Discovery of G Protein-Biased Antagonists against 5-HT7R

G protein-coupled receptors (GPCRs) mediate many physiological signals through G proteins and through β-arrestin–dependent pathways; ligands that preferentially activate one pathway over the other (biased ligands) can yield therapeutic benefits and fewer side effects. Among serotonin receptors, the 5-HT7 receptor (5-HT7R) is a Gs-coupled GPCR that elevates intracellular cAMP, is highly expressed in the central nervous system, and has been implicated in thermoregulation, circadian rhythm, cognition and neurodevelopmental disorders including autism spectrum disorders (ASD). Previous work identified β-arrestin–biased and conventional agonists/antagonists for 5-HT7R, but, according to the extracted text, a ligand with preferential activation of the Gs protein pathway had not been reported. Lee and colleagues set out to design, synthesise and pharmacologically evaluate a series of 6-chloro-2′-methoxy biphenyl derivatives bearing different amine moieties to identify Gs protein–biased ligands at 5-HT7R. The study combined medicinal chemistry, in vitro binding and functional assays (Gs-mediated cAMP production and β-arrestin recruitment), selectivity profiling across serotonin receptor subtypes, in vitro ADME (CYP inhibition and microsomal stability), in vivo pharmacokinetics in mice, and a behavioural test (self-grooming) in Shank3 transgenic mice to probe a potential link between 5-HT7R signalling bias and ASD-relevant stereotyped behaviour.

This work combined synthetic chemistry with in vitro pharmacology, ADME assays, in vivo pharmacokinetics and behavioural testing. Chemically, three scaffolds were prepared: 6-chloro-2′-methoxy-[1,1′-biphenyl]-3-yl-N-alkylmethanamine derivatives (series 1), the one-carbon-homologated ethyleneamines (series 2), and pyrrolidine derivatives (series 3). Representative synthetic routes included iodination, Suzuki cross-coupling to install the biphenyl core, reductive amination and SN2 reactions; full synthetic and characterisation procedures (NMR, LC/MS, HRMS, HPLC purity ≥95%) are reported in the Experimental Section. Binding and functional assays were performed in cell systems. Binding affinities (K i ) at 5-HT7R were measured by [3H]LSD radioligand competition using crude membrane fractions from transfected HEK293 cells; IC50 values were converted to K i. Gs protein–mediated activity was assessed by a luminescence cAMP assay (GloSensor) in HEK293 cells transiently transfected with human 5-HT7R and the GloSensor plasmid; compounds were tested over multiple concentrations and at least in triplicate. β-Arrestin recruitment was measured in HTLA cells (HEK293 derivative stably expressing a tTA-dependent luciferase reporter and a β-arrestin2–TEV fusion) transiently transfected with a 5-HT7R construct (Tango assay); again serial dilutions and triplicate determinations were used. Reported readouts included EC50 (concentration producing half-maximal response) and Emax (maximal observed efficacy), and relative functional bias was quantified using ΔΔlog(τ/KA) and bias factors compared with reference ligands (5-HT and E-55888). Selectivity across other serotonin receptor subtypes (5-HT1A, 1B, 1D, 1E, 2A, 2B, 2C, 3, 5A, 6) was assessed by radioligand binding (K i values). Drug–drug interaction potential and metabolic stability were probed with in vitro CYP450 inhibition assays (CYP1A2, 2C9, 2C19, 2D6, 3A4) and human liver microsomal stability assays, respectively. Pharmacokinetic (PK) parameters for lead compound 2b were obtained after intravenous (1 mg/kg) and intraperitoneal (10 mg/kg) dosing in male ICR mice; plasma was collected at multiple timepoints and analysed by LC–MS/MS, with non-compartmental PK analysis to estimate clearance, half-life, AUC and bioavailability. The self-grooming behavioural assay was performed in wild-type and Shank3 transgenic male mice (7–9 weeks old); drugs were administered intraperitoneally (2b at 5 mg/kg; SB269970 at 30 mg/kg) and total grooming duration and counts were recorded over 30 minutes. The extracted text does not clearly report animal group sizes or the number of replicates for behavioural cohorts.

