AZ3451

• Molecular Formula: C30H27BrN4O3
• Molecular Weight: 571.5 g/mol
• InChI: InChI=1S/C30H27BrN4O3/c1-18(20-5-3-2-4-6-20)35-26-12-9-21(30(36)33-22-10-7-19(16-32)8-11-22)13-25(26)34-29(35)23-14-27-28(15-24(23)31)38-17-37-27/h7-15,18,20H,2-6,17H2,1H3,(H,33,36)/t18-/m0/s1
• InChIKey: FJAOGFGHTPYADT-SFHVURJKSA-N

AZ3451: A Negative Allosteric Modulator of PAR2 with a Remote Binding Site for Inflammation and Pain Research

AZ3451 is a synthetic organic small molecule that functions as a negative allosteric modulator (NAM) of protease-activated receptor 2 (PAR2), a G-protein-coupled receptor implicated in inflammatory disease, pain, and cancer [1]. It was identified through DNA-encoded library screening against purified PAR2-StaR and optimized to a potent antagonist with a 1H-benzimidazole scaffold [1]. Unlike orthosteric or near-orthosteric PAR2 ligands, AZ3451 binds to a remote allosteric site outside the helical bundle, preventing the structural rearrangements required for receptor activation and downstream signaling [1]. Physicochemical characterization reports a molecular weight of 570.13 Da, XLogP of 7.46, and DMSO solubility of 97 mg/mL (169.74 mM) [2].

1 Allosteric PAR2 antagonism mechanism study fit

2 GPCR structural pharmacology workflow context

3 Selectivity profiling over PAR1 research models

• [1] Cheng RKY, Fiez-Vandal C, Schlenker O, et al. Structural insight into allosteric modulation of protease-activated receptor 2. Nature. 2017;545(7652):112-115. doi:10.1038/nature22309 View Source
• [2] IUPHAR/BPS Guide to Pharmacology. AZ3451 Ligand Page, GtoPdb Ligand ID: 9560. Accessed 2025. View Source

Why PAR2 Antagonists Are Not Interchangeable: Binding Site, Mechanism, and Potency Dictate Experimental Outcomes

PAR2 antagonists cannot be treated as a homogenous class for scientific procurement. AZ3451 binds to a remote allosteric site distinct from the orthosteric or near-extracellular pockets engaged by competitors such as AZ8838 or GB88 [1]. This difference in binding topography translates into fundamentally different antagonism mechanisms—non-competitive (allosteric) versus competitive (orthosteric-like)—which in turn affects the ability to block signaling induced by multiple endogenous proteases, tethered ligands, and biased agonists [2]. Furthermore, the >100-fold potency gap between AZ3451 and earlier-generation PAR2 antagonists such as AZ8838 or ENMD-1068 means that milligram-for-milligram substitution leads to vastly different target engagement in cellular and in vivo models [1][3]. Simply selecting any 'PAR2 antagonist' without verifying binding site, mode of antagonism, and functional potency introduces uncontrolled variables into experimental design.

Binding Site

Remote allosteric site distinct from orthosteric PAR2 antagonists may shift pharmacological profile

Mechanism

Non-competitive antagonism mode may not transfer to competitive antagonists like AZ8838

Potency Context

Functional potency context differs significantly from earlier-generation PAR2 antagonists

• [1] Cheng RKY, Fiez-Vandal C, Schlenker O, et al. Structural insight into allosteric modulation of protease-activated receptor 2. Nature. 2017;545(7652):112-115. doi:10.1038/nature22309 View Source
• [2] Kennedy AJ, Sundström L, Geschwindner S, et al. Protease-activated receptor-2 ligands reveal orthosteric and allosteric mechanisms of receptor inhibition. Commun Biol. 2020;3:782. doi:10.1038/s42003-020-01504-0 View Source
• [3] GtoPdb/IUPHAR-BPS Guide to Pharmacology. AZ3451 Ligand Page, GtoPdb Ligand ID: 9560. Accessed 2025. View Source

AZ3451: Quantified Differentiation Against Key PAR2 Antagonist Comparators

Non-Competitive Allosteric Antagonism: Schild Analysis Distinguishes AZ3451 from the Competitive Antagonist AZ8838

AZ3451 acts as a non-competitive antagonist at PAR2, whereas AZ8838, its closest structural analog from the same discovery program, is a competitive antagonist. Schild analysis in CHO-hPAR2 cells showed that for AZ8838, the Schild slope was not statistically distinguishable from unity (p > 0.05 vs hypothetical slope = 1), consistent with competitive antagonism. In contrast, AZ3451's Schild slope was significantly different from unity (p < 0.001 vs hypothetical slope = 1), confirming non-competitive, allosteric antagonism against both 2f-LIGRL-NH₂ peptide agonist and trypsin protease agonist [1]. This mechanistic distinction means AZ3451 provides insurmountable blockade regardless of agonist concentration.

