Optimization and anti-cancer properties of fluoromethylketones as covalent inhibitors for ubiquitin C-terminal hydrolase L1

molecules

Article

Optimization and Anti-Cancer Properties of

Fluoromethylketones as Covalent Inhibitors for Ubiquitin

C-Terminal Hydrolase L1

Aaron D. Krabill ¹, Hao Chen ¹, Sajjad Hussain ^2,3, Chad S. Hewitt ¹, Ryan D. Imhoff ¹, Christine S. Muli ¹,

Chittaranjan Das ^4,5, Paul J. Galardy ², Michael K. Wendt ^1,5,6and Daniel P. Flaherty ^1,5,6,

*

1

Department of Medicinal Chemistry and Molecular Pharmacology, College of Pharmacy, 575 Stadium Mall

Dr., West Lafayette, IN 47907, USA; akrabil@purdue.edu (A.D.K.); chen2334@purdue.edu (H.C.);

hewitt8@purdue.edu (C.S.H.); rimhoff@purdue.edu (R.D.I.); cmuli@purdue.edu (C.S.M.);

mwendt@purdue.edu (M.K.W.)

2

3

4

5

Division of Pediatric Hematology-Oncology, Mayo Clinic, 200 First St. Guggenheim 15,

Rochester, MN 55905, USA; hussain.sajjad@mayo.edu (S.H.); galardy.paul@mayo.edu (P.J.G.)

Department of Pediatric and Adolescent Medicine, Mayo Clinic, 200 First St. Guggenheim 15,

Rochester, MN 55905, USA

Department of Chemistry, College of Science, 560 Oval Dr.,

West Lafayette, IN 47907, USA; cdas@purdue.edu

Purdue Center for Cancer Research, Hanson Life Sciences Research Building, 201 University St.,

West Lafayette, IN 47907, USA

Purdue Institute for Drug Discovery, 720 Clinic Ln., West Lafayette, IN 47907, USA

Correspondence: dﬂaher@purdue.edu

6

ꢀꢁꢂꢀꢃꢄꢅꢆꢇ

*

ꢀꢁꢂꢃꢄꢅꢆ

Citation: Krabill, A.D.; Chen, H.;

Hussain, S.; Hewitt, C.S.; Imhoff, R.D.;

Muli, C.S.; Das, C.; Galardy, P.J.;

Wendt, M.K.; Flaherty, D.P.

Abstract: The deubiquitinating enzyme (DUB) UCHL1 is implicated in various disease states in-

cluding neurodegenerative disease and cancer. However, there is a lack of quality probe molecules

to gain a better understanding on UCHL1 biology. To this end a study was carried out to fully

characterize and optimize the irreversible covalent UCHL1 inhibitor VAEFMK. Structure-activity

relationship studies identiﬁed modiﬁcations to improve activity versus the target and a full cellular

characterization was carried out for the ﬁrst time with this scaffold. The studies produced a new

Optimization and Anti-Cancer

Properties of Fluoromethylketones as

Covalent Inhibitors for Ubiquitin

C-Terminal Hydrolase L1. Molecules

2021, 26, 1227. https://doi.org/

10.3390/molecules26051227

inhibitor, 34, with an IC₅₀value of 7.7 µM against UCHL1 and no observable activity versus the

closest related DUB UCHL3. The molecule was also capable of selectively inhibiting UCHL1 in

cells and did not demonstrate any discernible off-target toxicity. Finally, the molecule was used for

initial probe studies to assess the role of UCHL1 role in proliferation of myeloma cells and migration

behavior in small cell lung cancer cells making 34 a new tool to be used in the biological evaluation

of UCHL1.

Academic Editor: Steven Fletcher

Received: 12 January 2021

Accepted: 19 February 2021

Published: 25 February 2021

Keywords: covalent inhibitors; ﬂuoromethylketones; UCHL1 inhibitors; deubiquitinases

Publisher’s Note: MDPI stays neutral

with regard to jurisdictional claims in

published maps and institutional afﬁl-

iations.

1. Introduction

Ubiquitination of substrate proteins plays a role in a number of cellular pathways

including protein trafﬁcking, DNA damage response, and proteasomal degradation [1–3].

The formation of a covalent bond between the C-terminal carboxylate of ubiquitin (Ub)

and the primary amine on a lysine side chain of substrate proteins results in an isopep-

tide bond. This bond formation is catalyzed by a series of cascading E1 (Ub-activating),

E2 (Ub-conjugating), and E3 (Ub-ligating) enzymes, which often results in the addition

of Ub in the form of a poly-Ub chain in both linear and branched architectures [4–9].

The identity of these Ub-chains forms the basis of the complex Ub-signaling pathway

observed in eukaryotes.

The removal of Ub is catalyzed by deubiquitinases (DUBs), of which there are ap-

proximately 100 in the human genome [10]. The majority of these DUBs are cysteine

proteases, although a family of DUBs, the JAMM proteases, do require a zinc cofactor for

Copyright:

Licensee MDPI, Basel, Switzerland.

This article is an open access article

distributed under the terms and

conditions of the Creative Commons

Attribution (CC BY) license (https://

creativecommons.org/licenses/by/

4.0/).

Molecules 2021, 26, 1227. https://doi.org/10.3390/molecules26051227

https://ww w .mdpi.com/journal/molecules

Molecules 2021, 26, 1227

2 of 24

catalysis [

7

]. The cysteine protease DUBs are categorized into six sub-families: ubiqui-

tin speciﬁc proteases (USP), Machado-Josephin domain (MJD), Ovarian tumor proteases

(OTU), Ubiquitin C-terminal hydrolases (UCH), and the recently discovered MINDY and

ZUFSP DUBs [11

number of disease states including cancer [13

such DUBs have emerged as potential therapeutic targets [18–20].

,

12]. Dysregulation of protein ubiquitination status is implicated in a

–

15] and neurodegeneration [16 17], and as

,

Ubiquitin C-terminal Hydrolase L1 (UCHL1) is a 25 kDa DUB whose ectopic expres-

sion is linked to a number of cancers [21]. Normally found in the central and peripheral

nervous system, its physiological role under normal conditions is not well understood [22].

UCHL1 comprises approximately 1–5% of the total soluble protein in the brain [23] and

is implicated in the progression of neurodegenerative diseases [24

ered an oncogene in a number of cancers including lymphoma [28

cancer [30–32], glioblastoma [33], and myeloma [34 36] among others [37

expression of UCHL1 in these cancers often correlates with increased metastatic behavior

and poor patient prognosis [31 39 41]. Both genetic depletion of UCHL1 as well as trans-

fection to introduce a catalytically inactive UCHL1 mutant reduced the ability of various

cancer cell lines to metastasize [42 43]. This suggests that pharmacological inhibition of

–

27]. It is also consid-

29] small-cell lung

38]. Additionally,

,

–

,

–

,

UCHL1 is a potential therapeutic avenue for treatment of UCHL1-implicated cancers.

The previous gold standard UCHL1 probe was the reversible inhibitor LDN-57444 [44

]

(Figure 1), however, the utility of this molecule as a UCHL1 inhibitor has been questioned

recently as on-target engagement of UCHL1 in cellular environments has not been ob-

served [45]. An alternative strategy for targeting UCHL1 is via electrophilic conjugation of

the catalytic cysteine, Cys90. A recent set of reversible cyanamide-based covalent inhibitors

have been described represented by MT-19 [46] and IMP-1710 [45] (Figure 1). MT-19

displayed a UCHL1 IC₅₀value of 670 nM after 30-min pre-incubation while IMP-1710 dis-

played a value of 38 nM. However, these inhibitors may suffer from apparent non-selective

toxicity, particularly in non-UCHL1 expressing cells. Krabill et al. [46] showed MT-19

displayed anti-proliferative properties in KMS-12 cell lines that do not express UCHL1,

and IMP-1710 began to display cytotoxicity at 10

ﬁbroblasts [45].

µM in patient derived human bronchial

Figure 1. Previously reported UCHL1 inhibitors with biochemical IC₅₀values versus UCHL1.

The tripeptide benzyloxycarbonyl-Val-Ala-Glu(γ-methoxy)-ﬂuoromethylketone

(VAEFMK, ) (Figure 1) represents an alternative class of covalent UCHL1 inhibitor that

1

has remained unexplored. VAEFMK was originally discovered serendipitously as a hit

in a screen of halo-methylketone tripeptides against the herpes simplex virus cysteine

protease UL36 [47]. However, during counter-screening against a panel of human DUBs,

VAEFMK was shown to inhibit UCHL1. The molecule was co-crystallized with UCHL1 by

Davies et al., which remains the only ligand-bound UCHL1 complex reported to date [48].

However, there is little biochemical and cellular information reported for this scaffold.

Covalent peptides have long been utilized as inhibitors for cysteine proteases including

the known caspase inhibitor VADFMK [49]. The ﬂuoromethylketone moiety is less reactive

than chloromethylketone and cyanamide counterparts and is an irreversible inhibitor, as

opposed to the reversibility of the cyanamides. Thus, we hypothesized VAEFMK would

exhibit little off-target non-speciﬁc reactivity in cells, making it a suitable starting point

Molecules 2021, 26, 1227

3 of 24

for structure-activity relationship (SAR) studies. To this end, our group set out to fully

characterize VAEFMK as a UCHL1 inhibitor in biochemical and cellular assays as well as

carry out SAR studies on the scaffold to improve potency and understand the mechanism

of action. Successfully characterizing this inhibitor and conﬁrming on-target engagement

in a cellular environment would serve to validate this molecular scaffold as a novel probe

for UCHL1. The results of these studies are presented herein.

2. Results and Discussion

2.1. Synthesis of VAEFMK and Derivatives

VAEFMK

protocols [50 53]. In brief, to prepare the FMK containing intermediates (Scheme 1 ) the key

ﬂuorinating reagent was synthesized via transesteriﬁcation of dimethyl ﬂuoromalonate

with benzyl alcohol, resulting in dibenzyl ﬂuoromalonate . Hydrolysis of a single

benzyl group yielded key reagent 3-(benzyloxy)-2-ﬂuoro-3-oxopropanoic acid (MBF)

The reagents could be converted to the magnesium enolate and added to an activated

ester of either Boc-Glu(OMe)-OH (6a) or Boc-Glu(OtBu)-OH (6b), forming intermediates

7a . These intermediates were isolated and immediately hydrogenated using H₂and

Pd/C to yield the key building blocks Boc-Glu(OMe)-FMK 8a or Boc- Glu(OtBu)-FMK 8b

1 and analogs were synthesized using variations to previously published

–

2

3

4.

5

-

b

.

Scheme 1. Synthesis of ﬂuoromethylketone amino acid intermediates. Reagents and conditions: (a) benzyl alcohol (10 eq),

TsOH (0.1 eq), toluene, 75 ^◦C, 2 h; (

crude to next step; (

to next step; (f) H₂, Pd/C, toluene, RT, overnight.

b

) NaOH (1.05 e_◦q), i-PrOH, 45 ^◦C, 70 min; (

c

) isopropyl MgCl (2.0 eq), THF, 0 ^◦C, 1 h,

◦

d

)

6

(1.0 eq), CDI (1.1 eq), THF, 0 C, 1 h, crude to next step; (

e)

5

(2.0 eq), THF,

−

20 C—RT, 3.5 h, crude

The corresponding chloromethylketone (CMK) intermediate was prepared according

to Scheme 2 from 6a. The carboxylic acid was activated by isobutyl chloroformate to form

the mixed anhydride 9, followed by the nucleophilic addition of diazomethane, generated

in situ with Diazald^®, to form a diazoketone 10. Quenching the reaction with HCl in

1,4-dioxane provided the building block Boc-Glu(OMe)-CMK 11.

