Deciphering the allosteric binding mechanism of the human

Subscriber access provided by BOSTON COLLEGE

Article

Deciphering the allosteric binding mechanism of the human tropomyosin receptor kinase A (hTrkA) inhibitors Govindan Subramanian, Paul D. Johnson, Theresa Zachary, Nicole Roush, Yaqi Zhu, Scott J Bowen, Ann Janssen, Brian Duclos, Tracey Williams, Christopher Javens, Nancy Dekki Shalaly, Daniel Martinez Molina, Arthur J. Wittwer, and Jeffrey L. Hirsch ACS Chem. Biol., Just Accepted Manuscript • DOI: 10.1021/acschembio.9b00126 • Publication Date (Web): 06 May 2019 Downloaded from http://pubs.acs.org on May 6, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

Deciphering the allosteric binding mechanism of the human tropomyosin receptor kinase A (hTrkA) inhibitors Govindan Subramanian,*,† Paul D. Johnson,† Theresa Zachary,† Nicole Roush,† Yaqi Zhu,† Scott J. Bowen,† Ann Janssen,† Brian Duclos,† Tracey Williams,† Christopher Javens,† Nancy Dekki Shalaly,§ Daniel Martinez Molina,§ Arthur J. Wittwer,¶ and Jeffrey L. Hirsch¶ †Veterinary

Medicine Research & Development (VMRD), Zoetis, 333 Portage Street,

Kalamazoo, MI49007, USA §Pelago

Bioscience, Banvaktsvägen 20, 17148 Solna, Sweden

¶Confluence

Discovery Technologies, 4320 Forest Park Avenue, St. Louis, MO63108, USA

ABSTRACT: Access to cryptic binding pockets or allosteric sites on a kinase that present themselves when the enzyme is in a specific conformational state offer a paradigm shift in designing the next generation small molecule kinase inhibitors. The current work showcases an extensive and exhaustive array of in vitro biochemical and biophysical tools and techniques deployed along with structural biology efforts of inhibitor-bound kinase complexes to characterize and confirm the cryptic allosteric binding pocket and docking mode of the small molecule actives identified for hTrkA. Specifically, assays were designed and implemented to lock the kinase in a predominantly active or inactive conformation and the effect of the kinase inhibitor probed to understand the hTrkA binding and hTrkB selectivity. The current outcome suggests that inhibitors with a fast association rate take advantage of the inactive protein conformation and lock the kinase state by also exhibiting a slow off-rate. This in-turn shifts the inactive/active state protein conformational equilibrium cycle, affecting the subsequent downstream signaling.

Human tropomyosin receptor kinase (Trk) belongs to the receptor tyrosine kinase (RTK) subfamily within the ~500-member protein kinase superfamily.1,2,3

To date, three well

differentiated Trk subfamily members (TrkA, TrkB, TrkC)4 have been reported in the literature, with TrkA, being clinically relevant for diseases like chronic pain5 and oncological indications.6,7

1 ACS Paragon Plus Environment

ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Being a RTK, the overall topology of TrkA includes a multi-domain extracellular region spanning ~275 amino acid residues8 that serves as the receptor’s binding site for endogenous neurotrophin ligands, and more specifically, for the preferential binding of the -nerve growth factor (-NGF) ligand.9,10 A single -helical transmembrane segment and the juxtamembrane (JM) loop successively follow the extracellular topology onto the intracellular face, to tail, with another ~270 amino acid long kinase domain on the cytoplasmic end. It is well recognized that homodimers of the Trk subfamily members are responsible for the endogenous ligand binding and ensuing downstream signal transduction.11,12 Several lines of evidence suggest -NGF binding to TrkA activates the NGF/TrkA pathway and trigger the signaling cascades responsible for increased sensitivity to the nociceptors that result in chronic pain sensitization.13,14 Clinical proof-of-concept of -NGF neutralizing antibodies like Tanezumab, Fulranumab, Fasinumab, etc., elicit statistically significant pain relief and corroborate the NGF/TrkA pathway as a potential therapeutic avenue for chronic pain management in humans.15

Although sequestering -NGF using neutralizing antibodies is

currently the most advanced therapy for modulating pain symptoms, unexpected side effects currently caution the unconditional translation of the clinically validated antibodies into a therapeutic prescription for human patients.16,17,18 Thus, efforts to target the TrkA receptor and intervene in the pain signaling mechanism emerged as a potential alternative for chronic pain management. Though TrkA initially was identified as a proto-oncogene19,20 mutations to the TrkA receptor revealed congenital insensitivity to pain and lent strong support to go after the receptor for pain indications.21,22 Yet, no meaningful success has been reported in the literature for a neutralizing antibody that accesses the extracellular domains of the TrkA receptor.23 However, small molecule inhibition of the kinase domain at the 2 ACS Paragon Plus Environment

Page 2 of 38

Page 3 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


intracellular side is an appropriate intervention opportunity that can be utilized effectively based on successful therapeutics for other members of the kinase superfamily, especially the RTKs.24,25,26 Moreover, the kinase domain is the appropriate region to target Trk oncogenes, since several of the NTRK gene fusion products often replace the extracellular domain, but the TrkA kinase domain is always retained in the oncogenic fusion protein.27,28 In fact, preliminary reports offer a glimpse into the potential of TrkA kinase inhibitors to respond in a tumor-agnostic manner.29 Small molecule inhibitors like the recently approved Larotrectinib (LOXO-101), LOXO-195 in advanced clinical trials, and additional followers like Entretinib, Cabozantinib, DS-6051b, etc., in their early clinical phases for several oncological indications showcase the recent advances of hTrkA kinase domain modulation.30 Several other pan Trk small molecule kinase inhibitors in various discovery/development phases31 also lend strong support to target modulation affecting the tumor-agnostic disease outcomes. Despite the significant progress of TrkA inhibitors, liabilities arising from the apparent resistance of TrkA mutations, subtype selectivity issues, and the extent of kinome promiscuity are beginning to dampen the hopes of these compounds achieving their full therapeutic potential.32 Hence, a new generation of small molecule inhibitors that bind differently and selectively is needed33 as an alternate paradigm for TrkA target modulation. In general, the binding of small molecules to the kinase domain of the TrkA receptor can fall into three main categories. Type 1 inhibitors are compounds that occupy the catalytic ATP binding site and are competitive with ATP, while Type 2 inhibitors occupy the same site but are noncompetitive with respect to ATP.34 The former results from an active state conformation of the kinase domain while the latter arises from an inactive state conformation that restricts access to ATP. Allosteric inhibitors bind to the kinase domain as well, but away from the ATP site, and can effectively bind either conformation.35,36 The protein conformational versatility and the non3 ACS Paragon Plus Environment


