Elucidating the molecular logic of a metabotropic glutamate receptor heterodimer

Metabotropic glutamate (mGlu) receptors are class C GPCRs that form disulfide-linked dimers and modulate neuronal excitability and synaptic transmission. Eight mGlu subtypes are grouped by sequence, G-protein coupling, and pharmacology. Beyond homodimers, mGlu protomers heterodimerize in cells and brain tissue, altering pharmacology. mGlu₂/₄ heteromers are prominent in cortex, striatum, hippocampus, and olfactory bulb, and heteromerization likely influences circuit-specific physiology relevant to neuropsychiatric disorders. However, mixed populations of homo- and heterodimers complicate attribution of signaling. The authors’ CODA-RET approach isolates signaling from defined dimer species to dissect activation mechanisms and allosteric modulation. Motivated by divergent electrophysiological effects of mGlu₄ PAMs (e.g., Lu AF21934 vs PHCCC) at different synapses, the study tests the hypothesis that differential PAM activity arises from distinct actions at mGlu₄/₄ homomers versus mGlu₂/₄ heterodimers and seeks to define the molecular determinants of these differences.

Prior work established that mGlu receptors form heterodimers in heterologous systems and in vivo, with functional evidence for mGlu₂/₄ heteromers in multiple brain regions using nanobody-based energy transfer approaches. Electrophysiology showed that some mGlu₄ PAMs (VU0155041, Lu AF21934) potentiate responses at corticostriatal and thalamocortical synapses, whereas others (PHCCC, VU0418506) do not, though they can act at other synapses. Previous modeling and mutagenesis in mGlu₄/₄ homodimers suggested two overlapping allosteric sub-pockets in the 7TM domain: an upper (shallow) pocket (e.g., VU0155041) and a lower (deep) pocket (e.g., PHCCC). Prior studies on heteromer activation mechanisms yielded conflicting results regarding cis vs trans-activation, partly due to methodological differences (e.g., GABAB C-tail retention systems and chimeric G proteins). These gaps motivated a systematic comparison of homomer versus heteromer pharmacology and mechanism with tools that isolate defined dimer signaling.

• Assay to isolate dimer-specific signaling: CODA-RET using split RLuc8 fragments L1 and L2 fused to mGlu₂ and/or mGlu₄ C-termini to force complementation only in defined homo- or heterodimers. Gαᵢ with mVenus serves as BRET acceptor to report G-protein recruitment upon activation. Bystander BRET and forced association controls performed.
• Mutagenesis to map function: Allosteric pocket mutants in mGlu₄ 7TM domain to test docking predictions: R655Q (upper pocket), W798A (shared/stronger for lower pocket PAMs), V826M (lower pocket). Orthosteric site and G protein coupling mutations to control cis/trans activation paths: mGlu₄ T182A (agonist binding-deficient), F781A (G-protein coupling-deficient); mGlu₂ T168A (agonist binding-deficient), F756A (G-protein coupling-deficient). Configurations constructed to probe cis (agonist binds and G protein couples via same protomer) and trans (agonist binds one protomer, coupling via the other) activation in homomers and heteromers.
• Ligands: Orthosteric agonists L-AP4 (group III), DCG IV (group II), glutamate; mGlu₄ PAMs: VU0155041, Lu AF21934 (upper pocket), PHCCC, VU0418506, VU0364770, ADX88178 (lower pocket), VU0415374 (bridging both pockets); mGlu₂ PAMs: BINA, LY487379; mGlu₄ NAM: VU0448383. Competition experiments assessed overlap (e.g., VU0155041 vs VU0415374, ADX88178) in heteromers.
• Computational docking: Induced-fit docking using cryo-EM structure of human mGlu₄/₄-Gi (PDB 7E9H) and analysis of heteromer structure (PDB 8JD5). SiteMap identified pockets; Glide Induced Fit (Schrödinger) docked PAMs, with interaction fingerprint clustering to define residue interaction groups (upper-only, shared, lower-only).
• Downstream signaling: BRET-based cAMP inhibition assay (CAMYEL) in HEK293T with uncomplemented receptors to assess Gi functional output. Use of selective agonists and G-protein-blocking mutants isolates heterodimer-mediated cAMP signals indicative of trans-activation.
• Cell system and transfection: HEK293T cells maintained in DMEM; PEI transfection; specified DNA ratios for homo-/heterodimer expression; serum starvation before assays.
• CODA-RET protocol: Cells plated (~300k/well), ligand stimulation 5 min at 37°C, Coelenterazine H substrate, BRET read (525/485 nm). Dose-response fitting with fixed Hill slope = 1. Statistics via unpaired two-tailed t-tests (p<0.05 significant).
• cAMP protocol: Cells plated (~60k/well), forskolin (10 µM), ligand treatment 30 min at 37°C; BRET read via CAMYEL sensor.