Scheme 2. Synthesis of Boc-Glu(OMe)-CMK. Reagents and conditions: (a) N-methylmorpholine (1.3 eq), isobutyl chlorofor-

◦

mate (1.2 eq), THF, 0 C, 45 min; (

b

) Diazald^®(2.2 eq), carbitol (1 mL), ether (1 mL) 37% w/v KOH, 0 C, 2 hr; (

c) 4 M HCl in

◦

1,4-dioxane, until solution is not yellow, 0 C, 30 min.

These building blocks were then combined into the synthesis of the ﬁnal benzyloxy

(Cbz)-protected tripeptide analogs according to Scheme 3. Diversity at R₁and R₂amino

acid side chains were introduced at this point of the synthesis to generate the Cbz-protected

dipeptide methyl ester intermediates 12a

–t. The ester was then cleaved to provide the

free carboxylic acids 13a that were carried into the ﬁnal reaction crude. The analogous

–s

Molecules 2021, 26, 1227

4 of 24

intermediates 12t

The Boc-protected halomethylketone intermediates 8a

provide the free amines, which were then carried into the ﬁnal peptide coupling step and

paired with 13a to yield the ﬁnal electrophilic analogs. Several analogs, including the

–

v

were then fed into the synthesis of ﬁnal analogs as shown in Scheme 3

.

,

8b, and 11 were deprotected to

–

v

Glu-ﬂuoromethylketone derivative, required side-chain protecting groups to be removed

and were subjected to deprotection to arrive at the ﬁnal analog.

Scheme 3. Synthesis of Cbz-protected tripeptide halomethylketones. Reagents and conditions: (

a

) Cbz-carboxyl_◦ic acid

(1.0 eq), methyl ester amino acid (1.0 eq), N-methylmorpholine (4.0 eq), isobutyl chloroformate (1.5 eq), THF, 0 C–RT,

◦

overnight; (

b

) LiOH (1.2 eq), 4:1 THF:H₂O, 0 C, 3 h. (

c)

8a,

8b, or 11 (1.0 eq), 4.0 M HCl in 1,4-dioxane (7.4 eq), 30 min, RT,

carried into d; (

(1.5 eq), THF, 0 C, 45 min; ii. addition of Boc-deprotected 8a

d

) i. carbamate-protected dipeptide acid 12a

–

,

s

(0.8 eq), N-methylmorpholine (4.0 eq), isobutyl chloroformate

◦

8b, or 11, THF, 0 C–RT, overnight; (

e) For analogs that require

side-chain deprotection—precursor tripeptide intermediate (1.0 eq), 95:2.5:2.5 v:v:v TFA:TIPS:H₂O, 1 h, room temperature.

For further investigation in cellular engagement assays four analogs were synthesized

with modiﬁcation to the benzyloxy-capped end of the tripeptide. The ﬁrst set installed a

p-ethynylbenzyloxy capping group to be leveraged for click chemistry. The synthesis of

these analogs (Scheme S1) began by generating the p-ethynylbenzyl chloroformate from

4-ethynylbenzyl alcohol and triphosgene. This reagent was used to cap the corresponding

dipeptide methyl esters to provide 12u

–

v. These intermediates were fed into the synthetic

route outlined by Scheme 3. to generate the carboxylic acid intermediates 13u

–

v, which

were then coupled with the corresponding Glu(OMe) halomethylketones to arrive at

analogs 34–37.

2.2. Structure-Activity Relationship Studies for Tripeptide Halomethylketones

The in vitro inhibition of UCHL1 was determined by monitoring the cleavage of

rhodamine 110 from substrate Rhodamine110-ubiquitin (Ub-Rho) [54]. As a putative ir-

reversible inhibitor of UCHL1, the potency of

1 did increase with increasing length of

preincubation, indicative of the irreversible nature of inhibition (Figure 2). After approxi-

mately 112 min, the molecule had reached near maximum efﬁcacy, with the half maximal

inhibitory concentration (IC₅₀) value of 29 µM and 24 µM at 3-h incubation. Thus, a 3-h

preincubation period was selected for the structure-activity relationship (SAR) comparison

of all analogs to ensure full reaction time and for consistent comparison of activity against

UCHL1 within the scaffold. The initial molecule

toward two closely related UCH family DUB, UCHL3 and UCHL5, at concentrations up to

400 M [48]. This was conﬁrmed in our assay with no activity versus UCHL3 (Figure S1)

at the top concentration of 200 M. Therefore, during SAR we did not assess off-target inhi-

1 was reported to not possess any activity

µ

bition of UCHL3 or UCHL5, instead we chose to assess selectivity of our ﬁnal compounds

when the time came.

The in vitro efﬁcacy of covalent inhibitors can be broken down into the reversible

interactions that contribute to binding afﬁnity and the covalent reactivity of the elec-

trophilic “warhead” with the catalytic Cys. The ﬂuoromethylketone of

1 is very slow to

react with the active site cysteine, requiring approximately 2 h to reach nearly full inhibi-

tion. Chloromethylketones are more reactive than ﬂuoromethyl ketones, and have been

shown to be 3-fold more effective in matched molecular pair analogs [49]. Therefore, we

Molecules 2021, 26, 1227

5 of 24

reasoned that the efﬁcacy of

1

could be improved simply by increasing the reactivity of

the warhead. To test this hypothesis, the ﬂuoromethylketone of

1 was substituted for a

chloromethylketone, resulting in compound 14 (Table 1). Surprisingly, subsequent testing

revealed a complete loss of in vitro activity against UCHL1 for the chloromethylketone

analog. DTT is commonly used in this assay to ensure the active site cysteine remains in a

reduced state to carry out its catalytic function. However, it can also act as a bio-nucleophile

leading us to hypothesize that the excess DTT may be forming an adduct with the highly

reactive chloromethylketone and effectively impairing the activity of the molecule.

Figure 2. Time-dependent inhibition of UCHL1 by

1. IC₅₀curves for His-UCHL1 treated with

1

at

various points in time shows increasing potency with increasing length of incubation.

Table 1. UCHL1 inhibition data for VAEFMK analogs.

a

IC₅₀(µM) ^b

Cpd

X

R₁

R₂

R₃

R₄

1

F

Cl

F

Cl

Me

H

Val

Ala

Gly

Ala

D-Ala

Leu

Phe

Ser

Thr

Asn

Asp

Glu

Ala

Gly

Ala

H

24

>200

95

>200

76

23

13

185

100

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

Val

Me

D-Val

Gly

Ala

Leu

Phe

Ser

Thr

Asn

Asp

Val

Phe

Val

119

>200

100

>200

7.7

H

-C≡CH

62

11

>200

Amino acids are L unless otherwise speciﬁed as D. ^bExperiments were performed in technical triplicate and

a

averages are reported. IC₅₀values are from 3 h of preincubation with UCHL1. Errors provided as 95% conﬁdence

intervals reported in supporting material Table S1.

Molecules 2021, 26, 1227

6 of 24

Analysis of the UCHL1-

1

co-crystal structure (Figure 3A) shows a hydrogen bond is

observed between the carbonyl oxygen of the ﬂuoromethylketone and the amide nitrogen

of Cys90, as well as the ε²nitrogen of the UCHL1 Q84 side chain [46]. The amide carbonyl

oxygen between the glutamic acid methylester and alanine of

1

forms a hydrogen bond with

forms

backbone amide of Asn88. Finally, the

ε

²oxygen of the glutamic acid methyl ester of

1

a hydrogen bond with an Arg153 side chain guanidine proton. Thus, we hypothesized

that removal of the glutamic acid methylester from analog 15 to form the glutamate side

chain (16) may provide an opportunity to form a salt-bridge with Arg153. Unexpectedly,

this modiﬁcation resulted in a complete abrogation of activity, even after 6 h of incubation

with the enzyme. One possible explanation for the loss of efﬁcacy is the potential for

an intramolecular cyclization between the glutamate oxygen and the ﬂuoromethylketone

carbonyl, that could result in the formation of a six-membered lactone (Figure S2). A similar

observation has been documented for aspartic acid ﬂuoromethylketone, which undergoes

a pH dependent cyclization as observed by NMR [50]. This would need to be further

investigated to conﬁrm the cyclization in future work.

Figure 3. Crystal structure of VAEFMK

contacts (yellow dashed line) with critical residues in UCHL1 active site. (

of VAEFMK binding pose in UCHL1 active site. ( ) Point of view depicting that valine in VAEFMK

is directed toward cross-over loop amino acid residues Val158 and Asn159. ( ) Electrostatic surface

1

(green sticks) bound to UCHL1 (cyan). (A) VAEFMK

B

) Surface representation

C

D

map representation with VAEFMK bound (generated in PyMol). Red regions indicate negative

surface charge and blue indicates positive surface charge. Arrows indicate the vector in which the

D-amino acid side chains for valine and alanine would be directed from VAEFMK.

SAR at R₂began with substitution of the enantiomer D-valine (17) in place of the

natural L-valine found in

1 and this modiﬁcation led to a complete loss of activity (Table 1).

Additionally, removal of the valine side chain entirely (18) results in a complete abrogation

of activity against UCHL1 where as an alanine at R₂(19) rescued activity, though the

Molecules 2021, 26, 1227

7 of 24

molecule was 3-fold less effective (IC₅₀= 76

the chain length by substitution of a leucine (20, IC₅₀= 23

IC₅₀= 13 µM) led to improvement over the parent molecule . The trend observed at R₂to

µM) than parent compound

1. Increasing

µM) or phenylalanine (21,

1

be Phe > Leu > Val > Ala > Gly clearly indicates the preference for a bulky, hydrophobic

residue at this position and the need for the natural amino acid stereochemistry. Further

inspection of the VAEFMK-bound structure of UCHL1 (Figure 3 C) indicates that the valine

is directed toward the ﬂexible crossover loop of UCHL1 [55]. This would explain why

bulkier hydrophobic residues were preferred at R₂as there is space to accept such a

modiﬁcation. There is little room for any D-enantiomeric substituent at R₂as the vector

(Figure 3D, top arrow) representing the D-valine substitution in analog 17 would be

directed toward and clash with the surface of UCHL1 explaining the lack of activity for

this analog.

Insertion of polar side chains reduced UCHL1 activity. For example, the serine

containing derivative at R₂displayed much reduced activity with an IC₅₀= 185 µM.

However, increasing the lipophilicity of this polar side chain by incorporating a threonine,

which more resembles the steric size of the parent valine side chain, displayed an IC₅₀

value of 100 µM demonstrating that the lipophilic bulk can mitigate some of the loss

of activity due to incorporation of the polar hydroxyl group. Moving to an asparagine,

which is similar in size to a leucine but contains the polar amide side chain, maintained

activity comparable to the threonine with an IC₅₀value of 119 µM, however, when this

position was modiﬁed with an acidic aspartate side chain, which would be charged at

physiological pH, the molecule completely lost activity even at the highest concentrations

tested. The observations from this subset of analogs indicate that non-ionizable polar side

chains but do still exhibit some anti-UCHL1 activity but are not preferred, and that this

activity can be improved with addition of hydrophobic atoms. However, introduction of

an ionizable side chain abrogated activity of the scaffold. The electrostatic potential surface

representation (Figure 3D) shows that the surface that the side chains would be directed

toward carries a partial negative charge, possibly explaining the aversion to residues that

carry a negative ionizable charge as this may result in electrostatic repulsion from the

binding site.