ATP binding pockets can be instrumental in achieving small molecule inhibitor selectivity against the Trk subtypes and across the kinome. The high sequence identity of the kinase domain among the subfamily members (~76%) and the residues lining the active ATP binding site (>90% identity between hTrkA & hTrkB) makes it extremely difficult for Type 1 inhibitors to achieve selectivity across the subtypes, resulting in several inhibitors demonstrating pan-Trk activity.37 However, recent literature reveals early inroads into the identification of Type 238 and allosteric modulators39,40,41 that exhibit selectivity across the Trk subtypes and the kinome.41 Consequently, the current TrkA lead-finding strategy focused on the identification of allosteric inhibitors that demonstrate subtype specificity and hold potential for kinome selectivity. Compared to the traditional kinase inhibitors widely reported in the literature, until now, allosteric kinase inhibitors identified across the kinome are very sparse.35 Therefore, the kinase field suffers from a lack of generalized experimental and in silico computational approaches to identify the non-orthosteric inhibitors.42,43 As a result, a systematic effort to identify novel chemical leads is outlined along with attempts to compare their binding mechanism (Type 1 vs. allosteric) through elaborate biochemical, biophysical, and intact cellular assays, and final characterization of the same via structural biology efforts on representatives. The current work provides the basic framework for considering and selecting the appropriate in vitro screening assays or platforms to interrogate selective conformational states and explore unconventional binding pockets and interactions to achieve the desired target specificity.

RESULTS


Page 4 of 38

Page 5 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


In silico ligand-based modeling. Virtual screening of ~9 million compounds from the ZINC database (http://zinc.docking.org) via three-dimensional shape- and feature-based ROCS approach44 was initially pursued (supporting information Method 1S). This was followed by rank ordering the hits using ComboScore, application of Lipinski-like physicochemical property filters,45 visual inspection of top ranked hits, and preferred vendors with the constraint of availability for immediate purchase.

This workflow resulted in the initial in silico hit list

prioritization of ~1644 compounds with potential for in vitro evaluation using biochemical techniques. HTRF Assay.

A preliminary in vitro biochemical assay using the hTrkA kinase domain

(supporting information Experiment 1S) initiated via a single-point screening (in duplicate) at 10M compound concentration resulted in 290 positives with an average inhibition of ≥ 60%. The higher hit rate necessitated an additional single-point screening effort on the positives (in duplicate) by lowering the compound concentration to 1M resulting in 54 positives exhibiting ≥ 70% inhibition. A final dose-response of these positives yielded 6 actives with IC50 < 100nM and two novel lead chemical scaffolds represented by 1 and 2 (Scheme 1). Active benzamide, 1, and other analogs of this scaffold (supporting information Method. 1S) were deprioritized from further chemical exploration due to the near compound planarity that affect physicochemical properties like solubility.

Additionally, minimal avenues to explore structure-activity

relationships (SAR) served as a bottleneck for further progression of this chemical scaffold. The singleton substituted acetamide lead, 2, was further elaborated conservatively via medicinal chemistry SAR probing, resulting in analogs 3-8 (Scheme 1, supporting information Experiment 2S) along with a drastic modification to reduce the lipophilicity and polycyclic substituent nature, exemplified by 9. The consistently good hTrkA enzyme activity (Table 1) of the analogs provided the preliminary confidence to nominate this scaffold as a new chemical lead series. 5 ACS Paragon Plus Environment


Given the diversity of the lead chemical scaffolds, it became apparent that binding mode characterization was critical before any expanded lead optimization could be pursued, since the two positive control compounds 1046 and 11,47 representing the allosteric and Type 1 inhibitors reported in literature, did not provide discriminatory activities for the two different binding modes in this assay format.


Page 6 of 38

Page 7 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


N

O

O

O

N N

N H

X

N

N

N

O

O

N

Y N

N H

N

O

O

N

X

4

N

N

N H

2 (X = CH); Enamine [Z223603930] 3 (X = N)

1, ChemDiv (G771-0041)

N

N

N

N H

5 (X = CH; Y = H) 6 (X = N; Y = CH3) 7 (X = CH; Y = CH3)

N

O

O N

N H

O

N

8

O

S N

N H

9, Enamine (Z88589241)

O

N N

N

N

O N H

Cl

N

N

N

NH O

N H

F

N

N H

N

N H

F

11, AZ-23 [CAS#: 915720-21-7] 10, Array Biopharma [WO2015/175788A1] AxonMedchem (Axon1610) (CAS#: 1824664-89-2] Scheme 1. Structures of hTrk-A kinase domain inhibitors. 7 ACS Paragon Plus Environment

ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47

Page 8 of 38

Table 1. in vitro screening results for 1-11 using different assay formats. All results are reported in nM units, except for the SPR KD values (pM). Assay HTRF FLIPRTM Ca2+ Flux Caliper SPR CETSA NO a Cpd_ID Enz_IC50 Cel_IC50b Cel_IC50c Ina_IC50d Act_IC50e Ina_KDd Act_KDe Ina_EC50f Act_EC50g Cel_IC50h >20000 68.3 >20000 2400 >10000 46500 >10000 1 >10000 2.5 237.3 24 >1000 1600 >10000 >10000 2 3.7 797.0 3 44.3 >6242.6 2600 4 >10000 2.2 350.1 4000 5 66.5 434 1.4 12.2 77 >10000 4600 39500 >1000 6 >10000 0.2 25.1 2200 7 >10000 1.0 61.8 1750 8 44.1 >20000 85 >10000 6960 >10000 9 i 10.8 0.00138 10 2.2 4 >100 80.4 6400 16.5 6.8 0.669 0.124 0.3 4.9 7.3 1M) for all the compounds tested and hence an IC50 is not being reported for 6. The weak activity of 6 is also consistent with that obtained using CETSA (Table 1) and lends supporting evidence to the biophysical method developed here for characterizing allosteric kinase inhibition. 125 100

% Inhibition

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


75 50 25 0 0.0001

0.01

1

100

10000

[Compound] (nM)



Page 16 of 38

Figure 4. Effects of TrkA inhibitors 11 (blue), 10 (red), and 6 (green) on rat -NGF induced neurite outgrowth in PC12 cells (representative curves; error bars, standard deviation, n = 2).

The consistently superior potency of the HTS-identified inhibitors in a recombinant hTrkA overexpressing cell line as well as the physiologically-relevant unbiased primary PC12 cells under active and inactive kinase conditions furthered the confidence of the inhibitors being potentially bound to a non-orthosteric site of the kinase. Consequently, the X-ray structure of the inhibitor bound hTrkA kinase domain complex was pursued to confirm the allosteric binding site that would enable a structure-based lead optimization effort for the lead chemical series. X-ray of hTrkA-ligand complex: Crystallographic efforts pursued on 1, 6, and 9 using the extended hTrkA kinase domain that spans the JM region additionally yielded inhibitor bound complex structures at 2.19Å, 1.97Å, and 2.31Å resolutions respectively. The crystals contained one kinase domain monomer in the asymmetric unit and the electron density revealed an unambiguous binding configuration that included the ligand orientation and conformation. Some short loop regions could not be fully defined by the electron density (supporting information Table 4S) but were not detrimental to the ligand binding.

