ACS Medicinal Chemistry Letters

Alzheimers disease (AD) is characterized by the accumulation of amyloid-{beta} (A{beta}) peptides, which are a key factor in its pathogenesis. In this study, we present the design and evaluation of {gamma}-amino-L-proline peptides as metabolically stable, cell-penetrating molecules that can modulate amyloidogenic processing. We screened a library of {gamma}-peptides in primary neuronal cultures to determine their effects on endogenous A{beta}1-42 production, cytotoxicity, and {beta}-secretase (BACE1) activity. Comparative analysis of structurally related analogues enabled the identification of molecular features associated with A{beta}-lowering activity, establishing a qualitative structure-activity relationship. Peptide 33 (P33) emerged as a lead candidate, selectively reducing BACE1 activity without significantly inhibiting the homologous enzyme, BACE2. In vitro blood-brain barrier (BBB) assays revealed that P33 exhibits favorable transendothelial permeability. Intraperitoneal administration of P33 in APP/PS1 mice decreased A{beta} levels, reduced amyloid plaque burden, and improved performance in a behavioral recognition task without inducing cytotoxicity or systemic toxicity. These results define cis-{gamma}-amino-L-proline peptides as a bioorganically distinct and modular scaffold for the development of intracellular modulators of A{beta} production. HighlightsO_LI{gamma}LJAminoLJLLJproline peptides as metabolically stable modulators of A{beta} production. C_LIO_LIP33 showed BBB permeability and BACE1 inhibition in primary cortical neurons. C_LIO_LIIn APP/PS1 mice, P33 lowers amyloid burden and improves cognition. C_LIO_LIP33 shows good biocompatibility, supporting its therapeutic potential in AD C_LI

Alzheimers disease (AD) is a multifactorial disease with mixed pathologies. Consequentially, drugs targeting multiple pathological processes may offer synergistic benefits. While histone deacetylase (HDAC) inhibitors have demonstrated efficacy in alleviating AD-related pathologies in animal models, the neuroprotective Wnt/{beta}-catenin signaling pathway remains compromised in AD brain. CI-994 is a class I HDAC inhibitor containing N-(2-aminophenyl)-benzamide. Our recent studies indicate that CI-994 is also an activator of Wnt/{beta}-catenin signaling by stabilizing Wnt co-receptor LRP6. We herein use CI-994 as a scaffold to develop novel potent dual modulators of class I HDACs and Wnt/{beta}-catenin signaling for AD therapy. Our lead compound, W2A-28, selectively inhibits class I HDAC1, 2 and 3 with IC50 values of 0.51 M, 0.68 M, and 0.22 M, respectively, and shows no inhibitory activities on other HDACs. Furthermore, W2A-28 potently activates Wnt reporter activity with an EC50 value of 1.61 M in Wnt-3A-expressing HEK293 cells. As expected, activation of Wnt/{beta}-catenin signaling by W2A-28 is associated with elevated LRP6 protein level. Importantly, W2A-28 displays excellent microsomal stability in both mouse and human liver microsomal stability assays, alongside high permeability and a lack of active efflux in MDR1-MDCKII models. Critically, W2A-28 treatment significantly enhances histone acetylation, activates Wnt/{beta}-catenin signaling, and suppresses tau phosphorylation in AD patient-specific cerebral organoids carrying APOE {varepsilon}4/{varepsilon}4 or APOE {varepsilon}3/{varepsilon}4 with PSEN1 M146V mutation. Our findings position W2A-28 as a promising multi-target drug candidate for AD therapy.