The same modiﬁcations were attempted at R₃as were built in at R₂with limited

success. Similar to the R₂position, the substitution of the D-alanine (26) in place of the

parent L-alanine resulted in loss of activity. Only the glycine substituent (15) retained any

activity, albeit much reduced (IC₅₀= 95

µM) compared to the parent

1, while the remaining

analogs all exhibited IC₅₀values > 200

µM. Again, inspection of the ligand bound structure

explains the observed results as alanine at R₃is directed toward the UCHL1 surface in

a tight space that would not be amenable to increased steric bulk (Figure 3D). This is

evidenced by the fact that the glycine analog retains some activity, but also indicates the

methyl group present in the alanine derivative does provide some level of hydrophobic

binding. Finally, the direction in which the D-alanine would be projected toward also

would clash with UCHL1 surface offering an explanation for the loss of efﬁcacy with

this analog.

In preparation for cellular target engagement studies the benzyloxy capping group

was modiﬁed to install a para-alkyne handle amenable to click chemistry at R₄. Interestingly,

each derivative that had the 4-alkynylbenzyloxy installed was actually more potent than

the nearest neighbor benzyloxy analog. For example, the 4-alkyne containing analog 34

displayed an IC₅₀value of 7.7 µM compared to the matched molecular pair 1 for an increase

of approximately 3-fold activity versus UCHL1. The same trend was observed for 35, which

contained a glycine at R₃and the 4-alkyne at R₄, being about 1.5-fold more active than the

non-alkyne containing analog 15. The trend was not as noticeable when comparing 36, with

a phenylalanine at R₂and alkyne at R₄, to the non-alkyne derivative 21, although there was

slight improvement observed. As a means of comparison, an alkyne containing derivative

of the chloromethylketone (37) was also produced but this molecule was still inactive in

the in vitro assay. The improvement of activity by installation of the alkyne was somewhat

Molecules 2021, 26, 1227

8 of 24

surprising as the benzyloxy group in the crystal structure was found to be solvent-exposed

and not taking part in any speciﬁc interactions; nonetheless, the data suggests that there

may be opportunity to further explore SAR on the phenyl ring in future work.

In addition to the catalytic Cys90 residue, UCHL1 possesses ﬁve other cysteines.

Among the ﬁve alternative cysteine residues Cys152 has previously been reported to ex-

hibit nucleophilic activity toward endogenous electrophiles [56,57] as well as be involved

in trans-nitrosylation events [58], while Cys220 has been reported to play a role in farne-

sylation [59] and AKT signaling [35]. A common shortcoming to covalent inhibitors is

that the reactive electrophilic groups may partake in non-speciﬁc labeling of off-target

cysteines. To investigate if alternative cysteines are affected on UCHL1 we used analog 34

to conﬁrm that the ﬂuoromethylketone derivatives only conjugate Cys90. Molecule 34 was

incubated with both recombinant wild-type UCHL1 containing Cys90 and catalytically

inactive UCHL1 that possesses a C90A mutation, then a click reaction was performed

to append a ﬂuorophore for in-gel ﬂuorescence imaging. It was observed that analog 34

only formed a covalent adduct with the Cys90 of the WT-UCHL1 while no ﬂuorescent

bands were observed for the UCHL1^C90Atreated protein (Figure 4). Additionally, samples

were analyzed by mass spectrometry to increase sensitivity for detection of any non-Cys90

adducts. The mass spectrometry data unequivocally suggests that 34 reacts only with

Cys90 of UCHL1 as a single adduct + 486.00 Da was observed that would correspond to

34 without the ﬂuorine atom (Figure 4C). There were no adducts observed in the UCHL1

C90A samples incubated with 34 (Figure 4D). Table 2 summarizes the mass spectrometry

data from the samples tested. This data conﬁrms that the ﬂuoromethylketone electrophile

is sufﬁciently selective to the catalytic Cys90 on UCHL1.

Figure 4. Fluoromethylketone 34 selectively conjugates to the Cys90. (A) Recombinant WT-UCHL1 and catalytically inactive

C90A mutant followed click reaction with Cy5-Azide and analyzed by in-gel ﬂuorescence shows the alkyne-tagged analog

34 forms a covalent adduct with WT-UCHL1 but does not form an adduct with the catalytically inactive C90A mutant.

Coomassie stain provided as loading control. (B) Deconvoluted mass spectrum for WT-UCHL1 at 10

µM. (C) Deconvoluted

mass spectrum for WT-UCHL1 (10

µ

M) treated with 34 (20 M). ( ) Deconvoluted mass spectrum for C90A-UCHL1 (10

µ

D

µM) treated with 34 (20 µM).

To summarize thus far, our SAR studies revealed that L-amino acids were superior to

D-amino acids at R₂and R₃. Bulky, lipophilic side chains were preferred for maintaining or

Molecules 2021, 26, 1227

9 of 24

improving activity against UCHL1 at R₂, whereas any increase of steric bulk at R₃was detri-

mental to activity. Polar functional groups reduced activity at both locations. Finally, the

derivatives with alkynes at R₄all outperformed the corresponding 4-H analogs to varying

degrees, with molecule 34 being the most potent in terms of the single time-point efﬁcacy

versus UCHL1 with an IC₅₀value of 7.7

room for improvement in modifying the benzyloxy cap to the peptide. This molecule also

maintained no activity against UCHL3 at > 200 M (Figure S1), conﬁrming the superior

µM, lending evidence that there may be additional

µ

selectivity over the closest related DUB family member. It should be noted that while no

modiﬁcations were made to alter the Glu side chain beyond de-methylation of the ester,

previous studies have conﬁrmed that the well-known pan caspase inhibitor VADFMK [60],

which differs from VAEFMK by an Asp(OMe) in place of the Glu(OMe), is also inactive

against UCHL1 [48]. Taken together, it is apparent that the Glu(OMe) is crucial to position

the electrophile in just the right orientation to undergo the covalent reaction in the context

of the relatively inert ﬂuoromethylketone electrophile and the active site Cys90 of UCHL1.

Table 2. Observed deconvoluted masses for recombinant WT- and C90A-UCHL1 treated with 34.

Calculated Deconvoluted

Protein and Treatment

Adduct(s) Observed (Da)

Masses Observed (Da) ^a

25,235.782

UCHL1

none

+486.00

none

WT-UCHL1 + 34 (1:1)

WT-UCHL1 + 34 (1:2)

UCHL1 C90A

25,235.782 and 25,721.782

25,205.782

UCHL1 C90A + 34 (1:1)

25,205.782

none

UCHL1 C90A + 34 (1:2)

Average of duplicate samples.

25,205.782

none

a

2.3. Kinetic Analysis of Fluoromethylketone Analogs

Irreversible covalent inhibitors may be further characterized by determining the

reversible binding constant (K_I) as well as the rate of the covalent bond formation under

saturating conditions (k_inact) [61–63]. By determining the IC₅₀over the course of various

timepoints, the pseudo ﬁrst-order rate constant (k_obs) could be calculated. Plotting k_obsas a

function of inhibitor concentration generated a curve that could be ﬁt using the equation

Y = k_inact*X/(K_I+ X) (Figure 5) [64]. For parent compound

1, these values were determined

to be K_I= 60.04

µ

M and k_inact= 0.0173 s⁻¹, resulting in a k_inact/K_Ivalue of 288.05 M⁻¹s⁻¹

(Table 3). In comparison, the k_inact/K_Ivalues for the two more potent analogs from the SAR

studies were 810.0 M⁻¹s⁻¹and 208.4 M⁻¹s⁻¹for 34 and 36, respectively. The improvement

in both IC₅₀and k_inact/K_Ifor 34 appears to be attributed to an increase the afﬁnity of

the analog towards UCHL1 as the K_Iwas improved greater than 3-fold. The fact that 36

was essentially equipotent to

previous single time-point assays with a set preincubation of 3-h showed the molecule to

be 2-fold more potent than . However, this could likely be due to a difference in the assay

1 in the covalent kinetics was somewhat unexpected as the

1

conditions as far as preincubation times because covalent inhibitors become more potent

over time [63]. This underscores the importance of determining the k_inact/K_Ivalue for key

analogs to robustly determine inhibitor efﬁcacy and further conﬁrmed the superiority of 34

compared to the original parent molecule.

Molecules 2021, 26, 1227

10 of 24

Figure 5. Inactivation kinetics for FMK analogs. k_inact/K_Iplots for UCHL1 treated with

1

,

34, and 36

−1

(Values are 288.05 M⁻¹S⁻¹, 813.3 M⁻¹S⁻¹, and 209.9 M⁻¹S for 1, 34, and 36, respectively.

Table 3. Kinetics of covalent inhibition of UCHL1.

k_inact/K_I

R²

−1

Cpd

k_inact(s

)

K_I(µM)

−1

(M⁻¹s

)

1

34

36

0.0173

0.0136

0.0157

60.04

16.79

75.32

288.1

810.0

208.4

0.9623

0.9779

0.9478

2.4. Fluoromethylketone Analogs Preclude Binding of Ub to UCHL1

In addition to its role as a deubiquitinating enzyme other putative physiological

functions for UCHL1 under normal condition are hypothesized to be to (1) maintain a mono-

Ub pool within cells [65

the Ub-UCHL1 protein-protein interaction (PPI) [67

strong binding afﬁnity of Ub to UCHL1, it is believed that within cells UCHL1 is mostly

bound to free mono-Ub. The ligand-bound structure of bound to UCHL1 conﬁrms that

,

66] and (2) mask sites of activation on Ub through maintaining

,68]. Given this, and the relatively

1

the inhibitor does not bind to the Ub-binding site of UCHL1, but actually approaches

and binds to the face opposite of the Ub-site and interacts with the catalytic cysteine

(Figure 6A, PDB: 4DM9 [48] and PDB: 3KW5 [55]). This would seemingly still allow the

Ub-binding interface to be available to interact with mono-Ub. To determine if

abrogate the UCHL1:Ub interaction, Ub-binding studies were carried out using biolayer

interferometry (BLI). His-UCHL1 was preincubated with an excess of (2 mM) or DMSO

1 is able to

1

overnight at room temperature to ensure full covalent modiﬁcation of the Cys90 was

achieved. The dissociation constant (K_d) of Ub towards UCHL1 pre-treated with DMSO

treated control was determined and proved to be similar to previously published afﬁnities

(Figure 6 B) [69]. However, the presence of

the ability of UCHL1 to interact with Ub (Figure 6C). Taken together, these data suggest

that even though binds on the opposite face of UCHL1 compared to the Ub-binding

domain it still completely precludes Ub binding. This may be a factor of stabilizing

1 conjugated to the Cys90 completely abrogated

1

the crossover loop and, in turn, precluding Ub binding; however, further investigation

is needed to test this hypothesis. Nonetheless, it appears the inhibition of UCHL1 by

the ﬂuoromethylketone analogs would reduce the binding of mono-Ub within the cell to

UCHL1.

Molecules 2021, 26, 1227

11 of 24

Figure 6. Competition of binding for

surface) (PDB: 4DM9) and Ub (gray ribbon) bound to UCHL1 (PDB: 3KW5). Crossover loop of UCHL1 shown in red ribbon

for clarity. ( ) Association/dissociation and steady-state binding data for UCHL1 and 1:1 serial dilutions Ub in BLI assay

1 versus Ub. (A) Overlaid crystal structure of 1 (green sticks) bound to UCHL1 (cyan

B

buffer. (C) Association/dissociation and steady-state binding data for UCHL1 after preincubation with 1.

2.5. Cellular Efﬁcacy in Cancer Cell Lines

Levels of UCHL1 expression are correlated with increased metastatic behavior and

poor patient prognosis in a number of cancers. To evaluate 1 and analogs in a cellular con-

text, three previously validated cell lines were utilized based on their sensitivity to UCHL1

depletion: SW1271 (small-cell lung cancer), KMS11 and KMS12 (myeloma cell lines).

