Discovery of Novel Pyrazolo-Pyridone DCN1 Inhibitors Controlling

Subscriber access provided by Macquarie University

Article

Discovery of Novel Pyrazolo-Pyridone DCN1 Inhibitors Controlling Cullin Neddylation Ho Shin Kim, Jared T. Hammill, Daniel C. Scott, Yizhe Chen, Jaeki Min, Jonah Rector, Bhuvanesh Singh, Brenda A. Schulman, and R. Kiplin Guy J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b00410 • Publication Date (Web): 29 Aug 2019 Downloaded from pubs.acs.org on August 30, 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 44 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

Journal of Medicinal Chemistry

Discovery of Novel Pyrazolo-Pyridone DCN1 Inhibitors Controlling Cullin Neddylation

Ho Shin Kim1; Jared T. Hammill1; Daniel C. Scott2; Yizhe Chen1; Jaeki Min3; Jonah Rector1; Bhuvanesh Singh4; Brenda A. Schulman2,5; R. Kiplin Guy1*

1Department

of Pharmaceutical Sciences, University of Kentucky, Lexington, Kentucky 40508,

United States 2Department

of Structural Biology, St. Jude Children’s Research Hospital, Memphis,

Tennessee 38105, United States 3Department

of Chemical Biology and Therapeutics, St. Jude Children's Research Hospital,

Memphis, Tennessee 38105, United States 4Department

of Surgery, Laboratory of Epithelial Cancer Biology, Memorial Sloan Kettering

Cancer Center, New York, New York 10065, United States 5Department

of Molecular Machines and Signaling, Max Planck Institute of Biochemistry,

Martinsried, Germany

*Correspondence: [email protected]

1

ACS Paragon Plus Environment

Journal of Medicinal Chemistry 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

Page 2 of 44

Abstract Chemical control of cullin neddylation is attracting increased attention based largely on the successes of the NEDD8-activating enzyme (E1) inhibitor Pevonedistat. Recently reported chemical probes enable selective and time-dependent inhibition of downstream members of the neddylation tri-enzymatic cascade including the co-E3, DCN1. In this work, we report the optimization of a novel class of small molecule inhibitors of the DCN1-UBE2M interaction. Rational X-ray co-structure enabled optimization afforded a 25-fold improvement in potency relative to the initial screening hit. The potency gains are largely attributed to additional hydrophobic interactions mimicking the N-terminal acetyl group that drives binding of UBE2M to DCN1. The compounds inhibit the protein-protein interaction, block NEDD8 transfer in biochemical assays, engage DCN1 in cells, and selectively reduce the steady-state neddylation of Cul1 and Cul3 in two squamous carcinoma cell lines harboring DCN1 amplification. N-Acetyl Pocket O O HN N N

O O N

Optimization

NH

N N

25 times potency improvement

NH

F New scaffold HIT compound (1) TR-FRET IC50 = 5.1 µM

New optimized compound (27) TR-FRET IC50 = 0.2 µM

Introduction NEDD8 (neural precursor cell expressed developmentally down-regulated protein 8) is a ubiquitin-like protein (UBL) that is post-translationally appended to eukaryotic proteins in a process termed neddylation. A three-step enzymatic cascade carries out neddylation. The first step is the ATP-dependent formation of an E1-NEDD8 thioester intermediate. Next, the NEDD8 is transferred by a transthioesterification reaction from the E1 to a Cys on the E2. E3s then join the E2-NEDD8 complex and recruit specific target proteins. The E3 complex then catalyzes the formation of an isopeptide bond connecting NEDD8’s C-terminus to the ɣ-amino group of a lysine side-chain on the target protein (Figure 1). 2


Page 3 of 44 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 1. General Neddylation Cascade The cullin family of ubiquitination E3s are the most well-characterized substrates of neddylation. Upon neddylation, the cullins constellate the formation of the cullin-RING E3 UB ligases (CRLs), a family of ligases that has approximately 300 members.1, 2 The CRLs regulate diverse biological processes including cell cycle, signal transduction, DNA replication, and viral modulation.3-6 CRL dysfunction is implicated in a number of human diseases, including cancer.7-11 Drug discovery efforts targeting the CRLs and the associated proteasomal protein degradation machinery have been extensive and continue to grow.9,

12-14

The neddylation

pathway has been successfully targeted by MLN4924 (Pevonedistat), an inhibitor of NEDD8’s E1 enzyme, that completely blocks NEDD8 ligation to substrates. MLN4924 is currently being tested in oncology clinical trials.15 An inhibitor of the COP9 signalosome, responsible for deneddylation of the CRLs, also displays anti-tumor activity.16 We have reported the discovery of

inhibitors

of

DCN1

mediated

neddylation.17-19

Peptidomimetics

and

triazolo[1,5‑a]pyrimidine-based inhibitors of the DCN1-UBE2M interaction were also recently reported by others.20-22 All the previously reported inhibitors potently and selectively inhibit the DCN1/2-UBE2M protein interaction in biochemical assays, bind and thermally stabilize DCN1 in cells, and reduce the steady-state levels of cullin neddylation in a variety of cell lines including HCC95, CAL33, KYSE70, H2170, SK-MES-1, and MGC-803 cells.17-22 DCN1 (Defective in Cullin Neddylation 1) is also known as DCUN1D1, DCNL1, or SCCRO (Squamous Cell Carcinoma-related Oncogene); we use “DCN1” hereafter. DCN1 acts as a co-E3 together with RBX1 to stimulate the transfer of NEDD8 from its E2 (UBE2M) to the cullin 3



proteins.23 Humans express five isoforms of DCNs. In vitro, the five DCNs can cooperate with RBX1 to promote NEDD8 ligation from UBE2M or UBE2F to CULs 1-5, or with RBX2 to promote NEDD8 ligation from UBE2F to CULs 1-5.24 In mammals, the genetic redundancy is currently unclear although DCN1 is clearly not essential. Only DCN1 is broadly expressed across all tissues. DCN1 and DCN2 share >80% sequence homology (100% conservation in UBE2M binding pocket) and are the only isoforms expressed in both the cytosol and nucleus. DCN3, 4, and 5 share less than 35% sequence homology with DCN1 and appear to have more distinct tissue and cellular distributions.24-28 The high sequence homology, overlapping expression patterns, and similar subcellular distributions suggest that DCN1 and 2 may be redundant, while DCNs 3, 4, and 5 likely possess unique roles. DCN1 is the most well characterized isoform due to its common amplification as part of a large 3q26.3 amplicon in squamous cell carcinomas (SCC) and other tumors.29 DCN1 amplification in SCC negatively correlates with cause-specific survival suggesting that targeting DCN1 may be of clinical utility.28-35 Although other important cancer gene(s) may be present within the amplicon, several studies have demonstrated that DCN1 plays a critical role in tumor progression and metastasis, and drives selection for 3q amplification in SCC.32, 36-39 Recently, DCN1 has been reported to have a driving role in prostate cancer and depletion of DCN1 in LNCaP cells significantly reduced their proliferation, migration, and invasion capability.40 Similar results have been reported in cervical cancers.41 DCN1’s emerging roles in these diseases suggest it may present a new oncology drug discovery target.42, 43 A druggable pocket in DCN1 was initially revealed by crystallographic and biochemical studies showing how DCN1’s activation of the NEDD8 pathway requires binding to UBE2M. This interaction is driven by binding of UBE2M’s N-terminal acetyl-methionine and immediately proximal amino acid side chains to a ≈350 Å3 hydrophobic pocket on DCN1.19, 20, 23, 27 This interaction depends on the N-terminal acetylation of UBE2M, with the equilibrium dissociation constant for the N-terminally acetylated protein being >100-fold more potent relative to the non-acetylated protein. 4


Page 4 of 44

Page 5 of 44 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


We recently reported the optimization of a series of N-benzyl piperidines that occupy the UBE2M-binding pocket on DCN1.17, 18 While this series of compounds, represented by NAcMOPT (Figure 1), were the first reported small molecule inhibitors of DCN1, they suffered several limitations: 1) they do not access the N-acetyl pocket that controls the binding of the native substrate;27 2) they possess only one stereocenter, thus lacking three-dimensional character affording access the available sub-pockets within the binding pocket, and 3) they have a moderate murine half-life that requires relatively high and frequent dosing to maintain relevant concentrations in mouse models. A common challenge associated with developing new classes of chemical probes is determination of whether elicited cellular phenotypes are caused by engagement of the desired target or by non-specific interaction(s) driven by the inhibitor’s core structure. Although biological tools (CRISPR, siRNA, etc.) offer some insights, reduction of target protein levels and acute pharmacologic inhibition of specific protein interactions may have different consequences. Therefore, we sought a structurally unrelated class of small molecule inhibitors that might overcome the inherent liabilities of the piperidine class. This work focuses on a second class of inhibitors that contain a pyrazolo-pyridone core. The original lead was discovered during our initial high-throughout campaign. This class has two chiral centers thus imparting a greater degree of three-dimensional structure and affording opportunities for increased binding potency, target selectivity, and improved solubility. Our overall strategy was to first identify the minimum pharmacophore for this class, next define the structural drivers of potency for binding to DCN1, and then use data gathered about optimal substituents from the first-generation inhibitors to optimize binding. Here, new compounds were designed based on hypotheses generated from combining empirically derived SAR and examination of X-ray costructures. New analogs were tested for potency using our previously reported TR-FRET assay (vide infra). Key compounds were evaluated for cellular target engagement and neddylation activity by immunoblotting for steady-state levels of cullin neddylation using two cancer cell lines with amplified DCN1 expression (HCC95 and CAL33). Ultimately, these 5



Page 6 of 44

studies identified compound 27, which is 25-fold more potent in our biochemical TR-FRET assay than the original hit, thermally stabilizes DCN1 in cells, and selectively reduces steadystate levels of CUL1 and CUL3 neddylation in HCC95 and CAL33 cells. Results We reported19 the results of our high-throughput screen of roughly 600,000 compounds using a binding assay based on the TR-FRET signal between a biotinylated DCN1 protein, recognized by Terbium-linked streptavidin, and the helical stapled peptide derived from Nterminally acetylated UBE2M harboring a C-terminal AlexaFluor 488. This work identified several structurally divergent inhibitors of the targeted binding event with potency (IC50) ranging from 1 to 30 micromolar.19 Prior manuscripts discussed the optimization of the hit series leading to NAcM-OPT. This manuscript focuses on optimization of a series of pyrazolopyridones, represented by compound 1, discovered during our screen. Resynthesis of several analogs validated the potency of the pyrazolo-pyridone screening samples (Figure 2, compound 1). Compound 1 also inhibited the transfer of NEDD8 from the activated E2 to its cullin substrate in our pulse-chase assay23, 27 (Supplementary Figure C). An X-ray co-structure of the compound 1 with DCN1 showed that it bound to the targeted UBE2M binding pocket (Figure 2).

