Discovery of the Potent and Selective Inhaled Janus Kinase 1 Inhibitor AZD4604 and Its Preclinical Characterization.

JAK-STAT cytokines are critical in regulating immunity. Persistent activation of JAK-STAT signaling pathways by cytokines drives chronic inflammatory diseases such as asthma. Herein, we report on the discovery of a highly JAK1-selective, ATP-competitive series of inhibitors having a 1000-fold selectivity over other JAK family members and the approach used to identify compounds suitable for inhaled administration. Ultimately, compound 16 was selected as the clinical candidate, and upon dry powder inhalation, we could demonstrate a high local concentration in the lung as well as low plasma concentrations, suggesting no systemic JAK1 target engagement. Compound 16 has progressed into clinical trials. Using 16, we found JAK1 inhibition to be more efficacious than JAK3 inhibition in IL-4-driven Th2 asthma.

Characterization of Selective and Potent JAK1 Inhibitors Intended for the Inhaled Treatment of Asthma.

Purpose: Janus kinase 1 (JAK1) is implicated in multiple inflammatory pathways that are critical for the pathogenesis of asthma, including the interleukin (IL)-4, IL-5, IL-13, and thymic stromal lymphopoietin cytokine signaling pathways, which have previously been targeted to treat allergic asthma. Here, we describe the development of AZD0449 and AZD4604, two novel and highly selective JAK1 inhibitors with promising properties for inhalation. Methods: The effects of AZD0449 and AZD4604 in JAK1 signaling pathways were assessed by measuring phosphorylation of signal transducer and activator of transcription (STAT) proteins and chemokine release using immunoassays of whole blood from healthy human volunteers and rats. Pharmacokinetic studies performed on rats evaluated AZD0449 at a lung deposited dose of 52 µg/kg and AZD4604 at 30 µg/kg. The efficacy of AZD0449 and AZD4604 was assessed by evaluating lung inflammation (cell count and cytokine levels) and the late asthmatic response (average enhanced pause [Penh]). Results: Both compounds inhibited JAK1-dependent cytokine signaling pathways in a dose-dependent manner in human and rat leukocytes. After intratracheal administration in rats, both compounds exhibited low systemic exposures and medium-to-long terminal lung half-lives (AZD0449, 34 hours; AZD4604, 5 hours). Both compounds inhibited STAT3 and STAT5 phosphorylation in lung tissue from ovalbumin (OVA)-challenged rats. AZD0449 and AZD4604 also inhibited eosinophilia in the lung and reduced the late asthmatic response, measured as Penh in the OVA rat model. Conclusion: AZD0449 and AZD4604 show potential as inhibitors of signaling pathways involved in asthmatic immune responses, with target engagement demonstrated locally in the lung. These findings support the clinical development of AZD0449 and AZD4604 for the treatment of patients with asthma.

The Vital Role of Proteomics in Characterizing Novel Protein Degraders.

Mass spectrometry-based proteomics profiling is a discovery tool that enables researchers to understand the mechanisms of action of drug candidates. When applied to proteolysis targeting chimeras (PROTACs) such approaches provide unbiased perspectives of the binding, degradation selectivity, and mechanism related to efficacy and safety. Specifically, global profiling experiments can identify direct degradation events and assess downstream pathway modulation that may result from degradation or off-target inhibition. Targeted proteomics approaches can be used to quantify the levels of relevant E3 ligases and the protein of interest in cell lines and tissues of interest, which can inform the line of sight and provide insights on possible safety liabilities early in the project. Furthermore, proteomics approaches can be applied to understand protein turnover and resynthesis rates and inform on target tractability, as well as pharmacokinetics/pharmacodynamics understanding. In this perspective, we survey the literature around the impact of mass spectrometry-based proteomics in the development of PROTACs and present our envisioned proteomics cascade for supporting targeted protein degradation projects.

Regenerable Biosensors for Small-Molecule Kinetic Characterization Using SPR.

