for 1 h at 37 °C. Subsequently, 3 μL each of IP1-d2 and anti-IP1 cryptate in lysis buffer were added and incubated for 30 min at room temperature. IP1 levels were quantified using the IP-One HTRF kit on an EnVision multiplate reader (PerkinElmer), and data were normalized to the baseline ligand response. Dose–response curves were fitted using a three-parameter logistic model (log[agonist] vs. response, GraphPad Prism v8.0.2) with a fixed Hill slope of 1 to determine EC 50 values. pEC 50 and span for each curve were calculated by GraphPad Prism v8.0.2. ∆pEC 50 equals pEC 50 of GnRH to specific mutant minus pEC 50 of GnRH to WT. Data were present as mean values ± SEM; all experiments were performed at least three times independently; n.s., not significant; comparisons among multiple groups were performed by one-way ANOVA and differences with P < 0.05 were considered significant: * P < 0.05; ** P < 0.01; *** P < 0.001; **** P < 0.0001. The cell seeding and transfection followed the same method as the IP1 assay. Cells were seeded (12-well plate) and transiently transfected with 1.0 μg WT or mutated hGnRHR every hole. After 24 h of transfection, cells were washed once with phosphate-buffered saline (PBS) and then detached with 0.2% (w/v) EDTA in PBS. Cells were blocked with PBS containing 5% (w/v) bovine serum albumin (BSA) for 15 min at room temperature before incubating with primary anti-FLAG antibody (diluted with PBS containing 5% BSA at a ratio of 1:200, Abclonal) for 1 h at room temperature. Cells were then washed three times with PBS containing 1% (w/v) BSA and then incubated with anti-mouse Alexa-488-conjugated secondary antibody (diluted at a ratio of 1:1,000, Abclonal) at 4 ˚C in the dark for 1 h. After another three times of washing, cells were collected, and fluorescence intensity was quantified in a BD Accuri C6 flow cytometer system (BD Biosciences) through BD Accuri C6 software1.0.264.21 at excitation 488 nm and emission 519 nm. Approximately 10,000 cellular events per sample were collected to detect the surface expression of GnRHR mutants. The data were normalized to the WT receptor. Experiments were performed at least three times, data were presented as means ± SEM. Different agonist–GnRHR complexes were modeled using the builder module in Schrödinger’s software suite based on the GnRH–pGnRHR–G q complex, with the G protein components excluded. The resulting complexes were prepared using the Protein Preparation Wizard in Schrödinger’s Maestro, where hydrogens were added, and protonation states of residues were assigned via PROPKA3 ( 57 ). The structures were then energy-minimized with a 0.3 Å constraint applied to the heavy atoms. Before performing MMGBSA calculations, an implicit membrane was incorporated into the transmembrane region. MMGBSA calculations were carried out using the OPLS4 force field ( 58 ) and a generalized Born model with variable dielectric for the solvent ( 59 ). The elagolix-GnRHR (PDB: 7BR3) complex was first prepared using Schrödinger’s Maestro software. During the preparation, hydrogens were added to the system at pH 7.0, missing side chains were incorporated, and the structure was minimized to resolve any steric conflicts. Using these minimized structures as the starting point, the antagonists linzagolix, merigolix, relugolix, and sufugolix were docked into the GnRHR binding pocket using glide with standard precision. All experiments were performed at least three times independently. GraphPad Prism (version 8.0.2) was used to analyze the statistical results. All data are presented as the means ± SE of the means (means ± SEM). Comparisons among multiple groups were performed by one-way ANOVA. Differences with P < 0.05 were considered significant.