Cl O O HN N N

Gln114

Cl N

HN

NH

Compound 1 TR-FRET IC50 = 5.1  0.6 µM

N

O

NAcM-OPT TR-FRET IC50 = 0.3  0.1 µM

Figure 2. Left: Overlay of 1(orange):DCN1(gray) and NAcM-OPT(blue):DCN1 X-ray crystal co-structures (PDB 6P5W and 5V86), highlighting that 1 more efficiently occupies the N-acetyl pocket (yellow highlight). Hashed orange line represents a potential hydrogen-bonding 6


Page 7 of 44 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


interaction between the amide of 1 and backbone amide of Gln114. Right: 2D chemical structures of compound 1 and NAcM-OPT. Design and synthesis Based on validation of the screening hit, we embarked on a hit-to-lead campaign aimed at delivering an early lead that overcame the challenges associated with our first-generation inhibitors embodied by NAcM-OPT. Our initial compound design was based on insights derived from overlaying the X-ray co-structures of 1:DCN1 (PDB 6P5W), UBE2MNAc:DCN1 (PDB 5V83), and NAcM-OPT:DCN1 (PDB 5V86) (Figure 2) and is summarized in Figure 3. Compound 1 aligns well with NAcM-OPT in addressing three binding sub-pockets: the Ile, Leu, and hinge pockets. Therefore, initial substitutions in these pockets were informed by the structure-activity relationships (SAR) derived during the optimization of first-generation lead NAcM-OPT.17, 18 NAcM-OPT does not directly occupy the N-acetyl pocket (yellow highlight, Figure 2). However, its piperidine ring is engaged in a putative hydrogen bonding network with a bound water molecule within the N-acetyl pocket that strongly suggests fulfilling this pocket could strengthen binding. This hypothesis is further supported by characterization of analogs of the native binding peptide, which demonstrated that occupancy by even a small hydrophobic group (acetyl vs. formyl) can enhance affinity 10-fold.27 The X-ray co-structure of 1:DCN1 shows that compound 1 positions the central pyrazolo-pyridone ring closer to the Nacetyl pocket (Figure 2). Thus, we designed analogs to test whether substitution on the pyridone’s nitrogen could induce tighter binding by direct hydrophobic packing rather than indirect water-mediated binding (Figure 3). In addition, we explored the base pharmacophore for the pyrazolo-pyridone series by utilizing reverse amides in the hinge pocket to determine the significance of a predicted hydrogen bond between compound 1’s amide and Gln114 of DCN1 (Figures 2, 3). Compound 1 has four aromatic rings that could negatively affect its physiochemical properties and reduce its utility as a cellular probe to study the cullin neddylation.44 To address this issue, we tested whether smaller, more flexible substituents could replace these aryl rings. We also tried a variety of substituent modifications on the aryl 7



Page 8 of 44

ring to introduce additional electronic or hydrophobic interactions (Figure 3). The overall goal was to increase or maintain potency by targeted substitution of one ring while reducing total molecular weight by removing a less critical ring or substituent.

[Design] [N-Acetyl Pocket] • Mimic N-Acetyl

N-Acetyl Ile

O O HN N N

[Ile & Leu Pocket] •Reduce # of Aryl rings •Refine Hydrophobic interactions •Add electronic interactions

Hinge

NH [Hinge Pocket] •Explore induced pocket; multiple substitiution, swap amide bond

Leu

Figure 3. Design strategy and X-ray co-structure overlay of compound 1:DCN1 (PDB 6P5W) and UBE2M:DCN1 (PDB 5V83), highlighting the key regions targeted for optimization. Compounds were synthesized by a short and efficient three-step procedure consisting of preparation of the oxazolone intermediate45, pyrazolo-pyridone ring formation46, and substitution using alkylation or acylation (Scheme 1). Formation of the core pyrazolo-pyridone ring afforded a separable mixture of cis and trans diastereomers. Typically, the cis product appeared as the less polar spot by TLC (hexane/ethyl acetate). The cis and trans isomers were distinguished by 1H NMR and were assigned based on application of the Karplus equation to the 3JH-H vicinal proton-proton coupling for the C4 or C5 protons of pyrazolopyridone ring (Cis = 7-8 Hz, Trans = 9-11 Hz). The inactive trans-isomer could be converted to the active cis product by stirring with a catalytic amount of Lewis acid (SnCl2) in refluxing chlorobenzene. However, use of a large excess of Lewis acid (10+ equivalents) favors formation of dehydrated side products rather than the pyrazolo-pyridone ring. Alkyl substitution of the pyrazolo-pyridone amide proceeds under relatively mild conditions presumably due to the presence of adjacent electron-withdrawing groups that increase amide acidity. Most alkylations afforded a 5:1 mixture of N- vs. O-alkylation as determined by 2D HMBC NMR analysis (Supplementary Figure A). Finally, a series of reverse amides were synthesized from dimethyl malonate through an alternate route: condensation, hydrolysis, and 8


Page 9 of 44 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


amide coupling (Supplementary Scheme A).

O O HO

N H

O

R3

R4 N N N

H R1

O O 



C4

NH C5 R2

R2

a or b

O

R2

R4 O O

R3 R3

N N N

R3

O N

R1

NH2 c

N N

O O R3

R3

NH d or e R2

HN N N

R1

NH R2

Scheme 1. General synthesis of pyrazolo-pyridones. Reagents and conditions: (a) NaOAc, Ac2O, 85 - 90 oC, 1 - 2 h; (b) i) EDCI·HCl, DCM, r.t., 16 h; ii) Al2O3, 4Å MS, DCM, r.t., 16 h; (c) SnCl2, chlorobenzene, reflux, 16 h; (d) Cs2CO3, R-X (X= Br, I), DMF, r.t., 1 - 16 h; (e) AcCl, pyridine, DCM, r.t., 1h.

The key annulation reaction (Scheme 1, step c) was previously reported using a mixture of ethylene glycol and acetic acid as the solvent.46 In our hands, these conditions afforded low yields and complex mixtures, particularly formation of undesired ethylene glycol adducts. Switching from a nucleophilic polar protic solvent to a non-polar or polar aprotic solvent such as chlorobenzene, DMF, or NMP improved yields and suppressed formation of side products. As discussed below, the initial SAR showed that the cis-diastereomers were active and the trans-diastereomers inactive. Therefore, we optimized the reaction to favor formation of the cis-diastereomer. Addition of a catalytic amount of tin (II) chloride followed by refluxing in chlorobenzene overnight proved the most selective, affording a 3:1 cis to trans ratio, and moderate isolated yields of the pure cis product (30 - 40%).

Structure−Activity Relationships. The main goal of the first phase of the work was to define and optimize the core pharmacophore by testing requirements for key rings and functional groups. The secondary 9



goal was to define the best range of pendant substituents occupying the key sub-pockets. Our overall optimization strategy uses a standard recursive process of sequential hypothesisbased design of analog sets and testing of those sets in models of increasing complexity to define possible structure-activity relationships (SAR) that drive further analog design. All novel analogs were prepared and tested as racemates. Potency was assessed using our TR-FRET binding assay. We targeted four regions of the compounds, named according to the peptide ligand feature that they replace within the ligand-binding pocket, to systematically delineate the SAR: the Ile, N-acetyl, Leu, and hinge pockets (Figure 3). During this study, we prepared and tested over 140 analogs. However, to improve clarity and readability of the manuscript, only key compounds, which clearly delineate SAR, are reported in the main text. A complete list of all compounds prepared and evaluated during this study and supporting noted SAR trends can be found in the Supplementary Tables. First, we prepared roughly 30 compounds to test the inhibitory effects of the cis- and transdiastereomers about the core pyrazolo-pyridone ring. In general, only the cis diastereomers inhibited DCN1-UBE2M binding at or below the maximum tested concentration of 15 µM. All the trans isomers were inactive at the maximum tested concentration of 15 µM (Supplementary Table A). These results demonstrate the three-dimensional orientation of two aromatic rings is critical to activity. Therefore, subsequent work utilized only the purified cis-diastereomers and all data reported below represent that set of compounds.

10


Page 10 of 44

Page 11 of 44 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


O O HN

O O HN

N N

NH

N N

R

R

NH

F IC50 (M) (TR-FRET)

No

1

5.1 ± 0.6

10

13 ± 4

2

4.0 ± 0.5

3

1.4 ± 0.2

3

1.4 ± 0.2

11

No

R

IC50 (M) (TR-FRET)

R

3.5 ± 0.5 CF3

F 4

F

5

12

7.1 ± 2.6

I 3.5 ± 0.2

N

7.1 ± 0.9

Cl

13

>15

Cl

F >15

14

>15

7

3.0 ± 0.3

15

>15

8

3.1 ± 0.2

16

>15

9

1.3 ± 0.1

17

>15

6 N

N

Table 1. SAR of Leu and hinge pocket. IC50 values were generated using our TR-FRET binding assay and are represented as the mean of three replicates with errors reported as the standard deviation.