These regions were modeled

appropriately using the PRIME module within the MAESTRO software suite (Schrödinger, New York, NY) to build a full-length kinase domain structure including the JM region.

As

anticipated, the inhibitors bound to a site distinctly different and away from the ATP binding site, confirming the existence of an allosteric pocket (supporting information Figure 5S). Additionally, the activation loop (A-loop) presents the 668DFG670 motif reminiscent of the hTrkA kinase domain inactive conformation34,38 and occludes part of the orthosteric site for any ATP binding or exchange. These two structural observations provide the conclusive evidence of the allosteric inhibitor binding site and the kinase being locked in an inactive state conformation.


Page 17 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5. X-ray structure of the kinase domain of hTrkA bound to inhibitor 1 (top, PDB code: 6NPT), 6 (middle, PDB code: 6NSS) and 9 (bottom, PDB code: 6NSP). The 3D binding site 17 ACS Paragon Plus Environment


(left) rendition and the schematic 2D depiction (right) of the ligand…hTrkA kinase domain residues define the allosteric binding pocket.

In terms of the ligand-kinase binding, the ligand amide N-H makes the critical H-bond interaction with the activation loop aspartate (D668) side chain and locks the kinase A-loop conformation in a DFG-out orientation. The amide carbonyl, however, interacts with different JM residue backbone atoms and in diverse ways (Figure 5). For instance, the amide carbonyl in 1 makes an indirect contact with the H648 backbone carbonyl via a solvent water mediated Hbond network interaction. On the other hand, the amide carbonyl in 6 makes a direct H-bond interaction with the I490 backbone N-H group. The amide carbonyl in 9 is proximal to I491 but a binding interaction cannot be elucidated due to the poor electron density and refinement of several of the JM residues in the inhibitor bound complex. Such varied binding interaction patterns have also been reported in literature (supporting information Table 5S) and provide insights into the binding site plasticity.39 The bound ligands are also stabilized via van der Waals interactions with the hydrophobic residues lining the allosteric binding site (Figure 5). A common post translational modification in the RTKs is the phosphorylation of the tyrosine residues in the activation loop that stabilize the A-loop such that the substrate and ATP can bind to the kinase active state.50 In the solved x-ray structures of the complex, the Y676 is buried with its sidechain hydroxyl functionality H-bonding with the D596 side chain carboxylate, and hence non-available for transphosphorylation.

A similar H-bond pattern and stabilization

between Y680 and N655 prevents this tyrosine participation in kinase activation as well. This leaves only Y681 in the A-loop exposed to provide any kind of docking environment for the substrate or partner protein. Although unclear at this point, the A-loop residue S677 is also exposed and is found to be phosphorylated in the hTrkA…1 x-ray complex. Y496 in the JM 18 ACS Paragon Plus Environment

Page 18 of 38

Page 19 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


domain is reported to get phosphorylated,51 but the significance and relevance of the same could not be ascertained due to the poor electron density in the solved x-ray structures.

DISCUSSION Preliminary in vitro biochemical screening using HTRF pointed to a very successful in silico ligand-based modeling campaign, yielding novel chemical scaffolds hitherto unreported for kinases and particularly, hTrkA. Additionally, the chemical leads did not possess the signature H-bond donor/acceptor kinase fingerprint typically reported for Type 1 and Type 2 hingebinding kinase inhibitors.34 Although experimentally unproven at this screening stage, 1-9 were potentially considered allosteric leads for hTrkA kinase, since the compounds would not fit well in the ATP binding site when docked in silico using the GLIDE module available from Schrodinger’s MAESTRO software suite.

Consequently, additional assay development and

validation using a multitude of in vitro biochemical and biophysical techniques were initiated to delineate and confirm the hypothesized binding preference. To potentially avert any presumed HTRF assay technology limitation or artifact, an orthogonal Caliper method was used that measures kinase activity by monitoring the phosphorylation of a fluorescently-labeled peptide substrate. These initial experiments provided early insights into the inhibitor differentiation between the active (Type 1) and inactive (allosteric) hTrkA kinase conformations (Table 1). Following the IC50 determination, the Ki’s were determined for a select subset (supporting information Table 2S) with the outcome giving a firm assurance that the actives were indeed kinase inhibitors.



Page 20 of 38

These biochemical screening data provided cautiously optimistic confidence that the chemical leads could potentially be allosteric inhibitors. The biochemical affinities (Table 1) derived from two different techniques above prompted the use of other approaches to further characterize the hTrkA inhibition mechanism of the identified leads. Consequently, biophysical techniques were sought out to derive additional insights into the binding characteristics. The well-established SPR technique was pursued using both active and inactive hTrkA enzyme. Among the 3 compounds tested, the binding kinetics revealed a slower off-rate for 11 in the active form when compared against the results from the inactive extended kinase domain hTrkA enzyme (supporting information Table 3S). A nearly 2-fold slower off-rate (kd) for the inactive form was observed for 10, suggesting the preference of 10 to bind more tightly to the inactive enzyme. A similar pattern for 6 provided the needed boost for the lead to be a potential allosteric inhibitor for hTrkA enzyme. The above assays employed artificial or engineered systems and conditions with the concern that they may be biased and not translate appropriately in a near physiological environment. Hence, select compounds were screened using primary rat neuronal PC12 cells. The -NGF induced neurite outgrowth assay clearly demonstrated that the new lead, 6, was active, but with weaker affinity when compared against the literature tool, 10. However, the technique was not amenable to discriminate between the hTrkA conformational macrostates and the preferential binding of the kinase inhibitors selected for the primary cells assay.

