PD-1/PD-L1-IN-32

• Molecular Formula: C29H25ClFN3O3
• Molecular Weight: 518.0 g/mol
• SMILES: C1=CC=C(C(=C1)C2=CC=CC(=C2Cl)COC3=CC(=C(C=C3)CNCCO)OCC4=CC(=CN=C4)C#N)F
• InChI: InChI=1S/C29H25ClFN3O3/c30-29-23(4-3-6-26(29)25-5-1-2-7-27(25)31)19-36-24-9-8-22(17-33-10-11-35)28(13-24)37-18-21-12-20(14-32)15-34-16-21/h1-9,12-13,15-16,33,35H,10-11,17-19H2
• InChIKey: DSELNKYPNSPLEU-UHFFFAOYSA-N

PD-1/PD-L1-IN-32, also designated compound A56, is a synthetic small-molecule inhibitor of the programmed cell death-1/programmed cell death-ligand 1 (PD-1/PD-L1) protein–protein interaction (PPI). It was developed through computer-aided structural optimization of a 2-arylmethoxy-4-(2,2'-dihalogen-substituted biphenyl-3-ylmethoxy) benzylamine scaffold.

PD-1/PD-L1-IN-32 (A56): Potent Small-Molecule PD-1/PD-L1 Immune Checkpoint Inhibitor for Preclinical Cancer Immunotherapy Research

PD-1/PD-L1-IN-32, also designated compound A56, is a synthetic small-molecule inhibitor of the programmed cell death-1/programmed cell death-ligand 1 (PD-1/PD-L1) protein–protein interaction (PPI). It was developed through computer-aided structural optimization of a 2-arylmethoxy-4-(2,2'-dihalogen-substituted biphenyl-3-ylmethoxy) benzylamine scaffold [1]. A56 blocks the PD-1/PD-L1 interaction by inducing PD-L1 dimerization and stabilizing its closed conformation, thereby preventing PD-1 binding and reactivating T-cell-mediated antitumor immunity [2]. The compound exhibits an IC50 of 2.4 ± 0.8 nM in a homogeneous time-resolved fluorescence (HTRF) assay and has demonstrated significant tumor growth inhibition (TGI 55.20% at 50 mg/kg, i.g.) in an hPD-L1 MC38 humanized mouse model without overt toxicity [1].

• [1] Zhang H, Zhou S, Plewka J, et al. Design, Synthesis, and Antitumor Activity Evaluation of 2-Arylmethoxy-4-(2,2'-dihalogen-substituted biphenyl-3-ylmethoxy) Benzylamine Derivatives as Potent PD-1/PD-L1 Inhibitors. J Med Chem. 2023;66(15):10579-10603. doi:10.1021/acs.jmedchem.3c00731 View Source
• [2] Li S, et al. Molecular dynamics reveals novel small-molecule inhibitors block PD-1/PD-L1 by promoting PD-L1 dimer stability. Phys Chem Chem Phys. 2025;27(46):24932-24947. doi:10.1039/d5cp03584h View Source

Why PD-1/PD-L1-IN-32 Cannot Be Simply Replaced by Other Biphenyl-Based PD-L1 Inhibitors

Small-molecule PD-1/PD-L1 inhibitors sharing a biphenyl core, such as BMS-202, BMS-37, LP23, and HD10, exhibit wide-ranging potency (IC50 from ~3.1 nM to >50 nM), divergent binding modes, and variable species cross-reactivity that preclude simple functional interchange [1][2]. Unlike antibodies, these small molecules act by stabilizing distinct PD-L1 dimer conformations, and even subtle scaffold modifications can profoundly alter the dimerization interface, binding hotspot residues, and downstream T-cell activation profiles [2]. PD-1/PD-L1-IN-32 (A56) possesses a resolved co-crystal structure (3.5 Å) revealing a binding mode distinct from BMS-202 and other analogs, and it has been specifically validated in humanized murine models with quantitative pharmacokinetic profiling—data that are not transferable to other in-class compounds [1][3].