Mammalian GnRHRs share high sequence identity and lack a C-terminal tail, a feature present in nonmammalian GnRHRs ( SI Appendix , Fig. S1 ). Due to difficulties in obtaining stable wild-type (WT) hGnRHR protein using insect and 293 expression systems, we chose to study the pig GnRHR (pGnRHR), which closely resembles the human receptor. To investigate evolutionary aspects of this receptor, we also analyzed the African clawed frog GnRHR (fGnRHR), which represents a nonmammalian counterpart ( SI Appendix , Fig. S1 ). To produce high-quality GnRHR–G q complexes, we modified the receptor’s N terminus by adding a prolactin precursor sequence to enhance expression and purification, followed by a FLAG (DYKDDDDV) tag for affinity purification as we designed before ( 28 ). We utilized an engineered Gα q chimera based on the mini-Gα s scaffold, with its N terminus replaced by Gα i 1 sequences to facilitate scFv16 binding ( 29 ). Unless otherwise stated, “G q ” refers to this mini-Gα q chimera used throughout our structural studies. The GnRHR receptors were coexpressed with mini-Gα q , Gβγ, and scFv16 in High Five insect cells. Since the mammal GnRH decapeptide sequence (pGlu-His-Trp-Ser-Tyr-Gly-Leu-Arg-Pro-Gly-NH 2 ) exists across human, pig and African clawed frog ( SI Appendix , Fig. S2 ) ( 30 , 31 ), we used this ligand to activate the receptors, forming ligand-bound receptor-G protein complexes at the cell membrane. Affinity chromatography followed by size-exclusion chromatography yielded complex peaks, confirming the high purity and homogeneity of the complexes, making them suitable for cryo-EM analysis ( SI Appendix , Fig. S3 A and D ). Further cryo-EM studies successfully resolved the structures of the G q -coupled pig and African clawed fGnRHRs bound to GnRH at 3.18 Å and 2.67Å resolution, relatively ( Fig. 1 A – D and SI Appendix , Fig. S3 ). These high-resolution structures enabled precise placement of all key components, including GnRHR, the three G protein subunits, scFv16, and the ligand. In the fGnRHR structure, most side chains within the extracellular loops (ECLs), seven transmembrane helices (TMDs), and intracellular loops (ICLs) were well defined, except for ICL3. Similarly, the ICL1 and ICL3 regions of pGnRHR were unresolved ( Fig. 1 E and SI Appendix , Fig. S4 ). Structure features of fGnRHR and pGnRHRs bound to GnRH. ( A and B ) Orthogonal views of the cryo-EM density map and the corresponding atomic model of the GnRH–fGnRHR–miniG q complex. ( C and D ) Orthogonal views of the cryo-EM density map and the corresponding atomic model of the GnRH–pGnRHR–miniG q complex. ( E ) Structural differences between ICL2 and ECLs of fGnRHRs and pGnRHRs. fGnRHR is shown in blue and pGnRHR is shown in purple. Although sequence alignments indicate that the fGnRHR contains a C-terminal tail, this region was not visible in the structure, mirroring the pGnRHR. These findings suggest that the C-terminal tail may primarily influence ligand-dependent receptor phosphorylation and rapid desensitization rather than G protein coupling. Structurally, the two G q -coupled complexes displayed highly similar conformational arrangements, consistent with other GPCR-G protein complexes. The rmsd of the TMDs between the two structures was just 0.89 Å, highlighting their structural similarity. Subtle differences were observed in the ECLs and ICL2 ( Fig. 1 E ), which may reflect species-specific variations in receptor dynamics or ligand interactions. GnRH and its analogs adopt a distinctive “U”-shaped conformation, as first revealed by NMR and X-ray diffraction and confirmed in the cryo-EM structures presented here ( 32 ) ( Fig. 2 A ). Unlike other peptide-bound GPCR complexes—where peptides typically insert into the receptor’s transmembranedomain (TMD) via their N terminus, C terminus, or a cyclic middle segment—GnRH interacts with its receptor through both N- and C-terminal ends, leaving its cyclic middle segment exposed toward the extracellular space ( Fig. 2 A – C ). The conserved binding pocket of GnRHs across GnRHRs. ( A ) The U-shaped conformation of GnRH within the binding pocket of GnRHRs. ( B and C ) Side view of similar binding modes of GnRH in fGnRHRs and pGnRHRs. Cryo-EM density maps of GnRH are shown in salmon and green, relatively. ( D and E ) Conserved interactions in the binding pocket across different GnRHRs. ( F ) The