2.5.1. VAEFMK Efﬁcacy in Myeloma Cells

KMS11 cells endogenously express UCHL1 and are sensitive to genetic depletion of

UCHL1 using shRNA, whereas KMS12 cells do not express UCHL1 and are not sensitive

to genetic depletion These cells provide a good control for evaluating off-target effects

of UCHL1 inhibitors [36

,46]. Consistent with previously reported sensitivity to UCHL1

depletion [34], KMS11 cells dosed with 100

µM 1

exhibited reduced proliferation compared

to DMSO controls after 24, 48, 72 and 96 h timepoints (Figure 7A). Alternatively, KMS12

cells, which are not sensitive to UCHL1 genetic depletion, showed reduced growth at

the 24-h timepoint but regained proliferative ability at the 96-h timepoint. The KMS12

cells are less sensitive to

1 than KMS11 and the molecule clearly displays a differential

effect on KMS11 and KMS12 cells that would indicate they selectively inhibit UCHL1 in

KMS11 cells.

To conﬁrm on-target engagement in KMS11 cells, a Ub activity-based probe (ABP) gel-

shift assay was employed [70]. Using the hemagglutinin-tagged Ub-vinylmethylester (HA-

Ub-VME) ABP to examine the activity of

Cells were treated in dose-response with

treated with HA-Ub-VME. The HA-Ub-VME labels any remaining DUBs that have not been

inhibited by . Blotting for UCHL1 showed a dose-dependent decrease in HA-Ub-UCHL1

complex formation with increasing concentration of , conﬁrming on-target engagement

1

in KMS11 cells, an immunoblot was performed.

1

(0–250 M), washed with PBS, then lysed and

µ

1

of UCHL1 in cells (Figure 7B, top panel). The same blot was performed for UCHL3 to

conﬁrm the molecule does not inhibit the closely related DUB in cells. The blot shows

full formation of the HA-Ub-UCHL3 complex at all concentrations after preincubation

with

HA-Ub-VME reacts with all DUBs in lysates; thus, blotting for HA qualitatively reveals the

broader DUB selectivity of the inhibitor. This blot revealed that exhibited the expected

1 indicating the molecule does not inhibit UCHL3 in cells (Figure 7B, second panel).

1

dose-dependent decrease in HA-Ub-UCHL1 complex formation (Figure 7B, bottom panel).

However, a higher exposure of the blot did show another DUB corresponding to the

Molecules 2021, 26, 1227

12 of 24

molecular weight UCHL5 may also be inhibited at 250

µM (Figure S4). No other DUBs

in the KMS11 cell line revealed any dose-dependent changes in complex formation upon

treatment with suggesting selectivity for UCHL1 within the DUBs expressed in this cell

line and corroborating the observed UCHL1-dependent reduction in cell proliferation.

1

Figure 7. Cellular characterization of

KMS12 cells treated with (100 M) or DMSO control at 24-, 48-, 72-, and 96-h timepoints. N = 3, *, p < 0.05; **, p < 0.01;

***, p < 0.001. Statistical signiﬁcance calculated using student’s t-test ( ) Immunoblots data for KMS11 cells treated with

(0–250 M) then reacted with HA-Ub-VME. Blotting for UCHL1 (top panel) indicates the molecular weight shift of the

1 in KMS11 and KMS12 myeloma cells. (A) Cellular proliferation data for KMS11 and

1

µ

B

1

µ

HA-Ub-UCHL1 complex in the control lane and the dose-dependent decrease of this complex formation in KMS11 cells.

Blotting for UCHL3 (second panel from top) indicates formation of HA-Ub-UCHL3 complex at every dose. Blotting for

HA (bottom panel) conﬁrms only dose-dependent decrease of HA-Ub-UCHL1 complex formation and no dose-dependent

decrease of any other DUB.

2.5.2. Cell Viability and On-target Engagement of Analogs in SW1271 Small Cell Lung

Cancer Cells

UCHL1 is a key promoter of cell invasion and metastasis in small cell lung cancer

but not necessarily required for cell viability or proliferation [43

assess the efﬁcacy of targeting UCHL1 using the SW1271 small cell lung cancer cell line.

Treatment of SW1271 cells with ﬂuoromethylketones and 34 (100 M) has no effect on the

cell viability, whereas chloromethylketones 14 and 37 both kill the cells in a dose-dependent

manner with CC₅₀values of 0.156 M and 0.176 M, respectively (Figure 8A). The Ub-ABP

,71]. Thus, we chose to

1

µ

assay was used to conﬁrm on target engagement of UCHL1 within these cells. Analog

34 inhibited HA-Ub-VME complex formation with UCHL1 in a dose-dependent manner

(Figure 8B, top panel) and demonstrated selectivity over other DUBs in SW1271 cells as the

only band in the HA-blot to exhibit signiﬁcant dose-dependent signal reduction was the

Ub-UCHL1 band (Figure 8B, bottom panel).

Molecules 2021, 26, 1227

13 of 24

Figure 8. SW1271 small cell lung cancer cell viability and on target engagement of inhibitors. (

cell line in the presence of selected ﬂuoromethylketone and chloromethylketone analogs. (

A

) Cell viability of the SW1271

B) Ub-ABP assay showing on-

target engagement of ﬂuoromethylketone 34 with UCHL1 in a dose-dependent manner (top panel). HA immunoblot shows

selectivity for UCHL1 over other DUBs within SW1271 cells (bottom panel). Blotting for tubulin loading control shown on

same HA-blot (C) In-gel ﬂuorescence of SW1271 cells treated with ﬂuoromethyl ketone 34 and chloromethylketone 37 then

subjected to click reaction with Cy5-azide ﬂuorophore.

Leveraging the alkyne handle, a click reaction was performed with an azide containing

sulfo-cyanine5 ﬂuorophore (Cy5) to visualize proteins within the cells that were conjugated

to either the ﬂuoromethylketone 34 or the chloromethylketone 37. Cells were pretreated

with either analog then harvested and washed to removed unreacted inhibitor. The cells

were then lysed and the click reaction was performed on lysates. The ﬂuoromethylketone

displayed high selectivity in cells as the only dark band observed was that of a molecular

weight corresponding to UCHL1 (approx. 25 kDa) with very light bands at other molecular

weights that are likely a result of weak non-speciﬁc activity at other cysteines (Figure 8C).

Conversely, the chloromethylketone 37 was observed to non-speciﬁcally label multiple

proteins within the cell (Figure 8C). The degree of promiscuity observed for the more

reactive chloromethylketone warhead likely explains the cytotoxicity against the SW1271

cell line, while the less reactive and more selective ﬂuoromethylketone 34 did not reduce

SW1271 cell viability. Further evaluation using proteomic proﬁling is warranted to gain a

quantitative understanding on the potential for non-speciﬁcity at off-target cysteines.

2.5.3. Inhibition of UCHL1 Reduces Migratory Capability of SW1271 Cells

UCHL1 has been reported to regulate and contribute to cell invasiveness and metastatic

behavior in small cell lung cancer. To assess the role of UCHL1 in cell migration, a wound

healing assay was performed using control and UCHL1 depleted SW1721 cells. First,

shRNA knockdown of UCHL1 in SW1271 cells was performed and efﬁcient knockdown

was observed with shRNA 079 and 274 compared to the vector control pLKO (Figure 9A

,

vector and shRNA information in Table S3). While these cells did exhibit reduction of

viability (Figure 9B) the knockdown cell lines were stable and used in the wound healing

assay. Genetic depletion of UCHL1 reduced the ability of SW1271 cells to migrate by

approximately 20% compared to the pLKO control vector cells at 24 hrs (Figure 9C,D).

The same experiments were performed on wild type SW1271 cells treated with ﬂu-

oromethylketone analogs. Molecules

viability at 100 M to determine if any observed reduction in cell viability may inﬂuence the

migration assays (Figure 9E). Contrary to how the cells behaved in the shRNA knockdown

experiments, when dosed with molecules neither nor 34 displayed cytotoxic effect on the

SW1271 cells at these concentrations. However, consistent with shRNA results treatment

of SW1271 cells with 100 M of 34 also signiﬁcantly reduced the ability of the cells to

migrate compared to the DMSO control while the same reduction of migratory ability

was not observed for the less potent (Figure 9F,G). These results illustrate the utility of

1 and 34 were ﬁrst assessed for their effect on cell

µ

1

µ

1

Molecules 2021, 26, 1227

14 of 24

pharmacological inhibition using 34 compared to shRNA knockdown as the using the

probe allows one to separate the impact of cell viability from the antimigratory activity

of UCHL1. Due to the slow reactive nature of the inhibitor and the modest potency the

anti-migratory efﬁcacy of these analogs may be improved with longer duration of treatment

or increasing the frequency of treatment. Further studies are underway to elucidate the role

UCHL1 plays in the migratory cellular pathways using both genetic depletion and inhibitor

approaches. Nonetheless, these data indicate that selective pharmacologic inhibition of

UCHL1 reduced the invasive capabilities of SW1271 small cell lung cancer cells.

Figure 9. Effect of UCHL1 on migration of SW1271 cells. (

control (pLKO) and shRNA 079 and 274. ( ) Effect of shRNA knockdown on SW1271 cell viability compared to pLKO control.

(C) Wound healing cell images on SW1271 pLKO and UCHL1 knockdown cells at 0- and 24-h time points. Representative

replicates shown for each. ( ) Migration rate in % area covered by cells in 24-h period compared to pLKO control. ( ) Effect

of compounds and 34 on cell viability at 100 M dose compared to vehicle control (DMSO). ( ) Wound healing cell images

of wild type SW1271 cells treated with ﬂuoromethylketone inhibitors at 0 and 24-h time points. Representative replicates

shown. ( ) Migration rate in % area covered by cells in 24-h period after treatment with inhibitor. N = 6, * p < 0.05; **, p <

A) Knockdown of UCHL1 in SW1271 cells. Shown are vector

B

D

E

1

µ

F

G

0.01, *** p < 0.001. NS = not signiﬁcant. Statistical signiﬁcance calculated using two-tailed t-test.

3. Materials and Methods

3.1. Chemistry

3.1.1. General Experimental

¹H-, ¹³C- and ¹⁹F-NMR spectra were recorded on DRX500 or ARX-800 spectrometers

(Bruker, Billerica, MA, USA) in [D₆] DMSO or CDCl₃with or without the internal standard

of TMS at 0.05 or 0.1% v/v. The purity of all ﬁnal compounds was > 95% purity as

assessed by HPLC. Final compounds were analyzed on a 1200 series chromatograph

(Agilent, Santa Clara, CA, USA). The chromatographic methods used were either: (A)

Hypersil GOLD C18 column (Thermo Scientiﬁc, Waltham, MA, USA) 3 µM particle size,

Molecules 2021, 26, 1227

15 of 24

150 mm length, 4.6 mm ID) or (B) a Thermo Scientiﬁc Hypersil GOLD C18 column (3

µM particle size, 250 mm length, 4.6 mm ID). UV detection wavelength = 254 nm; ﬂow

rate = 1.0 mL min⁻¹; solvent = acetonitrile/water. Both organic and aqueous mobile

phase contain 0.1% v/v triﬂuoroacetic acid. The mass spectrometer used is a CMS-L

Compact Mass Spectrometer (Advion, Ithaca, NY, USA) with an ESI or an APCI ionization

source. Samples are submitted for analysis using either the atmospheric solids analysis

probe (ASAP) or ﬂow injection analysis (FIA). Compounds were prepared according to

the following protocols and are detailed below. Intermediates 13a (Catalog #V4659, AK

Scientiﬁc, Union City, CA, USA) and 13b (Catalog #7975AH, AK Scientiﬁc) were purchased

from commercial vendors. These intermediates were left in Scheme 3 for continuity.

3.1.2. General Procedure for synthesis of tripeptide halomethylketones

Synthesis for analog 21 presented. All other analogs and alternative synthetic proce-

dures reported in Supplementary Materials.