• [1] Zhang H, et al. J Med Chem. 2023;66(15):10579-10603. doi:10.1021/acs.jmedchem.3c00731 View Source
• [2] Li S, et al. Phys Chem Chem Phys. 2025;27(46):24932-24947. doi:10.1039/d5cp03584h View Source
• [3] Awadasseid A, Wang R, Zhang K, et al. Comparative druggability of PD-1/PD-L1 inhibitors (A56, HD10): Cross-species differences in PD-L1 exposure and efficacy. Bioorg Med Chem. 2025. doi:10.1016/j.bmc.2025.117093 View Source

PD-1/PD-L1-IN-32 (A56) vs. Closest Analogs: Quantified Differentiation for Procurement Decision-Making

PD-1/PD-L1 Inhibitory Potency: A56 vs. BMS-202, BMS-37, LP23, and HD10

PD-1/PD-L1-IN-32 (A56) demonstrates an IC50 of 2.4 ± 0.8 nM against the human PD-1/PD-L1 interaction in an HTRF assay, representing the most potent compound within its chemical series [1]. By comparison, BMS-202—the prototypical biphenyl PD-L1 inhibitor—exhibits an IC50 of approximately 18–40.4 nM in comparable HTRF formats [2][3]. BMS-37 displays an IC50 range of 18–200 nM . LP23, a non-arylmethylamine analog, shows an IC50 of 16.7 nM [4]. HD10, a closely related analog from the same research group, achieves an IC50 of 3.1 nM [5]. Thus, A56 is 7.5- to 17-fold more potent than BMS-202, 7.5- to 83-fold more potent than BMS-37, 7.0-fold more potent than LP23, and 1.3-fold more potent than HD10.

Evidence Dimension	Human PD-1/PD-L1 inhibitory activity (IC50)
Target Compound Data	IC50 = 2.4 ± 0.8 nM (A56 / PD-1/PD-L1-IN-32)
Comparator Or Baseline	BMS-202: IC50 = 18 nM (BindingDB) / 40.4 nM (hPD-L1); BMS-37: IC50 = 18–200 nM; LP23: IC50 = 16.7 nM; HD10: IC50 = 3.1 nM
Quantified Difference	A56 is 1.3× more potent than HD10; 7.5–17× more potent than BMS-202; ≥7× more potent than BMS-37 and LP23
Conditions	HTRF assay measuring PD-1/PD-L1 protein–protein interaction blockade (in vitro, cell-free)

Why This Matters

Lower IC50 translates to lower compound quantities required for equivalent target engagement in cellular and in vivo experiments, potentially reducing cost per experiment and off-target exposure at pharmacologically effective concentrations.

• [1] Zhang H, et al. J Med Chem. 2023;66(15):10579-10603. View Source
• [2] BindingDB BDBM50239947: BMS-202 IC50 = 18 nM (HTRF). View Source
• [3] Awadasseid A, et al. Bioorg Med Chem. 2025. BMS-202 hPD-L1 IC50 = 40.4 nM. View Source
• [4] Liu Y, et al. Discovery and Crystallography Study of Novel Biphenyl Ether and Oxadiazole Thioether-Based Small-Molecule PD-1/PD-L1 Inhibitors. J Med Chem. 2023. LP23 IC50 = 16.7 nM. View Source
• [5] Zhang H, et al. Design, synthesis, and evaluation of antitumor activity of 2-arylmethoxy-4-(2-fluoromethyl-biphenyl-3-ylmethoxy) benzylamine derivatives as PD-1/PD-L1 inhibitors. Bioorg Med Chem. 2024. HD10 IC50 = 3.1 nM. View Source

Cross-Species PD-L1 Binding Affinity: A56 vs. BMS-202 and HD10

In a direct comparative study, A56 displayed binding affinities of IC50 = 2.4 nM for human PD-L1 and retained substantial affinity for murine PD-L1, whereas BMS-202 showed a human PD-L1 IC50 of 40.4 nM and a markedly weaker interaction with murine PD-L1 [1]. HD10 exhibited an IC50 of 3.1 nM for human PD-L1 [1]. The cross-species binding capability of A56 underlies its robust antitumor activity in both humanized (hPD-L1 knock-in) and syngeneic mouse models, while BMS-202's poor murine PD-L1 binding limits its utility in standard syngeneic tumor models [1].

Evidence Dimension	Human and murine PD-L1 binding affinity (IC50)
Target Compound Data	A56: human PD-L1 IC50 = 2.4 nM; murine PD-L1 = active (binding confirmed)
Comparator Or Baseline	BMS-202: human PD-L1 IC50 = 40.4 nM; murine PD-L1 = weak binding; HD10: human PD-L1 IC50 = 3.1 nM
Quantified Difference	A56 binds human PD-L1 with 16.8-fold higher affinity than BMS-202; A56/human IC50 is 1.3-fold better than HD10
Conditions	Cell-free binding assay; recombinant human PD-L1 and murine PD-L1 proteins

Why This Matters

Compounds lacking murine PD-L1 cross-reactivity (e.g., BMS-202) cannot be reliably evaluated in standard syngeneic mouse models, necessitating costly humanized models. A56's dual-species activity supports broader preclinical model applicability and reduces total experimental costs.