Consequently, significant assay

development was undertaken with the label-free CETSA technique as the same has not been explored, evaluated, or reported for characterizing non-orthosteric binding of small molecules in literature. Dose-response of the Type 1 inhibitor (11) using CETSA yielded an IC50 of ~120nM at 55oC (active kinase) and very minimal inhibition at the highest tested inhibitor concentration of 100M under 40oC (inactive kinase) temperature conditions. This clearly suggests that there 20 ACS Paragon Plus Environment

Page 21 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


are at least two protein conformational populations in these cells. This was further validated using the allosteric inhibitor, 10, that resulted in an IC50 of ~ 80nM when screened at 40oC and is >6M at 55oC (Table 1 and supporting information Experiment 5S). A similar observation was noted in the dose-response curves of 6 as well, albeit with weaker affinities. This should be expected since the inhibitors are still chemical leads that have not yet been optimized like 10. Furthermore, the measured solubility of the inhibitors may also be attributed to the low activities of 1-9 when compared to 10 and 11, and the additional requirement that the compounds need to possess reasonable cell permeability (supporting information Table 6S) for appropriate target engagement. Preliminary crystallographic efforts utilizing the standard kinase domain construct and an allosteric inhibitor did not yield any structural solution for the bound complex. However, inclusion of the JM domain residues to the kinase domain resulted in complex X-ray structures bound with allosteric inhibitors. In fact, the allosteric ligand was sandwiched between the Aloop and the JM domain residues. The significance of including the JM domain and its relevance to allosteric inhibitor binding has also been disclosed recently for hTrkA39 and reported for other RTKs like EGFR.52 Subsequent superposition of the allosteric complexes solved here with the reported x-ray structure of Type 1 inhibitor (11)53 bound to hTrkA revealed that the allosteric binding site is unlikely to exist when the kinase is in an active state conformation. This results from the different disposition of the A-loop residues that allow these residues to occupy the allosteric binding pocket region and potentially push the JM domain residues outward. The cryptic allosteric pocket, therefore, is accessible only when the kinase is in an inactive state conformation. Any inhibitor that docks to this allosteric site would stabilize the inactive form of the kinase and impose additional barriers to the kinase activation since the A-loop tyrosine residues are not easily available for transphosphorylation. The unstructured JM domain fold is 21 ACS Paragon Plus Environment


also suggestive of this region being more flexible and adaptive to the ligand environment. Comparison of the current x-ray structures to that reported in recent literature39-41 also illustrates the side-chain flexibility of F646 between an open and closed state, adaptive JM domain residues with different folds near the binding site (supporting information Table 5S and Analysis 1S) to accommodate the diverse chemical scaffolds occupying this region. Most commercially available kinome panels use screening formats tailored for Type 1 inhibitors that preferentially bind to the kinase active state. All evidence for the current leads suggests the inhibitors preferentially bind the kinase inactive state (Table 1). Although a kinome panel screen would predict the leads to be highly specific and non-promiscuous, such data is deemed potentially misleading due to the use of abbreviated kinase domain sequences and the screen mostly performed under kinase active-state conditions. Limited biochemical enzyme assays using extended kinase domain and FLIPR™ cell-based assays performed using the full-length kinase for hTrkA and hTrkB, however, gave encouraging signs of selectivity against the sub-type (Table 1 and supporting information Figure 3S). Finally, three out of the four reported mutations in the hTrkA kinase domain implicated in resistance mechanisms54,55 are in the ATP site and weakens the Type-1 inhibitors binding to the orthosteric site (supporting information Figure 6S). Visual inspection of the allosteric inhibitor bound hTrkA complexes suggest that none of these mutations would have a detrimental effect on inhibitor binding. Early ADME evaluation of the inhibitors was also undertaken as part screening funnel (supporting information Table 6S). The outcome clearly suggested that none of the designed lead analogs exhibited a profile worth progressing towards in vivo pharmacokinetic and efficacy studies due to their poor kinetic solubility, weak MDCK permeability, and high dog/rat liver microsomal lability. Although the in vitro assays and binding mechanisms were clarified, a


Page 22 of 38

Page 23 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


multifactorial lead optimization effort is required to maneuver the hurdles in the identified chemical series.

CONCLUSIONS AND PERSPECTIVE The current work showcases a systematic workflow encompassing multiple orthogonal techniques and technologies to decipher the small molecule inhibitor binding targeting a nonorthosteric site in hTrkA. The allosteric site can be accessed for small molecule binding only when the JM region is included along with the traditional kinase domain in the experimental setup. This site is cryptic but is easily accessible for exploration when the kinase equilibrium shifts towards an inactive or an autoinhibited state. In vitro assays using the Caliper, Biacore, and CETSA technologies clearly provide convincing evidence to this end. The absolute selectivity of the representative inhibitors to hTrkB kinase domain via the Caliper technology and to the full length hTrkB protein in the cell-based FLIPR™ assay provide early insights into the selectivity profile across the kinome.

Additional structural biology evidence from x-ray solutions of

representative analogs (1, 6, 9) support the small molecule inhibitor binding to the hTrkA kinase inactive state conformation. The binding site residues and the ligand interactions are also not with reported residues that confer resistance apprehensions. Instead of focusing efforts on identifying the next generation Type 1 inhibitors that overcome the predominant resistance mechanisms, the current effort achieved small molecule allosteric inhibitor leads with subtype selectivity as well. The identified leads exhibit fast association kinetics that preferentially bind and lock the inactive state kinase conformation, offering an alternate paradigm to prosecute the target.

Evaluating their potential as disease modifiers or therapeutics could change the

conventional wisdom and provide the critical breakthroughs for treating diseases.



The multi-step kinase activation process allows the enzyme to adapt a series of dynamic macrostates providing conformational plasticity to the protein that could be targeted for small molecule inhibition. The one-size fits all kinase screening philosophy currently employed in the literature fails to uncover the hidden opportunities presented by these targets, as most assays are designed to identify or evaluate active state kinase inhibitors (e.g. kinome profiling). In vitro screening assays need to be developed or tailored specifically to capitalize on cryptic (protein conformation specific) and allosteric binding sites that provide the best prospect of achieving the selectivity shortcomings of Type 1 kinase inhibitors. Targeting non-orthosteric binding sites also provides the additional advantage of identifying inhibitors much less susceptible to resistance mutations that emerge with treatment/time. Similarly, arresting or shifting the conformational equilibrium towards the inactive state via allosteric tight binders with slow dissociation can affect the downstream signaling. The JM domain residue, Y496, near the allosteric binding site is a known phosphotyrosine docking site involved in signal transduction.56 Further experiments are warranted to understand how small molecule binding in this region alters the binding of adaptor/effector proteins.