bar plot of ΔpEC 50 of important residues in the binding pocket of hGnRHR. ΔpEC 50 = pEC 50 of GnRH to a specific mutant GnRHR–pEC 50 of GnRH to WT GnRHR, which are all measured using IP1 accumulation assays. * P < 0.05; ** P < 0.01; *** P < 0.001; **** P < 0.0001. Data were shown as mean ± SEM from three independent experiments, which were performed in triplicates; significance was determined with one-way ANOVA. N.A., no active receptors; dashed-border rectangles, inapplicable pEC 50 of mutations causing significant reduction of ligand potency that were not able to be calculated. ( G ) 2D illustration of GnRH in the conserved binding pockets and corresponding interactions. ( H and I ) Species–specific interactions in the pockets of fGnRHRs and pGnRHRs, respectively. Side chains of residues are displayed in sticks. Hydrogen bonds are depicted as black dashed lines. In the pGnRHR complex, GnRH establishes extensive interactions with the receptor’s N terminus, TM1-7, and ECLs, excluding ECL1. High-resolution cryo-EM density maps provided detailed insights into these interactions. For instance, the first amino acid, pyroglutamate (pGlu1), forms key interactions with R38 1.35 , K121 3.32 , Y283 6.51 , and F309 7.39 in TM1, TM3, TM6, and TM7 ( Fig. 2 D ). The second residue, histidine (His2), contacts M125 3.36 , T215 5.42 , Y283 6.51 , and Y284 6.52 in TM3, TM5, and TM6, while the third residue, tryptophan (Trp3), interacts with Y283 6.51 , W291 6.59 , and F309 7.39 from TM6 and TM7 ( Fig. 2 D ). This pattern continues along the GnRH sequence, with residues such as tyrosine (Tyr5) and glycine (Gly10) forming critical hydrophilic and hydrophobic interactions with receptor residues in TM2, TM3, and TM6 ( Fig. 2 E ). Sequence alignment across 11 mammalian species, including hGnRHR, reveals that these key interaction residues in pGnRHR are completely conserved, suggesting that this binding mode is likely preserved across species ( SI Appendix , Fig. S1 ). Supporting this, functional studies on hGnRHR showed that disrupting these conserved interactions significantly impaired GnRH-induced receptor activation, confirming the critical role of these contacts in GnRH recognition and receptor activation in humans ( Fig. 2 F and SI Appendix , Fig. S5 and Tables S1 and S2 ). Despite moderate sequence identity between pig and African clawed fGnRHRs, structural comparisons reveal a highly overlapped GnRH binding site in both receptors. Several conserved interactions are evident; for instance, pGlu1 engages R56 1.35 , K133 3.32 , Y294 6.51 , and F322 7.39 in fGnRHR similarly to R38 1.35 , K121 3.32 , Y283 6.51 , and F309 7.39 in pGnRHR ( Fig. 2 D ). Other residues, such as His2, Trp3, Tyr5, Pro9, and Gly10, also maintain conserved contacts with key residues such as Y301 6.58 , W302 6.59 , and F322 7.39 in fGnRHR ( Fig. 2 D , E , and G ). However, there are also some interactions unique to these two receptors as illustrated in Fig. 2 H and I , which reflect the species-specific adaptations in the receptor’s interaction network. For example, the N terminus of pGnRHR interacts with GnRH while no corresponding cryo-EM density was observed in the fGnRHR map ( SI Appendix , Fig. S6 ). Additionally, residues such as L308 ECL3 and S318 7.35 in fGnRHR differ in type but correspond to V297 ECL3 and N305 7.35 in pGnRHR, still contributing to GnRH interactions through unique adaptations ( Fig. 2 H and I ). Although some interaction networks are species-specific, the highly conserved binding site and consistent U-shaped conformation of GnRH across both complexes strongly suggest an evolutionary conservation of receptor engagement mechanisms. These findings highlight the universal principles governing GnRH recognition by GnRHRs while accommodating species-specific variations that may fine-tune receptor function. Comparing the GnRH-bound active structures of GnRHR with the antagonist-bound inactive hGnRHR structure (PDB: 7BR3) reveals key molecular mechanisms underlying ligand-induced receptor activation ( Fig. 3 A ) ( 27 ). Superimposing these conformations reveals significant structural rearrangements in the receptor’s N terminus, ECLs, and TMD, with the most striking changes observed in the N terminus. In the inactive state, the N terminus extends deeply into the orthosteric pocket, occupying the agonist binding site. Upon GnRH binding, this