A key activity in small-molecule drug discovery is the characterization of compound-target interactions. Surface plasmon resonance (SPR) is a flexible technique for this purpose, with a wide affinity range (micromoles to picomoles), low protein requirements, and the ability to characterize the kinetics of compound binding. However, a key requirement of SPR is the immobilization of the target protein to the surface of the sensor chip. The most commonly used immobilization techniques (covalent immobilization, streptavidin-biotin) are irreversible in nature, which can afford excellent baseline stability but impose limitations throughput for slowly dissociating compounds or unstable targets. Reversible immobilization (e.g., His-tag-Ni-NTA) is possible but typically precludes accurate quantification of slow dissociation kinetics due to baseline drift.Here we present our investigation of three immobilization strategies (dual-His-tagged target protein, His-tagged streptavidin, and switchavidin) that combine the robustness of irreversible immobilization with the flexibility of reversible immobilization. Each has its own advantages and limitations, and while a universal immobilization procedure remains to be found, these strategies add to the immobilization toolbox that enables previously out-of-scope applications. Such applications are highlighted in two examples that greatly increased throughput for the kinetic characterization of potent kinase inhibitors and kinetic profiling of covalent inhibitors.

Proteolysis-targeting chimeras in drug development: A safety perspective.

Proteolysis-targeting chimeras are a new drug modality that exploits the endogenous ubiquitin proteasome system to degrade a protein of interest for therapeutic benefit. As the first-generation of proteolysis-targeting chimeras have now entered clinical trials for oncology indications, it is timely to consider the theoretical safety risks inherent with this modality which include off-target degradation, intracellular accumulation of natural substrates for the E3 ligases used in the ubiquitin proteasome system, proteasome saturation by ubiquitinated proteins, and liabilities associated with the "hook effect" of proteolysis-targeting chimeras This review describes in vitro and non-clinical in vivo data that provide mechanistic insight of these safety risks and approaches being used to mitigate these risks in the next generation of proteolysis-targeting chimera molecules to extend therapeutic applications beyond life-threatening diseases.

Macrocyclic Modalities Combining Peptide Epitopes and Natural Product Fragments.

"Hot loop" protein segments have variable structure and conformation and contribute crucially to protein-protein interactions. We describe a new hot loop mimicking modality, termed PepNats, in which natural product (NP)-inspired structures are incorporated as conformation-determining and -restricting structural elements into macrocyclic hot loop-derived peptides. Macrocyclic PepNats representing hot loops of inducible nitric oxide synthase (iNOS) and human agouti-related protein (AGRP) were synthesized on solid support employing macrocyclization by imine formation and subsequent stereoselective 1,3-dipolar cycloaddition as key steps. PepNats derived from the iNOS DINNN hot loop and the AGRP RFF hot spot sequence yielded novel and potent ligands of the SPRY domain-containing SOCS box protein 2 (SPSB2) that binds to iNOS, and selective ligands for AGRP-binding melanocortin (MC) receptors. NP-inspired fragment absolute configuration determines the conformation of the peptide part responsible for binding. These results demonstrate that combination of NP-inspired scaffolds with peptidic epitopes enables identification of novel hot loop mimics with conformationally constrained and biologically relevant structure.

Cryo-EM structure of the activated RET signaling complex reveals the importance of its cysteine-rich domain.

Signaling through the receptor tyrosine kinase RET is essential during normal development. Both gain- and loss-of-function mutations are involved in a variety of diseases, yet the molecular details of receptor activation have remained elusive. We have reconstituted the complete extracellular region of the RET signaling complex together with Neurturin (NRTN) and GFRα2 and determined its structure at 5.7-Å resolution by cryo-EM. The proteins form an assembly through RET-GFRα2 and RET-NRTN interfaces. Two key interaction points required for RET extracellular domain binding were observed: (i) the calcium-binding site in RET that contacts GFRα2 domain 3 and (ii) the RET cysteine-rich domain interaction with NRTN. The structure highlights the importance of the RET cysteine-rich domain and allows proposition of a model to explain how complex formation leads to RET receptor dimerization and its activation. This provides a framework for targeting RET activity and for further exploration of mechanisms underlying neurological diseases.

A class of highly selective inhibitors bind to an active state of PI3Kγ.

We have discovered a class of PI3Kγ inhibitors exhibiting over 1,000-fold selectivity over PI3Kα and PI3Kß. On the basis of X-ray crystallography, hydrogen-deuterium exchange-mass spectrometry and surface plasmon resonance experiments we propose that the cyclopropylethyl moiety displaces the DFG motif of the enzyme away from the adenosine tri-phosphate binding site, inducing a large conformational change in both the kinase- and helical domains of PI3Kγ. Site directed mutagenesis explained how the conformational changes occur. Our results suggest that these cyclopropylethyl substituted compounds selectively inhibit the active state of PI3Kγ, which is unique to these compounds and to the PI3Kγ isoform, explaining their excellent potency and unmatched isoform selectivity that were confirmed in cellular systems. This is the first example of a Class I PI3K inhibitor achieving its selectivity by affecting the DFG motif in a manner that bears similarity to DFG in/out for type II protein kinase inhibitors.