Dibenzyl 2-ﬂuoromalonate (3)

A mixture of dimethyl 2-ﬂuoromalonate 2 (2.0 g, 13 mmol, 1.0 eq), benzyl alcohol (7.0 g,

65 mmol, 4.9 eq,), p-toluenesulfonic acid monohydrate (150 mg, 0.80 mmol, 0.06 eq) in

toluene (5.6 mL) were placed in a round bottom ﬂask outﬁtted with a Dean-Stark apparatus

◦

and was heated with stirring to 75 C in vacuo (27 mm of Hg) until all of the toluene

had distilled. Then (75 mm Hg, 112 C) for an additional 5 h. The mixture was cooled

to 75 ^◦C and isopropanol (15 mL) was added followed by hexanes (30 mL). The mixture

was placed in the freezer and allowed to crystalize overnight. The product was ﬁltered,

washed with hexanes (2

×

30 mL), and dried overnight in vacuo, yielding

3

as a white solid

1

(3.0 g, 9.9 mmol, 76%). H NMR (500 MHz, Chloroform-d)

δ 7.64–7.28 (m, 10H), 5.37 (d,

J = 48.0 Hz, 1H), 5.25 (s, 4H); APCI-MS: m/z 301.1 [M − H]⁻.

3-(Benzyloxy)-2-ﬂuoro-3-oxopropanoic acid (4)

Intermediate

3 (2.0 g, 6.6 mmol, 1.0 eq) was suspended in isopropanol (12 mL) and

◦

heated to 45 C. 1.0 M aqueous NaOH (6.9 mL, 6.9 mmol, 1.1 eq) was added dropwise

over 1 h. After an additional 10 min, the solution was concentrated to 5 mL and water

(2.5 mL) was added. The pH was adjusted to 9 using saturated sodium bicarbonate and

washed with DCM (5

×

5 mL) to remove benzyl alcohol. The pH of the solution was

adjusted to 2.2 using 5 M HCl, and di-isopropyl ether (5 mL) was used to extract the

product. The pH of the aqueous layer was adjusted to 1.9 using 5 M HCl, and further

extracted with di-isopropyl ether (5 mL). The combined extracts were washed with brine

(5 mL), dried over MgSO₄, ﬁltered, and concentrated in vacuo at 35 ^◦C to provide an oily

residue. This was triturated with hexanes (7 mL) overnight with stirring to give a solid.

The solid was ﬁltered and dried under vacuum to yield

4 as a white solid (0.42 g, 2.0 mmol,

1

30%). H NMR (500 MHz, Chloroform-d)

5.32 (d, J = 1.1 Hz, 2H).

δ 7.38 (d, J = 2.6 Hz, 5H), 5.40 (d, J = 47.9 Hz, 1H),

Methyl (S)-4-((tert-butoxycarbonyl)amino)-6-ﬂuoro-5-oxohexanoate (8a)

To vial #1 was added 4 (1.3 g, 6.3 mmol, 1.2 eq) and THF (2 mL/mmol). This was cooled

to 0 ^◦C before adding 2.0 M isopropylmagnesium ch_◦loride in THF (6.30 mL, 12.5 mmol,

2.40 eq). The white suspension was stirred for 1 h at 0 C to generate the magnesium salt

to be used for the next reaction.

5

To vial #2 was added (S)-2-((tert-butoxycarbonyl)amino)-5-methoxy-5-oxopentanoic

acid 6a (1.4 g, 5.2 mmol, 1.0 eq) and THF (25 mL). The solution was cooled to 0 ^◦C before

addition of 1,1⁰-carbonyldiimidazole (0.85 g, 5.2 mmol, 1.0 eq). The reaction was stirred

◦

for 1 h at 0 C. After 1 h vial #2 was cooled to

of magnesium salt

−

20 C followed by dropwise addition

◦

5

from vial #1. The mixture was stirred for 45 min at

−

20 C, and

then allowed to warm to room temperature and stir for 3.5 h. The reaction contents were

then poured onto 1.0 M HCl, extracted with EtOAc, washed in succession with saturated

Molecules 2021, 26, 1227

16 of 24

sodium bicarbonate and brine then dried over sodium sulfate. The organic layer was

then ﬁltered and concentrated under reduced pressure to provide crude 1-benzyl 7-methyl

(4S)-4-((tert-butoxycarbonyl)amino)-2-ﬂuoro-3-oxoheptanedioate 7a.

The residue was dissolved in toluene and Pd/C (0.10 g, 0.94 mmol, 0.18 eq) was added

followed by bubbling with hydrogen gas through the solution for 10 min. The bubbling

was stopped and the reaction was allowed to stir under hydrogen atmosphere overnight

at room temperature. The mixture was then ﬁltered, concentrated, and puriﬁed via ﬂash

chromatography (5–60% EtOAc in hexanes) to yield 8a (0.85 g, 3.1 mmol, 59%) as a white

1

solid. H NMR (500 MHz, Chloroform-d)

δ 5.21 (s, 1H), 5.15–4.91 (m, 2H), 4.62 (s, 1H), 3.68

(s, 3H), 2.57–2.31 (m, 2H), 2.27–2.16 (m, 1H), 1.88 (dq, J = 14.8, 7.4 Hz, 1H), 1.44 (s, 9H).

APCI-MS: m/z 278.0 [M + H]⁺.

Methyl ((benzyloxy)carbonyl)-L-phenylalanyl-L-alaninate (12g)

To a solution of ((benzyloxy)carbonyl)-L-phenylalanine (0.72 g, 2.4 mmol, 1.0 eq) and

N-methylmorpholine (0.97 g, 9.6 mmol, 4.0 e_◦q) in anhydrous THF was added isobutyl

chloroformate (0.49 g, 3.6 mmol, 1.5 eq) at 0 C. The methyl L-alaninate hydrochloride

(0.33 g, 2.4 mmol, 1.0 eq) was then added and the reaction was allowed to warm slowly to

room temperature overnight. The next day the reaction mixture was diluted with EtOAc

(25 mL) and washed in succession with 1.0 M HCl (50 mL), 5% sodium carbonate solution

(50 mL), brine (50 mL), then dried over sodium sulfate before ﬁltering and concentrating in

vacuo. The residue was recrystallized in a DCM:Hexanes solution and ﬁltered, yielding

1

12g as a white crystalline solid (0.61 g, 1.6 mmol, 66%). H NMR (500 MHz, DMSO-d₆)

δ

8.49 (d, J = 7.1 Hz, 1H), 7.49 (d, J = 8.8 Hz, 1H), 7.37–7.02 (m, 10H), 4.88 (s, 2H), 4.39–4.10

(m, 2H), 3.58 (s, 3H), 2.96 (dd, J = 13.9, 3.7 Hz, 1H), 2.67 (dd, J = 13.7, 11.0 Hz, 1H), 1.28 (d,

J = 7.3 Hz, 3H); APCI-MS: m/z 385.1 [M + H]⁺.

((Benzyloxy)carbonyl)-L-phenylalanyl-L-alanine (13g)

The dipeptide methyl ester 12g (0.61 g, 1.6 mmol, 1.0 eq) was dissolved in THF:H₂O

(4:1, 10 mL) followed by addition of LiOH (0.046 g, 1.9 mmol, 1.2 eq) at 0 ^◦C. The mixture

◦

was stirred at 0 C for 3 h before acidifying with 1.0 M HCl (2 mL) to < pH 3. The aqueous

phase was extracted with EtOAc (3

dried over sodium sulfate, ﬁltered, and concentrated to yield 13g as a white solid (0.52 g,

×

5 mL) and the organic layers were combined and

1.4 mmol, 87%) which was taken forward crude to the next step.

Cbz-L-Phe-L-Ala-L-Glu(OMe)-ﬂuoromethylketone (21)

To vial #1 was added 8a (0.050 g, 0.18 mmol, 1.0 equiv) and 4.0 M HCl in 1,4-dioxane

(1.0 mL, 4.0 mmol, 22 eq). This was allowed to stir for 30 min at room temperature and

monitored for completion by TLC, which indicated the formation of a very polar product.

Upon completion, the solvent was removed in vacuo, and the oil was triturated with diethyl

ether for 10 min and carefully decanted and dried in vacuo. The resulting Boc-removed

Glu(OMe)-ﬂuoromethylketone oil was diluted in THF and used directly in the next step.

To vial #2 was added 13g (0.054 g, 0.14 mmol, 0.8 eq) in THF under argon atmosphere.

To this solution was added isobutyl chloroformate _◦(0.036 mL, 0.27 mmol, 1.5 eq) and N-

methylmorpholine (0.079 mL, 0.72 mmol, 4.0 eq) at 0 C under argon atmosphere. This was

◦

allowed to stir for 45 min at 0 C after which the deprotected Glu(OMe)-ﬂuoromethylketone

◦

in THF from vial #1 was added dropwise at 0 C and allowed to stir for 15 min before

warming to room temperature slowly overnight. The next day the reaction was diluted

with EtOAc and washed in succession with 1 N HCl, 5% sodium bicarbonate, and brine.

The organic layer was collected and dried over sodium sulfate, ﬁltered, and concentrated in

vacuo to provide the crude product. This residue was then recrystallized in DCM:Hexanes

overnight, ﬁltered, and dried in vacuo to yield 21 as a white solid (0.033 g, 0.063 mmol,

1

35%). H NMR (800 MHz, DMSO-d₆)

δ 8.33 (d, J = 7.4 Hz, 1H), 8.27 (d, J = 6.8 Hz, 1H), 7.51

(d, J = 8.5 Hz, 1H), 7.35–7.14 (m, 11H), 5.25–5.07 (m, 2H), 4.93 (s, 2H), 4.40–4.15 (m, 3H), 3.57

(s, 3H), 3.01 (dd, J = 14.0, 3.7 Hz, 1H), 2.78–2.64 (m, 1H), 2.40–2.26 (m, 2H), 2.12–1.98 (m,

Molecules 2021, 26, 1227

17 of 24

1H), 1.86–1.63 (m, 1H), 1.26 (d, J = 7.1 Hz, 3H). ¹³C NMR (200 MHz, DMSO-d₆)

δ

203.7 (d,

J = 14 Hz), 173.3, 173.1, 172.0, 156.3, 138.5, 137.4, 129.6, 128.7, 128.4, 128.1, 127.8, 126.6, 84.3

(d, J = 178 Hz), 65.6, 56.3, 54.7, 51.8, 48.7, 37.7, 29.7, 24.6, 18.0. ESI-MS: m/z 530.0 [M + H]⁺

HPLC t_R= 11.974 min (Column B); HPLC Purity: 98.2%.

;

3.2. Biological Evaluation

3.2.1. Fluorescence-based Ub-Rho Deubiquitinase Activity Assay

The assay was performed in black 384-well plates (Catalog #12566624, Fisher, Waltham,

MA, USA). DUB stock solutions were diluted in reaction buffer (50 mm Tris pH 7.6, 0.5 mm

EDTA, 5 mm DTT, 0.1 % (w/v) BSA) to a concentration of 2.5 nM for UCHL1 or 0.025 nM

for UCHL3. Stock solutions of Ub-Rhodamine 110 (U-555, Boston Biochem, Cambridge,

MA, USA); 125 nM for UCHL1 assay, and 250 nM for UCHL3 assay) were prepared in the

same buffer. A 10 mM stock solution was made for each inhibitor in DMSO, then a dilution

of 600

was added 20

inhibitor concentrations of ranging from 0.78

well. These were allowed to incubate, while sealed, for the 3 h at room temperature. After

incubation 20 L of each Ub-Rho stock solution was added to the respective wells for each

µ

M in reaction buffer was made followed eight by 1:1 serial dilutions. To each well

L of DUB stock solution and 10 L of inhibitor solutions for nine ﬁnal

M–200 M along with a DMSO only control

µ

DUB to yield ﬁnal concentration of 50 nM for UCHL1 or 100 nM for UCHL3, respectively.