region undergoes a substantial outward rotation, functioning as a “lid” that interacts directly with the ligand. In the pGnRHR structure, only part of the N terminus is visible in the cryo-EM density map, suggesting that upon ligand binding, it adopts a highly dynamic conformation outside the TMD pocket ( Fig. 3 B ). This reorganization is accompanied by additional movements in the receptor’s extracellular regions, including inward shifts of about 4.0 Å in TM1 and TM7 and an outward displacement of 3.3 Å in TM2. These extracellular conformational changes propagate to the intracellular side, where they induce significant structural rearrangements characterized by a substantial outward shift of TM6, a smaller inward shift of TM7, and a pronounced 9.1 Å displacement of TM5 ( Fig. 3 C ). Together, these ligand-induced conformational changes fully activate GnRHR, enabling downstream G q protein signaling. Activation mechanism of GnRHR. ( A ) Structural superposition of active GnRHRs and the inactive hGnRHR (PDB: 7BR3) from the side view. ( B ) Different modes of N terminus of inactive human and active pGnRHRs. The N terminus of inactive hGnRHR is highlighted in red. The movement directions of the N terminus is highlighted as a black arrow. ( C ) Top and cytoplasmic views of active GnRHRs and the inactive GnRHR. The movement directions of TMs of active GnRHRs relative to those of inactive GnRHR are highlighted as red arrows. The distances of the movements are measured by pig and hGnRHRs. ( D ) The insertion of the N-terminal segment of GnRH is deeper into the TMD core than the C terminus. ( E ) Activation of GnRHRs induced by the His2 of GnRH. ( F – H ) Conformational changes of the microswitches upon GnRHR activation such as DRS ( F ), PAF ( G ), and DPxxY ( H ) motifs. The rotation directions of residue side chains upon pGnRHR activation compared with the antagonist-bound hGnRHR are indicated by black arrows. In the active state, the N-terminal segment of the U-shaped GnRH ligand penetrates deeper into the TMD core than the C terminus, playing a pivotal role in receptor activation ( Fig. 3 D ). Further analysis reveals that the side chain of histidine at position 2 (His2) in GnRH does not directly interact with the conserved toggle switch residue W 6.48 , which is crucial for class A GPCR activation. Instead, His2 interacts with residues Y283 (Y294) 6.51 and Y284 (Y295) 6.52 , inducing conformational shifts in these residues compared to the inactive receptor. This movement facilitates the rotation of W280 (W291) 6.48 , triggering the outward movement of TM6. Simultaneously, a significant rotation of F309 (F322) 7.39 from TM7 occurs, enabling an interaction with Trp3, which drives the inward movement of the extracellular end of TM7 and shifts the cytoplasmic end inward ( Fig. 3 C and E ). These coordinated conformational changes are transmitted to conserved “microswitch” residues (DRS, PAF, and DPxxY motifs), ultimately driving GnRHR activation ( Fig. 3 F and G ). Alanine substitutions at Y283 6.51 , Y284 6.52 , and F309 7.39 significantly impair GnRH-induced hGnRHR activation, underscoring the critical role of these residues in ligand-mediated receptor activation ( Fig. 2 F and SI Appendix , Fig. S5 ). GnRHR is central to reproductive hormone regulation, and a large number of GnRHR agonists and antagonists are widely used in clinical settings to treat reproductive, oncologic, and endocrine disorders ( 12 , 20 ). Agonists are generally designed based on the sequence of the endogenous GnRH hormone, with a few key modifications to enhance stability and efficacy ( 19 ) ( SI Appendix , Fig. S7 ). Typically, the sixth glycine (Gly) residue is substituted with D-type amino acids, such as D-tryptophan (D-Trp), D-tert-butylserine (D-tBu-Ser), D-alanine (D-Ala), or D-tert-butylglycine (D-tBu-Gly), along with a modification at the peptide’s C-terminus such as acetylation or -NHNHCONH2 (Goserelin) ( 17 , 33 ) ( Fig. 4 A and B ). Substitution of the sixth glycine residue with D-amino acids has been reported to stabilize the type II’ β-turn conformation of GnRH analogs ( 34 ) ( Fig. 4 B ). In line with this, structural comparison between our cryo-EM model of GnRH and the previously reported X-ray crystal structure of Triptorelin (PDB: 4D5M) reveals a prominent type II’ β-turn spanning residues Tyr5 to Arg8 ( SI Appendix , Fig. S8 ). This observation suggests that incorporating a D-amino acid at position 6 promotes a preorganized, active-like conformation of the peptide, which may facilitate enhanced receptor engagement and activation. SAR of GnRHR ligands. ( A and B ) 2D and 3D illustrations of engineered agonists of GnRHR with the sixth glycine substituted with D-type amino acids, along with a C-terminal modification. ( C ) Extra interactions formed by the side chain of the substituted D-type amino acids with N terminus, TM5-7, ECL2, or ECL3 of the pGnRHR by modeling. ( D ) Representative docking poses of GnRHR antagonists in the binding pocket of hGnRHR. ( E ) SAR principles of GnRHR ligands. The rotation of important residues upon activation compared with the inactive GnRHR are indicated by black arrows. We modeled the binding conformation of nine GnRH analogs, which have all been on market, based on the cryo-EM structure of GnRH-pGnRHR using Schrödinger ( SI Appendix , Fig. S7 ). According to the modeling, we find that analogs with the sixth position replaced by a large D-amino acid may have more interactions with the N terminus, TM5-7, ECL2, or ECL3, which may contribute to the binding ability to the receptor ( Fig. 4 A – C ). Our Molecular Mechanics Generalized Born Surface Area (MMGBSA) calculations support this hypothesis, showing that these modified peptides have lower conformational energy compared to native GnRH with the pGnRHR model ( SI Appendix , Fig. S9 ). Functional assays further support this finding: the four representative analogs—Alarelin, Buserelin, Nafarelin, and Triptorelin—all of which possess larger side chains at position 6, form additional contacts with the receptor. Mutations at these interacting residues markedly reduce their activation potency toward GnRHR ( Fig. 4 C and SI Appendix , Fig. S10 and Table S2 ). Notably, mutation of two key aromatic residues—H199 ECL2 and Y290 6.58 —located near the binding pocket for the sixth residue, differentially affected agonist potency. Alanine substitution of these residues (H199 ECL2 A and Y290 6.58 A) results in markedly reduced potency of native GnRH. However, the activity of the D-amino acid-containing analogs is less severely impacted ( SI Appendix , Fig. S10 ), suggesting that the bulky D-amino acid at position 6 promotes a more robust and active conformation, potentially compensating for the loss of key receptor interactions and highlighting a synergistic relationship between peptide folding and receptor activation. Finally, because endogenous peptide hormones like GnRH typically have very short half-lives and limited drug viability, the replacement with D-type amino acids and the C-terminal acetylation were introduced to extend the half-life, common modifications in peptide-based drugs ( 35 , 36 ). For the small-molecular antagonists, we docked linzagolix, merigolix, sufugolix, and relugolix based on the antagonist-bound hGnRHR structure (PDB: 7BR3) ( 37 – 40 ) ( Fig. 4 D and SI Appendix , Fig. S7 ). Like elagolix, these antagonists occupy an overlapping binding site within the TMD pocket but bind closer to TM5-7, leaving room for the deep insertion of the receptor’s N terminus ( Fig. 4 D ). This binding configuration contrasts with GnRH, which occupies the central orthosteric pocket, causing the N terminus to rotate outward upon activation. Despite similar binding depths, antagonists and agonists induce distinct conformations in critical residues, resulting in active or inactive states of GnRHR ( Fig. 4 E ). Key residues Y283 6.51 and F309 7.39 play crucial roles in modulating receptor activation or inactivation. In the GnRH-bound (active) state, Y283 6.51 rotates toward GnRH, stabilizing the interaction, whereas in antagonist-bound structures, Y283 6.51 rotates away, suggesting that agonist binding stabilizes Y283 6.51 near the ligand. This rotation of Y283 6.51 affects neighboring residues, including Y284 6.52 and the toggle switch W280 6.48 , which together promote either an active or inactive conformation ( Fig. 4 E ). Additionally, F309 7.39 interacts hydrophobically with GnRH in the active state but clashes with antagonist ligands, disrupting its alignment and inducing further shifts in adjacent F313 7.43 and TM7 ( Fig. 4 E ). The milder effect of F309 7.39 A vs. Y283 6.51 A in IP1 assays likely reflects their distinct roles: Y283 6.51 anchors GnRH mainly