Schild Analysis

Head-to-head

AZ3451: p < 0.001 vs unity (non-competitive). AZ8838: p > 0.05 vs unity (competitive).

Mechanistic differentiation context

CHO-hPAR2 FLIPR; 2f-LIGRL-NH₂ & trypsin agonists

Evidence Dimension	Mode of antagonism determined by Schild slope deviation from unity
Target Compound Data	AZ3451: Schild slope statistically distinguishable from 1 (p < 0.001); non-competitive antagonist
Comparator Or Baseline	AZ8838: Schild slope not statistically distinguishable from 1 (p > 0.05); competitive antagonist
Quantified Difference	Mechanistic classification confirmed by Schild analysis p-values (p < 0.001 for AZ3451 vs p > 0.05 for AZ8838); qualitative distinction: non-competitive vs competitive
Conditions	CHO-hPAR2 cells; calcium mobilization FLIPR assay; agonists: 2f-LIGRL-NH₂ and trypsin; n ≥ 3 independent experiments; Gaddum/Schild EC50 global fitting competitive model

Why This Matters

Non-competitive allosteric antagonism enables AZ3451 to inhibit PAR2 signaling regardless of agonist concentration or identity, providing insurmountable blockade—a feature competitive orthosteric antagonists like AZ8838 cannot achieve.

• [1] Kennedy AJ, Sundström L, Geschwindner S, et al. Protease-activated receptor-2 ligands reveal orthosteric and allosteric mechanisms of receptor inhibition. Commun Biol. 2020;3:782. doi:10.1038/s42003-020-01504-0 View Source

Binding Affinity Superiority: AZ3451 Kd of 13.5 nM vs AZ8838 Kd of 344 nM

In the same 2017 Nature study that characterized both compounds, AZ3451 demonstrated substantially tighter binding to PAR2 than AZ8838. AZ3451 bound PAR2 with a Kd of 13.5 ± 3 nM measured by Biacore surface plasmon resonance (SPR), whereas AZ8838 exhibited a Kd of 344 ± 74 nM measured by radioligand saturation binding with [³H]-AZ8838 [1][2]. This represents an approximately 25-fold affinity advantage for AZ3451. The binding site for AZ3451 was mapped by X-ray crystallography to a remote allosteric pocket outside the helical bundle, distinct from the occluded near-extracellular pocket occupied by AZ8838 [1].

Binding Affinity (Kd)

Head-to-head

AZ3451: 13.5 ± 3 nM vs AZ8838: 344 ± 74 nM (~25-fold difference)

Reported target occupancy context

Biacore SPR (AZ3451); Radioligand binding (AZ8838)

Evidence Dimension	PAR2 equilibrium binding affinity (Kd)
Target Compound Data	AZ3451: Kd = 13.5 ± 3 nM (Biacore SPR, n = 3)
Comparator Or Baseline	AZ8838: Kd = 344 ± 74 nM (radioligand saturation binding with [³H]-AZ8838)
Quantified Difference	~25-fold higher affinity for AZ3451 (13.5 nM vs 344 nM)
Conditions	PAR2-StaR purified protein; Biacore SPR for AZ3451; radioligand saturation binding for AZ8838; both datasets from Cheng et al. Nature 2017 (PMID: 28445455)

Why This Matters

The 25-fold higher binding affinity means that at equivalent concentrations, AZ3451 achieves substantially greater target occupancy, enabling complete receptor blockade at lower compound concentrations and reducing the risk of off-target effects from high-concentration compound exposure.