We have developed and characterized a potent and cell active tau tubulin kinase 1 and 2 (TTBK1 and TTBK2) inhibitor, 13. Compound 13 demonstrates in-cell, kinome-wide selectivity, and potently inhibits both TTBK1 and TTBK2. As part of our medicinal chemistry campaign, we also identified a structurally similar negative control, compound 5, which lacks in-cell affinity for TTBK1 and TTBK2. Based on their substrates, which include TDP-43, tau, and tubulin, TTBK1 and TTBK2 inhibition has been pursued as a therapeutic approach for Alzheimers disease, frontotemporal lobe dementia, and amyotrophic lateral sclerosis. TTBK2 is also an effector of ciliogenesis, acting in concert with CEP164, CP110, and CEP83 to initiate the biogenesis of primary cilia. The development of selective chemical tools for these kinases facilitates investigation into TTBK1/2-mediated pathways and potential disease-altering ramifications linked to their pharmacological perturbation.

Homeodomain-interacting protein kinase 4 (HIPK4) remains an understudied member of the dark kinome. While genetic knockout studies suggest roles for HIPK4 in spermiogenesis and cutaneous squamous cell carcinoma, whether these cellular functions can be recapitulated by pharmacological inhibition remains to be determined. However, such investigations have been hampered by a lack of high-quality chemical tools. To address this, we employed a rational design strategy utilizing macrocyclization of a bosutinib-based scaffold. Systematic optimization led to the discovery of AZ137 (28e), a potent and selective HIPK4 inhibitor (IC50 = 11 nM; cellular EC50 = 76 nM). AZ137 exhibits exceptional selectivity across three comprehensive orthogonal panels, high solubility, and no detectable cytotoxicity. Its cellular activity was confirmed in cell-based assays of HIPK4-dependent F-actin remodeling. Together with a negative control compound, this probe set provides a foundational framework for the validating HIPK4 as a therapeutic target and a high-quality resource to elucidate its roles in normal physiology and disease. For Table of Contents Only O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=106 SRC="FIGDIR/small/720179v1_ufig1.gif" ALT="Figure 1"> View larger version (28K): org.highwire.dtl.DTLVardef@12438borg.highwire.dtl.DTLVardef@11083beorg.highwire.dtl.DTLVardef@1395fb4org.highwire.dtl.DTLVardef@1ba3db8_HPS_FORMAT_FIGEXP M_FIG C_FIG

Ligand optimization is central to drug discovery as hundreds of analogs might be designed and synthesized between an initial hit and a therapeutic candidate. The efficiency of this process is unclear, at least partly because there is no random background for optimization against which to compare. Such a random background might emerge from synthetically accessible but otherwise systematic random small substitutions across starting ligands, measuring likelihood of achieving a substantial improvement in affinity/potency or other property by any single perturbation. Recent literature and ligand-affinity/potency databases suggest that perhaps 10% of analogs with minor modifications improve upon a parents potency substantially (by [≥]10-fold), but this number is clouded by reporting bias, intentional improvement, and inter-group reproducibility. To begin to establish a background expectation for ligand optimization, we comprehensively and systematically modified 18 lead molecules across six targets with single atom changes; 257 compounds were synthesized. Unexpectedly, 11.2% of these random small perturbation analogs improved potency by [≥]10-fold over their parents. Conversely, these more potent analogs typically had worse in vitro pharmacokinetics (e.g. reduced metabolic stability, lower plasma free fraction). While it was possible to find analogs where the potency increase compensated for inferior exposure and half-life, resulting in more potent compounds in vivo, overall a frustrated landscape for ligand optimization is revealed. This study begins to establish a background expectation for ligand potency optimization and offers a simple strategy to do so. It also begins to quantify the challenges confronting the field in moving beyond in vitro potency.