Second, we examined the steric and electronic requirements of the Leu pocket (Table 1). Introduction of a p-fluorine (3) increased potency relative to the unsubstituted benzene ring 11



(2). A survey of other p-halogen substitutions (Cl or Br) or the p-methyl substitution (1) revealed that they were less potent than the p-fluorine (3) analog (Supplementary Table B1). Introduction of either the p-cyano (42) or p-methoxy (43) group completely abolished activity (Supplementary Table B-1). Shifting the fluorine to the m-position of the phenyl ring (4) reduced potency by 5-fold. However, the m-, p-difluoro analog (38) retained the activity of 3. Together, these results suggest that the p-fluorine substitution is critical for binding and other small electron-withdrawing substituents, such as cyano, are not viable replacements. This trend was reinforced by the inactive isosteric pyridine analog (6), whose activity could be partially rescued by addition of a p-fluoro substituent (5). We also investigated the effects of more flexible substituents. Replacement of the phenyl ring with a benzyl group (7) retained the potency of the unsubstituted phenyl analog (2). Replacement of the aryl ring with a variety of linear (propyl or butyl) or cyclic (cyclohexyl and cyclopentyl) alkyl groups was well tolerated (8, 9, and 44-46, Supplementary Table B-1). However, introduction of the shorter, isopropyl substituent (47) removed activity. The data suggest that occupation of the hydrophobic Leu pocket is a critical driver of potency but that the pocket is relatively forgiving, within the constraints of a fixed steric bulk. Thus, the Leu pocket affords a site that could be further manipulated to modulate the physiochemical properties of the inhibitors. Due to its superior potency and inert chemical nature, the remainder of our SAR study fixed the substituent targeting the Leu pocket as the p-F–phenyl. Surveying possible phenyl ring substituents for the hinge pocket (Table 1, Supplementary Table B-2) revealed that m-substitution of the phenyl ring was critical for potency. Removal of the m-methyl from the phenyl ring (10) reduced potency by 10-fold. In addition, changing the position of methyl to either the p- (14) or o- (15) position completely abrogated activity. The introduction of bulky m-substituents, such as tert-butyl (17), was not favorable. Replacement of m-CH3 with the less metabolically labile m-CF3 group (11) resulted in a two-fold reduction in potency. Replacement with smaller halogens (F (53), Cl (54)) was not tolerated but larger, roughly isosteric, halogens (Br (55) and I (12)) retained some activity (Supplementary Table 12


Page 12 of 44

Page 13 of 44 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


B-2). This sharp SAR about the hinge pocket is consistent with the SAR studies of NAcMOPT.17, 18 Similar to the binding of NAcM-OPT, the X-ray co-structure of 1:DCN1 revealed that the hinge pocket undergoes a structural rearrangement to produce a deep narrow hydrophobic cavity composed of alkyl and aromatic residues (Ile86, Phe89, Val102, Ile105, Ala106, Phe109, Ala111, Phe117, Phe122, Phe164) that tightly interacts with the phenyl ring. Rather unexpectedly, the 3,4-dichloro phenyl analog (13), which mimics NAcM-OPT, was completely devoid of activity. This result suggests that the pyrazolo-pyridone analog does not present the aryl ring to the induced hinge pocket in the same way that NAcM-OPT does. Examination of the X-ray co-structures revealed that both series have a key H-bond with the amide backbone of Gln114 on DCN1, which lock the compounds in a specific orientation. However, while NAcM-OPT participates as an H-bond donor through its urea, compound 1 interacts as an H-bond acceptor through the oxygen of its amide. The highly restrictive steric and electronic requirements of this pocket are further illustrated by the complete loss of potency upon introduction of either a m-F-phenyl (57) or a pyridine replacement (16). Converting the phenyl group to an aliphatic butyl chain (60) also results in loss of activity (Supplementary Table B-2). Taken together, the SAR reveals that the hinge pocket requires a phenyl ring with a small hydrophobic substituent, most efficiently met by a m-methyl substituted phenyl ring. Finally, we tested whether the reverse amide could maintain the key hydrogen bonding interaction with Gln114 or switch compound 1 from an H-bond acceptor to an H-bond donor, as observed with NAcM-OPT. Investigating five analogs, including the m-methyl (128) and m-, p-dichloro phenyl (131) derivatives, we found that the reverse amide was not tolerated (Supplementary figure D-1). Further optimization studies fixed the substituent targeting the Leu pocket as the p-F–phenyl and the substituent targeting the hinge pocket as m-tolyl.

13



O O R

HN

R2 O O N R1 NH N N

NH

N N

Page 14 of 44

F

F IC50 (M) (TR-FRET)

IC50 (M) (TR-FRET)

No

3

1.4 ± 0.2

27

0.2 ± 0.1

18

>15

28

4.1 ± 0.8

>15

29

>15

30

21

1.2 ± 0.2

31

0.2 ± 0.1

22-a, b

0.7 ± 0.1 (TLC-up) 0.6 ± 0.1 (TLC-down)

32-a, b

1.2 ± 0.1 (TLC-up) 0.5 ± 0.2 (TLC-down)

23

0.5 ± 0.1

33

2.3 ± 0.8

24

0.3 ± 0.1

34

0.8 ± 0.1

25

2.0 ± 0.1

35

0.2 ± 0.1

No

19

R

N

20

H

R1

R2

O

7.1 ± 1.7 O

5.6 ± 0.8

Table 2. SAR of Ile and N-acetyl pocket. IC50 values were generated using our TR-FRET binding assay and are represented as the mean of three replicates with errors reported as the standard deviation.

Next, we explored the SAR of substituents reaching into the Ile pocket (Table 2, Supplementary Table B-3). Introduction of a methylene between the phenyl and pyrazole rings (18) reduced activity 10-fold. Thus, there is a limitation on the size of the substituent 14


Page 15 of 44 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


targeting this pocket. Introduction of a fluoro substitution at either the o- (61) or p- (62) positions had little effect on potency (Supplementary Table B-3). Since one of our design goals was to reduce the aryl ring count, we investigated replacements of the phenyl ring. Replacements of aryl ring with aliphatic chain were successful in many cases (22 – 25). Linear alkyl chains three (21) or two (65) carbons in length proved roughly equipotent to the phenyl derivative (3). Introduction of a linear four-carbon chain (63) reduced potency three-fold. The unsubstituted (20) and methylated (66) pyrazoles were inactive (Supplementary Table B-3). Introduction of substituted alkyl chains (22, 23, and 24), designed to mimic the Ile residue displayed two - five times better potency than the phenyl derivative (3). Introduction of a methyl group alpha to the pyrazole afforded an additional chiral center and the resulting diastereomers were separable by silica gel chromatography. In the case of the 2-butyl derivative, the diastereomers proved roughly equally potent (22a, b). However, in the cases of the 1-cyclopropyl-2-propane (70a, b) and 1,1,1-trifluoro-2-propane (71a, b) derivatives, one isomer was two to three times more potent. The cyclopropyl compound (25) was roughly equipotent to the phenyl (3). However, the cyclopentyl (24) and cyclohexyl (72) derivatives were two- and three-fold more potent respectively. The replacement of the phenyl ring with a polar pyridine (19) or addition of a polar atom to the linear chain (26) significantly decreased potency. Thus, demonstrating a strong preference for hydrophobic substituents in the Ile pocket. We also examined the effects of substituting the other pyrazole nitrogen. All such analogs were inactive at the maximum tested concentration of 15 µM (Supplementary figure D-1). Therefore, the alignment of pyrazolo-pyridone scaffold’s amide bond and pyrazole’s substituent are both crucial for binding. To understand requirements for binding to the N-acetyl pocket, we investigated whether substitution of the pyridone ring nitrogen (Table 2, Supplementary Table B-4) could mimic the N-terminal acetyl of the native UBE2M substrate. Alkyl substitution significantly increased potency with a relative order of no-substituent < propyl < methyl < ethyl (Supplementary Table B-4). In addition to efficiently occupying the N-acetyl pocket, alkyl substitution of the 15



pyridone improves peripheral hydrophobic interactions by pushing the p-F-phenyl substituent deeper into the Leu pocket (Figure 7). The five- to ten-fold observed improvement in potency afforded by substitution of the pyridone ring is consistent with previous binding studies of the native substrate. Careful characterization of the UBE2M-DCN1 interaction showed that Nterminal acetylation of UBE2M enhances its binding affinity for DCN1 ten-times relative to Nterminal formylation.27 Thus, occupancy of the N-acetyl binding pocket on DCN1 by even a small hydrophobic group can significantly enhance binding. Interestingly, introduction of acyl groups (29, 30), directly mimicking UBE2M’s N-terminal acetyl, gave compounds with poor potency. The compound reorientation, induced by substituting the pyridone, subtly effected the SAR about the Ile pocket. Linear alkyl and aryl Ile substituents remained roughly equally potent. However, sterically hindered alkyl Ile substituents (32 - 34) showed decreased potency relative to the unmodified pyridone. These results imply a steric interaction between the two adjacent substituents at the Ile and the N-acetyl pocket that induce an unfavorable change in their orientation. The cyclopropyl group (35) was an exception to this trend and increased potency about ten-fold relative to its parent (25). O-alkylated isomers were also prepared but proved inactive at the maximum tested concentration of 15 µM (Supplementary Table C).

Biochemical and cellular profiling of compound 27 As discussed above, DCN1 is one of the five human paralogues (DCN1-5), which display high structural similarity. In biochemical assays, all five isoforms can stimulate cullin neddylation.24, 42 The selectivity of 27 was assessed using a pulse-chase assay to determine the effects on NEDD8 transfer stimulated independently by each of the five DCN isoforms. At the concentration employed for our cellular studies (10 µM), compound 27 was highly selective, inhibiting DCN1 and DCN2, which are 100% identical in the N-Ac-Met binding pocket. DCN3, DCN4, or DCN5 stimulated cullin neddylation were not inhibited at 10 µM (Figure 4). The observed selectivity can be explained by subtle differences between the binding pockets on 16


Page 16 of 44

Page 17 of 44 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


DCN1/2 and DCNs 3-5.19 The high specificity of 27 amongst the highly homologous family of enzymes implies it is likely to be selective against other less-related binding pockets, as has been demonstrated for NAcM-OPT.19

Figure 4. The DCN isoform selectivity of compound 27 and the known DCN1/2 selective inhibitor NAcM-OPT (positive control). Pulse-chase assay monitoring transfer of NEDD8 from UBE2M to CUL2 in the presence of 10 µM inhibitor or DMSO as tested against all five of the DCN family members. Note different reaction times, based on differences in stimulation by the different DCN family members. “N8” refers to NEDD8, and “~” refers to covalent complex. A subset of inhibitors was tested for kinetic aqueous solubility (pH = 7.4 buffer). The solubility of the compounds ranged from poor (≤ 1 µM) to good (> 60 µM) (Table 3). Notably, switching the aryl substituent targeting the Ile pocket to an aliphatic group significantly improved solubility (3 vs. 65, Table 3). Next, the compounds were evaluated for oxidative metabolic stability using murine microsomes. For microsomal modeling, we chose a drug concentration of 0.8 µM, which is routinely achieved in plasma by small molecule inhibitors and is roughly comparable to the biochemical EC95 of our more potent compounds. Analogs with alkyl substitution of the pyrazole ring were more rapidly metabolized than their phenyl substituted counterparts (27 vs. 31, Table 3). Compound 3 and compound 27 showed roughly comparable (within 2-fold) murine microsomal stability to that previously reported for NAcMOPT. Based on the combination of potency, solubility, and microsomal stability, compound 27 was selected as the top compound for further cellular characterization. Given the goal of this work was to generate a novel class of cell active chemical probes, 17



we investigated the top compounds’ activity in cell-based studies. First, we tested compound effects on monolayer proliferation using a non-transformed fibroblast (BJ) cell line lacking DCN1 amplification. At the maximum tested concentration of 25 µM, none of the compounds inhibited cell growth (Table 3). Next, we assessed growth inhibition using a squamous cell carcinoma cell line (HCC95) known to express high levels of the DCN1 protein. Again, the compounds did not inhibit cell growth at 25 µM (Table 3).