A Comparative Study of Fluorescence Assays in Screening for BRD4.

Fluorescence assay technologies are commonly used in high-throughput screening because of their sensitivity and ease of use. Different technologies have their characteristics and the rationale for choosing one over the other can differ between projects because of factors such as availability of reagents, assay performance, and cost. Another important factor to consider is the assay susceptibility to artifacts, which is almost as important as the ability of the assay to pick up active compounds. Spending time and money on false positives or missing the opportunity to build chemistry around false negatives is something that every drug project tries to avoid. We used a BET family Bromodomain, BRD4(1), to explore the outcome of a screening campaign using three fluorescent assay technologies as primary assays. A diverse 7,038 compound set was screened in fluorescence lifetime, fluorescence polarization, and homogeneous time-resolved fluorescence to look at primary hit rates, compound overlap, and hit confirmation rates. The results show a difference between the fluorescence assay technologies with three separate hit lists and some overlap. The confirmed hits from each assay were further evaluated for translation into cells (NanoBRET™). Most of the actives confirmed in cells originated from compounds that overlapped between the assays. In addition, a well-annotated set of compounds with undesirable mechanism of inhibition was screened against BRD4(1) to compare the ability to discriminate true hits from artifact compounds. The results indicate a difference between the assays in their ability to generate false positives and negatives.

Inhibition of low-density lipoprotein receptor degradation with a cyclic peptide that disrupts the homodimerization of IDOL E3 ubiquitin ligase.

Cellular uptake of circulating cholesterol occurs via the low density lipoprotein receptor (LDLR). The E3 ubiquitin ligase IDOL is a mediator of LDLR degradation, with IDOL homodimerization thought to be required for its activity. To probe the possibility of modulating LDLR levels with an inhibitor of IDOL homodimerization, we screened a SICLOPPS library of 3.2 million cyclic peptides for compounds that disrupt this protein-protein interaction. We identified cyclo-CFFLYT as the lead inhibitor, and improved its activity through the incorporation of non-natural amino acids. The activity of the optimized cyclic peptide was assessed in hepatic cells, with a dose-dependent increase in LDLR levels observed in the presence of our IDOL homodimerization inhibitor.

Structure and biophysical characterization of the human full-length neurturin-GFRa2 complex: A role for heparan sulfate in signaling.

Neurturin (NRTN) provides trophic support to neurons and is considered a therapeutic agent for neurodegenerative diseases, such as Parkinson's disease. It binds to its co-receptor GFRa2, and the resulting NRTN-GFRa2 complex activates the transmembrane receptors rearranged during transfection (RET) or the neural cell adhesion molecule (NCAM). We report the crystal structure of NRTN, alone and in complex with GFRa2. This is the first crystal structure of a GFRa with all three domains and shows that domain 1 does not interact directly with NRTN, but it may support an interaction with RET and/or NCAM, via a highly conserved surface. In addition, biophysical results show that the relative concentration of GFRa2 on cell surfaces can affect the functional affinity of NRTN through avidity effects. We have identified a heparan sulfate-binding site on NRTN and a putative binding site in GFRa2, suggesting that heparan sulfate has a role in the assembly of the signaling complex. We further show that mutant NRTN with reduced affinity for heparan sulfate may provide a route forward for delivery of NRTN with increased exposure in preclinical in vivo models and ultimately to Parkinson's patients.

Unified Software Solution for Efficient SPR Data Analysis in Drug Research.

Surface plasmon resonance (SPR) is a powerful method for obtaining detailed molecular interaction parameters. Modern instrumentation with its increased throughput has enabled routine screening by SPR in hit-to-lead and lead optimization programs, and SPR has become a mainstream drug discovery technology. However, the processing and reporting of SPR data in drug discovery are typically performed manually, which is both time-consuming and tedious. Here, we present the workflow concept, design and experiences with a software module relying on a single, browser-based software platform for the processing, analysis, and reporting of SPR data. The efficiency of this concept lies in the immediate availability of end results: data are processed and analyzed upon loading the raw data file, allowing the user to immediately quality control the results. Once completed, the user can automatically report those results to data repositories for corporate access and quickly generate printed reports or documents. The software module has resulted in a very efficient and effective workflow through saved time and improved quality control. We discuss these benefits and show how this process defines a new benchmark in the drug discovery industry for the handling, interpretation, visualization, and sharing of SPR data.