• mGlu₄ PAM subsets differ in heteromer efficacy: VU0155041 and Lu AF21934 potentiate both mGlu₄/₄ homomers and mGlu₂/₄ heteromers; PHCCC, VU0418506, VU0364770, and ADX88178 potentiate mGlu₄/₄ but are inactive at mGlu₂/₄.
• Two overlapping allosteric sub-pockets validated: Docking on cryo-EM mGlu₄/₄ revealed an upper pocket (extracellular side) and a lower pocket (deeper in 7TM). VU0155041/Lu AF21934 bind upper; PHCCC/VU0418506/VU0364770/ADX88178 bind lower. Interaction fingerprints grouped residues into upper-specific, shared, and lower-specific sets.
• Mutational validation in mGlu₄/₄: R655Q abolished VU0155041/Lu AF21934 PAM effects with minimal impact on lower-pocket PAMs; W798A reduced VU0155041 and abolished lower-pocket PAMs; V826M spared upper-pocket PAMs but blocked lower-pocket PAMs. The same pattern held for VU0155041 on mGlu₂/₄ heteromers, indicating upper-pocket binding drives heteromer PAM efficacy.
• VU0415374 binds both pockets but is inactive at heteromers: It potentiated mGlu₄/₄ but had no effect on mGlu₂/₄; it competitively reduced VU0155041’s enhancement in mGlu₂/₄, indicating binding without positive modulation in the heteromer context.
• Orthosteric activation modes: mGlu₄/₄ homodimers display both cis- and trans-activation by orthosteric agonists (L-AP4, glutamate). In mGlu₂/₄ heteromers, both protomers signal efficiently via cis-activation by their selective agonists (L-AP4 for mGlu₄; DCG IV for mGlu₂), but orthosteric agonists alone produce little to no trans-activation in either direction.
• Upper-pocket mGlu₄ PAMs enable interprotomer communication in heteromers: VU0155041 enhances mGlu₄ cis-activation (L-AP4) and acts as a trans-PAM to enhance mGlu₂ cis-activation (DCG IV). Critically, VU0155041 enables robust trans-activation from mGlu₄ to mGlu₂ (L-AP4 driving Gi coupling via mGlu₂). The effect requires an intact G-coupling site on mGlu₂ and is reproduced with Lu AF21934.
• Asymmetry of trans-activation: Neither orthosteric agonists nor VU0155041 enabled substantial trans-activation from mGlu₂ to mGlu₄. mGlu₂ PAMs (BINA, LY487379) boosted mGlu₂ cis-activation but did not produce trans-activation to mGlu₄ or affect L-AP4-driven responses.
• Downstream signaling corroboration: In CAMYEL cAMP assays designed to isolate heteromer output, L-AP4 inhibited cAMP only in the presence of VU0155041 (but not PHCCC or BINA), confirming VU0155041-enabled trans-activation from mGlu₄ to mGlu₂ at the level of Gi signaling. No DCG IV-driven cAMP inhibition via mGlu₂-to-mGlu₄ trans-activation was detected with or without PAMs.
• Mechanistic implications: Data suggest the lower pocket either becomes inaccessible, lowers PAM affinity, or decouples from activation in the mGlu₂/₄ heteromer. Upper-pocket PAMs promote interprotomer allostery that is otherwise blunted in heteromers, revealing an inherent signaling asymmetry.

The study resolves how specific allosteric binding sites in mGlu₄ determine heteromer pharmacology and activation. The upper 7TM pocket supports PAM activity in both mGlu₄/₄ and mGlu₂/₄, enabling cis enhancement at mGlu₄ and trans-communication to mGlu₂, including trans-activation. In contrast, lower-pocket PAMs, although active in homomers, fail to modulate mGlu₂/₄, likely due to altered pocket accessibility or coupling within the heteromer. Orthosteric agonists alone drive independent cis-activation of each protomer in the heteromer with minimal trans-activation, unlike homodimers that display robust cis and trans modes. VU0155041 and Lu AF21934 reveal a latent capacity for interprotomer allostery in heteromers, but with directionality: mGlu₄-to-mGlu₂ communication is enabled, whereas mGlu₂-to-mGlu₄ remains poor with the tested PAMs, indicating asymmetry. Docking and mutagenesis substantiate pocket-specific roles and suggest that subtle structural differences (e.g., TM3 side chain orientations) in the heteromer may impede lower-pocket function. Discrepancies with prior reports likely arise from differences in receptor constructs (e.g., GABAB C-tail systems, chimeric G proteins) that may alter coupling or interfaces. Overall, the work clarifies the molecular logic of mGlu₂/₄ signaling and provides a framework for heteromer-selective pharmacology.

This work demonstrates that mGlu₄ PAMs segregate by allosteric binding pocket with distinct consequences for mGlu₂/₄ heteromer function: upper-pocket PAMs (VU0155041, Lu AF21934) are active in both homomers and heteromers and uniquely enable mGlu₄-to-mGlu₂ trans-activation and trans-PAM effects, whereas lower-pocket PAMs (PHCCC, VU0418506, VU0364770, ADX88178) fail to modulate mGlu₂/₄. Orthosteric agonists activate mGlu₂/₄ via cis mechanisms with minimal trans-activation unless an upper-pocket PAM is present. These findings explain divergent PAM effects observed across synapses and highlight an intrinsic asymmetry in heteromer signaling. Future directions include: identifying heteromer-selective modulators that spare homomers; structural studies to define PAM-induced interprotomer conformational changes and pocket accessibility in heteromers; and applying CODA-RET-guided screens with strategic mutations to discover precision therapeutics targeting specific mGlu heterodimers.

• Heterodimer isolation in downstream cAMP assays is indirect and relies on selective agonists and G-protein–blocking mutations; complete exclusion of homomer contributions is inferred but not directly visualized.
• Structural interpretations of pocket accessibility in heteromers are limited by uncertainty in side-chain positions and by the finite number of available cryo-EM states; direct structures of heteromers bound to different PAM classes would strengthen conclusions.
• The study uses HEK293T overexpression systems; native neuronal environments, receptor stoichiometries, and accessory proteins may influence heterodimer interfaces and allosteric coupling.
• Only a subset of PAMs/NAMs was tested; other chemotypes or pocket binders could reveal additional modes of trans-allostery, including potential mGlu₂-to-mGlu₄ effects.
• Trans-activation from mGlu₂ to mGlu₄ was not detected under conditions tested; low-sensitivity or context-dependent signaling pathways not probed here could conceivably reveal such effects.