18


Page 18 of 44

Page 19 of 44 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


No

Structure

R1

R2 R1

Proliferation EC50 (M) (Normal Fibroblasts - BJ) (HCC 95)

R2

t(1/2) /hr (mouse)

CLint' (ml/min/kg) (mouse)

Avg. Sol. (µM) (Kinetic)

O O

N

NH

N N

F 3

H

27

65

H

31

24

H

35

123

N

N N

O

> 25

> 25

1.1 ± 0.1

50.3 ± 5.0

5.5 ± 1.0

> 25

> 25

1.2 ± 0.1

44.9 ± 3.7

16.6 ± 0.6

> 25

> 25

0.6 ± 0.1

89.7 ± 4.7

92.8 ± 2.7

> 25

> 25

25

> 25

0.5 ± 0.1

116.4 ± 10.7

17.1 ± 1.4

> 25

> 25

25

> 25

ND

ND

ND

> 25

> 25

0.3 ± 0.1

160.4 ± 5.0

1.1 ± 1.0

> 25

> 25

ND

ND

ND

> 25

> 25

ND

ND

ND

O

N H

F O O HN

NH

N N

1

O O HN

NH

N N

87

F

141

N N

N

O N H

O

F

Table 3. Kinetic solubility, microsomal stability, and cell proliferation effects in normal BJ fibroblasts and malignant HCC95 cells. Values for the proliferation, half-life, clearance and solubility are represented as means plus or minus standard deviation calculated from at least one independent experiment, run in triplicate. ND indicates values were not determined. A key feature of any chemical probe is that it must engage its intended target in cells. We used the cellular thermal shift assay (CETSA)47 to demonstrate that 27 engages DCN1 in 19



HCC95 cells. Figure 5 demonstrates that the DCN1 protein is unstable at 52 °C in HCC95 cells treated with DMSO. However, in the presence of 10 µM of 27 or the known DCN1/2 inhibitor NAcM-OPT, the thermal stability of the DCN1 protein is enhanced. Compound 27 appears to more strongly enhance the thermal stability of DCN1 at 52 °C relative to NAcMOPT (Figure 5). This suggests that compound 27 effectively engages DCN1 in HCC95 cells.

Figure 5. Enhancement of DCN1 thermal stability by compounds 27 and NAcM-OPT (positive control) but not by DMSO (negative control). HCC95 cells were treated with either DMSO or 10 μM of the indicated compound for 1 h, heated at the indicated temperature for 3 minutes, lysed, and blotted with the indicated antibodies.

To confirm the cellular effects on neddylation, compound 27 and an inactive trans-isomer (87) were tested for inhibition of steady-state cullin neddylation as detected by immunoblot analysis. HCC95 and CAL33, two squamous cancer cell lines highly expressing DCN1 and reported to be sensitive to DCN1 inhibition,19 were chosen for analysis (Figure 6). As previously observed for other DCN1/2 selective inhibitors,17-21 compound 27 and NAcM-OPT but not the negative control (87) nor DMSO selectively inhibited steady-state cullin neddylation. In both cell lines, the most pronounced effects were observed on the neddylation of CUL1 and CUL3 but more subtle effects were observed for CUL4A. In concordance with our previous 20


Page 20 of 44

Page 21 of 44 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


studies,19 while complete blockade of neddylation by MLN4924 elicits similar cellular responses across cell lines, acute pharmacologic inhibition of DCN1-mediated neddylation show differential effects. In CAL33 cells, both NAcM-OPT and 27 display more pronounced, albeit weak, inhibition effects on CUL2 neddylation compared to HCC95. Conversely, the marginal effects on CUL5 neddylation are more robust for both compounds in HCC95 relative to CAL33. Unexpectedly, there appears to be subtle differences in the potency of NAcM-OPT and 27 in the two cell lines. In HCC95, 27 appears to more potently inhibit cullin neddylation than NAcM-OPT. In direct contrast, in CAL33, NAcM-OPT appears more potent than 27.

Consistent with all known small molecule inhibitors of the DCN1-UBE2M interaction,17-21 acute pharmacologic inhibition of DCN1-mediated neddylation results in a smaller reduction in neddylation compared to inhibition of the neddylation activating enzyme (E1) by MLN4924 or other NAE inhibitors. The difference in activity results from MLN4924’s irreversible inhibition of the sole E1 enzyme in the neddylation cascade. Thus, no NEDD8 transfer to the E2 and subsequently to the target protein (i.e. cullins) can occur. Alternatively, the residual neddylation activity observed for the DCN1 inhibitors is caused by DCN1-independent neddylation and likely represents the basal activity of RBX1, the other co-E3 enzyme that works with DCN1 to stimulate neddylation.19 The most stark phenotypic consequence of these different mechanisms of action, is the effect on gross substrate stabilization. Therefore, with short exposures of western blots that enable direct comparison to MLN4924, small molecule inhibition of the DCN1-UBE2M interaction does not elicit gross substrate stabilization (S.E. in Figure 6). Longer exposures, which result in saturation of the signal for substrates with MLN4924 treatment, permit the visualization of modest steady-state stabilization of some canonical CUL1/CUL3 substrates. HCC95 cells appear more responsive and subtle stabilization of several known CUL1/CUL3 substrates (p27, p21, cyclin E, and NRF2) can be detected at longer exposures (L.E. in Figure 6). Alternatively, in CAL33, even at longer exposures, only p27 and NRF2 stabilization is observed. 21



The observed stabilization of NRF2 could also be explained in a DCN1 independent manner if the tested compounds induced reactive oxygen species (ROS). Thus, ROS induction was measured for each compound in dose-response using HCC95 cells and the cell permeant reagent 2’,7’–dichlorofluorescin diacetate (DCFDA, also known as H2DCFDA and as DCFH-DA), a fluorogenic dye that measures hydroxyl, peroxyl, and other ROS activity. Even at 30 µM, neither 27, 87, nor NAcM-OPT have any measurable effects on reactive oxygen species suggesting the observed effects on NRF2 is not induced by ROS (Supplementary Figure D).

22


Page 22 of 44

Page 23 of 44 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 6. Western blot for inhibition of cellular neddylation at 48 hours by DMSO, MLN4924 23



(positive control, 1 µM), 87 (negative control, 10 µM), 27 (10 µM), and NAcM-OPT (10 µM) as indicated by steady-state levels of CUL1~NEDD8. “N8” refers to NEDD8, and “~” refers to covalent complex.

Discussion and Conclusions In conclusion, a series of pyrazolo-pyridone analogs were investigated as novel inhibitors of the DCN1-UBE2M interaction and cullin neddylation. The efficient synthetic routes outlined in Scheme 1 enabled rapid access to analogs to define the SAR and permitted gram scale preparation of 27. Comparative analysis of multiple X-ray co-crystal structures enabled rational drug design to access the N-acetyl pocket, providing a 10-fold boost in potency. Empirical systematic investigations of the structure-activity relationships led to a refinement of peripheral interactions. Together, these studies yielded compound 27, which is with 25-fold more potent than the HTS hit compound (1). The SAR studies revealed four critical drivers for potency (Figure 7): (1) the stereochemistry of the pyrazolo-pyridones, with the cis isomer being much more active than the trans isomer; (2) filling the N-acetyl binding pocket, with the ethyl substitution on pyrazolo-pyridone ring increases potency 5 – 10 fold; (3) replacement of either aryl ring targeting the Ile or Leu pocket with saturated alkyl chains does not compromise potency; and (4) the hinge pocket affords the sharpest SAR and even subtle electronic or steric deviation from the meta-substituted aryl significantly compromise potency. A highresolution X-ray co-crystal structure of compound 27 bound to DCN1 confirmed the binding mode and our SAR hypotheses that the ethyl group on pyrazolo-pyridone ring efficiently fills the N-acetyl pocket and pushes the compound deeper into the Leu pocket (Figure 7).