• [1] Cheng RKY, Fiez-Vandal C, Schlenker O, et al. Structural insight into allosteric modulation of protease-activated receptor 2. Nature. 2017;545(7652):112-115. doi:10.1038/nature22309 View Source
• [2] GtoPdb/IUPHAR-BPS Guide to MALARIA PHARMACOLOGY. AZ3451 Ligand Activity Charts. Accessed 2025. View Source

Functional Potency Gap: AZ3451 Is >400-Fold More Potent Than AZ8838 in FLIPR Calcium Assays

In cellular functional assays measuring PAR2-mediated calcium mobilization (FLIPR) in 1321N1 cells overexpressing wild-type human PAR2, AZ3451 inhibited SLIGRL-NH₂-induced calcium release with an IC50 of 5.4 nM (pIC50 = 8.27) [1]. By comparison, AZ8838 inhibited the same response with an IC50 of 2,300 nM (pIC50 = 5.64) [2], representing approximately a 425-fold potency difference. In the IP-One accumulation assay, AZ3451 showed an IC50 of 23 nM (pIC50 = 7.64), while AZ8838 showed an IC50 of 1,500 nM (pIC50 = 5.82), a 65-fold difference [1][2]. These data demonstrate that AZ3451 is consistently 1.8–2.6 log units more potent than AZ8838 across multiple functional readouts.

Functional Potency

Cross-study

AZ3451 IC50 = 5.4 nM vs AZ8838 IC50 = 2,300 nM (425-fold difference in FLIPR)

Assay potency context review

1321N1-hPAR2 cells; SLIGRL-NH₂ agonist

Evidence Dimension	Functional antagonism potency in cellular assays
Target Compound Data	AZ3451: FLIPR IC50 = 5.4 ± 1.1 nM, pIC50 8.27 (n = 11); IP-One IC50 = 23 ± 1 nM, pIC50 7.64 (n = 72)
Comparator Or Baseline	AZ8838: FLIPR IC50 = 2,300 nM, pIC50 5.64; IP-One IC50 = 1,500 nM, pIC50 5.82
Quantified Difference	425-fold greater potency in FLIPR calcium assay (5.4 nM vs 2,300 nM); 65-fold greater potency in IP-One assay (23 nM vs 1,500 nM)
Conditions	1321N1 cells overexpressing wild-type human PAR2; FLIPR calcium accumulation assay vs EC80 of SLIGRL-NH₂ (8 µM); IP-One accumulation assay vs SLIGRL-NH₂; data from Cheng et al. Nature 2017

Why This Matters

A 425-fold potency gap means that 1 mg of AZ3451 provides equivalent functional PAR2 blockade to approximately 425 mg of AZ8838 in this cellular system—a critical consideration for experimental design, compound procurement quantity, and in vivo dosing feasibility.

• [1] GtoPdb/IUPHAR-BPS Guide to MALARIA PHARMACOLOGY. AZ3451 Ligand Activity Charts. Accessed 2025. View Source
• [2] GtoPdb/IUPHAR-BPS Guide to IMMUNOPHARMACOLOGY. AZ8838 Ligand Activity Charts. Accessed 2025. View Source

PAR2 Selectivity: >20,000-Fold Window Over PAR1 in β-Arrestin-2 Recruitment

AZ3451 exhibits a well-characterized selectivity profile across the PAR family. In β-arrestin-2 recruitment assays, AZ3451 inhibited PAR2 with an IC50 of <2.5 nM, showed weak activity at PAR4 (IC50 = 380 nM), and was essentially inactive at PAR1 (IC50 > 50,000 nM) [1]. This corresponds to a >20,000-fold selectivity window for PAR2 over PAR1 and a 150-fold window over PAR4. This selectivity is structurally rationalized by the unique remote allosteric binding site of AZ3451: key residue differences between PAR2 and PAR1 (e.g., His135²·⁶⁴ in PAR2 vs Tyr162²·⁶⁴ in PAR1) prevent AZ3451 from binding to the PAR1 allosteric pocket [2]. Earlier PAR2 antagonists such as GB88 also show PAR2 selectivity but with substantially lower potency (IC50 ~2 µM), and their comprehensive PAR1/PAR4 selectivity data are less complete [3].