Plates were read immediately and ﬂuorescence of cleaved Rhodamine 110 ﬂuorophore

was monitored at

λ

ex

= 485 nm,

λ

em

= 535 nm continuously for 20 min on a Synergy Neo2

instrument (BioTek, Winooski, VT, USA) The raw data was loaded into GraphPad Prism 8

(GraphPad Software, San Diego, CA, USA; www.graphpad.com, accessed on 12 January

2021) and the slope of the linear portion of the ﬂuorescence vs. time curves was calculation

for each inhibitor concentration and % activity of the enzyme was determined compared to

DMSO-treated controls. The % activity was plotted as a function of inhibitor concentration

and the data was ﬁtted with non-linear regression analysis to calculate the IC₅₀values.

Errors are reported as the 95% conﬁdence interval calculated from experiments collected in

technical triplicate.

3.2.2. Time-Dependent Inhibitor Analysis

This assay was performed the same as described above with the exception that instead

of 3-h pre-incubation followed by Ub-Rho addition the Ub-Rho substrate was added at

20-min increments following addition of inhibitor to wells beginning at 0 min and ending

at 120 min. Plates were read continuously for 20 min following addition of Ub-Rho. The

% activity of each inhibitor concentration was the plotted vs. time of pre-incubation and

the slope of this plot is the pseudo ﬁrst-order rate constant (k_obs) for inhibition at each

concentration. The k_obswas plotted as a function of inhibitor concentration and then ﬁt to

the equation Y = k_inactX/(K_I+ X) to calculate k_inact, K_Iand k_inact/K_Ias described by Resnick

et al. [64].

3.2.3. Ub Binding Afﬁnity Analysis Using Biolayer Interferometry

This assay was carried out according to our previously published protocol using

an Octet RED384 biolayer interferometer (ForteBio, Fremont, CA, USA) [69]. A solution

containing 5 µM His-UCHL1 was incubated with 2 mM VAEFMK 1 or DMSO overnight in

reaction buffer (50 mM Tris pH 7.6, 0.5 mM EDTA, 5 mM DTT) at room temperature before

buffer exchanging into water using Zeba spin desalting columns (Thermo Scientiﬁc, catalog

no. 89882). The concentration of His-UCHL1 was determined by A₂₈₀on a NanoDrop

system (Thermo Scientiﬁc), after which His-UCHL1 was diluted into BLI buffer [1

×

PBS

containing 0.05% (v/v) Tween 20 and 0.1% (w/v) bovine serum albumin (BSA)]. The

concentration of Ub was determined by the BCA assay and diluted to top concentrations

into BLI buffer, and 1:1 serial dilutions were completed. The top concentration of Ub

was 2

µM. 40 µL of each solution was added to a 384-well tilted-bottom plate (part no.

185080, Molecular Devices, San Jose, CA, USA). The Ni-NTA biosensor was dipped ﬁrst

Molecules 2021, 26, 1227

18 of 24

into BLI buffer (initial baseline, 60 s), then into the His-UCH protein wells (loading step,

300 s), then into BLI buffer alone (baseline step, 60 s), then into the lowest concentration

of UbV (association step, 120 s), and then into buffer alone (dissociation step, 100 s).

A reference sensor loaded with protein was dipped into wells containing only buffer to

adjust for protein-buffer signals. The association-dissociation was repeated with increasing

concentrations of UbV. All measurements were taken at 30 ^◦C.

Octet RED384 Data Analysis Software (ForteBio, version 9.0.0.15) was used to collect

and analyze the raw data for the association and dissociation curves. After subtraction of a

reference sensor (loaded sensors dipped into wells containing only buffer), averages of the

association responses (in nanometer response signal from 110 to 115 s) were calculated and

plotted as a function of UbV concentration in Prism 8. These data were ﬁt to a nonlinear

regression one-site speciﬁc binding model to determine a K_d.

3.2.4. Protein Expression and Puriﬁcation

UCHL1, UCHL1^C90A, and UCHL3 (Expressed in pET15b by GenScript) for biochemical

assays were grown in LB growth medium at 37 ^◦C to an optical density of 0.6–0.8. After

0.1 mm IPTG induction at 17 ^◦C for 18 h the bacteria were lysed and pelleted at 15,000g,

and clariﬁed lysate was puriﬁed on HisPur Ni-NTA resin (Thermo Scientiﬁc) according to

manufacturer’s instructions.

3.2.5. Copper-Catalyzed Click Reaction and In-Gel Fluorescence for Wild-Type and

C90A UCHL1

Solutions containing 1 mg/mL recombinant UCHL1^WTor UCHL1^C90A(from pET15b)

(49

block buffer (Thermo Scientiﬁc # 37538). After 3 h, 6

was added to each sample, consisting of: 3 L 1.7 mM THPTA in 1:4 DMSO/tBuOH

(100 M ﬁnal concentration), 1 L 50 mM CuSO₄(1 mM ﬁnal concentration), 1 L TCEP

HCl (25 M ﬁnal concentration), and 1 uL Cy5-Azide (#AZ118, Click Chemistry Tools,

Scottsdale, AZ, USA; 25 M ﬁnal concentration). The samples were allowed to react for

2 h at room temperature protected from light before precipitating the sample by adding

600 L MeOH, 150 L chloroform, and 400 L water, vortexing between each addition.

The samples were pelleted by centrifuging at 170,000 g for 5 min. The upper aqueous

layer was removed without disturbing the precipitate, and 450 L MeOH was added

again, followed by centrifuging at 17,000 g for 5 min and decanting the solvent (2X). The

µL) was incubated with 10

µM inhibitor for 3 h at room temperature in PBS starting

µL of a freshly prepared cocktail

µ

×

µ

×

samples were allowed to air dry for 30 min before resuspending in 1

×

Laemmli buffer and

boiling at 90 ^◦C for 5 min. The samples were then analyzed by SDS-PAGE and imaged on

an Odyssey system (LI-COR, Lincoln, NE, USA) and subsequently Coomassie stained.

3.2.6. Mass Spectrometry Analysis of 34 Adducts on UCHL1

Solutions containing 10

L) was incubated with either 10

µ

M recombinant UCHL1^WTor UCHL1^C90A(from pET15b)

M or 20 M inhibitor for 3 h at room temperature in

(49

µ

PBS starting block buffer (Thermo Scientiﬁc # 37538). To quench reaction, chilled acetone

was added to the sample to achieve an overall 80% acetone solution. After overnight storage

at

−

20 ^◦C, suspension was centrifuged at 14,000

×

g at 4 ^◦C for 15 min. Supernatant was

discarded, and pellet was dried by vacuum centrifugation for 1 h. Pellet was reconstituted

in 50/50 water/acetonitrile with 0.1% formic acid. 25 pmol per sample was analyzed by

LC/MS (Agilent 1260 Inﬁnity II with a ZORBAX Rapid Resolution High Deﬁnition 300Å

Stable Bond C3, 2.1

mass spectrometer in positive ion mode.

×

100 mm, 1.8 µm column) attached to an Agilent 6129 quadrupole

The column was held at 45 ^◦C. Mobile solution A was 0.1% formic acid in water, and

mobile phase B was 0.1% formic acid in acetonitrile. The gradient used was hold at 5% B

for 5 min, increase linearly to 100% B for 10, and then hold at 95% B for 5 min. The mass

data were collected at a range of 500–1000 m/z.

Raw data were processed using MestReNova. For all samples, the deconvoluted mass

was calculated with a charged state range from 27 to 36, m/z range of 670–900 Da, and

Molecules 2021, 26, 1227

19 of 24

deconvoluted mass range from 25,000–26,000 Da. Representative data had an abundance

threshold of 10–20% for charged state deconvolution calculation. For C90A mutant, this

abundance threshold was lowered to 5% to detect the presence of any reacted protein. 34

adduct is expected to be +486.00 Da after displacement of ﬂuorine.

3.2.7. KMS Cell Proliferation Assay

The proliferation of KMS11 and KMS12 cells was monitored using an IncuCyte devise

(Sartorius, Goettingen, Germany; ww w .essenbioscience.com, accessed on 31 August 2020).

Cells (3

×

10⁴per well; 200

µL) were seeded in triplicate in 96 well plates the night prior to

the addition of compounds. The IncuCyte was set to collect images from ﬁve ﬁelds per

well at 4-h intervals by using the nuclear dye Nuclight Red (Sartorius). At each timepoint,

the number of cells per well was deter- mined by taking the mean of the ﬁve locations in

each well. The relative proliferation at each timepoint was normalized, with 100% set as

the number of cells observed for incubated with DMSO (0.1%) controls at 72 h.

3.2.8. KMS Cell Activity-Based Probe Target Engagement Assay

Twelve hours prior to the addition of compounds to the cells, 1.2 million KMS11 or

KMS12 cells mL⁻¹per well were seeded in a 12 well plate. Cells were incubated with

inhibitor for 4 h before washing the cells three time with PBS to remove excess compound.

Cells were harvested, pelleted, and lysed in 60

µL of lysis buffer (0.5%NP-40 in 50 mM Tris

pH 7.4). HA-Ub-VME was then added to a ﬁnal concentration of 1

µM and incubated for

1 h at room temperature. An equal volume of 2X Laemmli sample buffer was added, and

the samples were heated to 95 ^◦C before separating by gel electrophoresis and analyzed

by immunoblot.

3.2.9. SW1271 Cell Viability Assay

Luciferase-constructed cancer cells were seeded in 96-well plates at 10,000 cells/well

with the compounds at the indicated concentrations. DMSO was used as vehicle control.

Luminescence readings were acquired after adding 1% D-luciferin (GoldBio, St. Louis,

MO, USA) after 24 h. Cell viability was determined by the ratio of luminescence reading

compared to the one at 0 h, and normalized to vehicle control.

3.2.10. SW1271 Cell Activity-Based Probe Target Engagement Assay

Cells were treated with 100

µM inhibitor or DMSO for 4 h before being washed,

scraped, and collected. Cell pellets were thawed on ice before resuspending in 200

µL

lysis buffer (50 mM Tris pH 7.6, 150 mM NaCl, 5 mM MgCl₂, 0.5 mM EDTA, 5 mM DTT,

2 mM ATP, 0.5% NP-40, 10% v/v glycerol) and incubating on ice for 30 min with vigorous

vortexing every 10 min. Cell lysate was clariﬁed by centrifugation (17,000

and the supernatant was collected, and the concentration was determined by Bradford

(performed according to manufacturer’s protocol). Cells were normalized to 2 mg mL⁻¹

To each reaction mixture was added 36 L of lysate and 2 L of HA-Ub-VME or lysis buffer.

×

g for 10 min)

.

µ

This was incubated for 30 min at room temperature before quenching with 15

Laemmli buffer and heating to 95 C for 5 min. Samples were subjected to SDS-PAGE

and immunoblot.