24


Page 24 of 44

Page 25 of 44 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


[SAR summary] [N-Acetyl Pocket] 5-10 times potency improvement Best: Ethyl O O N N N Aryl  Alkyl >> H Hydrophobic

NH Tight induced fit Best: m-Tolyl

Tyr 181

F Aryl  Alkyl Best: p-Fluorophenyl

Gln114

Figure 7. SAR summary and X-ray crystal structure of compound 27(yellow):DCN1(gray) (PDB 6P5V). The ethyl substituent is represented as space filling spheres to highlight the key driver of potency. Hashed orange line represents the 2.6 Å hydrogen-bonding interaction between the compound’s hinge amide and backbone amide of DCN1 Gln114. DCN1 Tyr181 is shown as sticks to highlight the hydrophobic interaction and potential electrostatic interaction. A subset of compounds was profiled for kinetic aqueous solubility (pH = 7.4 buffer), murine microsomal stability, and their effects on the cellular proliferation of both normal (BJ) and transformed (HCC95) mammalian cell lines (Table 3). The lack of growth toxicity and range of solubility and stability values observed for a variety of biochemically potent analogs suggest that future rounds of optimization are likely to improve in vitro pharmacokinetic characteristics. Together, these studies suggested compound 27 had the best combination of properties and thus was selected for further profiling. Compound 27 binds to the targeted pocket on DCN1 (Figure 7) and potently and selectively inhibits DCN1/2 mediated cullin neddylation in biochemical assays (Table 2 and Figure 4). Cell-based studies demonstrated that compound 27 engages cellular DCN1 (Figure 5) and selectively reduces the level of steady-state CUL1 and CUL3 neddylation in HCC95 and CAL33 cells (Figure 6). As seen previously, acute pharmacologic inhibition of the DCN1-UBE2M interaction has different consequences from complete blockade of the NEDD8 pathway by MLN4924 and does not grossly stabilize canonical CRL substrates. 25



In HCC95 cells, compound 27 appears to better enhance the thermal stability of DCN1 relative to NAcM-OPT (Figure 5), which translates to more potent inhibition of cellular neddylation. Thus, supporting our original design hypothesis that fulfilling the N-acetyl pocket could afford more potent inhibitors of cullin neddylation in HCC95 cells. Overall, 27 overcomes two of the three noted limitations of NAcM-OPT by efficiently occupying the N-acetyl pocket and incorporating additional three-dimensional character. The promising preliminary ADMETox data suggests future optimization studies, with a focus on metabolic stability, may yield compounds suitable for single daily oral dosing in murine models and overcome the final liability of NAcM-OPT. Perhaps most intriguingly, the increase in potency observed in HCC95 cells does not translate to increased potency relative to NAcM-OPT in CAL33 cells. These results demonstrate that two different chemical classes with similar biochemical potencies can exhibit differential, cell line specific, effects on steady-state cullin neddylation. It also further supports the hypothesis that having multiple structural classes of inhibitors blocking the DCN1-UBE2M interaction is crucial to permit interrogation of the function of DCN1-mediated neddylation. Currently, it is unknown when neddylation of a particular cullin would require DCN1/2 activity or why different chemical classes of inhibitors would elicit selective cellular effects. A broad range of factors could be responsible for dictating this selectivity such as expression levels and cellular sub-localization of DCN1 and the other components of the neddylation pathway. In the long-term, developing multiple chemical scaffolds of DCN1 inhibitors may prove clinically relevant for different sub-types of cancers. Alternatively, combinations of various chemotypes of DCN1 inhibitors may afford simultaneous access to different subcellular pools of DCN1, effecting a broader spectrum of activity across multiple cancers. The emerging role of DCN1 in multiple tumor types and in vitro data for the DCN1 inhibitors suggests that they may offer an orthogonal, more selective, means to inhibit the NEDD8 pathway relative to the clinically validated NAE inhibitor MLN4924. Future studies, outside the scope of this work, will focus on elucidating the underlying mechanisms responsible for when 26


Page 26 of 44

Page 27 of 44 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


cellular neddylation of a particular cullin would involve DCN1 activity and why different structural scaffolds of DCN1 inhibitors would elicit cell line specific effects.

Experimental Section

TR-FRET Assay.19 The TR-FRET assay was carried out in black 384-well microtiter plates at a final volume of 20 μM per well. To screen library compounds, the assay cocktail was prepared as a mixture of 50 nM biotin-DCN1, 20 nM Ac-UBE2M12-AlexaFluor488, 2.5 nM Tb streptavidin (ThermoFisher) in assay buffer (25 mM HEPES, 100 mM NaCl, 0.1% Triton X-100, 0.5 mM DTT, pH 7.5). The assay cocktail was then incubated for 1 h at room temperature and distributed using a WellMate instrument (Matrix). Compounds to be screened were added to assay plates from DMSO stock solutions by pin transfer using 50SS pins (V&P Scientific). The assay mixture was incubated for 1 h at room temperature prior to measuring the TR-FRET signal with a PHERAstar or Clariostar plate reader (BMG Labtech) equipped with excitation modules at 337 nm and emissions at 490 and 520 nm. We set the integration start to 100 μs and the integration time to 200 μs. The number of flashes was set to 100. The relative fluorescence (Ex/Em = 520:490) was used for TR-FRET signal calculations. Assay end points were normalized from 0% (DMSO only) to 100% inhibition (unlabeled competitor peptide) for hit selection and curve fitting. All compounds were tested in triplicate or more.

General Cell Culture BJ (ATCC®, CRL-2522™), a normal human foreskin fibroblast cell line was purchased from the American Type Culture Collection (ATCC) and cultured in Eagle's Minimum Essential Medium (EMEM) medium (ATCC, 30-2003) supplemented with 10% FBS (ATCC, 30-2020). HCC95, a squamous cell lung carcinoma line, was cultured in RPMI-1640 medium with 2 mM L-glutamine (ATCC, 30-2001) supplemented with 10% fetal bovine serum (ATCC, 30-2020). 27



CAL33 (DSMZ no. ACC 447), a tongue squamous cell carcinoma, was purchased from the Leibniz-Institut DSMZ - Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH and was cultured in Dulbecco's Modified Eagle's Medium (DMEM) (ATCC® 30-2002™) supplemented with 10% fetal bovine serum (ATCC, 30-2020). Cells were routinely tested for mycoplasma contamination using the MycoAlert Mycoplasma Detection Kit (Lonza, LT07-218).

Evaluation of Cell Proliferation Effects Exponentially growing cells were plated in 384-well polystyrene, white, flat bottom, low flange, tissue culture treated assay plates (Corning, 3570) using a Matrix Wellmate liquid dispenser (Thermo Fisher), and incubated overnight at 37 ºC in a humidified 5% CO2 incubator. Each cell line was plated to a previously validated density to ensure logarithmic growth. For BJ, the cells were plated to 1000/cells per well in 30 microliters of complete media. For HCC95, the cells were plated to 600/cells per well in 30 microliters of complete media. Test articles were prepared as stock solutions in DMSO. Drug plates were prepared by serially diluting (ten, 3fold dilutions) stock solutions into a 384-well clear polypropylene plate (Corning, 3657). Test articles were manually transferred (73 nL) from the drug plate to the assay plate by hydrodynamic pin transfer using a pin tool (AFIX384FP1, V&P Scientific) adapted for manual transfer (BGPK, VP 381D-N, V&P Scientific) and equipped with FP1S50 pins (V&P Scientific). After drugging, the plates were incubated at 37 ºC in a humidified 5% CO2 incubator for 72 hours, and then cell viability was measured using Cell TiterGlo (Promega, G7573) according to the manufacturer’s recommendation. Luminescence was measured on a Clariostar plate reader (BMG Labtech). Assay endpoints were normalized from 0% (DMSO only) to 100% inhibition (58 µM Idarubicin) and fit to a semi-log plot using at least n = 3 technical replicates and n = 2 biological replicates. The curve fit was performed using the Collaborative Drug Discovery software and the standard Hill equation.

Cellular Thermal Shift Assay. 28


Page 28 of 44

Page 29 of 44 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


The cellular thermal shift assay (CETSA) was performed according to the previously published procedure.47 Briefly, 500,000 HCC95 cells were aliquoted in 100 μL of PBS and DMSO or the indicated compound (10 μM) was added (0.1 % final DMSO). Mixtures were incubated on ice for 1 h, and the cells were washed three times with PBS buffer. Cell pellets were resuspended in 50 μL of PBS and heated for 3 min at the indicated temperature in a thermocycler. Cells were lysed by three rounds of freezing in liquid nitrogen and thawing on ice. After pelleting at 20,000 rpm for 20 min at 4 oC, equal amounts of supernatant were removed and blotted with the indicated antibodies.

Western Blot Analysis Exponentially growing HCC95 or CAL33 cells were plated in 6-well plates at 0.4 x 106 cells/well in 2 ml of media and incubated overnight at 37 °C in a humidified 5% CO2 incubator. 24 and 48 hrs. after plating, the media was aspirated and replenished with 2 ml fresh media containing either 4 μL of DMSO or a 500x compound DMSO stock solution. The cells were harvested after 72 hr via trypsinization, thoroughly washed with PBS, pelleted, and stored at -80 °C. Cell pellets were thawed on ice and lysed by resuspension in 30-40 μl of lysis buffer [50 mM Tris, 150 mM NaCl, 0.5% NP-40, 0.1% SDS, 6.5 M Urea, 2 mM 1,10-orthophenanthroline, 1X Halt Protease and Phosphatase inhibitor cocktail (ThermoFisher), 0.25 kU Universal Nuclease (ThermoFisher), pH 7.5]. Cell suspensions were incubated on ice for 25 minutes with occasional mixing by pipetting up and down. Lysates were cleared by centrifugation at 13,000 rpm for 20 minutes and the supernatant collected. The protein concentration of total cell lysate was determined by BCA assay (Pierce) using BSA as a control. Cell lysates were diluted into 2X SDS-PAGE sample buffer such that 25 μg of total protein was loaded per well. Samples were heated at 95 °C for 2 minutes, briefly cleared by pulse centrifugation, separated on 4-12% NuPAGE gels (Invitrogen), and transferred to PVDF membranes (BIO-RAD) at 100 V for 90 minutes at 4 °C. Membranes were blocked for 1 hour in blocking buffer consisting of 1x TBS, 0.1% Tween-20, and 5% blotting grade non-fat dry milk (BIO-RAD). Primary antibodies were 29



Page 30 of 44

prepared in blocking buffer and incubated with membranes overnight at 4 °C with rocking, followed by extensive washing in 1x TBS, 0.1% Tween-20. Secondary antibodies were prepared in blocking buffer according to the manufactures recommendations and incubated with membranes for 1 hour at room temperature. After extensive washing, membranes were developed with SuperSignal West Pico Chemiluminescent substrate (ThermoFisher) and developed by film exposure (HyBlot CL, Denville scientific).

Intracellular ROS Assay Intracellular reactive oxygen species (ROS) were measured using the DCFDA / H2DCFDA Cellular ROS Assay Kit (Abcam) according to the manufacturer’s recommendation. Briefly, HCC95 cells were plated at 25,000 cells/well in 100 µL media into sterile black clear bottom 96-well plates and incubated overnight at 37 °C in a humidified 5% CO2 incubator. The following day, the media was removed, the wells were washed with 100 µL/well of sterile PBS, stained by adding 100 µL/well of the diluted DCFDA solution, and incubated at 37 °C for 45 min in the dark. Next, the wells were washed twice with sterile PBS and replenished with 100 µL of drugged buffer. The cells were incubated for 5 hours at 37 ˚C in the dark. The relative fluorescence signal (Ex/Em=485/535 nm) was measured using a Clariostar plate reader (BMG Labtech). The data was analyzed using GraphPad Prism version 8.00 for Windows, GraphPad Software, La Jolla California USA, www.graphpad.com.