The use of covalent warheads targeting the catalytic cysteine has been a cornerstone in coronavirus main protease (Mpro) inhibitor development, where various electrophilic motifs have been used including aldehydes, nitriles, ketoamides, and hydroxymethyl ketones (HMKs). Recent efforts have been mostly centered around nitrile warheads, given the success of compounds like Nirmatrelvir and Ensitrelvir in the clinic. However, finding and advancing alternative chemotypes with differentiating chemical and pharmacological profiles is essential for future pandemic preparedness. Among such alternatives, HMKs hold special interest because they balance reduced intrinsic electrophilicity with an excellent selectivity profile. Nevertheless, early HMK-based compounds, such as the clinical-stage Mpro inhibitor PF-00835231, suffered from poor oral bioavailability and therefore required intravenous administration, with or without prodrug derivatization of the hydroxyl group. Here, we describe our efforts in advancing the HMK field via the discovery of mCMX110, a lead that has superior potency, increased unbound exposure in vivo, and favorable oral bioavailability in preclinical studies. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=105 SRC="FIGDIR/small/725542v1_ufig1.gif" ALT="Figure 1"> View larger version (22K): org.highwire.dtl.DTLVardef@abe1c9org.highwire.dtl.DTLVardef@746a08org.highwire.dtl.DTLVardef@dd5861org.highwire.dtl.DTLVardef@1d572c7_HPS_FORMAT_FIGEXP M_FIG C_FIG

Neuroactive steroids modulate GABAA and NMDA receptors allosterically, typically requiring specific structural features for their activity. In this study, we characterize YX84, a novel neuroactive steroid bearing a 3{beta} sulfate and p-trifluoroacetylbenzyl alcohol attached in an ether linkage to a hydroxyl group at steroid carbon 17. This compound and similar analogues exhibit an atypical pharmacological profile, with three distinct actions at GABAA receptors. First, YX84 is a full agonist, with EC50 near 1 {micro}M and comparable efficacy to GABA at GABAA receptors in native hippocampal neurons. It presents as a full agonist relative to GABA at 4/{delta} subunit-containing receptors. Second, YX84 acts as a slow-onset, potent positive allosteric modulator (PAM) of GABAA receptors at concentrations below those that gate a response. Finally, YX84 exhibits rapid desensitizing and/or blocking kinetics; voltage dependence is consistent with a contribution of channel block. Structure- activity relationship analyses reveal that both functional groups are essential for gating activity, while classical requirements such as carbon 3 hydroxyl stereoselectivity and carbon 5 reduction are dispensable. YX84 also modestly inhibits NMDA receptor currents, suggesting weak negative allosteric modulation. Behavioral assays show that intraperitoneal administration of YX84 (30 mg/kg) does not impair sensorimotor function, unlike allopregnanolone. These findings identify YX84 as a structurally distinct neuroactive steroid with dual receptor activity and favorable behavioral tolerability, offering a promising scaffold for therapeutic development targeting excitatory/inhibitory imbalance in neuropsychiatric disorders if pharmacokinetic considerations can be overcome.

The LC3/GABARAP protein family is a promising target for selective inhibition of autophagy and for targeted protein degradation. LC3/GABARAP proteins are challenging targets for small-molecule drug development due to their long, shallow binding grooves. In this work, we evaluate multiple approaches to stabilizing the extended structure of the native binding motif, producing N-methylated peptides and stapled peptides with low nanomolar affinity. A crystal structure and molecular dynamics simulations support a model where the N-methylation pre-organizes the motif into an extended, strand-like structure. N-methylation allowed minimization of the binding motif to a tetrapeptide that retained sub-micromolar affinity while minimizing charge and overall molecular weight. The truncated, N-methylated tetrapeptide showed moderate passive permeability. These results highlight more drug-like space for the development of LC3/GABARAP ligands with high affinity and selectivity.