µ

L of 4X

◦

3.2.11. Copper-Catalyzed Click Reaction and In-Gel Fluorescence of Treated SW1271 Cells

Cells were treated with 100

scraped, and lysed in 200 L of lysis buffer (1

HALT) for 30 min on ice with vigorous vortexing every 10 min. Samples were then pelleted

at 17,000 g for 10 min before collecting the supernatant and measuring concentration

using BCA. Samples were normalized to 2 mg/mL and stored at

80 ^◦C overnight. The

following day, samples were thawed. While thawing, a click cocktail was prepared (for

each sample, the cocktail consists of: 3 L of 1.7 mM THPTA in 1:4 tBuOH/DMSO, 1

of 50 mM CuSO4 in water, 1 L of 1.25 mM Cy5-N3 in DMSO, and 1 L of 50 mM TCEP

µM inhibitor or DMSO for 4 h before being washed,

µ

×

PBS, 1% v/v Triton X-100, 0.1% w/v SDS, 1X

×

−

µ

µL

µ

Molecules 2021, 26, 1227

20 of 24

in water freshly prepared). 50 µL of each sample was added to a new tube, and 6 µL

of click cocktail was also added to these tubes. The samples were allowed to react for

2 h at room temperature protected from light before precipitating the sample by adding

600

The samples were pelleted by centrifuging at 17,000

layer was removed without disturbing the precipitate, and 450

again, followed by centrifuging at 17,000 g for 5 min (2X) and decanting the solvent. The

µL MeOH, 150

µL Chloroform, and 400

µ

L water, vortexing between each addition.

g for 5 min. The upper aqueous

L MeOH was added

×

µ

×

samples were allowed to air dry for 30 min before resuspending in 4X Laemmli buffer and

◦

boiling at 90 C for 5 min. The samples were then analyzed by SDS-PAGE and imaged on a

Licor Odyssey and subsequently Coomassie stained.

3.2.12. shRNA UCHL1 Knockdown in SW1271 Cells

pLKO.1 TRC vectors with shRNAs targeting human UCHL1 were packaged into

lentiviral particles and used for stable transduction of the SW1271 cells via puromycin

selection. The shRNA sequences are listed in Supplementary Table S3. The pLKO.1 vector

with non-targeting, scrambled shRNA was used as a control. Depletion of UCHL1 was

determined via immunoblotting.

3.2.13. Scratch Wound Healing Cell Migration Assay

Wild-type or shRNA treated UCHL1 knockdown SW1271 cells were seeded in 12-well

plates at 1 million cells/well. The cells were cultured overnight to grow as a conﬂuent single

layer. The cross-style wounds were created using pipet tips to scratch on the single layer of

cells. The compounds were prepared in full growth media at the desired concentrations

and added after the wounds were created. DMSO was used as the vehicle control. Photos

of the wounds were captured immediately after adding treatments at 0 h. Photos of the

wounds were captured at 0 and 24 h. The areas of cross-style wounds were quantiﬁed

using Image J 1.52a software. The migration rate (%) was deﬁned as the percentage of area

at each timepoint compared to the one at 0 h.

4. Conclusions

In summary, we have presented an SAR study and biological characterization of an

irreversible covalent inhibitor of UCHL1. Starting with the parent compound VAEFMK

1

it was determined that bulky, lipophilic moieties are preferred in the valine position,

while alteration of the alanine position was not tolerated. Additionally, it was determined

that unnatural D-amino acids abrogate activity. The SAR at these two positions could be

rationalized by observations of the VAEFMK-bound UCHL1 crystal structure. Installing

an alkyne handle to the benzyloxy capping group of the peptide improved activity when

compared to three nearest neighbor analogs suggesting future SAR work on the benzyloxy

may improve potency. To round out SAR it was determined that modiﬁcation of the

ﬂuoromethylketone electrophilic site to a more reactive chloromethylketone surprisingly

lost activity versus UCHL1 in vitro. Furthermore, when the chloromethylketone derivative

was tested in cellular assays it was toxic and shown to non-speciﬁcally label numerous

proteins. Conversely, the ﬂuoromethylketones were selective for only UCHL1 over the

closely related UCHL3. Ultimately, the scaffold was improved from a UCHL1 IC₅₀value of

24

and only slight inhibition of UCHL5 at 250

UCHL1 selectively in two different cancer relevant cells lines. Molecule

µ

M for VAEFMK

1

to 7.7

µ

M for 34 with no activity against the closest related UCHL3

M. Finally, the probes were shown to engage

displayed anti-

µ

1

proliferative properties versus KMS11 myeloma cells dependent on UCHL1 while the

compound displayed no effect on KMS12 cells that do not express UCHL1. Analog 34

displayed anti-migratory properties in SW1271 small cell lung cancer cell lines consistent

with observed behavior from genetic depletion of UCHL1. While the probe displays

modest potency toward UCHL1 it is quite selective in cells and allowed for probing UCHL1

without reduction in cell viability as opposed to the shRNA knockdowns, which displayed

some reduced viability. A potential way to improve the cellular activity of the scaffold

Molecules 2021, 26, 1227

21 of 24

may be to alter the Glu(OMe) moiety. This ester is likely cleaved by cellular esterases

leaving the carboxylate side chain, which was shown to be inactive in vitro. The cellular

inhibition is likely a competition between inhibitor engagement with UCHL1 and the

ester being cleaved. Further optimization of this residue is ongoing. Alternative UCHL1

inhibitors from the cyanopyrrolidine scaffold, although selective in vitro, still exhibit off-

target toxicity in the micromolar range while 34 does not exhibit any cellular toxicity up to

100 µM. Analog 34 is yet another option to selectively probe UCHL1 and offers additional

capabilities in that an alkyne is already installed to be leveraged for various chemical

biology applications.

Supplementary Materials: The following are available online, Scheme S1: Synthesis of p-ethynyl-

benzyloxy-protected dipeptides, Table S1: UCHL1 inhibition provided as 95% conﬁdence interval,

Figure S1: Dose-response curves for analogs 1 (A) and 34 (B) against UCHL3 Figure S2: Proposed

cyclization of glutamate-ﬂuoromethylketone, Figure S3: Dark-exposure HA-immunoblot data for

KMS11, Table S2: Cell lines and culture conditions, Table S3: Detailed information for shRNA vectors.

Synthetic procedures and characterization of molecules also provided.

Author Contributions: Conceptualization, A.D.K., C.D. and D.P.F.; Synthesis, A.D.K. and R.D.I.; data

collection and analysis, A.D.K., H.C., S.H., C.S.H., R.D.I., C.S.M.; writing—original draft preparation,

A.D.K. and D.P.F.; writing—review and editing, A.D.K., H.C., S.H., C.S.H.; R.D.I., C.S.M., C.D., P.J.G.,

M.K.W. and D.P.F.; supervision, P.J.G., M.K.W., D.P.F.; project administration, D.P.F.; All authors have

read and agreed to the published version of the manuscript.

Funding: The authors gratefully acknowledge the support of the Purdue Interdepartmental NMR

Facility as a shared resource from the Purdue Center for Cancer Research, NIH grant P30 CA023168.

The authors are also grateful for the support from the Purdue University Center for Cancer Research

Small Grants Program, in particular the Phase 1 Concept Award. Finally, the authors acknowledge

the Purdue University College of Pharmacy for additional support.

Institutional Review Board Statement: Not applicable.

Informed Consent Statement: Not applicable.

Data Availability Statement: Data is contained within the article or Supplementary Material.

Acknowledgments: Not applicable.

Conﬂicts of Interest: The funders had no role in the design of the study; in the collection, analyses,

or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

Sample Availability: Samples of the compounds are available from the corresponding author

upon request.

References

1.

Acconcia, F.; Sigismund, S.; Polo, S. Ubiquitin in trafﬁcking: The network at work. Exp. Cell Res. 2009, 315, 1610–1618. [CrossRef]

[PubMed]

2.

3.

Jackson, S.P.; Durocher, D. Regulation of DNA Damage Responses by Ubiquitin and SUMO. Mol. Cell 2013, 49, 795–807. [CrossRef]

Myung, J.; Kim, K.B.; Crews, C.M. The ubiquitin-proteasome pathway and proteasome inhibitors. Med. Res. Rev. 2001, 21, 245–273

[CrossRef] [PubMed]

.

4.

Schulman, B.A.; Wade Harper, J. Ubiquitin-like protein activation by E1 enzymes: The apex for downstream signalling pathways.

Nat. Rev. Mol. Cell Biol. 2009, 10, 319–331. [CrossRef]

5.

6.

7.

8.

Ye, Y.; Rape, M. Building ubiquitin chains: E2 enzymes at work. Nat. Rev. Mol. Cell Biol. 2009, 10, 755–764. [CrossRef]

Deshaies, R.J.; Joazeiro, C.A.P. RING Domain E3 Ubiquitin Ligases. Annu. Rev. Biochem. 2009, 78, 399–434. [CrossRef]

Komander, D. The emerging complexity of protein ubiquitination. Biochem. Soc. Trans. 2009, 37, 937–953. [CrossRef] [PubMed]

Rahighi, S.; Ikeda, F.; Kawasaki, M.; Akutsu, M.; Suzuki, N.; Kato, R.; Kensche, T.; Uejima, T.; Bloor, S.; Komander, D.; et al.

Speciﬁc Recognition of Linear Ubiquitin Chains by NEMO Is Important for NF-κB Activation. Cell 2009, 136, 1098–1109. [CrossRef]

[PubMed]

9.

Haakonsen, D.L.; Rape, M. Branching Out: Improved Signaling by Heterotypic Ubiquitin Chains. Trends Cell Biol.

2019, 29, 704–716. [CrossRef]

10. Komander, D.; Clague, M.J.; Urbé, S. Breaking the chains: Structure and function of the deubiquitinases. Nat. Rev. Mol. Cell Biol.

2009, 10. [CrossRef]

Molecules 2021, 26, 1227

22 of 24

11. Abdul Rehman, S.A.; Kristariyanto, Y.A.; Choi, S.Y.; Nkosi, P.J.; Weidlich, S.; Labib, K.; Hofmann, K.; Kulathu, Y. MINDY-1

Is a Member of an Evolutionarily Conserved and Structurally Distinct New Family of Deubiquitinating Enzymes. Mol. Cell

2016, 63, 146–155. [CrossRef]

12. Hewings, D.S.; Heideker, J.; Ma, T.P.; AhYoung, A.P.; El Oualid, F.; Amore, A.; Costakes, G.T.; Kirchhofer, D.; Brasher, B.;

Pillow, T.; et al. Reactive-site-centric chemoproteomics identiﬁes a distinct class of deubiquitinase enzymes. Nat. Commun.

2018, 9, 1162. [CrossRef]

13. Jara, J.H.; Frank, D.D.; Özdinler, P.H. Could Dysregulation of UPS be a Common Underlying Mechanism for Cancer and

Neurodegeneration? Lessons from UCHL1. Cell Biochem. Biophys. 2013, 67, 45–53. [CrossRef] [PubMed]

14. Hussain, S.; Zhang, Y.; Galardy, P. DUBs and cancer: The role of deubiquitinating enzymes as oncogenes, non-oncogenes and

tumor suppressors. Cell Cycle 2009, 8, 1688–1697. [CrossRef]

15. Lim, K.-H.; Baek, K.-H. Deubiquitinating enzymes as therapeutic targets in cancer. Curr. Pharm. Des. 2013, 19, 4039–4052.

[CrossRef]

16. Atkin, G.; Paulson, H. Ubiquitin pathways in neurodegenerative disease. Front. Mol. Neurosci. 2014, 7, 63. [CrossRef]

17. Todi, S.V.; Paulson, H.L. Balancing act: Deubiquitinating enzymes in the nervous system. Trends Neurosci. 2011, 34, 370–382.

[CrossRef]

18. Kemp, M. Recent Advances in The Discovery of Deubiquitinating Enzyme Inhibitors, 1st ed.; Elsevier B.V.: Amsterdam, The Netherlands,

2016.

19. Schauer, N.; Magin, R.S.; Liu, X.; Doherty, L.; Buhrlage, S. Advances in Discovering Deubiquitinating Enzyme (DUB) Inhibitors.