Synthesis yielded 16 target compounds across series 1–3 with high purities. In binding assays, 15 of the 16 compounds bound 5-HT7R with K i values ranging from 1.6 to 110 nM. Compound 2a had the best binding affinity (K i = 1.6 nM), comparable to reference agonist E-55888 (K i = 1.3 nM); 2b showed K i = 2.8 nM. Functional profiling separated balanced agonists from Gs-biased ligands. In the cAMP (Gs) assay, many compounds showed potency and efficacy: examples include 1a (EC50 = 460 nM, Emax = 72%), 2a (EC50 = 73 nM, Emax = 77%), and 2b (EC50 = 180 nM, Emax = 91%). In the β-arrestin (Tango) assay, fewer compounds were active; three balanced agonists were identified (1a, 2a, 2e), while six compounds (1b, 1c, 2b, 2c, 2f, 2g) showed activity in cAMP production but no detectable β-arrestin recruitment and were therefore classified as Gs protein–biased in this series. Among these, 2b was highlighted as the best Gs-biased ligand on the basis of potency and efficacy in the cAMP assay (EC50 = 180 nM; Emax = 91%), with 1b as a secondary Gs-biased example (EC50 = 770 nM; Emax = 71%). Quantitative bias assessment using ΔΔlog(τ/KA) yielded a bias factor for 2b of 461, indicating strong Gs bias relative to the reference ligand E-55888, which had a bias factor <1 and thus was relatively biased toward β-arrestin in these analyses. Selectivity testing against other 5-HT receptor subtypes showed variable profiles. Compounds 1a–1e generally lacked selectivity over 5-HT2B (K i values near low tens of nM), raising potential safety considerations, whereas many compounds in series 2 exhibited improved selectivity over 5-HT2B; notably, 2a and 2b displayed more than 40-fold selectivity over 5-HT2B. Some compounds had poor selectivity over 5-HT1D. In in vitro ADME assays, 2b exhibited minimal inhibition of five major CYP isoforms (remaining enzyme activities >88.7%), suggesting low potential for CYP-mediated drug–drug interactions in these assays. Microsomal stability testing showed 71.5% remaining concentration for 2b after incubation with human liver microsomes, indicating relative metabolic stability in vitro. PK studies in male ICR mice indicated rapid systemic clearance after intravenous administration of 2b (CL = 477.7 mL/min/kg; T1/2 = 0.5 h; AUC_last = 32.9 h·ng/mL). After intraperitoneal dosing, a longer half-life (T1/2 = 1.0 h) and a larger AUC_last (442.3 h·ng/mL) were reported, yielding an apparent bioavailability of about 134% (authors present this value; no further discussion of causes is given in the extracted text). In the self-grooming behavioural assay, administration of 2b (5 mg/kg, i.p.) markedly increased grooming duration in Shank3 transgenic mice (t = 553 s) relative to saline; co-administration of the selective 5-HT7R antagonist SB269970 (30 mg/kg, given 10 min before 2b) reduced the grooming duration toward saline levels (reported t = 70.8 s in WT and 166 s in TG mice), indicating that the 2b-induced increase in grooming was blocked by a 5-HT7R antagonist. SB269970 alone did not produce significant changes compared with saline in the reported measures (WT t = 29.3 s; TG t = 181 s). The extracted text does not report statistical tests, p-values or the number of animals per group in the behavioural results.

Lee and colleagues interpret their data as the identification of a novel chemical series of 6-chloro-2′-methoxy biphenyl derivatives that yield both balanced and Gs protein–biased ligands at 5-HT7R. They emphasise that, prior to this work, a Gs protein–biased ligand for 5-HT7R had not been reported, and that compound 2b represents a potent Gs-biased agonist as judged by cAMP efficacy and the calculated bias factor (bias factor = 461 versus E-55888). Structure–activity observations are reported: mono-ethyl substitution on the amine appears to favour Gs bias in this scaffold, whereas mono-methyl substitution (as in 2a) produced a balanced agonist. In terms of translational relevance, the authors link the pharmacology to behaviour by showing that systemic administration of the Gs-biased agonist 2b increased self-grooming in Shank3 transgenic mice, an ASD-relevant stereotypy model, and that the effect was reversed by the selective 5-HT7R antagonist SB269970. From these findings the investigators suggest that 5-HT7R signalling, and specifically Gs-mediated activation, may modulate stereotyped behaviours and could be considered in future work targeting ASD symptoms. Safety and developability issues are addressed to a degree: many compounds showed acceptable selectivity profiles, with series 2 showing improved selectivity over the 5-HT2B receptor that is associated with valvulopathy; 2b showed little CYP inhibition in vitro and reasonable microsomal stability. The authors also report PK parameters indicating relatively rapid systemic elimination, and they present bioavailability data for 2b after intraperitoneal dosing. The extracted text notes that more experiments are required to clarify the association between 5-HT7R and autism, which the authors acknowledge as a limitation of the current behavioural findings. No additional limitations or statistical detail are presented in the extracted Results/Discussion text.

Lee and colleagues conclude that they synthesised a set of 6-chloro-2′-methoxy biphenyl derivatives and identified multiple 5-HT7R ligands with distinct functional profiles: three balanced agonists (1a, 2a, 2e) and six Gs protein–biased ligands (1b, 1c, 2b, 2c, 2f, 2g). Compound 2b emerged as the best Gs protein–biased ligand in the series (EC50 = 180 nM; Emax = 91%). In vivo, administration of 2b increased self-grooming in Shank3 transgenic mice, an effect that was reversed by the selective 5-HT7R antagonist SB269970. On the basis of these results the authors suggest that 5-HT7R is associated with ASD-relevant stereotypy and could be a therapeutic target for treating such behaviours, while noting that further studies are needed.