Despite the development of vaccines and antivirals, coronavirus disease 2019 (COVID-19) continues to affect populations worldwide. Given the high mutation rate of the SARS-CoV-2 virus and reports of drug resistance, there is a continued need for new therapeutic options. SARS-CoV-2 main protease (Mpro) is essential for viral replication and is a conserved target among coronaviruses. Most known Mpro inhibitors target the active site, although allosteric sites have already been identified. In this study, we conducted a virtual screening of 2,060 compounds targeting an allosteric site of SARS-CoV-2 Mpro. From this screen, 41 computational hits and analogs were selected and evaluated using biochemical assays against SARS-CoV-2 Mpro. Among them, compound 25, a semicarbazone, demonstrated a half-maximal inhibitory concentration (IC50) of 99 M. Additionally, two thiosemicarbazone analogs (compounds 50 and 51) inhibited SARS-CoV-2 Mpro with IC50 values of 61 M and 70 M. Biochemical assays suggest that these compounds act as noncovalent competitive inhibitors of SARS-CoV-2 Mpro. Molecular dynamics simulations revealed that compound 25 is unstable at the allosteric site of SARS-CoV-2 Mpro but forms stable and favorable interactions at the active site, supporting its potential as a competitive inhibitor, a finding subsequently confirmed by biochemical assays. Our structure-based computational and biochemical approach identified semicarbazone and thiosemicarbazone scaffolds as promising candidates for the development of reversible SARS-CoV-2 Mpro inhibitors.

The chikungunya virus (CHIKV) outbreak imposes a significant burden on healthcare systems and raises an urgent need for effective antiviral therapies. So far there are no specific drugs against CHIKV. A CHIKV macrodomain is critical for virulence and counteracts the host immune response, representing a promising antiviral drug target. Here, we describe small molecule inhibitors targeting the CHIKV macrodomain. Compound 1 (MDOLL-0273) was identified through a high-throughput screening using a fluorescence resonance energy transfer (FRET)-based assay, and its inhibitory activity was validated through multiple orthogonal assays. Compound 1 has a dual thiobarbiturate-indole scaffold and exhibits an IC50 of 8.9 {micro}M. X-ray crystallography revealed that the inhibitor occupies an adenine binding site of the macrodomain and extends into a novel cryptic pocket. Notably, the inhibitor shows high selectivity for the CHIKV macrodomain over a panel of human and viral ADP-ribosyl binding and hydrolyzing proteins. Structure-activity relationship studies and medicinal chemistry efforts provide a promising starting point for further hit optimization.

TEA domain (TEAD) proteins bind co-activator Yes-associated protein (YAP) to regulate the expression of target genes of the Hippo pathway. The TEAD*YAP protein-protein interaction is not druggable, but TEADs possess a unique and deep palmitate pocket with a highly conserved cysteine located outside the TEAD*YAP protein-protein interaction interface. Here, we screen a fragment library of acrylamide electrophiles and identify a fragment that forms an adduct with the conserved palmitate pocket cysteine and inhibits TEAD4 binding to YAP. Synthesis of a focused set of derivatives and time- and concentration-dependent studies with four TEADs provide reaction rates and binding constants. Co-crystal structures of fragments bound to TEAD2 and TEAD3 reveal reaction at the conserved palmitate pocket cysteine but also at another less conserved cysteine located in the palmitate pocket of TEAD2 closer to the TEAD*YAP interface. These fragments provide a starting point for the development of allosteric acrylamide small-molecule covalent TEAD*YAP inhibitors.