PAR2 Selectivity

Class-level

>20,000-fold over PAR1; 150-fold over PAR4 in β-arrestin-2 assay

Supports isoform-selectivity review

β-arrestin-2 recruitment; PAR1 co-expression tissues

Evidence Dimension	PAR isoform selectivity measured by β-arrestin-2 recruitment assay
Target Compound Data	AZ3451: PAR2 IC50 < 2.5 nM; PAR4 IC50 = 380 nM; PAR1 IC50 > 50,000 nM
Comparator Or Baseline	Class-level baseline: most earlier PAR2 antagonists (GB88 IC50 ~2 µM, ENMD-1068 IC50 ~1.6 µM–5 mM) lack comprehensive quantitative PAR1/PAR4 selectivity profiling
Quantified Difference	>20,000-fold selectivity for PAR2 vs PAR1; 150-fold selectivity for PAR2 vs PAR4
Conditions	β-arrestin-2 recruitment assay in engineered cell lines; data from Cheng et al. Nature 2017 (PMID: 28445455)

Why This Matters

The >20,000-fold PAR2/PAR1 selectivity window ensures that AZ3451 can be used to interrogate PAR2-specific biology without confounding PAR1 engagement—critical because PAR1 is the most widely expressed PAR family member, and its inadvertent antagonism would confound phenotypic interpretation in tissues co-expressing both receptors.

• [1] Bertin Bioreagent / Cayman Chemical. AZ3451 product overview, CAT N° 29671. Accessed 2025. View Source
• [2] Cheng RKY, Fiez-Vandal C, Schlenker O, et al. Nature. 2017;545:112-115. Extended Data Figure 1: Comparison between PAR2 and PAR1 allosteric pockets. doi:10.1038/nature22309 View Source
• [3] Suen JY, Cotterell A, Lohman RJ, et al. Modulating human proteinase activated receptor 2 with a novel antagonist (GB88) and agonist (GB110). Br J Pharmacol. 2012;165(5):1413-1423. doi:10.1111/j.1476-5381.2011.01610.x View Source

Multi-Pathway Suppression in Chondrocytes: Simultaneous Inhibition of P38/MAPK, NF-κB, and PI3K/AKT/mTOR

In primary rat chondrocytes stimulated with IL-1β (10 ng/mL), AZ3451 (10 µM) suppressed phosphorylation of P38 (MAPK pathway), IKKα/β, IκBα, and p65 (NF-κB pathway), and PI3K, AKT, and mTOR (PI3K/AKT/mTOR pathway) [1]. The study provided mechanistic benchmarking through combination experiments: when AZ3451 was co-administered with the P38 inhibitor SB203580, the PI3K/AKT inhibitor LY294002, or the NF-κB inhibitor JSH-23, each combination exerted better protection against IL-1β-induced collagen II and aggrecan degradation than AZ3451 alone, demonstrating additive efficacy [1]. Downstream, AZ3451 reduced expression of cartilage-degrading enzymes (MMP1, MMP13, ADAMTS5) and inflammatory mediators (iNOS, COX2), while restoring matrix proteins (Aggrecan, Collagen II) and the chondrogenic transcription factor Sox9 [1].

Chondrocyte Pathways

Cross-study

Suppressed P38/MAPK, NF-κB, PI3K/AKT/mTOR; Additive protection with single-node inhibitors

Reported pathway-response context

IL-1β rat chondrocytes; 10 µM AZ3451

Evidence Dimension	Multi-pathway signaling suppression and cartilage matrix protein protection in IL-1β-stimulated chondrocytes
Target Compound Data	AZ3451 (10 µM): suppressed P38, IKKα/β, IκBα, p65, PI3K, AKT, and mTOR phosphorylation; reduced MMP1, MMP13, ADAMTS5, iNOS, COX2 expression; increased Aggrecan, Collagen II, Sox9 (western blot quantification in source Figures 2, 5, 6)
Comparator Or Baseline	Single-pathway inhibitors: P38 inhibitor SB203580, NF-κB inhibitor JSH-23, PI3K/AKT inhibitor LY294002—each partially protected matrix; AZ3451 + each inhibitor combination provided additive protection over AZ3451 alone
Quantified Difference	AZ3451 engages three signaling pathways simultaneously; combination with single-pathway inhibitors yields additive protection over AZ3451 monotherapy (statistically significant difference, p < 0.05)
Conditions	Primary rat chondrocytes (passages 1–3); IL-1β 10 ng/mL stimulation; AZ3451 at 10 µM; western blot quantification; n ≥ 3 independent experiments

Why This Matters

Multi-pathway engagement distinguishes AZ3451 from single-node kinase inhibitors and supports its utility as a chemical probe for osteoarthritis research where P38, NF-κB, and PI3K/AKT pathways converge on cartilage degradation, enabling simultaneous interrogation of all three signaling nodes.