• [1] Awadasseid A, et al. Comparative druggability of PD-1/PD-L1 inhibitors (A56, HD10). Bioorg Med Chem. 2025. doi:10.1016/j.bmc.2025.117093 View Source

In Vivo Antitumor Efficacy: A56 Tumor Growth Inhibition vs. LP23 and Clinical Candidate INCB086550

PD-1/PD-L1-IN-32 (A56) achieved a tumor growth inhibition (TGI) rate of 55.20% at 50 mg/kg (oral gavage) in the hPD-L1 MC38 humanized mouse model over a treatment period, with no observable toxicity [1]. In a separate study, LP23 demonstrated a TGI of 88.6% at 30 mg/kg (i.p.) in the B16-F10 syngeneic melanoma model ; however, this model employs murine PD-L1, making cross-study comparisons confounded by species differences in PD-L1 engagement. The clinical-stage small molecule INCB086550 (PD-1/PD-L1-IN-8) has reported IC50 values of 1.9–4.9 nM across species in HTRF assays , and Phase I/II clinical data demonstrated pharmacodynamic modulation of PD-L1 but did not meet efficacy endpoints for accelerated approval, underscoring that nanomolar potency alone does not guarantee translatable in vivo efficacy. A56's combination of single-digit nanomolar potency with in vivo TGI and a clean toxicity profile in a humanized model distinguishes it from investigational compounds that lack reported in vivo validation.

Evidence Dimension	In vivo tumor growth inhibition
Target Compound Data	TGI = 55.20% at 50 mg/kg p.o. (A56 in hPD-L1 MC38 humanized mouse model)
Comparator Or Baseline	LP23: TGI = 88.6% at 30 mg/kg i.p. (B16-F10 murine model); INCB086550: clinical PD data available but no approved efficacy endpoints
Quantified Difference	A56 demonstrates oral efficacy in a humanized PD-L1 model; LP23 uses intraperitoneal dosing in a syngeneic model with murine PD-L1 (non-comparable). INCB086550: no publicly reported TGI in preclinical models.
Conditions	hPD-L1 MC38 humanized mouse model; oral gavage (50 mg/kg); tumor volume measured throughout treatment

Why This Matters

A56 is one of very few small-molecule PD-L1 inhibitors with publicly disclosed oral in vivo efficacy in a humanized PD-L1 model and a clean toxicity profile, whereas many comparators either lack reported in vivo data, require intraperitoneal administration, or have shown disappointing clinical translatability despite potent biochemical IC50 values.

• [1] Zhang H, et al. J Med Chem. 2023;66(15):10579-10603. View Source

Pharmacokinetic Profile in Sprague-Dawley Rats: Oral Bioavailability and Dose-Exposure Relationship

Pharmacokinetic analysis of PD-1/PD-L1-IN-32 (A56) was performed in Sprague-Dawley rats, yielding quantitative parameters across intravenous, oral, and intraperitoneal routes . Following oral administration (40 mg/kg), A56 achieved a Cmax of 21,504.90 μg/L, Tmax of 1.75 h, elimination half-life (t1/2) of 1.72 h, AUC(0–t) of 63,579.08 μg·h/L, and an oral bioavailability of 114.09% . Intravenous dosing (8 mg/kg) yielded a t1/2 of 0.99 h, clearance (CL) of 0.60 L/h/kg, and volume of distribution (V1) of 1.09 L/kg . By comparison, BMS-202 has been reported to have poor oral bioavailability and a shorter half-life, contributing to limited in vivo utility without formulation optimization. INCB086550, while orally bioavailable, has only limited publicly available preclinical PK parameters. The >100% bioavailability value for A56 suggests potential enterohepatic recirculation or nonlinear pharmacokinetics that may sustain effective plasma concentrations beyond what simple half-life calculations indicate.