Intrinsically disordered proteins (IDPs) represent major yet challenging therapeutic targets in neurodegenerative disease due to their conformational heterogeneity and aggregation-prone behavior. Tau protein is a prototypical IDP that forms pathological aggregates in Alzheimers disease and related tauopathies. Despite extensive clinical efforts, tau-directed monoclonal antibodies have demonstrated limited efficacy. Concurrently, single-domain antibodies (nanobodies) have been gaining importance because of their small size and membrane penetrating capabilities. New design paradigms are therefore required for nanobodies to enable precise targeting of disease-relevant conformations. Here, we describe a biophysical modelling and AI-guided nanobody discovery targeting the VQIVYK motif of tau, which constitutes the structural core of neurofibrillary tangles in Alzheimers Disease. Biophysical modelling-based target analysis identified low-energy conformational states of VQIVYK. These conformational insights were used to guide AI-driven nanobody design of CDR3 loops. Starting from a nanobody scaffold, we generated 145 candidate nanobodies through systematic backbone sampling and neural network-guided sequence design, followed by multi-dimensional computational prioritization. Two candidates demonstrated robust binding to synthetic full tau protein in ELISA binding assays, achieving binding indices of 148.9% and 140%, relative to reference controls. Notably, one candidate also exhibited strong reactivity in post-mortem Alzheimers disease human brain tissue, with a binding index of 236.1%, exceeding that of the positive control (222.9%). Structural analysis indicates that our nanobodies engineered CDR3 engages VQIVYK through optimized aromatic and hydrophobic interactions. Together, these findings establish a proof-of-concept for biophysics-guided, AI-guided nanobody engineering against IDPs and identifies them as a promising lead for tau-targeted single domain antibody development.

Protein tyrosine kinases are important regulators of cell signaling, and aberrant kinase activity contributes to many human diseases, including cancers. All protein tyrosine kinases share a highly-conserved ATP binding pocket but diverge in their substrate binding sites in order to mediate distinct signaling events. Many potent and efficacious ATP-competitive tyrosine kinase inhibitors have been developed, however it remains challenging to achieve on-target selectivity across different kinases and target specific disease mutants, given the high degree of conservation in the ATP-binding pocket. By contrast, the variable substrate-binding site offers an opportunity for selective inhibition, provided molecules can be targeted to this site. Here, we present a modular strategy to design selective, peptide-based covalent inhibitors of tyrosine kinases with a distinct binding mode from existing ATP-competitive inhibitors. Using Src kinase as a model system, we demonstrate that Src-selective reactivity can be achieved by first designing an optimized substrate peptide and then strategically positioning an electrophile on the peptide to target a non-conserved cysteine on the kinase. We show that substrate-derived covalent peptides can inhibit kinase activity, bind simultaneously with an ATP-competitive inhibitor, and even inhibit the activity of kinases bearing a common drug resistance mutation. We further explore the application of this approach to develop an inhibitor of the cancer-relevant fibroblast growth factor receptor 1 kinase that shows selectivity for an oncogenic mutant over the wild-type enzyme. Our modular strategy to generate selective covalent peptides targeting protein tyrosine kinases provides a promising framework for future chemical probe and drug development efforts.

Large-conductance calcium-activated potassium (BKCa) channels are ubiquitously expressed in mammalian cells and regulate electrical activity, intracellular calcium signaling, and cell survival. Although BKCa dysfunction has been linked to multiple diseases, the number of selective channel modulators is limited. In this study, we characterize dibenzoylmethane (DBM), a plant-derived compound isolated from Hottonia palustris, as a novel inhibitor of BKCa channel activity in both plasma membrane and mitochondrial BKCa. Electrophysiological recordings revealed that DBM lowers the open probability of BKCa channels in a concentration-dependent fashion and markedly reduces mean open time, leading to a pronounced flickering behavior - hallmarks of pore-targeted blockade. Competition experiments demonstrated that DBM antagonizes the effect of paxilline, a high-affinity pore-binding inhibitor, suggesting overlapping binding sites. Molecular dynamics simulations further supported this hypothesis, showing that several DBM molecules can block the pore by employing {pi}-{pi} interactions with each other and pore residues. On top of the pore, the carbonyl groups of DBM block the nearest potassium ion in the selectivity filter. The presence of DBM induces the removal of water molecules from the pore. To assess the structural requirements for activity, we tested three DBM analogs: phenyl-1,3-butanedione (PBD), trans-chalcone (T-Ch), and (E)-1,3-diphenylprop-2-en-1-ol (DPE). T-Ch and DPE inhibited BKCa channels with comparable efficacy to DBM, whereas PBD was significantly less potent. These results indicate that diphenyl substitution and structural rigidity are critical determinants of inhibitory activity. Our findings position DBM and its analogs as promising chemical scaffolds for the development of selective BKCa channel modulators with potential pharmacological applications.