via polar interactions essential for activation, while F309 7.39 ’s hydrophobic contacts and π–π interactions with Trp3 and Tyr5 fine-tune TM7 dynamics, with its loss partially compensated by adjacent residues (e.g., Y283 6.51 , Y290 6.58 , and W291 6.59 ). These structural observations suggest three SAR principles for GnRHR ligands: 1) agonists position centrally in the orthosteric pocket, displacing the N terminus, while antagonists bind near TM5-TM7, accommodating the N terminus; 2) binding depth alone does not determine receptor activation state; and 3) the residues Y283 6.51 is a primary driver of receptor activation and F309 7.39 acts as a modulator, both playing critical roles in the receptor activation process ( Fig. 4 D and E ). Together, these ligand-dependent mechanisms provide insights that can guide the design of more effective GnRHR-targeting drugs. The assembly of pGnRHR and fGnRHR bound to the G q protein reveals a similar overall configuration, yet several unique features distinguish the two complexes. Notably, the C-terminal end of the α5 helix in the Gα q subunit exhibits a small shift of 2.2 Å between the complexes, while the N-terminal end of the αN helix shows a more substantial variation of approximately 8.5 Å when aligning the two receptors ( Fig. 5 A ). This variation suggests a receptor-specific interaction mode with the downstream G q protein in the two species. G q protein coupling by GnRHRs. ( A ) Deviation of α5 and αN helix of G q protein in assembly with different GnRHRs. ( B and C ) Conserved interaction of hydrophilic and hydrophobic interactions between GnRHRs and G q protein. ( D and E ) Species-specific interactions of pig ( D ) and frog ( E ) GnRHRs and G q protein. ( F ) The hydrophobic and hydrophilic interfaces formed by ICL2 of GnRHRs and the αN helix of Gα q . Side chains of residues are displayed in sticks. Hydrogen bonds are depicted as black dashed lines. Despite these interspecies structural differences, further comparisons highlight that the insertion depth of the α5 helix of Gα q into the receptor—which is typically conserved across various receptors—is shallower in the GnRHR structures ( SI Appendix , Fig. S11 ) ( 28 , 29 , 41 – 43 ). This limited insertion suggests a reduced interaction interface between GnRHRs and Gα q , potentially contributing to the observed instability of the GnRHR–G q complex during purification. Consistent with this, the cryo-EM maps lack densities corresponding to the ICL regions (ICL1 and ICL3 of pGnRHR and ICL3 of fGnRHR), further indicating potential flexibility or instability in these areas ( Fig. 1 E and SI Appendix , Fig. S4 ). Two major interfaces between GnRHR and Gα q were identified in the structures: The first interface occurs between the cytoplasmic cavity of the receptor helices and the α5 helix of Gα q , while the second interface is located at the receptor’s ICL2 and the V-shaped structure formed by the αN helix and the remaining α5 helix of Gα q ( Fig. 5 A ). At the first interface, the C-terminal α5 helix of Gα q forms direct interactions with residues from receptor TM3, TM5, and TM6, with additional interactions occurring between the α5 helix and TM2 of fGnRHR ( Fig. 5 B – E ). Structural inspections indicate that many interaction residues are highly conserved. For example, residues Y356 and N352 from the C-terminus of α5 form hydrogen bonds with D 3.49 and A 3.53 of the two receptors, respectively ( Fig. 5 B ), while L353 forms hydrophobic interactions with I 3.54 and A 6.29 ( Fig. 5 C ). Additionally, N357 from the α5 helix forms hydrogen bonds with R264 6.32 from pGnRHR ( Fig. 5 D ), while Y356 and L358 form additional hydrogen bonds with H87 2.38 and R275 6.32 ( Fig. 5 E ) from fGnRHR, which are species-specific interactions. At the second interface, the side chain of L147 (L159) from receptor ICL2 inserts into a hydrophobic core formed by residues L34, V192, F194, F341, and I348 of Gα q . Besides hydrophobic interactions, polar interactions between receptor ICL2 and R31, R32 from the αN helix of Gα q were also observed ( Fig. 5 F ). These interactions are consistently present between GPCRs and Gα q and play a pivotal role in stabilizing the GPCR-G q complex. Together, these structural observations suggest that GnRHRs have evolved a largely conserved coupling mode with other GPCRs to engage downstream G q proteins.