METHODS Caliper enzyme assay. A Caliper LC3000 instrument (Perkin Elmer) with a LabChip EZReader 12-sipper chip (Perkin Elmer, 760404) was used to separate and quantify the peptide substrate and phospho-peptide reaction product. The separation buffer contained 100 mM HEPES, 10 mM EDTA, 0.0005% Tween 20, 0.1% Coating Reagent 3 (Perkin Elmer, 760050), and 1% DMSO, with a downstream voltage of -800, upstream voltage of -2750, and screen pressure of -1.1 psi. CSKtide was purchased from Anaspec (AS-63842), Inactive TrkA was purchased from Life Technologies (PV-4114) and was activated by incubating 1 µM with 2 mM 24 ACS Paragon Plus Environment

Page 24 of 38

Page 25 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ATP in assay buffer (20 mM HEPES, pH 7.4, 10 mM MgCl2, 0.01% BSA, 0.0005% Tween 20, 1 mM DTT) at room temperature for 2 h, after which small aliquots were stored at -80°C. For IC50 assays, compounds were diluted in 100% DMSO to 500x the highest concentration to be tested. A D300 dispenser (Tecan) was used to deliver 80 nL of compound or DMSO control to the assay plate (Axygen/Corning, VWR, 47743-468). Compounds were tested using duplicate 11-point dose response curves with 3-fold changes in concentration between doses. TrkA enzyme (20 µL) was added and incubated for 20 min at room temperature before addition of 20 µL ATP/CSKtide mixture. Final concentrations (in assay buffer) were 0.2% DMSO, 1 µM CSKtide, 100 µM ATP, 0.5 or 1 nM active TrkA or 20 nM inactive TrkA, and variable amounts of compound. After 2 h at room temperature, reactions were terminated by the addition of 40 µL of stop buffer (20 mM HEPES, pH 7.4, 30 mM EDTA) and the phospho-peptide product quantified on the Caliper LC3000. Under these conditions there was typically 20%-30% conversion of substrate to product in the absence of inhibitor. Results were expressed as % inhibition relative to no enzyme and no inhibitor controls and IC50 values calculated using GraFit software (Erithacus Software, Limited). For incomplete S-curves, plateaus were fixed at either 100% or 0% inhibition as appropriate.

Single cycle kinetics (SCK) binding assay using Biacore T200. 40 µg/ml each of ATP treated (Carna Biosciences, Inc, cat# 08-486-20N) and non-ATP treated (Carna Biosciences, Inc, cat# 08-486-23N) biotinylated-TrkA were captured on SA sensor at 4 °C in fresh-made, pre-cooled TTNM buffer (50 mM Tris-HCL pH 7.5, 0.05% P20, 150 mM NaCl and 5 mM MgCl2) with flow rate 10 µl/min aimed surface density of 6000 RU. Then 10 µg/ml biocytin was injected 3 x 60 s at flow rate 10 µl/min to block remaining streptavidin. Switch temperature to 22°C, pre-run TTNM with 1% DMSO until surface is stable. Compounds were diluted to final concentrations 25 ACS Paragon Plus Environment


in TTNM buffer with 1% DMSO. Using SCK method, compounds were tested at 0.078, 0.31, 1.25, 5 and 20 µM with 100 s contact time and 1200 s dissociation time at flow rate of 30 µl/min. TTNM buffer with 1% DMSO buffer was also injected before every cycle as reference control. Data were analysed using Biacore Evaluation Software by 1:1 binding and residence time () were then calculated based on dissociation rate.

Cellular thermal shift assay (CETSA) Melt and shift curves. Melt and shift curves were generated in intact PC12 cells first for the assay development compounds 10 and 11. The cells were collected and pelleted by centrifugation and thereafter washed in HBSS and re-suspended in HBSS to a cell density of 40 million cells/ml. The compound was diluted to 20μM final concentration in HBSS, giving a DMSO concentration of 0.2%. Equal volumes of cell suspension and compound at 2x final concentration were mixed to yield a final concentration of 10 μM compound and 20 million cell/ml that was incubated for 60 min at 37°C with gentle continuous rotation. 0.1% DMSO was used as a negative control. Cell viability was determined before and after incubation with compound. The cell-compound suspension was then aliquoted into PCR strips (30l) and followed by a 12-step heat challenge between 37°-70°C for 3 minutes, after which the samples were kept on ice. Cells were lysed by addition of 10l LB4, supplemented with 100x blocking buffer, while shaking at 500 rpm for 30 min.

Isothermal concentration-response curves. Intact PC12 cells were pelleted by centrifugation, washed and re-suspended in HBSS at a cell density of 40 million cells/ml. The cell suspension was divided into 40 µl aliquots and an equal 26 ACS Paragon Plus Environment

Page 26 of 38

Page 27 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


volume of HBSS containing 2X of the intended final compound concentration was added, resulting in a final cell concentration of 20 million cells/ml. A 7 step concentration response series (100M to 10nM) of the compounds was prepared in 1% DMSO. The cells were incubated with compound series at 37°C for 60 min, with gentle mixing. The aliquots were heated to 40°C, 52°C and 55° as determined from the established melt and shift curves, for 3 minutes. Thereafter, the cells were lysed by addition of 10 l LB4 (4x) supplemented with 100x blocking reagent and kept shaking at 500rpm for 30min. The HTRF intensities were obtained by measuring the intensity at two wavelengths 665nm and 620nm (BMG, Clariostar). The obtained intensities were plotted as the ratio count (665nm/620nm) normalized to a relevant temperature count as specified. The normalized intensities were analyzed and plotted using GraphPad Prism software. Data points are shown as mean values with error bars indicating ± SEM. No error bars are shown if SEM is smaller than the symbol.

PC-12 neurite outgrowth assay. TrkA inhibitors were functionally evaluated in a neurite outgrowth assay using primary rat neuronal PC-12 cells that express both TrkA and its coreceptor p75. ATCC modified RPMI 1640 medium (Life Technologies, Carlsbad, CA) supplemented with 5% fetal bovine serum (FBS, Life Technologies, Carlsbad, CA) and 10% heat-inactivated horse serum (HI-HS, Life Technologies, Carlsbad, CA) were used to maintain the primary PC-12 cells (American Type Culture Collection [ATCC], Rockville, MD) and the same was incubated at 37°C with 5% CO2. PC-12 cells were removed from culture flasks with a cell scraper (Corning Inc., Corning, NY) and cell suspension centrifuged at 200 x g for 10 minutes on the experiment day. At the end of centrifugation, the supernatant was discarded and cell pellet re-suspended in 5 mL assay medium: ATCC modified RPMI 1640 supplemented with 27 ACS Paragon Plus Environment


1% HI-HS, 0.5% FBS, and 10 µg/ml gentamicin. A 10mL syringe outfitted with a 22 gauge x 1.5inch needle (Becton Dickinson, Franklin Lakes, NJ) was used to gently aspirate the resulting cell suspension, and nearly 15 – 20 times, to break up the cell clusters prior to counting and further dilution of the suspension.

Collagen IV coated 96-well microplates (Corning Inc.,

Corning, NY) was used to perform the PC-12 neurite outgrowth assay by preincubating 3,000 cells per well with select TrkA inhibitors, solubilized in DMSO, at the indicated concentrations for 30 minutes before adding 50 ng/ml of the recombinant rat β-NGF (R & D Systems Inc., Minneapolis, MN). The final assay DMSO concentration was between 0.1% – 0.25%. An IncuCyte ZOOM live-cell imaging system and its NeuroTrack analysis software (Essen Biosciences, Ann Arbor, MI) was used to evaluate the effects of select TrkA inhibitors on rat βNGF induced neurite outgrowth, performed after a 7-day culture period at 37°C in the presence of 5% CO2.