Human Neutrophil Elastase (HNE) plays a vital role in several inflammatory diseases, however its role in the tumour microenvironment and the potential in cancer treatment is still unrevealed. Considering the potential of {beta}-lactams as HNE inhibitors, the present work describes the development of a synthetic strategy to obtain two different types (Type I and Type II) of quenched activity-based probes (qABPs), using a {beta}-lactam ring as a warhead and BODIPY-FL as a fluorophore. The two types differ in mechanism and relative position between the fluorophore and the quencher moiety. The qABPs synthesized presented IC50 values against HNE lower than 0.5 {micro}M, and high selectivity compared with homologous serine hydrolases. Type II qABPs showed a more efficient turn-on mechanism, and selectively targeted HNE in different cell lysates. The qABP 22 was internalized in U937 cells and in human neutrophils and successfully targeted HNE in both.

Continuous emergence of SARS-CoV-2 variants carrying mutations in Spike presents a significant challenge for durable antiviral agents. Here we screen for random 13-amino acid non-mimetic macrocyclic peptides that bind to Spike and identify PA-001 that inhibits SARS-CoV-2 infection with high potency at 0.23-2.9 nM as 50% inhibitory concentration (IC50). PA-001 bound to Spike S2 subunit and inhibited the membrane fusion during virus entry. Through drug-resistant selection, we revealed that PA-001 targeted the fusion peptide proximal region (FPPR) in S2, which has not been recognized as a drug target to date. Consistent with its highly conserved amino acid sequences beyond strains, PA-001 exhibited broad antiviral activity against all tested SARS-CoV-2 variants, in contrast to clinically-approved S1-targeting antibodies that lost activity to Omicron variants. PA-001 suppressed SARS-CoV-2 propagation and disease progression in mouse- and hamster-infection models, both by administration prophylactically and therapeutically. Combination therapy with remdesivir further enhanced antiviral profiles. In clinical phase-I trial, PA-001 was well-tolerated and showed high systemic exposure, with 4,300-10,300-fold concentration of IC50 as maximum plasma concentration by single administration to healthy volunteers. These evidence propose FPPR as an unexpected antiviral drug target accessible by macrocyclic peptides and identify PA-001 as a potent anti-SARS-CoV-2 fusion inhibitor.

G protein-coupled receptors (GPCRs) are central regulators of human physiology and disease, classifying them as relevant targets for therapeutic interventions. As transmembrane proteins, they convert extracellular signals into intracellular responses through agonist-induced conformational changes. Understanding how agonists stabilize active receptor conformations is decisive for rational drug design. In this study, we used the endogenous ligand pancreatic polypeptide (PP) and the cyclic hexapeptides UR-AK95c and UR-AK86c as molecular tools to determine key interactions critical for Y4R activation, which plays a crucial role in metabolic diseases. Guided by molecular docking, we systematically replaced Y4R residues and assessed activation. The in vitro and in silico studies delineated a key Y4R activation interface centered around the conserved C-terminal RXRY-NH2 motif of the peptides and, separately, identified receptor residues with distinct peptide-specific functional effects. Next, we performed an ultra-large library screening (ULLS) and experimentally validated three predicted hits as selective Y4R agonists that engage in a substantial subset of the identified critical receptor contacts. This study demonstrates how GPCR activation interface knowledge can be translated into the discovery of novel small-molecule agonists and outlines a general strategy for advanced GPCR drug discovery.