Discovery of Second Generation Reversible Covalent DPP1 Inhibitors Leading to an Oxazepane Amidoacetonitrile Based Clinical Candidate (AZD7986).

A novel series of second generation DPP1 inhibitors free from aorta binding liabilities found for earlier compound series was discovered. This work culminated in the identification of compound 30 (AZD7986) as a highly potent, reversible, and selective clinical candidate for COPD, with predicted human PK properties suitable for once daily human dosing.

A comprehensive company database analysis of biological assay variability.

Analysis of data from various compounds measured in diverse biological assays is a central part of drug discovery research projects. However, no systematic overview of the variability in biological assays has been published and judgments on assay quality and robustness of data are often based on personal belief and experience within the drug discovery community. To address this we performed a reproducibility analysis of all biological assays at AstraZeneca between 2005 and 2014. We found an average experimental uncertainty of less than a twofold difference and no technologies or assay types had higher variability than others. This work suggests that robust data can be obtained from the most commonly applied biological assays.

Pharmacokinetics and the drug-target residence time concept.

The concept of drug-target residence time has been in focus in recent drug discovery literature. However, few studies consider the combined effect of pharmacokinetics (PK) and binding kinetics (BK) on the duration of effect of a drug. Using a simple model that takes both PK and BK into account, we found that prolongation of binding owing to a long drug-target residence time can only occur when the binding dissociation is slower than the PK elimination. Data for several drugs and/or drug candidates in the literature indicate that the opposite is observed, that is, they have a slower elimination compared with dissociation. These observations greatly reduce the usability of drug-target residence times for estimating the duration of effect of a drug in vivo.

Mechanistic and kinetic characterization of hepatitis C virus NS3 protein interactions with NS4A and protease inhibitors.

The mechanism and kinetics of the interactions between ligands and immobilized full-length hepatitis C virus (HCV) genotype 1a NS3 have been characterized by SPR biosensor technology. The NS3 interactions for a series of NS3 protease inhibitors as well as for the NS4A cofactor, represented by a peptide corresponding to the sequence interacting with the enzyme, were found to be heterogeneous. It may represent interactions with two stable conformations of the protein. The NS3-NS4A interaction consisted of a high-affinity (K(D) = 50 nM) and a low-affinity (K(D) = 2 µM) interaction, contributing equally to the overall binding. By immobilizing NS3 alone or together with NS4A it was shown that all inhibitors had a higher affinity for NS3 in the presence of NS4A. NS4A thus has a direct effect on the binding of inhibitors to NS3 and not only on catalysis. As predicted, the mechanism-based inhibitor VX 950 exhibited a time-dependent interaction with a slow formation of a stable complex. BILN 2061 or ITMN-191 showed no signs of time-dependent interactions, but ITMN-191 had the highest affinity of the tested compounds, with both the slowest dissociation (k(off)) and fastest association rate, closely followed by BILN 2061. The k(off) for the inhibitors correlated strongly with their NS3 protease inhibitory effect as well as with their effect on replication of viral proteins in replicon cell cultures, confirming the relevance of the kinetic data. This approach for obtaining kinetic and mechanistic data for NS3 protease inhibitor and cofactor interactions is expected to be of importance for understanding the characteristics of HCV NS3 functionality as well as for anti-HCV lead discovery and optimization.

Structure-activity relationships of HCV NS3 protease inhibitors evaluated on the drug-resistant variants A156T and D168V.

BACKGROUND: HCV infections are a serious threat to public health. An important drug target is the NS3 protease, for which several inhibitors are in clinical trials. Because of the high mutation rate of the virus, resistance against any HCV-specific drug is likely to become a substantial problem. Structure-activity data for the major resistant variants are therefore needed to guide future designs of protease inhibitors. METHODS: The inhibitory potency of tripeptide NS3 protease inhibitors, with either a P2 proline or phenylglycine, in combination with different P3 and P1-P1' groups, was assessed in enzyme activity assays using the full-length NS3 protein with known resistance-conferring substitutions A156T or D168V. The results obtained from these variants were compared with the inhibition of the wild-type enzyme. Molecular modelling was used to rationalize the biochemical results. RESULTS: Inhibitors combining the P2 proline and P1 (1R,2S)-1-amino-2-vinylcyclopropyl-carboxylic acid (vinylACCA) lost much of their potency on the resistant variants. Exchange of the P2 proline for phenylglycine yielded inhibitors that were equipotent on the wild-type and on the A156T and D168V variants. The same result was obtained from the combination of either the P2 residue with a norvaline or an aromatic scaffold in the P1 position. CONCLUSIONS: The combination of a substituted P2 proline and P1 vinylACCA appears to be the main problem behind the observed resistance. Molecular modelling suggests an enforced change in binding conformation for the P2 proline-based inhibitors, whereas the phenylglycine-based inhibitors retained their wild-type binding conformation in the substituted forms of the enzyme.