Drugs targeting GnRHR have been pivotal in treating reproductive disorders and cancers ( 44 ). This study provides a high-resolution structural analysis of GnRHR activation, shedding light on both conserved and species-specific mechanisms of ligand recognition, receptor activation, and G protein coupling. By resolving cryo-EM structures of pig and African clawed fGnRHRs bound to GnRH, we elucidate the molecular basis of receptor function and highlight evolutionary adaptations that preserve critical physiological roles while accommodating species-specific variations. Our findings reveal a conserved U-shaped conformation of GnRH that facilitates precise interactions with key residues such as K 3.32 , Y 6.51 , and Y 6.52 in both pig and hGnRHRs. These interactions underscore an evolutionary strategy to maintain high-affinity binding and receptor activation across species. Structural comparisons with fGnRHR confirm the universality of this ligand binding mode despite moderate sequence divergence. Species-specific differences, particularly in ECL interactions and unique receptor–ligand contact points, reflect evolutionary fine-tuning of GnRHR function. These adaptations may optimize receptor activation dynamics and ligand specificity to meet distinct physiological demands. The absence of a visible C-terminal tail in both species, despite its presence in nonmammalian receptors, suggests a limited role in G protein coupling but a potential involvement in receptor phosphorylation and desensitization ( 45 ). Ligand-induced GnRHR activation involves significant structural rearrangements, including displacement of the N terminus and outward rotation of TM6, critical for G protein coupling. These conformational changes align with established GPCR activation models, providing a detailed framework for understanding GnRHR signaling. At the receptor-G protein interface, conserved contacts with the α5 helix of Gα q highlight shared principles of GPCR-G protein coupling, while variations in coupling dynamics may modulate signaling outcomes. SAR analysis of GnRHR-targeting drugs reveal the molecular basis of ligand efficacy and specificity. Substitutions in the endogenous GnRH sequence, such as replacing the sixth glycine with D-type amino acids, enhance ligand stability and binding affinity by stabilizing the U-shaped conformation. Docking models of agonists and antagonists offer further insights into activation and inactivation mechanisms. Agonists like Leuprorelin occupy the orthosteric pocket centrally, displacing the receptor’s N terminus to stabilize the active state. In contrast, antagonists such as linzagolix and relugolix bind closer to TM5-TM7, maintaining the receptor in its inactive conformation. Structural and functional analyses indicate that substitution of Gly6 with D-amino acids stabilizes a type II’ β-turn conformation, as observed in both cryo-EM and crystal structures of GnRH analogs. This pre-organized, active-like conformation likely enhances receptor engagement and confers greater functional robustness against mutations near the pocket of peptide position 6. These insights have significant implications for drug discovery. Leveraging conserved binding sites and understanding ligand-specific interactions can guide the design of selective and biased ligands that fine-tune GnRHR activity. Such ligands could preferentially modulate specific signaling pathways, minimizing side effects while maximizing therapeutic efficacy in conditions such as endometriosis, prostate cancer, and infertility. In conclusion, this study advances our understanding of GnRHR biology by elucidating conserved and unique aspects of receptor–ligand interactions, activation mechanisms, and G protein coupling. These structural insights provide a solid foundation for developing next-generation GnRHR-targeting therapeutics with improved specificity and efficacy, addressing unmet needs in reproductive and endocrine health. Future work should explore receptor dynamics in vivo and expand the structural toolkit to encompass diverse receptor states and ligand types, further enhancing the translational potential of this research.

Appendix 01 (PDF)