Anzhi Wei

Structures of adnectin/protein complexes reveal an expanded binding footprint.

Adnectins are targeted biologics derived from the tenth type III domain of human fibronectin (¹⁰Fn3), a member of the immunoglobulin superfamily. Target-specific binders are selected from libraries generated by diversifying the three ¹⁰Fn3 loops that are analogous to the complementarity determining regions of antibodies. The crystal structures of two Adnectins were determined, each in complex with its therapeutic target, EGFR or IL-23. Both Adnectins bind different epitopes than those bound by known monoclonal antibodies. Molecular modeling suggests that some of these epitopes might not be accessible to antibodies because of the size and concave shape of the antibody combining site. In addition to interactions from the Adnectin diversified loops, residues from the N terminus and/or the β strands interact with the target proteins in both complexes. Alanine-scanning mutagenesis confirmed the calculated binding energies of these β strand interactions, indicating that these nonloop residues can expand the available binding footprint.

Tetrahydroquinoline derivatives as potent and selective factor XIa inhibitors.

Antithrombotic agents that are inhibitors of factor XIa (FXIa) have the potential to demonstrate robust efficacy with a low bleeding risk profile. Herein, we describe a series of tetrahydroquinoline (THQ) derivatives as FXIa inhibitors. Compound 1 was identified as a potent and selective tool compound for proof of concept studies. It exhibited excellent antithrombotic efficacy in rabbit thrombosis models and did not prolong bleeding times. This demonstrates proof of concept for the FXIa mechanism in animal models with a reversible, small molecule inhibitor.

Phenylimidazoles as Potent and Selective Inhibitors of Coagulation Factor XIa with in Vivo Antithrombotic Activity

Novel inhibitors of FXIa containing an (S)-2-phenyl-1-(4-phenyl-1H-imidazol-2-yl)ethanamine core have been optimized to provide compound 16b, a potent, reversible inhibitor of FXIa (Ki = 0.3 nM) having in vivo antithrombotic efficacy in the rabbit AV-shunt thrombosis model (ID50 = 0.6 mg/kg + 1 mg kg(-1) h(-1)). Initial analog selection was informed by molecular modeling using compounds 11a and 11h overlaid onto the X-ray crystal structure of tetrahydroquinoline 3 complexed to FXIa. Further optimization was achieved by specific modifications derived from careful analysis of the X-ray crystal structure of the FXIa/11h complex. Compound 16b was well tolerated and enabled extensive pharmacologic evaluation of the FXIa mechanism up to the ID90 for thrombus inhibition.

Discovery of Novel P1 Groups for Coagulation Factor VIIa Inhibition Using Fragment-Based Screening

A multidisciplinary, fragment-based screening approach involving protein ensemble docking and biochemical and NMR assays is described. This approach led to the discovery of several structurally diverse, neutral surrogates for cationic factor VIIa P1 groups, which are generally associated with poor pharmacokinetic (PK) properties. Among the novel factor VIIa inhibitory fragments identified were aryl halides, lactams, and heterocycles. Crystallographic structures for several bound fragments were obtained, leading to the successful design of a potent factor VIIa inhibitor with a neutral lactam P1 and improved permeability.

Design and Synthesis of Phenylpyrrolidine Phenylglycinamides As Highly Potent and Selective TF-FVIIa Inhibitors

Inhibitors of the Tissue Factor/Factor VIIa (TF-FVIIa) complex are promising novel anticoagulants that show excellent efficacy and minimal bleeding in preclinical models. On the basis of a zwitterionic phenylglycine acylsulfonamide 1, a phenylglycine benzylamide 2 was shown to possess improved permeability and oral bioavailability. Optimization of the benzylamide, guided by X-ray crystallography, led to a potent TF-FVIIa inhibitor 18i with promising oral bioavailability, but promiscuous activity in an in vitro safety panel of receptors and enzymes. Introducing an acid on the pyrrolidine ring, guided by molecular modeling, resulted in highly potent, selective, and efficacious TF-FVIIa inhibitors with clean in vitro safety profile. The pyrrolidine acid 20 showed a moderate clearance, low volume of distribution, and a short t 1/2 in dog PK studies.

Discovery of nonbenzamidine factor VIIa inhibitors using a biaryl acid scaffold

In this Letter, we describe the synthesis of several nonamidine analogs of biaryl acid factor VIIa inhibitor 1 containing weakly basic or nonbasic P1 groups. 2-Aminoisoquinoline was found to be an excellent surrogate for the benzamidine group (compound 2) wherein potent inhibition of factor VIIa is maintained relative to most other related serine proteases. In an unanticipated result, the m-benzamide P1 (compounds 21a and 21b) proved to be a viable benzamidine replacement, albeit with a 20-40 fold loss in potency against factor VIIa.

Atropisomer Control in Macrocyclic Factor VIIa Inhibitors

Incorporation of a methyl group onto a macrocyclic FVIIa inhibitor improves potency 10-fold but is accompanied by atropisomerism due to restricted bond rotation in the macrocyclic structure, as demonstrated by NMR studies. We designed a conformational constraint favoring the desired atropisomer in which this methyl group interacts with the S2 pocket of FVIIa. A macrocyclic inhibitor incorporating this constraint was prepared and demonstrated by NMR to reside predominantly in the desired conformation. This modification improved potency 180-fold relative to the unsubstituted, racemic macrocycle and improved selectivity. An X-ray crystal structure of a closely related analogue in the FVIIa active site was obtained and matches the NMR and modeled conformations, confirming that this conformational constraint does indeed direct the methyl group into the S2 pocket as designed. The resulting rationally designed, conformationally stable template enables further optimization of these macrocyclic inhibitors.

Pyridine and pyridinone-based factor XIa inhibitors.

The structure-activity relationships (SAR) of six-membered ring replacements for the imidazole ring scaffold is described. This work led to the discovery of the potent and selective pyridine (S)-23 and pyridinone (±)-24 factor XIa inhibitors. SAR and X-ray crystal structure data highlight the key differences between imidazole and six-membered ring analogs.

Novel phenylalanine derived diamides as Factor XIa inhibitors.

The synthesis, structural activity relationships (SAR), and selectivity profile of a potent series of phenylalanine diamide FXIa inhibitors will be discussed. Exploration of P1 prime and P2 prime groups led to the discovery of compounds with high FXIa affinity, good potency in our clotting assay (aPPT), and high selectivity against a panel of relevant serine proteases as exemplified by compound 21. Compound 21 demonstrated good in vivo efficacy (EC50=2.8μM) in the rabbit electrically induced carotid arterial thrombosis model (ECAT).

Orally bioavailable pyridine and pyrimidine-based Factor XIa inhibitors: Discovery of the methyl N-phenyl carbamate P2 prime group

Pyridine-based Factor XIa (FXIa) inhibitor (S)-2 was optimized by modifying the P2 prime, P1, and scaffold regions. This work resulted in the discovery of the methyl N-phenyl carbamate P2 prime group which maintained FXIa activity, reduced the number of H-bond donors, and improved the physicochemical properties compared to the amino indazole P2 prime moiety. Compound (S)-17 was identified as a potent and selective FXIa inhibitor that was orally bioavailable. Replacement of the basic cyclohexyl methyl amine P1 in (S)-17 with the neutral p-chlorophenyltetrazole P1 resulted in the discovery of (S)-24 which showed a significant improvement in oral bioavailability compared to the previously reported imidazole (S)-23. Additional improvements in FXIa binding affinity, while maintaining oral bioavailability, was achieved by replacing the pyridine scaffold with either a regioisomeric pyridine or pyrimidine ring system.