Microorganisms in the human gut influence the efficacy and metabolism of host-targeted small molecule therapeutics, including the frontline Parkinsons disease drug levodopa (L-dopa). Previous work has identified a mechanism-based inhibitor of gut bacterial decarboxylases that degrade L-dopa, -fluoromethyltyrosine (AFMT). However, early experiments with AFMT in rodent models suggested undesirable in vivo metabolism by host tyrosine hydroxylase, producing a metabolite likely to worsen Parkinsons phenotypes and prevent application as an L-dopa co-treatment. Here, we demonstrate oxidation of AFMT in vitro by recombinant human tyrosine hydroxylase. We then develop AFMT analogs that retain activity against bacterial decarboxylases but have reduced susceptibility to host hydroxylation. Suitable arenes for inhibitor design were identified using assays with commercially available noncanonical amino acids, which revealed aryl difluorination as a promising modification. Difluoroaryl AFMT derivatives are less prone to degradation by tyrosine hydroxylase in vitro yet still inhibit L-dopa metabolism by bacterial decarboxylases. This work exemplifies how substrate reactivity can streamline design of mechanism-based enzyme inhibitors, as well as how constraints posed by the host can be incorporated during development of microbiome-targeted therapeutics. The compounds reported here are promising starting points for future studies in animal models and further exploration of gut bacterial effects on L-dopa treatment efficacy.

Targeting protein-protein interactions (PPIs) with small molecules is historically challenging due to shallow, solvent-exposed interfaces that lack classical binding pockets. Furthermore, employing traditional structure-based virtual screening (SBVS) across ultra-large chemical spaces to find novel chemotypes imposes prohibitive computational bottlenecks. Here, we report the first prospective, real-world application of the PyRMD2Dock platform, an AI-enforced SBVS workflow that integrates machine learning and standard docking available within the PyRMD Studio suite. To target the structurally demanding immune receptor CD28, a chemically diverse subset of 2.4 million molecules from the Enamine REAL Diversity Space was docked into a cleft adjacent to the canonical ligand interface. These data were used to train 672 classification models, and the optimized model rapidly screened the remaining [~]46 million compounds. Following interaction filtering and clustering, 232 highly prioritized ligands were identified. Experimental validation of 150 purchased candidates yielded a remarkable hit rate, identifying multiple direct CD28 binders. Lead compounds 100 and 104 exhibited submicromolar affinity (Kd = 343.8 nM and 407.1 nM, respectively), potent CD28-CD80 disruption, and functional blockade in cellular reporter assays. Furthermore, these compounds successfully reduced cytokine secretion in primary human tumor-PBMC and epithelial tissue co-culture models. This study validates PyRMD2Dock as a highly scalable, effective protocol for mining massive chemical libraries to discover small-molecule modulators of challenging immune receptor interfaces.

The glucose-dependent insulinotropic polypeptide receptor (GIPR) is an attractive therapeutic target for metabolic disorders, with GIPR antagonism emerging as a promising strategy for obesity and type 2 diabetes. However, developing functional antibodies against GPCRs remains challenging due to their complex architecture and conformational dynamics. Here, we employed AlfaBodY, an iterative active learning platform integrating structural and sequence information, to in silico design human anti-GIPR antibodies. Through four rounds of optimization, we generated antibodies with high binding affinities. Lead candidates AB106-131 (KD 1.2 nM) and AB106-156 (KD 1.7 nM) exhibited 7 to 10-fold higher affinity than 2G10 (KD 12 nM) while maintaining comparable antagonistic activity in a cAMP reporter assay (IC50 4[~]5 nM). In diet-induced obese mice, AB106-156 alone induced weight loss comparable to that of semaglutide ([~] -15%), while preserving lean mass and achieving sustained weight control after treatment withdrawal. Co-administration with the GLP-1 receptor agonist semaglutide produced synergistic weight reduction (-25.4%) and markedly attenuated the fat-mass rebound observed with semaglutide alone. Our results demonstrate that AI-driven design can generate potent anti-GIPR antibodies with favourable in vivo efficacy, supporting further development of GIPR antagonist for obesity and related metabolic disorders. The AlfaBodY platform enables faster development of more efficacious biologic drugs.