• [1] He L, Pan Y, Yu J, et al. Protease-activated receptor 2 (PAR-2) antagonist AZ3451 as a novel therapeutic agent for osteoarthritis. Aging. 2019;11(24):12532-12545. doi:10.18632/aging.102586 View Source

In Vivo Cartilage Protection: AZ3451 Reduces OARSI Scores in Rat Post-Traumatic Osteoarthritis Model

In a rat post-traumatic osteoarthritis (PTOA) model induced by surgical destabilization of the anterior cruciate ligament (ACLT), intra-articular injection of AZ3451 significantly ameliorated cartilage degradation compared to untreated PTOA controls [1]. Histological analysis using H&E, Toluidine blue, and Safranin O/Fast Green staining after 8 weeks revealed that AZ3451-treated animals retained richer proteoglycan content and exhibited less superficial articular cartilage erosion [1]. The OARSI histopathology score was significantly reduced in the AZ3451 treatment group compared to the PTOA group (p < 0.05; Figure 7B) [1]. Immunohistochemistry confirmed that AZ3451 decreased MMP13 and Cleaved-Caspase 3 expression while increasing Beclin1 (autophagy marker) in cartilage tissue in vivo, consistent with the in vitro findings [1]. While head-to-head in vivo comparisons with other PAR2 antagonists are not yet published, these data are directionally consistent with genetic evidence that PAR2-deficient mice show reduced cartilage degradation in experimental OA models [1].

Cartilage Protection

Supporting evidence

Reduced OARSI scores vs PTOA control (p < 0.05); Decreased MMP13 & Cleaved-Caspase 3

Model-response endpoint context

Rat ACLT model; Intra-articular injection; 8-week endpoint

Evidence Dimension	In vivo cartilage protection assessed by OARSI histopathology scoring
Target Compound Data	AZ3451 intra-articular injection significantly reduced OARSI scores vs untreated PTOA group (p < 0.05); decreased MMP13 and Cleaved-Caspase 3; increased Beclin1 expression in cartilage tissue (Figure 7A–F)
Comparator Or Baseline	Untreated PTOA (ACLT surgery) group showing significant superficial cartilage erosion and proteoglycan loss; PAR2 knockout mouse literature showing reduced cartilage degradation in experimental OA
Quantified Difference	Statistically significant reduction in OARSI scores (p < 0.05; exact scores reported in Figure 7B of source publication); qualitative consistency with PAR2 genetic ablation phenotype
Conditions	Male 8-week-old Sprague-Dawley rats; ACLT surgery-induced PTOA model; 8-week post-surgery endpoint; intra-articular AZ3451 injection; histological scoring using OARSI grading system

Why This Matters

In vivo cartilage protection data provide translational relevance for AZ3451 procurement by osteoarthritis research programs and validate that the multi-pathway modulation observed in vitro translates to tissue-level efficacy in a clinically relevant surgical OA model.

• [1] He L, Pan Y, Yu J, et al. Protease-activated receptor 2 (PAR-2) antagonist AZ3451 as a novel therapeutic agent for osteoarthritis. Aging. 2019;11(24):12532-12545. doi:10.18632/aging.102586 View Source

AZ3451: Recommended Research Application Scenarios Based on Differentiated Evidence

Osteoarthritis and Cartilage Degeneration: From In Vitro Pathway Analysis to In Vivo Preclinical Models

AZ3451 is the best-characterized PAR2 antagonist for osteoarthritis research, with published evidence spanning IL-1β-stimulated primary chondrocyte signaling through to in vivo ACLT rat models. The compound's ability to simultaneously suppress P38/MAPK, NF-κB, and PI3K/AKT/mTOR pathways, reduce cartilage-degrading enzymes (MMP13, ADAMTS5), and restore matrix proteins (Collagen II, Aggrecan) makes it the preferred chemical probe for studies investigating PAR2-dependent cartilage catabolism [1]. Its validated in vivo efficacy via intra-articular injection at the dose used in He et al. (2019) provides a benchmark for preclinical OA pharmacology studies seeking to translate in vitro pathway modulation to tissue-level outcomes.