Chemistry Experimental General. All NMR data was collected at room temperature in CDCl3 or (CD3)2SO on a 400 or 500 MHz Bruker or Agilent instrument. Chemical shifts (δ) are reported in parts per million (ppm) with internal CHCl3 (δ 7.26 ppm for 1H and 77.0 ppm for ppm for 1H and 39.5 ppm for

13C),

13C),

internal DMSO (δ 2.50

or internal TMS (δ 0.0 ppm for 1H and 0.0 ppm for

13C)

as

the reference. 1H NMR data are reported as follows: chemical shift, multiplicity (s = singlet, bs = broad singlet, d = doublet, t = triplet, q = quartet, p = pentet, sext = sextet, sep = septet, m= 30


Page 31 of 44 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


multiplet, dd = doublet of doublets, dt = doublet of triplets, td = triplet of doublets, qd = quartet of doublets), coupling constant(s) (J) in Hertz (Hz), and integration. Flash column chromatography was performed using a Biotage Isolera One and Biotage KP-SIL SNAP cartridges. All the tested compounds were characterized using

1H

NMR and HRMS.

Compounds that proved critical to our chemical analysis were further characterized using 13C NMR. All compounds were confirmed to be ≥ 95% purity prior to testing. Purity was assessed using an ultraperformance liquid chromatography mass spectrometry (Waters, MA) equipped with a PDA detector and a single quadruple detector. A BEH-C18 column (1.7 μm, 2.1 × 50mm2) was used. The flow rate was 0.7 mL/min, and the gradient started with 90% A (0.1% formic acid in H2O), changed to 95% B (0.1% formic acid in acetonitrile), and then returned to 90% A. The mass spectrometer was operated in the positive-ion mode with electrospray ionization. Integration was performed using Masslynx software 4.2

General Procedure A. Oxazolone Key Intermediate cyclization Under a nitrogen atmosphere, sodium acetate (1.0 mmol) and aldehyde (1.0 mmol) were added to a solution of carboxylic acid (1.0 mmol) in acetic anhydride (3.0 mmol). The resulting mixture was stirred at 85 oC for 2 h, cooled to room temperature, quenched with ice in ethanol, and stirred at room temperature overnight. The resulting solid was filtered, washed with water and diethyl ether, and dried under reduced pressure.

B. Oxazolone Key Intermediate with aliphatic chain B.1. Glycine coupling Carbonyl chloride (1.0 mmol) was added to a stirred solution of glycine (1.0 mmol) in 1N aqueous sodium hydroxide (3.0 ml), dropwise.

The reaction mixture was stirred at room

temperature overnight. Then, the pH of the mixture was adjusted to 1-2 with 1N aqueous HCl. The resulting solution was extracted with CH2Cl2, dried over MgSO4, filtered, concentrated 31



under reduced pressure, and purified by flash chromatography.

B.2. EDC-cyclization Under a nitrogen atmosphere, the amine (1.0 mmol), EDCI·HCl (1.3 mmol) and DIPEA (1.3 mmol) were added to a solution of the carboxylic acid (1.0 mmol) in CH2Cl2 (3 ml). The reaction mixture was stirred at room temperature overnight. The resulting solution was extracted with CH2Cl2, dried over MgSO4, filtered, concentrated under reduced pressure, and purified by flash chromatography.

B.3. Al2O3-condensation Under a nitrogen atmosphere, the aldehyde (0.5 mmol) was added dropwise to a suspension of oxazolone (0.1 mmol), activated molecular sieves 4Å (1.0 g), and activated aluminum oxide (1.0 mmol) in anhydrous CH2Cl2 (3 ml). The reaction mixture was stirred at room temperature for 6 h. The resulting solid was filtered, filtered through a pad of Celite® to remove molecular sieves and Al2O3. The filtrate was dried under reduced pressure and then purified by flash chromatography.

C. Dihydropyridinone-cyclization Under a nitrogen atmosphere, the oxazolone (0.5 mmol) and tin (II) chloride (0.05 mmol) were added to a solution of the amine (0.5 mmol) in chlorobenzene (1.0 ml). The reaction mixture was refluxed overnight. After cooling to room temperature, the mixture was extracted with EtOAc, washed with brine, dried over MgSO4, filtered, concentrated under reduced pressure, and purified by flash chromatography. Note: Compound 20 (Ile-R1= H) is derived from compound 23 (Ile-R1= t-butyl) that was dissolved in chlorobenzene/TFA (1:1, 0.3 M) and refluxed for 2 h. When the starting material was consumed, the mixture was neutralized with saturated aqueous sodium bicarbonate, extracted with EtOAc, dried over MgSO4, filtered, concentrated under reduced pressure, and 32


Page 32 of 44

Page 33 of 44 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


purified by flash chromatography.

D. Dihydropyridinone-amide alkylation Under a nitrogen atmosphere, cesium carbonate (1.1 mmol) and the alkyl halide (1.1 mmol) were added to a solution of the dihydropyridinone (0.1 mmol) in DMF (1 ml). The reaction mixture was stirred overnight at room temperature or heated at 100 oC for 3 h. Then the mixture was extracted with EtOAc, washed with brine, dried over MgSO4, filtered, concentrated under reduced pressure, and purified by flash chromatography.

E. Dihydropyridinone-amide acylation Under a nitrogen atmosphere, pyridine (0.075 mmol) and the acyl chloride (0.075 mmol) were added to a solution of the dihydropyiridinone (0.05 mmol) in CH2Cl2 (1.0 mL). The reaction mixture was stirred at room temperature overnight. Then the mixture was extracted with EtOAc, washed with brine, dried over MgSO4, filtered, concentrated under reduced pressure, and purified by flash chromatography.

F. Malonate-condensation Under a nitrogen atmosphere, 4-fluorobenzaldehyde (1.0 mmol), acetic acid (0.1 mmol) and piperidine (0.1 mmol) were added to a solution of dimethyl malonate (1.0 mmol) in toluene (2.0 ml). The reaction mixture was refluxed overnight. The crude mixture was concentrated under reduced pressure to remove the toluene, diluted with EtOAc, washed with brine, dried over MgSO4, filtered, concentrated under reduced pressure, and purified by flash chromatography.

G. Amide coupling (Reverse amide) The amine (0.1 mmol) was added to a solution of the ester (0.1 mmol) in chlorobenzene (0.3 ml). The reaction mixture was refluxed overnight. Then the mixture was extracted with EtOAc, 33



Page 34 of 44

washed with brine, dried over MgSO4, filtered, concentrated under reduced pressure, and purified by flash chromatography.

Spectral characterization of key compounds.

3-methyl-N-(rel-(4S,5S)-3-methyl-6-oxo-1-phenyl-4-(p-tolyl)-4,5,6,7-tetrahydro-1Hpyrazolo[3,4-b]pyridin-5-yl)benzamide (Compound 1) Compound 1 was synthesized by general procedure A, C and D starting from 4methylbenzaldehyde, (3-methylbenzoyl)glycine and 3-methyl-1-phenyl-1H-pyrazol-5-amine. 1H

NMR (400 MHz, Chloroform-d) δ 9.33 (s, 1H), 7.52 – 7.44 (m, 3H), 7.42 – 7.35 (m, 1H),

7.35 – 7.32 (m, 1H), 7.32 – 7.23 (m, 3H), 7.07 – 6.96 (m, 3H), 6.89 (d, J = 7.9 Hz, 2H), 6.54 (d, J = 6.1 Hz, 1H), 5.27 (dd, J = 7.5, 6.1 Hz, 1H), 4.75 (d, J = 7.5 Hz, 1H), 2.40 (s, 3H), 2.28 (s, 3H), 2.13 (s, 3H). 13C NMR (101 MHz, Chloroform-d) δ 169.3, 167.5, 147.2, 138.6, 137.5, 137.5, 136.0, 134.3, 133.9, 132.7, 129.6 (2C), 129.5 (2C), 128.6, 128.1 (2C), 127.9, 127.6, 124.2, 122.8 (2C), 104.1, 55.2, 38.0, 21.5, 21.1, 12.1. HRMS (ESI+) m/z calcd for C28H27N4O2+ [M+H]+ 451.2129 found 451.2136. N-(rel-(4S,5S)-7-ethyl-4-(4-fluorophenyl)-3-methyl-6-oxo-1-phenyl-4,5,6,7-tetrahydro1H-pyrazolo[3,4-b]pyridin-5-yl)-3-methylbenzamide (Compound 27) Compound 27 was synthesized by general procedure A, C and D starting from 4fluorobenzaldehyde, (3-methylbenzoyl)glycine and 3-methyl-1-phenyl-1H-pyrazol-5-amine. 1H


7.01 – 6.87 (m, 5H), 5.20 (dd, J = 7.1, 5.6 Hz, 1H), 4.77 (d, J = 7.0 Hz, 1H), 4.05 – 3.82 (m, 1H), 3.17 (dq, J = 14.0, 7.0 Hz, 1H), 2.39 (s, 3H), 2.14 (s, 3H), 0.99 (t, J = 7.1 Hz, 3H).

13C

NMR (101 MHz, Chloroform-d) δ 168.0, 167.3, 161.0 (d, J = 246.8 Hz), 147.0, 139.1, 139.0, 138.6, 133.7, 132.7, 132.5, 132.4, 130.0, 129.9, 129.6 (2C), 128.8, 128.6, 127.7, 125.3, 124.0, 115.5, 115.3, 105.9, 55.7, 39.2, 36.7, 21.3, 12.7, 11.9. HRMS (ESI+) m/z calcd for C29H28FN4O2+ [M+H]+ 483.2191, found 483.2193. 34


Page 35 of 44 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-((4S,5R)-4-(4-fluorophenyl)-3-methyl-6-oxo-1-phenyl-4,5,6,7-tetrahydro-1Hpyrazolo[3,4-b]pyridin-5-yl)-3-methylbenzamide (Compound 87) Compound 87 was synthesized by general procedure A, C and D starting from 4fluorobenzaldehyde, (3-methylbenzoyl)glycine and 3-methyl-1-phenyl-1H-pyrazol-5-amine. 1H


7.30 (m, 3H), 7.25 – 7.16 (m, 2H), 6.98 (t, J = 8.6 Hz, 2H), 6.33 (d, J = 8.7 Hz, 1H), 5.08 (dd, J = 12.2, 8.7 Hz, 1H), 4.25 (d, J = 12.2 Hz, 1H), 2.29 (s, 3H), 1.58 (s, 3H).