The outgrowth of the neurite was assessed by measuring the neurite length

(mm/cell-body cluster). Neurite outgrowth in the presence and absence of rat β-NGF was used to define the maximal and minimal response.

The effect of the TrkA inhibitors are then

expressed as a percentage of the minimal and maximal responses. The resulting percent inhibition/neutralization data at the selected concentrations was plotted with GraphPad Prism 7 (GraphPad software, San Diego, CA) for determining the dose-response IC50 using a 4-parameter curve fit.

Structural biology. Amino acids 483-796 of human TrkA (UniProt entry P04629) were cloned into vector pFastBac1 in frame with a TEV-cleavable GST-tag. Baculovirus was produced following the Bac-to-Bac-manual and the protein was expressed in Sf9 cells using disposable bioreactors. The protein for crystallization was purified by using a three-step chromatography procedure on a 20 mL GSTPrep-column (GE Health Care), TEV-cleavage followed by a second 28 ACS Paragon Plus Environment

Page 28 of 38

Page 29 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


pass over the GSTPrep, and a final gel filtration step on a Superdex75 26/60 column (GE Health Care). Pure protein was pooled after SDS-PAGE analysis and concentrated to ~20mg/mL in 20mM Tris/HCl, pH=8.5, 50mM NaCl, 2mM DTT, 0.5 mM PMSF. The protein at a concentration of 10 mg/ml was mixed with 2mM of the ligand (diluted from 100 mM DMSO stock solution) for 1 hour on ice and crystallized at 4°C (hanging drop) from: 0.10M KH2PO4/ 0.10M NaH2PO4/ 0.10M MES/NaOH pH = 6.00 and 1.90M NaCl (cryo: 25% glycerol in reservoir), or 18% (w/v) PEG 3350, 0.20M CaCl2, 0.10M, MES pH = 6.50 (cryo: direct). X-ray diffraction data were collected at a temperature of 100K at the Swiss Light Source (beamline PXI/X06SA) using a PILATUS 6M detector. Data were integrated, scaled and merged using XDS and the structure was refined with REFMAC5. Manual model completion was carried out using Coot. The quality of the final model was verified using PROCHECK and the other validation tools available through Coot.

ASSOCIATED CONTENT Supporting Information The supporting information is available on the ACS publications website free of charge at DOI: 10.1021/acschemSix supplemental tables, 6 supplemental figures, 1 scheme, 1 method, and 2 detailed experimental protocols (PDF). Accession Codes X-ray data deposited as PDB entries 6NPT (1), 6NSS (6), and 6NSP (9) AUTHOR INFORMATION Corresponding Author * Tel.: 1-269-359-9528. E-mail: [email protected] 29 ACS Paragon Plus Environment


Page 30 of 38

ORCID Govindan Subramanian: 0000-0002-6040-0415 Author Contributions G.S. conceived and executed the full work, performed in silico modeling, and wrote a significant part of the manuscript. P.D.J. synthesized the compounds; N.R. and T.Z. performed the in vitro enzyme and cell assays. Design and execution of the experiments are Y.Z. for the Biacore; S.J.B. for the neurite outgrowth; A.J.W. and J.L.H. for the Caliper; N.D.S. and D.M.M. for the CETSA assays. All authors read and contributed to the manuscript. Notes The authors declare the following competing financial interests. G.S. P.D.J. T.Z. N.R. Y.Z. S.J.B. A.J. B.D. T.W. C.J. were/are current employees of Zoetis. N.D.S. and D.M.M. are employees and shareholders of Pelago Bioscience AB. A.J.W, and J.L.H., are employees of Confluence Discovery Technologies ACKNOWLEDGEMENTS The authors thank the Veterinary Medicine Research & Development leadership, Zoetis, for funding this work. GS thanks A. Gonzales, G. Bainbridge, S. Kamerling, C. Haines, M. Cox, M. Tory, P. Misiak, and S. S. K. So for support and guidance throughout the research work. Outsourced structural biology work was performed under the guidance of S. Steinbacher, Proteros Biostructures GmbH, Germany. REFERENCES (1) Skaper, S. D. (2012) The neurotrophin family of neurotrophic factors:

An overview.

Methods Mol. Biol. 846, 1-12. (2) Blume-Jensen, P., and Hunter, T. (2001) Oncogenic kinase signaling. Nature. 411, 355-365.


Page 31 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(3) Hanks, S. K., and Hunter, T. (1995) The eukaryotic protein kinase superfamily: kinase (catalytic) domain structure and classification. FASEB J. 9, 576-596. (4) Huang, E. J., and Reichardt, L. F. (2003) Trk receptors: roles in neuronal signal transduction. Annu. Rev. Biochem. 72, 609-642. (5) Hefti, F. F., Rosenthal, A., Walicke, P. A., Wyatt, S., Vergara, G., Shelton, D. L., and Davies, A. M. (2006) Novel class of pain drugs based on antagonism of NGF. Trends Pharmacol. Sci. 27, 85-91. (6) Vaishnavi, A., Le A. T., and Doebele, R. C. (2015) TRKing down an old oncogene in a new era of targeted therapy. Cancer Discov. 5, 25-34. (7) Wang, T., Yu, D., and Lamb, M. L. (2009) Trk kinase inhibitors as new treatments for cancer and pain. Expert Opin. Ther. Pat. 19, 305-319. (8) Schneider, R., and Schweiger, M. (1991) A novel modular mosaic of cell adhesion motifs in the extracellular domains of the neurogenic Trk and trkB tyrosine kinase receptors. Oncogene. 6, 1807-1811 (9) Klein, R., Jing, S. Q., Nanduri, V., O’Rourke, E., and Barbacid, M. (1991) The Trk protooncogene encodes a receptor for nerve growth factor. Cell. 65, 189-197. (10) Kaplan, D. R., Martin-Zanca, D., and Parada, L. F. (1991) Tyrosine phosphorylation and tyrosine kinase activity of the Trk proto-oncogene product induced by NGF. Nature. 350, 158160. (11) Shen, J., and Maruyama, I. N. (2011) Nerve growth factor receptor TrkA exists as a preformed, yet inactive, dimer in living cells. FEBS Lett. 585, 295-299. (12) Maruyama, I. N. (2014) Mechanisms of activation of receptor tyrosine kinases: monomers or dimers. Cells. 3, 304-330.