GPCR Structural Pharmacology: Differentiating Allosteric from Orthosteric PAR2 Modulation

AZ3451 is distinguished by its structurally validated binding to a remote allosteric site outside the PAR2 helical bundle—a binding mode confirmed by X-ray crystallography and orthogonal binding assays (Biacore SPR, Kd = 13.5 nM) [1]. This makes AZ3451 uniquely suited for studies investigating allosteric modulation mechanisms at PAR2 and for direct comparison with orthosteric/competitive antagonists such as AZ8838. The Schild analysis data confirming non-competitive antagonism (p < 0.001 vs hypothetical slope of 1) provide a robust framework for differentiating allosteric from competitive pharmacology in functional assays [2]. Researchers designing experiments to distinguish orthosteric from allosteric PAR2 antagonism can use AZ3451 and AZ8838 as mechanistically defined comparator pair.

Oral Cancer Pain Research: Nanoparticle-Encapsulated AZ3451 for Endosomal PAR2 Targeting

AZ3451, when encapsulated in PAMAM-Chol nanoparticles (PAMAM-Chol-AZ NPs), demonstrated significantly enhanced efficacy over free AZ3451 in reversing PAR2 activation at both the plasma membrane and early endosomes, and provided greater and longer-lasting antinociception in three preclinical mouse models of oral cancer pain [1]. This application scenario is specifically supported by evidence that free AZ3451 alone has limited ability to access PAR2 in endosomal compartments, whereas nanoparticle encapsulation enables endosomal targeting—a critical advantage for studies of sustained GPCR signaling from intracellular compartments. Procurement of AZ3451 for nanoparticle formulation studies is recommended for research groups investigating the role of endosomal PAR2 signaling in cancer pain.

PAR2-Specific Pathway Dissection Requiring High Selectivity Over PAR1

For experiments requiring unambiguous attribution of biological effects to PAR2 rather than PAR1—particularly in tissues where both receptors are co-expressed (e.g., vasculature, fibroblasts, sensory neurons, and certain cancer microenvironments)—AZ3451 provides a >20,000-fold selectivity window over PAR1 in β-arrestin-2 recruitment assays [1]. This selectivity is structurally rationalized by the distinct remote allosteric binding pocket that is incompatible with key PAR1 residues (His135²·⁶⁴ in PAR2 replaced by Tyr162²·⁶⁴ in PAR1) [2]. While selectivity should be verified in each investigator's own assay system, AZ3451 currently represents the most comprehensively characterized PAR2/PAR1 selectivity profile among commercially available small-molecule PAR2 antagonists.

Application

Selection Property

Validation Focus

Osteoarthritis and Cartilage Degeneration Studies

Multi-pathway modulation profile

P38/MAPK, NF-κB, and PI3K/AKT/mTOR pathway-response endpoints

GPCR Structural Pharmacology Research

Allosteric binding mechanism

Orthosteric vs. allosteric antagonism differentiation assays

Oral Cancer Pain Model Research

Nanoparticle formulation context

Endosomal PAR2 targeting and antinociception endpoint models

PAR2-Specific Pathway Dissection

High selectivity window over PAR1

PAR1/PAR2 co-expression model selectivity review

• [1] He L, Pan Y, Yu J, et al. Protease-activated receptor 2 (PAR-2) antagonist AZ3451 as a novel therapeutic agent for osteoarthritis. Aging. 2019;11(24):12532-12545. doi:10.18632/aging.102586 View Source
• [2] Cheng RKY, Fiez-Vandal C, Schlenker O, et al. Structural insight into allosteric modulation of protease-activated receptor 2. Nature. 2017;545(7652):112-115. doi:10.1038/nature22309 View Source
• [3] Kennedy AJ, Sundström L, Geschwindner S, et al. Protease-activated receptor-2 ligands reveal orthosteric and allosteric mechanisms of receptor inhibition. Commun Biol. 2020;3:782. doi:10.1038/s42003-020-01504-0 View Source
• [4] Thai NQ, Lam DK, Thai J, et al. PAR2 on oral cancer cells and nociceptors contributes to oral cancer pain that can be relieved by nanoparticle-encapsulated AZ3451. Biomaterials. 2024;314:122874. doi:10.1016/j.biomaterials.2024.122874 View Source
• [5] Cheng RKY, Fiez-Vandal C, Schlenker O, et al. Nature. 2017;545:112-115. Extended Data Figure 1: Comparison between PAR2 and PAR1 allosteric pockets. doi:10.1038/nature22309 View Source