13C

NMR (126 MHz,

Chloroform-d) δ 168.0, 167.9, 162.3 (d, J = 246.5 Hz), 147.2, 138.5, 137.1, 135.7, 134.3, 133.7, 132.6, 130.3, 130.2, 130.0 (2C), 128.4, 128.1, 127.8, 124.0, 123.0 (2C), 115.9, 115.7, 101.9, 56.8, 43.0, 21.3, 13.0. HRMS (ESI+) m/z calcd for C27H24FN4O2+ [M+H]+ 455.1878, found 455.1883. 4-(4-fluorobenzylidene)-2-(m-tolyl)oxazol-5(4H)-one (Compound 144) Compound 144 was synthesized by general procedure A starting from 4-fluorobenzaldehyde and (3-methylbenzoyl)glycine. 1H

NMR (500 MHz, DMSO-d6) δ 8.21 – 8.12 (m, 2H), 7.98 – 7.88 (m, 2H), 7.40 – 7.32 (m, 2H),

7.16 – 7.06 (m, 3H), 2.40 (s, 3H). 13C NMR (126 MHz, DMSO-d6) δ 167.3, 164.3 (d, J = 167.6 Hz), 163.0, 139.3, 135.3, 135.2, 134.9, 133.2, 130.6, 129.8, 129.7, 128.6, 125.7, 125.5, 116.7, 116.6, 21.3. LRMS (ESI+) m/z calcd for C17H13FNO2+ [M+H]+ 282.3, found 282.3.

Ancillary Information Supporting Information: Characterization details for the other synthetic compounds as well as additional information regarding the X-ray crystallography and biological studies are described in Supporting Information. Chemical characterization and purity data; Expanded SAR tables including all compounds; Additional synthetic scheme; Pulse-chase data for compound 1; Reactive Oxygen Species data; Data collection and refinement statistics for 35



Lysozyme-DCN1P:1 and Lysozyme-DCN1P:27 (PDF) Molecular formula strings and TR-FRET potency (CSV)

Accession Codes: Atomic coordinates and experimental data for the X-ray co-crystal structures of 1 (PDB: 6P5W), NAcM-OPT (PDB: 5V86), and 27 (PDB: 6P5V) in complex with Lysozyme-DCN1P are available from the RCSB Protein Data Bank (www.rcsb.org). Authors will release the atomic coordinates and experimental data upon article publication.

Corresponding Author Information *E-mail: [email protected]. Phone: 901-595-5714. Fax: 901-595-5715

Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

Funding Sources BAS and DCS: NIH R37GM069530; P30CA021765; J.T.H., NIH F32GM113310; ALSAC/St. Jude.

Acknowledgment We thank the College of Pharmacy NMR Center (University of Kentucky) for NMR support. We acknowledge the High Throughput Biosciences Center, Medicinal Chemistry Center, Compound Management, and High Throughput Analytical Chemistry Centers in Chemical Biology and Therapeutics; Hartwell Center for use of their personnel and facilities. We thank the staff at the ALS 8.2.1 and Sercat 22-ID beamlines at the Advanced Light Source and 36


Page 36 of 44

Page 37 of 44 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


Advanced Photon Source.

Abbreviations Used: DCN1, defective in cullin neddylation 1; UBL, ubiquitin-like protein; UBE2M, NEDD8conjugating enzyme Ubc12; SCCRO, squamous cell carcinoma-related oncogene; SCC, Squamous cell carcinoma; CRL, cullin ring ligases; NEDD8, Neural Precursor Cell Expressed, Developmentally Down-Regulated 8; SAR, structure-activity relationship; SPR, structureproperty relationship; TR-FRET, time-resolved fluorescence resonance energy transfer; HTS, high-throughput screen; t-Bu, tert-butyl; CETSA, cellular thermal shift assay; ROS, reactive oxygen species; NAE, NEDD8-activating enzyme; RBX1, Ring-Box 1, E3 Ubiquitin Protein Ligase; CUL1, cullin 1; AF, AlexaFluor 488.

37



References 1.

Duda, D. M.; Scott, D. C.; Calabrese, M. F.; Zimmerman, E. S.; Zheng, N.; Schulman, B. A. Structural regulation of cullin-RING ubiquitin ligase complexes. Current Opinion in Structural Biology 2011, 21, 257-264.

2.

Zimmerman, E. S.; Schulman, B. A.; Zheng, N. Structural assembly of cullin-RING ubiquitin ligase complexes. Current Opinion in Structural Biology 2010, 20, 714-721.

3.

Petroski, M. D.; Deshaies, R. J. Function and regulation of cullin-RING ubiquitin ligases. Nat. Rev. Mol. Cell Biol. 2005, 6, 9-20.

4.

Dubiel, W.; Dubiel, D.; Wolf, D. A.; Naumann, M. Cullin 3-based ubiquitin ligases as master regulators of mammalian cell differentiation. Trends in Biochemical Sciences 2018, 43, 95-107.

5.

Strand, N. S.; Allen, J. M.; Ghulam, M.; Taylor, M. R.; Munday, R. K.; Carrillo, M.; Movsesyan, A.; Zayas, R. M. Dissecting the function of cullin-RING ubiquitin ligase complex genes in planarian regeneration. Developmental Biology 2018, 433, 210-217.

6.

Zheng, N.; Shabek, N. Ubiquitin ligases: Structure, function, and regulation. Annual Review of Biochemistry 2017, 86, 129-157.

7.

Lu, A.; Pfeffer, S. R. A CULLINary ride across the secretory pathway: more than just secretion. Trends Cell Biology 2014, 24, 389-399.

8.

Cui, D.; Xiong, X.; Zhao, Y. Cullin-RING ligases in regulation of autophagy. Cell Division 2016, 11, 8.

9.

Bulatov, E.; Ciulli, A. Targeting cullin-RING E3 ubiquitin ligases for drug discovery: structure, assembly and small-molecule modulation. Biochemical Journal 2015, 467, 365-386.

10.

Zhou, L.; Zhang, W.; Sun, Y.; Jia, L. Protein neddylation and its alterations in human cancers for targeted therapy. Cellular Signalling 2018, 44, 92-102.

11.

Ying, J.; Zhang, M.; Qiu, X.; Lu, Y. Targeting the neddylation pathway in cells as a potential therapeutic approach for diseases. Cancer Chemother Pharmacol 2018, 81, 38


Page 38 of 44

Page 39 of 44 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


797-808. 12.

Bielskiene, K.; Bagdoniene, L.; Mozuraitiene, J.; Kazbariene, B.; Janulionis, E. E3 ubiquitin ligases as drug targets and prognostic biomarkers in melanoma. Medicina (Kaunas) 2015, 51, 1-9.

13.

Huang, X.; Dixit, V. M. Drugging the undruggables: exploring the ubiquitin system for drug development. Cell Research 2016, 26, 484-498.

14.

Weathington, N. M.; Mallampalli, R. K. Emerging therapies targeting the ubiquitin proteasome system in cancer. J. Clin. Invest. 2014, 124, 6-12.

15.

MLN4924

Clinical

Trials.

https://clinicaltrials.gov/ct2/results?cond=&term=MLN4924&cntry=&state=&city=&dis t=, (accessed May 31, 2019). 16.

Schlierf, A.; Altmann, E.; Quancard, J.; Jefferson, A. B.; Assenberg, R.; Renatus, M.; Jones, M.; Hassiepen, U.; Schaefer, M.; Kiffe, M.; Weiss, A.; Wiesmann, C.; Sedrani, R.; Eder, J.; Martoglio, B. Targeted inhibition of the COP9 signalosome for treatment of cancer. Nat. Commun. 2016, 7, 13166.

17.

Hammill, J. T.; Bhasin, D.; Scott, D. C.; Min, J.; Chen, Y.; Lu, Y.; Yang, L.; Kim, H. S.; Connelly, M. C.; Hammill, C.; Holbrook, G.; Jeffries, C.; Singh, B.; Schulman, B. A.; Guy, R. K. Discovery of an orally bioavailable inhibitor of defective in cullin neddylation 1 (DCN1)-mediated cullin neddylation. J. Med. Chem. 2018, 61, 2694-2706.

18.

Hammill, J. T.; Scott, D. C.; Min, J.; Connelly, M. C.; Holbrook, G.; Zhu, F.; Matheny, A.; Yang, L.; Singh, B.; Schulman, B. A.; Guy, R. K. Piperidinyl ureas chemically control defective in cullin neddylation 1 (DCN1)-dediated cullin neddylation. J. Med. Chem. 2018, 61, 2680-2693.

19.

Scott, D. C.; Hammill, J. T.; Min, J.; Rhee, D. Y.; Connelly, M.; Sviderskiy, V. O.; Bhasin, D.; Chen, Y.; Ong, S.-S.; Chai, S. C. Blocking an N-terminal acetylationdependent protein interaction inhibits an E3 ligase. Nature Chemical Biology 2017, 13, 850-857. 39



20.

Zhou, H.; Lu, J.; Liu, L.; Bernard, D.; Yang, C. Y.; Fernandez-Salas, E.; Chinnaswamy, K.; Layton, S.; Stuckey, J.; Yu, Q.; Zhou, W.; Pan, Z.; Sun, Y.; Wang, S. A potent small-molecule inhibitor of the DCN1-UBC12 interaction that selectively blocks cullin 3 neddylation. Nat. Commun. 2017, 8, 1150.

21.

Zhou, H.; Zhou, W.; Zhou, B.; Liu, L.; Chern, T. R.; Chinnaswamy, K.; Lu, J.; Bernard, D.; Yang, C. Y.; Li, S.; Wang, M.; Stuckey, J.; Sun, Y.; Wang, S. High-affinity peptidomimetic inhibitors of the DCN1-UBC12 protein-protein interaction. J. Med. Chem. 2018, 61, 1934-1950.