(13) Hirose, M., Kuroda, Y., and Murata, E. (2016) NGF/TrkA signaling as a therapeutic target for pain. Pain Pract. 16, 175-182. (14) Norman, B. H., and McDermott, J. S. (2017) Targeting the nerve growth factor (NGF) pathways in drug discovery. Potential applications to new therapies for chronic pain. J. Med. Chem. 60, 66-88. (15) Yeh, J.-F., Akinci, A., Al Shaker, M., Chang, M. H., Danilov, A., Guileen, R., Johnson, K., Kim, Y. C., El-Shafei, A., Skljarevski, V., Duenas, H., and Tassanawipas, W. (2017) Monoclonal antibodies for chronic pain:

A practical review of mechanisms and clinical

applications. Mol. Pain. 13, 1-14. (16) Garber, K. (2011) Fate of novel painkiller mAbs hangs in balance. Nat. Biotechnol. 29, 173-174. (17) Bannwarth, B., and Kostine, M. (2017) Nerve growth factor antagonists: Is the future of monoclonal antibodies becoming clearer. Drugs. 77, 1377-1387. (18) Taneja, A., Pasqua, O. D., and Danhof, M. (2017) Challenges in translational drug research in neuropathic and inflammatory pain: the prerequisites for a new paradigm. Eur. J. Clin. Pharmacol. 73, 1219-1236. (19) Lamballe, F., Klein, R., and Barbacid, M. (1991) The trk family of oncogenes and neurotrophin receptors. Princess Takamatsu Symp. 22, 153-170. (20) Nakagawa, A. (2001) Trk receptor tyrosine kinases: A bridge between cancer and neural development. Cancer Lett. 169, 107-114. (21) Indo, Y., Tsuruta, M., Hayashida, Y., Karim, M. A., Ohta, K., Kawano, T., Mitsubuchi, H., Tonoki, H., Awaya, Y., and Matsuda, I. (1996) Mutations in the TRKA/NGF receptor gene in patients with congenital insensitivity to pain with anhidrosis. Nat. Genet. 13, 485-488.


Page 32 of 38

Page 33 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(22) Indo, Y. (2001) Molecular basis of congenital insensitivity to pain with anhidrosis (CIPA): mutations and polymorphisms in TRKA (NTRK1) gene encoding the receptor tyrosine kinase for nerve growth factor. Hum. Mutat. 18, 462-471. (23) Ugolini, G., Marinelli, S., Covaceuszach, S., Cattaneo, A., and Pavone, F. (2007) The function neutralizing anti-TrkA antibody MNAC13 reduces inflammatory and neuropathic pain. Proc. Nat. Acad. Sci. 104, 2985-2990. (24) Wu, P., Nielsen, T. E., and Clausen, M. H. (2016) Small-molecule kinase inhibitors: an analysis of FDA-approved drugs. Drug Discov. Today. 21, 5-10. (25) Berndt, N., Karim, R. M., and Schönbrunn, E. (2017) Advances of small molecule targeting of kinases. Curr. Opin. Chem. Biol. 39, 126-132. (26) Carles, F., Bourg, S., Meyer, C., and Bonnet, P. (2018) PKIDB: A curated, annotated and updated database of protein kinase inhibitors in clinical trials. Molecules, 23, 908. (27) Amatu, A., Sartore-Bianchi, A., and Siena, S. (2016) NTRK gene fusions as novel targets of cancer therapy across multiple tumor types. ESMO Open. 1, e000023. (28) Lange, A. M., and Lo, H.-W. (2018) Inhibiting TRK proteins in clinical cancer therapy. Cancers. 105, 1-15. (29) Kummar, S., and Lassen, U. N. (2018) TRK inhibition: A new tumor-agnostic treatment strategy. Target. Oncol. 13, 545-556. (30) Khotskaya, Y. B., Holla, V. R., Farago, A. F., Mills Shaw, K. R., Meric-Bernstam, F., and Hong, D. S. (2017) Targeting TRK family proteins in cancer. Pharmacol. Ther. 173, 58-66. (31) Yan, W., Lakkaniga, N. R., Carlomagno, F., Santoro, M., McDonald, N. Q., Lv, F., Gunaganti, N., Frett, B., and Li, H. Y. (2019) Insights into current tropomyosin receptor kinase (TRK)

inhibitors:

development

and

clinical

application.

http://pubs.acs.org/doi/10.1021/acs.jmedchem.8b01092. 33 ACS Paragon Plus Environment

J.

Med.

Chem.

62,


(32) Cocco, E., Scaltriti, M., and Drilon, A. (2018) NTRK fusion-positive cancers and TRK inhibitor therapy. Nat. Rev. Clin. Oncol. 15, 731-747. (33) Bhullar, K. S., Lagaron, N. O., McGowan, E. M., Parmar, I., Jha, A., hubbard, B. P., and Rupasinghe, H. P. V. (2018) Kinase-targeted cancer therapies: Progress, challenges and future directions. Mol. Cancer. 17, 48. (34) Roskoski, R. Jr. (2016) Classification of small molecules protein kinase inhibitors based upon the structures of their drug-enzyme complexes. Pharmacol. Res. 103, 26-48. (35) Wu, P., Clausen, M. H., and Nielsen, T. E. (2015) Allosteric small-molecule kinase inhibitors. Pharmacol. Ther. 156, 59-68. (36) Behera, P. M., and Dixit, A. (2017) The story of kinase inhibitors development with special reference to allosteric site. In: Grover A. (eds) Drug design: Principles and Application. Springer, Singapore, 57-68. (37) Bertrand, T., Kothe, M., Liu, J., Dupuy, A., Rak, A., Berne, P. F., Davis, S., Gladysheva, T., Valtre, C., Crenne, J. Y., and Mathieu, M. (2012) The crystal structures of TrkA and TrkB suggest key regions for achieving selective inhibition. J. Mol. Biol. 423, 439-453. (38) Stachel, S. J., Sanders, J. M., Henze, D. A., Rudd, M. T., Su, H. P., Li, Y., Nanda, K. K., Eqbertson, M. S., Manley, P. J., Jones, K. L., Brnardic, E. J., Green, A. Grobler, J. A., Hanney, B., Leitl, M., Lai, M. T., Munshi, V., Murphy, D., Rickert, D., Krasowska-Zoladek, A., Daley, C., Zuck, P., Kane, S. A., and Bilodeau, M. T. (2014) Maximizing diversity from a kinase screen: Identification of novel and selective pan-Trk inhibitors for chronic pain. J. Med. Chem. 57, 5800-5816. (39) Su, H.-P., Rickert, K., Burlein, C., Narayan, K., Bukhtiyarova, M., Hurzy, D. M., Stump, C. A., Zhang, X., Reid, J., Krasowska-Zoladek, A., Tummala, S., Shipman, J. M., Kornienko, M., Lemaire, P. A., Krosky, D., Heller, A., Achab, A., Chamberlin, C., Saradjian, P., Sauvagnat, B., 34 ACS Paragon Plus Environment

Page 34 of 38

Page 35 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Yang, X., Ziebell, M. R., Nickbarg, E., Sanders, J. M., Bilodeau, M. T., Carroll, S. S., Lumb, K. J., Soisson, S. M., Henze, D. A., and Cooke, A. J.