Evidence Dimension	Oral pharmacokinetics in rats
Target Compound Data	F = 114.09%; Cmax = 21,504.90 μg/L; Tmax = 1.75 h; t1/2 = 1.72 h; AUC(0–t) = 63,579.08 μg·h/L (40 mg/kg p.o.)
Comparator Or Baseline	BMS-202: poor oral bioavailability (class-level inference, limited published PK data); INCB086550: oral PK data available only through clinical trial disclosures, not comparable preclinical format
Quantified Difference	A56 exhibits complete oral bioavailability; comprehensive PK parameters publicly available for cross-study comparison and experimental dose design.
Conditions	Sprague-Dawley rats; iv (8 mg/kg), ig (40 mg/kg), ip (10 mg/kg); plasma concentration measured over time

Why This Matters

Well-characterized PK parameters enable researchers to design dose regimens and time points with confidence, avoiding wasted compound and animal resources on pharmacokinetic optimization. A56's high oral bioavailability supports repeated oral dosing in chronic tumor models, a practical advantage over injectable comparators or compounds with poor absorption.

Structural Differentiation: A56 Co-Crystal Structure Reveals a Binding Mode Distinct from BMS-202 and BMS-37

The X-ray co-crystal structure of A56 bound to the PD-L1 dimer at 3.5 Å resolution revealed a novel binding mode that differs fundamentally from the established binding poses of BMS-202 and BMS-37 [1]. While BMS-202 and BMS-37 occupy the canonical PD-L1 dimer interface pocket and induce dimer formation through symmetric interactions, A56 engages a distinct set of hotspot residues (including Y56(B), M115(A/B), Y123(A), and K124(A)) that confer superior binding stability, as demonstrated by molecular dynamics simulations [2]. The A56/PD-L1 system involves the largest number of hotspot residues among the inhibitors BMS-202, LP23, and A56, contributing to the highest structural rigidity and correlated motions within the PD-L1 dimer [2]. Additionally, 1H NMR titration confirmed tight binding with a KD < 1 μM [1]. This structural differentiation means that A56 is not a simple potency-improved analog of BMS-202 but a mechanistically distinct stabilizer of the PD-L1 dimer's closed conformation.

Evidence Dimension	PD-L1 binding mode and hotspot residue engagement
Target Compound Data	A56: co-crystal structure at 3.5 Å; engages Y56(B), M115(A/B), Y123(A), K124(A); KD < 1 μM (NMR); highest structural rigidity among tested inhibitors
Comparator Or Baseline	BMS-202: canonical biphenyl binding mode; fewer hotspot residues engaged; LP23: distinct non-arylmethylamine binding mode; BMS-37: similar to BMS-202
Quantified Difference	A56 engages more hotspot residues than BMS-202 and LP23; produces the highest structural rigidity (DCCM and PCA analysis)
Conditions	X-ray crystallography (3.5 Å); 1H NMR titration; molecular dynamics simulations (AMBER force field, explicit solvent)

Why This Matters

A structurally validated, distinct binding mode provides a rational basis for A56's superior potency and binding stability. For researchers developing structure–activity relationship (SAR) studies or PROTAC designs, the available co-crystal structure enables rational modification without the ambiguity of docking to homology models or low-resolution structures.

• [1] Zhang H, et al. J Med Chem. 2023;66(15):10579-10603. View Source
• [2] Li S, et al. Phys Chem Chem Phys. 2025;27(46):24932-24947. View Source

Safety Profile: A56 Demonstrates No Observable Toxicity in Humanized Mouse Model at Efficacious Dose

In the hPD-L1 MC38 humanized mouse model, PD-1/PD-L1-IN-32 (A56) administered at 50 mg/kg orally produced significant tumor growth inhibition (TGI 55.20%) without obvious toxicity against normal mouse function [1]. In contrast, BMS-37 displays non-specific cytotoxicity against modified Jurkat T cells with an EC50 of 3–6 μM , indicating a narrower therapeutic window. Several biphenyl-based inhibitors from the BMS series have been reported to exhibit off-target cellular toxicity at concentrations near their PD-L1 inhibitory range, complicating preclinical efficacy interpretation. The absence of observable toxicity at a therapeutically effective dose suggests that A56 possesses an improved selectivity profile relative to earlier-generation biphenyl PD-L1 inhibitors, although formal safety pharmacology studies (hERG, CYP inhibition, AMES) have not been publicly disclosed and should be conducted prior to advanced preclinical development.