22.

Wang, S.; Zhao, L.; Shi, X.-J.; Ding, L.; Yang, L.; Wang, Z.-Z.; Shen, D.; Tang, K.; Li, X.-J.; Mamun, M. A. A.; Li, H.; Yu, B.; Zheng, Y.-C.; Wang, S.; Liu, H.-M. Development of highly potent, selective, and cellular active Triazolo[1,5-a]pyrimidine-Based Inhibitors Targeting the DCN1–UBC12 Protein–Protein Interaction. J. Med. Chem. 2019, 62, 2772–2797.

23.

Scott, D. C.; Sviderskiy, V. O.; Monda, J. K.; Lydeard, J. R.; Cho, S. E.; Harper, J. W.; Schulman, B. A. Structure of a RING E3 trapped in action reveals ligation mechanism for the ubiquitin-like protein NEDD8. Cell 2014, 157, 1671-1684.

24.

Monda, J. K.; Scott, D. C.; Miller, D. J.; Lydeard, J.; King, D.; Harper, J. W.; Bennett, E. J.; Schulman, B. A. Structural conservation of distinctive N-terminal acetylationdependent interactions across a family of mammalian NEDD8 ligation enzymes. Structure 2013, 21, 42-53.

25.

Kurz, T.; Ozlu, N.; Rudolf, F.; O'Rourke, S. M.; Luke, B.; Hofmann, K.; Hyman, A. A.; Bowerman, B.; Peter, M. The conserved protein DCN-1/Dcn1p is required for cullin neddylation in C. elegans and S. cerevisiae. Nature 2005, 435, 1257-1261.

26.

Keuss, M. J.; Thomas, Y.; McArthur, R.; Wood, N. T.; Knebel, A.; Kurz, T. Characterization of the mammalian family of DCN-type NEDD8 E3 ligases. J. Cell Sci. 2016, 129, 1441-1454.

27.

Scott, D. C.; Monda, J. K.; Bennett, E. J.; Harper, J. W.; Schulman, B. A. N-terminal 40


Page 40 of 44

Page 41 of 44 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


acetylation acts as an avidity enhancer within an interconnected multiprotein complex. Science 2011, 334, 674-678. 28.

Fu, W.; Sun, J.; Huang, G.; Liu, J. C.; Kaufman, A.; Ryan, R. J.; Ramanathan, S. Y.; Venkatesh, T.; Singh, B. Squamous Cell Carcinoma-related Oncogene (SCCRO) family members regulate cell growth and proliferation through their cooperative and antagonistic effects on cullin neddylation. The Journal of Biological Chemistry 2016, 291, 6200-6217.

29.

Sarkaria, I.; P, O. c.; Talbot, S. G.; Reddy, P. G.; Ngai, I.; Maghami, E.; Patel, K. N.; Lee, B.; Yonekawa, Y.; Dudas, M.; Kaufman, A.; Ryan, R.; Ghossein, R.; Rao, P. H.; Stoffel, A.; Ramanathan, Y.; Singh, B. Squamous cell carcinoma related oncogene/DCUN1D1 is highly conserved and activated by amplification in squamous cell carcinomas. Cancer Research 2006, 66, 9437-9444.

30.

Kim, A. Y.; Bommelje, C. C.; Lee, B. E.; Yonekawa, Y.; Choi, L.; Morris, L. G.; Huang, G.; Kaufman, A.; Ryan, R. J.; Hao, B.; Ramanathan, Y.; Singh, B. SCCRO (DCUN1D1) is an essential component of the E3 complex for neddylation. The Journal of Biological Chemistry 2008, 283, 33211-33220.

31.

Sarkaria, I. S.; Pham, D.; Ghossein, R. A.; Talbot, S. G.; Hezel, M.; Dudas, M. E.; Ebright, M. I.; Chuai, S.; Memoli, N.; Venkatraman, E. S.; Miller, V. A.; Kris, M. G.; Zakowski, M. F.; Rusch, V. W.; Singh, B. SCCRO expression correlates with invasive progression in bronchioloalveolar carcinoma. The Annals of thoracic surgery 2004, 78, 1734-1741.

32.

Sarkaria, I.; P, O. c.; Talbot, S. G.; Reddy, P. G.; Ngai, I.; Maghami, E.; Patel, K. N.; Lee, B.; Yonekawa, Y.; Dudas, M.; Kaufman, A.; Ryan, R.; Ghossein, R.; Rao, P. H.; Stoffel, A.; Ramanathan, Y.; Singh, B. Squamous cell carcinoma related oncogene/DCUN1D1 is highly conserved and activated by amplification in squamous cell carcinomas. Cancer Research 2006, 66, 9437-9444.

33.

Estilo, C. L.; P, O. C.; Ngai, I.; Patel, S. G.; Reddy, P. G.; Dao, S.; Shaha, A. R.; Kraus, 41



D. H.; Boyle, J. O.; Wong, R. J.; Pfister, D. G.; Huryn, J. M.; Zlotolow, I. M.; Shah, J. P.; Singh, B. The role of novel oncogenes squamous cell carcinoma-related oncogene and phosphatidylinositol 3-kinase p110alpha in squamous cell carcinoma of the oral tongue. Clinical Cancer Research 2003, 9, 2300-2306. 34.

P, O. c.; Sarkaria, I.; Talbot, S. G.; Reddy, P.; Dao, S.; Ngai, I.; Shaha, A.; Kraus, D.; Shah, J.; Rusch, V.; Ramanathan, Y.; Singh, B. SCCRO (DCUN1D1) induces extracellular matrix invasion by activating matrix metalloproteinase 2. Clinical Cancer Research 2008, 14, 6780-6789.

35.

Talbot, S. G.; P, O. c.; Sarkaria, I. S.; Ghossein, R.; Reddy, P.; Ngai, I.; Cordeiro, C. N.; Wong, R. J.; Kris, M. G.; Rusch, V. W.; Singh, B. Squamous cell carcinoma related oncogene regulates angiogenesis through vascular endothelial growth factor-A. Ann. Surg. Oncol. 2004, 11, 530-534.

36.

Shemorry, A.; Hwang, C. S.; Varshavsky, A. Control of protein quality and stoichiometries by N-terminal acetylation and the N-end rule pathway. Molecular Cell 2013, 50, 540-551.

37.

Chen, Y.; McGee, J.; Chen, X.; Doman, T. N.; Gong, X.; Zhang, Y.; Hamm, N.; Ma, X.; Higgs, R. E.; Bhagwat, S. V.; Buchanan, S.; Peng, S. B.; Staschke, K. A.; Yadav, V.; Yue, Y.; Kouros-Mehr, H. Identification of druggable cancer driver genes amplified across TCGA datasets. Plos One 2014, 9, e98293.

38.

Huang, G.; Singh, B. Coamplification and cooperation: toward identifying biologically relevant oncogenes. Clinical Cancer Research 2013, 19, 5549-5551.

39.

Estilo, C. L.; P, O. C.; Ngai, I.; Patel, S. G.; Reddy, P. G.; Dao, S.; Shaha, A. R.; Kraus, D. H.; Boyle, J. O.; Wong, R. J.; Pfister, D. G.; Huryn, J. M.; Zlotolow, I. M.; Shah, J. P.; Singh, B. The role of novel oncogenes squamous cell carcinoma-related oncogene and phosphatidylinositol 3-kinase p110alpha in squamous cell carcinoma of the oral tongue. Clinical Cancer Research 2003, 9, 2300-2306.

40.

Zhang, Z. H.; Li, J.; Luo, F.; Wang, Y. S. Clinical significance of SCCRO (DCUN1D1) 42


Page 42 of 44

Page 43 of 44 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 prostate cancer and its proliferation-inhibiting effect on Lncap cells. European Review for Medical and Pharmacological Sciences 2017, 21, 4283-4291. 41.

Jiang, Y.; Hou, R.; Li, S.; Li, S.; Dang, G. MicroRNA-302 inhibits cell migration and invasion in cervical cancer by targeting DCUN1D1. Experimental and Therapeutic Medicine 2018, 16, 1000-1008.

42.

Keuss, M. J.; Thomas, Y.; McArthur, R.; Wood, N. T.; Knebel, A.; Kurz, T. Characterization of the mammalian family of DCN-type NEDD8 E3 ligases. Journal of Cell Science 2016, 129, 1441-1454.

43.

Zhang, Z.; Li, J.; Luo, F.; Wang, Y. Clinical significance of SCCRO (DCUN1D1) in prostate cancer and its proliferation-inhibiting effect on Lncap cells. European Review for Medical and Pharmacological Sciences 2017, 21, 4283-4291.

44.

Ritchie, T. J.; Macdonald, S. J. The impact of aromatic ring count on compound developability--are too many aromatic rings a liability in drug design? Drug Discovery Today 2009, 14, 1011-1020.

45.

Izumi, S.; Kobayashi, Y.; Takemoto, Y. Catalytic Asymmetric Synthesis of antialpha,beta-diamino acid derivatives. Organic Letters 2016, 18, 696-699.

46.

Shi, F.; Zhang, J.; Tu, S.; Jia, R.; Zhang, Y.; Jiang, B.; Jiang, H. An efficient synthesis of new class of pyrazolo [3, 4‐b] pyridine‐6‐one derivatives by a novel cascade reaction. Journal of Heterocyclic Chemistry 2007, 44, 1013-1017.

47.

Martinez Molina, D.; Jafari, R.; Ignatushchenko, M.; Seki, T.; Larsson, E. A.; Dan, C.; Sreekumar, L.; Cao, Y.; Nordlund, P. Monitoring drug target engagement in cells and tissues using the cellular thermal shift assay. Science 2013, 341, 84-87.

43



Page 44 of 44

Table of Contents graphic N-Acetyl Pocket O O HN N N

O O N

Optimization

NH

N N

25 times potency improvement

NH

F New scaffold HIT compound (1) TR-FRET IC50 = 5.1 µM

New optimized compound (27) TR-FRET IC50 = 0.2 µM

44


Discovery of Novel Pyrazolo-Pyridone DCN1 Inhibitors Controlling

Recommend Documents