(2017) Structural characterization of

nonactive site, TrkA-selective kinase inhibitors. Proc. Nat. Acad. Sci. 114, E297-E306. (40) Furuya, N., Momose, T., Katsuno, K., Fushimi, N., Muranaka, H., Handa, C., Ozawa, T., and Kinoshita, T. (2017) The juxtamembrane region of TrkA kinase is critical for inhibitor selectivity. Bioorg. Med. Chem. Lett. 27, 1233-1236. (41) Bagal, S. K., Omoto, K., Blakemore, D. C., Bungay, P. J., Bilsland, J. G., Clarke, P. J., Corbett, M. S., Cronin, C. N., Cui, J. J., Dias, R., Flanagan, N. J., Greasley, S. E., Grimley, R., Johnson, E., Fengas, D., Kitching, L., Kraus, M. L., McAlpine, I., Nagata, A., Waldron, G. J., and Warmus, J. S. (2019) Discovery of allosteric, potent, subtype selective, and peripherally restricted TrkA kinase inhibitors. J. Med. Chem. 62, 247-265. (42) Wang, W., Krosky, D., and Ahn, K. (2017) Discovery of inactive conformation-selective kinase inhibitors utilizing cascade assays. Biochemistry, 56, 4449-4456. (43) Wilson, L. J., Linley, A., Hammond, D. E., Hood, F. E., Coulson, J. M., MacEwan, D. J., Ross, S. J., Slupsky, J. R., Smith, P. D., Eyers, P. A., and Prior, I. A. (2018) New perspectives, opportunities, and challenges in exploring the human protein kinome. Cancer Res. 78, 15-29. (44)

Hawkins, P.C.D., Skillman, A.G., and Nicholls, A. (2007) Comparison of shape-matching

and docking as virtual screening tools. J. Med. Chem. 50, 74-82. (45) Lipinski, C. A., Lombardo, F., Dominy, B. W., and Feeney, P. J. (1997) Experimental and computational approaches to estimate solubility and permeability in drug discovery and development setting. Adv. Drug Deliv. Rev. 23, 3-25. (46) Shelley, A., Andrews, S. W., Baer, B., Crane, Z., Liu, W., Watson, D. J., and Malen, P. L. (2015) Preparation of 1-((3S,4R)-4-(3-fluorophenyl)-1-(2-methoxyethyl)pyrrolidine-3-yl)-3-(4-



Page 36 of 38

methyl-3-(2-methylpyrimidin-5-yl)-1-phenyl-5-yl)urea as TrkA kinase inhibitor. PCT Int. Appl. WO2015/175788A1. (47) Wang, T., Lamb, M. L., Scott, D. A., Wang, H., Block, M. H., Lyne, P. D., Lee, J. W., Davies, A. M., Zhang, H. J., Zhu, Y., Gu, F., Han, Y., Wang, B., Mohr, P. J., Kaus, R. J., Josey, J. A., Hoffmann, E., Thress, K., Macintyre, T., Wang, H., Omer, C. A., and Yu, D. (2008) Identification of 4-aminopyrazolylpyrimidines as potent inhibitors of Trk kinases.

J. Med.

Chem. 51, 4672-4684. (48) Forsell, P., Almqvist, H., Hillertz, P., Akerud, T., Otrocka, M., Eisele, L., Sun, K., Andersson, H., Trivedi, S., Wollberg, A. R., Dekker, N., Rottici, D., and Sandberg, K. (2012) The use of TrkA-PathHunter assay in high-throughput screening to identify compounds that affect nerve growth factor signaling. J. Biomol. Screen. 18, 659-669. (49) Martinez Molina, D., Jafari, R., Ignatushchenko, M., Seki, T., Larsson, E. A., Dan, C., Sreekumar, L., Cao, Y., and Nordlund, P. (2013) Monitoring drug target engagement in cells and tissues using the cellular thermal shift assay. Science. 341, 84-87. (50) Bradshaw, R. A., Chalkley, R. J., Biarc, J., and Burlingame, A. L. (2013) Receptor tyrosine kinase signaling mechanisms: Devolving TrkA responses with phosphoproteomics. Adv. Biol. Regul. 53, 87-96. (51) Biarc, J., Chalkley, R. J., Burlingame, A. L., and Bradshaw, R. A. (2012) The induction of serine/threonine protein phosphorylations by a PDGFR/TrkA chimera in stably transfected PC12 cells. Mol. Cell Proteomics. 11, 15-30. (52) Red Brewer, M., Choi, S. H., Alvarado, D., Moravcevic, K., Pozzi, A., Lemmon, M. A., and Carpenter, G. (2009) The juxtamembrane region of the EGF receptor functions as an activation domain. Mol. Cell. 34, 641-651.


Page 37 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(53) Wang, T., Lamb, M. L., Block, M. H., Davies, A. M., Han, Y., Hoffmann, E., Ioannidis, S., Josey, J. A., Liu, Z. Y., Lyne, P. D., MacIntyre, T., Mohr, P. J., Omer, C. A. Sjogren, T., Thress, K., Wang, B., Yu, D., and Zhang, H. J. (2012) Discovery of disubstituted imidazo[4,5b]pyridines and purines as potent TrkA inhibitors. ACS Med. Chem. Lett. 3, 705-709. (54) Drillon, A., Nagasubramanian, R., Blake, J. F., Ku, N., Tuch, B. B., Ebata, K., Smith, S., Lauriault, V., Kolakowski, G. R., Brandhuber, B. J., Larsen, P. D., Bouhana, K. S., Winski, S. L., Hamor, R., Wu, W. I., Parker, A., Morales, T. H., Sullivan, F. X., DeWolf, W. E., Wollenberg, L. A., Gordon, P. R., Douglas-Lindsay, D. N., Scaltriti, M., Benayed, R., Raj, S., Hanusch, B., Schram, A. M., Jonsson, P., Berger, M. F., Hechtman, J. F., Taylor, B. S., Andrews, S., Rothenberg, S. M., and Hyman, D. M. (2017) A next-generation Trk kinase inhibitor overcomes acquired resistance to prior Trk kinase inhibition in patients with Trk fusionpositive solid tumors. Cancer Discov. 7, 963-972. (55) Cocco, E., Scaltriti, M., and Drilon, A. (2018) NTRK fusion-positive cancers and TRK inhibitor therapy. Nat. Rev. Clin. Oncol. 15, 731-747 (56) Biarc, J., Chalkley, R. J., Burlingame, A. L., and Bradshaw, R. A. (2013) Dissecting the roles of tyrosines 490 and 785 of TrkA protein in the induction of downstream protein phosphorylation using chimeric receptors. J. Biol. Chem. 288, 16606-16618.



268x128mm (72 x 72 DPI)

ACS Paragon Plus Environment

Page 38 of 38

Deciphering the allosteric binding mechanism of the human

Recommend Documents