Hepatitis C virus NS3 protease is activated by low concentrations of protease inhibitors.

The nonstructural protein 3 (NS3) of hepatitis C virus (HCV) is a bifunctional enzyme with a protease and a helicase functionality located in each of the two domains of the single peptide chain. There is little experimental evidence for a functional role of this unexpected arrangement since artificial single domain forms of both enzymes are catalytically competent. We have observed that low concentrations of certain protease inhibitors activate the protease of full-length NS3 from HCV genotype 1a with up to 100%, depending on the preincubation time and the inhibitor used. The activation was reduced, but not eliminated, by increased ionic strength, lowered glycerol concentration, or lowered pH. In all cases, it was at the expense of a significant loss of activity. Activation was not seen with the artificial protease domain of genotype 1b NS3 fused with a fragment of the NS4A cofactor. This truncated and covalently modified enzyme form was much less active and exhibited fundamentally different catalytic properties to the full-length NS3 protease without the fused cofactor. The most plausible explanation for the activation was found to involve a slow transition between two enzyme conformations, which differed in their catalytic ability and affinity for inhibitors. Equations derived based on this assumption resulted in better fits to the experimental data than the equation for simple competitive inhibition. The mechanism may involve an inhibitor-induced stabilization of the helicase domain in a conformation that enhances the protease activity, or an improved alignment of the catalytic triad in the protease. The proposed mnemonic mechanism and derived equations are viable for both these explanations and can serve as a basic framework for future studies of enzymes activated by inhibitors or other ligands.

Beta-amino acid substitutions and structure-based CoMFA modeling of hepatitis C virus NS3 protease inhibitors.

In an effort to develop a new type of HCV NS3 peptidomimetic inhibitor, a series of tripeptide inhibitors incorporating a mix of alpha- and beta-amino acids has been synthesized. To understand the structural implications of beta-amino acid substitution, the P(1), P(2), and P(3) positions of a potent tripeptide scaffold were scanned and combined with carboxylic acid and acyl sulfonamide C-terminal groups. Inhibition was evaluated and revealed that the structural changes resulted in a loss in potency compared with the alpha-peptide analogues. However, several compounds exhibited muM potency. Inhibition data were compared with modeled ligand-protein binding poses to understand how changes in ligand structure affected inhibition potency. The P(3) position seemed to be the least sensitive position for beta-amino acid substitution. Moreover, the importance of a proper oxyanion hole interaction for good potency was suggested by both inhibition data and molecular modeling. To gain further insight into the structural requirements for potent inhibitors, a three-dimensional quantitative structure-activity relationship (3D-QSAR) model has been constructed using comparative molecular field analysis (CoMFA). The most predictive CoMFA model has q(2)=0.48 and r(pred)(2)=0.68.

Effects on protease inhibition by modifying of helicase residues in hepatitis C virus nonstructural protein 3.

This study of the full-length bifunctional nonstructural protein 3 from hepatitis C virus (HCV) has revealed that residues in the helicase domain affect the inhibition of the protease. Two residues (Q526 and H528), apparently located in the interface between the S2 and S4 binding pockets of the substrate binding site of the protease, were selected for modification, and three enzyme variants (Q526A, H528A and H528S) were expressed, purified and characterized. The substitutions resulted in indistinguishable K(m) values and slightly lower k(cat) values compared to the wild-type. The K(i) values for a series of structurally diverse protease inhibitors were affected by the substitutions, with increases or decreases up to 10-fold. The inhibition profiles for H528A and H528S were different, confirming that not only did the removal of the imidazole side chain have an effect, but also that minor differences in the nature of the introduced side chain influenced the characteristics of the enzyme. These results indicate that residues in the helicase domain of nonstructural protein 3 can influence the protease, supporting our hypothesis that full-length hepatitis C virus nonstructural protein 3 should be used for protease inhibitor optimization and characterization. Furthermore, the data suggest that inhibitors can be designed to interact with residues in the helicase domain, potentially leading to more potent and selective compounds.