Evidence Dimension	In vivo tolerability at efficacious dose
Target Compound Data	A56: no obvious toxicity at 50 mg/kg p.o. (hPD-L1 MC38 humanized mouse model); TGI = 55.20%
Comparator Or Baseline	BMS-37: EC50 = 3–6 μM (Jurkat T cell cytotoxicity); BMS-202: limited in vivo data due to poor PK; INCB086550: clinical immune-related adverse events reported
Quantified Difference	A56 achieves in vivo efficacy without evident toxicity at the same dose, whereas BMS-37 shows cellular toxicity in the low micromolar range.
Conditions	hPD-L1 MC38 humanized mouse model; 50 mg/kg i.g.; body weight, behavior, and gross organ appearance monitored

Why This Matters

A favorable therapeutic window at the preclinical stage reduces the risk of attrition during lead optimization and provides confidence that observed antitumor effects are immune-mediated rather than driven by direct cytotoxicity. For procurement decisions, this supports selection of A56 for long-term in vivo efficacy studies where compound tolerability is critical.

• [1] Zhang H, et al. J Med Chem. 2023;66(15):10579-10603. View Source

PD-1/PD-L1-IN-32 (A56): Optimal Experimental and Industrial Application Scenarios Based on Quantitative Evidence

Preclinical In Vivo Efficacy Studies Requiring Oral Dosing in Humanized PD-L1 Models

PD-1/PD-L1-IN-32 (A56) is uniquely suited for long-term oral dosing studies in humanized PD-L1 mouse models, having demonstrated a TGI of 55.20% at 50 mg/kg p.o. in the hPD-L1 MC38 model without observable toxicity [1]. Its high oral bioavailability (F = 114.09% in rats) and favorable Cmax (21,504.90 μg/L) support once-daily oral gavage protocols . Unlike BMS-202, which exhibits poor murine PD-L1 binding and limited oral bioavailability, A56's cross-species activity enables direct comparison between humanized and syngeneic models, facilitating mechanistic studies of species-specific immune responses [2].

PROTAC Design Using A56 as a PD-L1-Targeting Warhead

A56 has been successfully employed as a PD-L1-targeting warhead in PROTAC design, leading to the development of compounds 9i and 9j, which induce PD-L1 internalization and degradation via both proteasomal and lysosomal pathways while retaining PD-1/PD-L1 inhibitory activity [3]. The availability of a co-crystal structure (3.5 Å) [1] and well-characterized binding mode with defined hotspot residues (Y56, M115, Y123, K124) [4] enables rational linker attachment without disrupting PD-L1 engagement, a critical advantage over inhibitors like BMS-37 or LP23, for which fewer structural data are publicly available.

Structure-Based Drug Design and Molecular Dynamics Benchmarking

The A56/PD-L1 co-crystal structure at 3.5 Å resolution and accompanying molecular dynamics data provide a robust platform for structure-based optimization of next-generation PD-L1 inhibitors [1][4]. A56's engagement of the largest number of hotspot residues among BMS-202, LP23, and A56, combined with its highest structural rigidity in MD simulations, makes it an ideal benchmark compound for computational studies aimed at improving binding free energy or pharmacokinetic properties [4]. Researchers performing docking screens or free energy perturbation (FEP) calculations can use A56 as a well-characterized reference ligand.

In Vitro Functional T-Cell Activation Assays Requiring High-Potency PD-1/PD-L1 Blockade

With an IC50 of 2.4 ± 0.8 nM in HTRF assays [1] and confirmed cell-based hPD-1/hPD-L1 blocking activity [1], A56 is appropriate for in vitro T-cell activation and tumor cell co-culture experiments where maximal PD-L1 blockade is required at low compound concentrations to minimize solvent toxicity. A56's potency advantage over BMS-202 (IC50 18–40.4 nM) [2] and LP23 (IC50 16.7 nM) [5] translates to lower DMSO concentrations in cell culture media (typically ≤0.1% at IC90), reducing vehicle-related artifacts in functional immune assays.

• [1] Zhang H, et al. J Med Chem. 2023;66(15):10579-10603. View Source
• [2] Awadasseid A, et al. Bioorg Med Chem. 2025. View Source
• [3] Design, synthesis, anti-tumor activity and mechanism of novel PROTACs as degraders of PD-L1 and inhibitors of PD-1/PD-L1 interaction. Bioorg Med Chem. 2024. doi:10.1016/j.bmc.2024.117093 View Source
• [4] Li S, et al. Phys Chem Chem Phys. 2025;27(46):24932-24947. View Source
• [5] Liu Y, et al. J Med Chem. 2023. LP23 IC50 = 16.7 nM. View Source