J. Med. Chem. 2019. [CrossRef] [PubMed]

20. Gong, B.; Radulovic, M.; Figueiredo-Pereira, M.E.; Cardozo, C. The Ubiquitin-Proteasome System: Potential Therapeutic Targets

for Alzheimer’s Disease and Spinal Cord Injury. Front. Mol. Neurosci. 2016, 9, 4. [CrossRef]

21. Fang, Y.; Fu, D.; Shen, X.Z. The potential role of ubiquitin c-terminal hydrolases in oncogenesis. Biochim. Biophys. Acta Rev. Cancer

2010, 1806, 1–6. [CrossRef] [PubMed]

22. Bishop, P.; Rocca, D.; Henley, J.M. Ubiquitin C-terminal hydrolase L1 (UCH-L1): Structure, distribution and roles in brain function

and dysfunction. Biochem. J. 2016, 473, 2453–2462. [CrossRef] [PubMed]

23. Day, I.N.M.; Thompson, R.J. UCHL1 (PGP 9.5): Neuronal biomarker and ubiquitin system protein. Prog. Neurobiol.

2010, 90, 327–362. [CrossRef] [PubMed]

24. Leroy, E.; Boyer, R.; Auburger, G.; Leube, B.; Ulm, G.; Mezey, E.; Harta, G.; Brownstein, M.J.; Jonnalagada, S.; Chernova, T.; et al.

The ubiquitin pathway in Parkinson’s disease. Nature 1998, 395, 451–452. [CrossRef]

25. Janda, E.; Isidoro, C.; Carresi, C.; Mollace, V. Defective autophagy in Parkinson’s disease: Role of oxidative stress. Mol. Neurobiol.

2012, 46, 639–661. [CrossRef] [PubMed]

26. Nishikawa, K.; Li, H.; Kawamura, R.; Osaka, H.; Wang, Y.L.; Hara, Y.; Hirokawa, T.; Manago, Y.; Amano, T.; Noda, M.; et al.

Alterations of structure and hydrolase activity of parkinsonism-associated human ubiquitin carboxyl-terminal hydrolase L1

variants. Biochem. Biophys. Res. Commun. 2003, 304, 176–183. [CrossRef]

27. Setsuie, R.; Wada, K. The functions of UCH-L1 and its relation to neurodegenerative diseases. Neurochem. Int. 2007, 51, 105–111.

[CrossRef]

28. Hussain, S.; Bedekovics, T.; Liu, Q.; Hu, W.; Jeon, H.; Johnson, S.H.; Vasmatzis, G.; May, D.G.; Roux, K.J.; Galardy, P.J.

UCH-L1 bypasses mTOR to promote protein biosynthesis and is required for MYC-driven lymphomagenesis in mice. Blood

2018, 132, 2564–2574. [CrossRef] [PubMed]

29. Bedekovics, T.; Hussain, S.; Feldman, A.L.; Galardy, P.J. UCH-L1 is induced in germinal center B cells and identiﬁes patients with

aggressive germinal center diffuse large B-cell lymphoma. Blood 2016, 127, 1564–1574. [CrossRef] [PubMed]

30. Carolan, B.J.; Heguy, A.; Harvey, B.G.; Leopold, P.L.; Ferris, B.; Crystal, R.G. Up-regulation of expression of the ubiquitin

carboxyl-terminal hydrolase L1 gene in human airway epithelium of cigarette smokers. Cancer Res. 2006, 66, 10729–10740.

[CrossRef]

31. Shimada, Y.; Kudo, Y.; Maehara, S.; Matsubayashi, J.; Otaki, Y.; Kajiwara, N.; Ohira, T.; Minna, J.D.; Ikeda, N. Ubiquitin C-terminal

hydrolase-L1 has prognostic relevance and is a therapeutic target for high-grade neuroendocrine lung cancers. Cancer Sci.

2020, 111, 610–620. [CrossRef] [PubMed]

32. Chen, G.; Gharib, T.G.; Huang, C.-C.; Thomas, D.G.; Shedden, K.A.; Taylor, J.M.G.; Kardia, S.L.R.; Misek, D.E.; Giordano, T.J.;

Iannettoni, M.D. Proteomic analysis of lung adenocarcinoma: Identiﬁcation of a highly expressed set of proteins in tumors.

Clin. Cancer Res. 2002, 8, 2298–2305. [PubMed]

33. Sanchez-Diaz, P.C.; Chang, J.C.; Moses, E.S.; Dao, T.; Chen, Y.; Hung, J.Y. Ubiquitin carboxyl-Terminal esterase L1 (UCHL1) is

associated with stem-like cancer cell functions in pediatric high-grade glioma. PLoS ONE 2017, 12, e0176879. [CrossRef] [PubMed ]

34. Hussain, S.; Bedekovics, T.; Chesi, M.; Bergsagel, L.P.; Galardy, P.J. UCHL1 is a biomarker of aggressive multiple myeloma

required for disease progression. Oncotarget 2015, 5. [CrossRef]

35. Hussain, S.; Bedekovics, T.; Ali, A.; Zaid, O.; May, D.G.; Roux, K.J.; Galardy, P.J. A cysteine near the C-terminus of UCH-L1 is

dispensable for catalytic activity but is required to promote AKT phosphorylation, eIF4F assembly, and malignant B-cell survival.

Cell Death Discov. 2019, 1–9. [CrossRef]

Molecules 2021, 26, 1227

23 of 24

36. Otsuki, T.; Yata, K.; Takata-Tomokuni, A.; Hyodoh, F.; Miura, Y.; Sakaguchi, H.; Hatayama, T.; Hatada, S.; Tsujioka, T.; Sato, Y.

Expression of protein gene product 9 5 (PGP9 5)/ubiquitin-C-terminal hydrolase 1 (UCHL-1) in human myeloma cells. Br. J.

Haematol. 2004, 127, 292–298. [CrossRef] [PubMed]

·

37. Gu, Y.Y.; Yang, M.; Zhao, M.; Luo, Q.; Yang, L.; Peng, H.; Wang, J.; Huang, S.K.; Zheng, Z.X.; Yuan, X.H.; et al. The de-ubiquitinase

UCHL1 promotes gastric cancer metastasis via the Akt and Erk1/2 pathways. Tumor Biol. 2015, 36, 8379–8387. [CrossRef]

[PubMed]

38. Liu, S.; González-Prieto, R.; Zhang, M.; Geurink, P.P.; Kooij, R.; Iyengar, P.V.; van Dinther, M.; Bos, E.; Zhang, X.;

Le Dévédec, S.E.; et al. Deubiquitinase activity proﬁling identiﬁes UCHL1 as a candidate oncoprotein that promotes

TGFβ-induced breast cancer metastasis. Clin. Cancer Res. 2020, 26, 1460–1473. [CrossRef]

39. Jang, M.J.; Baek, S.H.; Kim, J.H. UCH-L1 promotes cancer metastasis in prostate cancer cells through EMT induction. Cancer Lett.

2011, 302, 128–135. [CrossRef]

40. Ma, Y.; Zhao, M.; Zhong, J.; Shi, L.; Luo, Q.; Liu, J.; Wang, J.; Yuan, X.; Huang, C. Proteomic proﬁling of proteins associated with

lymph node metastasis in colorectal cancer. J. Cell. Biochem. 2010, 110, 1512–1519. [CrossRef]

41. Miyoshi, Y.; Nakayama, S.; Torikoshi, Y.; Tanaka, S.; Ishihara, H.; Taguchi, T.; Tamaki, Y.; Noguchi, S. High expression of ubiquitin

caboxy-terminal hydrolase-L1 and -L3 mRNA predicts early recurrence in patients with invasive breast cancer. Cancer Sci.

2006, 97, 523–529. [CrossRef]

42. Goto, Y.; Zeng, L.; Yeom, C.J.; Zhu, Y.; Morinibu, A.; Shinomiya, K.; Kobayashi, M.; Hirota, K.; Itasaka, S.; Yoshimura, M.; et al.

UCHL1 provides diagnostic and antimetastatic strategies due to its deubiquitinating effect on HIF-1

[CrossRef] [PubMed]

α. Nat. Commun. 2015, 6, 6153.

43. Kim, H.J.; Kim, Y.M.; Lim, S.; Nam, Y.K.; Jeong, J.; Kim, H.-J.; Lee, K.-J. Ubiquitin C-terminal hydrolase-L1 is a key regulator of

tumor cell invasion and metastasis. Oncogene 2009, 28, 117–127. [CrossRef]

44. Liu, Y.; Lashuel, H.; Choi, S.; Xing, X.; Case, A.; Ni, J.; Yeh, L.-A.; Cuny, G.D.; Stein, R.L.; Lansbury, P.T. Discovery of Inhibitors that

Elucidate the Role of UCH-L1 Activity in the H1299 Lung Cancer Cell Line. Chem. Biol. 2003, 10, 837–846. [CrossRef] [PubMed]

45. Panyain, N.; Godinat, A.; Lanyon-Hogg, T.; Lachiondo-Ortega, S.; Will, E.J.; Soudy, C.; Mason, K.; Elkhalifa, S.; Smith, L.M.;

Harrigan, J.A.; et al. Discovery of a potent and selective covalent inhibitor and activity-based probe for the deubiquitylating

enzyme UCHL1, with anti-ﬁbrotic activity. J. Am. Chem. Soc. 2020. [CrossRef]

46. Krabill, A.D.A.D.; Chen, H.; Hussain, S.; Feng, C.; Abdullah, A.; Das, C.; Aryal, U.K.; Post, C.B.C.B.; Wendt, M.K.M.K.;

Galardy, P.J.; et al. Ubiquitin C-Terminal Hydrolase L1: Biochemical and Cellular Characterization of a Covalent Cyanopyrrolidine-

Based Inhibitor. ChemBioChem 2020, 21, 712–722. [CrossRef] [PubMed]

47. Kattenhorn, L.M.; Korbel, G.A.; Kessler, B.M.; Spooner, E.; Ploegh, H.L. A Deubiquitinating Enzyme Encoded by HSV-1 Belongs

to a Family of Cysteine Proteases that Is Conserved across the Family Herpesviridae. Mol. Cell 2005, 19, 547–557. [CrossRef]

48. Davies, C.W.; Chaney, J.; Korbel, G.; Ringe, D.; Petsko, G.A.; Ploegh, H.; Das, C. The co-crystal structure of ubiquitin carboxy-

terminal hydrolase L1 (UCHL1) with a tripeptide ﬂuoromethyl ketone (Z-VAE(OMe)-FMK). Bioorganic Med. Chem. Lett.

2012, 22, 3900–3904. [CrossRef] [PubMed]

49. Powers, J.C.; Asgian, J.L.; James, K.E. Irreversible Inhibitors of Serine, Cysteine, and threonine proteases. Chem. Rev. 2002

[CrossRef] [PubMed]

.

50. Palmer, J. Process for Forming a Fluoromethyl Ketone. U.S. Patent No. 5,210,272, 11 May 1993.

51. Witte, M.D.; Descals, C.V.; De Lavoir, S.V.P.; Florea, B.I.; Van Der Marel, G.A.; Overkleeft, H.S. Bodipy-VAD-Fmk, a useful tool to

study yeast peptide N-glycanase activity. Org. Biomol. Chem. 2007, 5, 3690–3697. [CrossRef]

52. Morris, T.S.; Frormann, S.; Shechosky, S.; Lowe, C.; Laid, M.S.; Gauss-mﬁller, V.; Purcell, R.H.; Emerson, S.U.; Vederas, J.C.;

Malcolm, B.A. In Vitro and Ex Vivo Inhibition of Hepatitis A Virus 3C Proteinase by a Peptidyl Monoﬂuoromethyl Ketone. Bioorg.

Med. Chem. 1997, 5, 797–807. [CrossRef]

53. Misaghi, S.; Pacold, M.E.; Blom, D.; Ploegh, H.L.; Korbel, G.A. Using a small molecule inhibitor of peptide: N-glycanaseto probe

its role in glycoprotein turnover. Chem. Biol. 2004, 11, 1677–1687. [CrossRef] [PubMed]

54. Hassiepen, U.; Eidhoff, U.; Meder, G.; Bulber, J.F.; Hein, A.; Bodendorf, U.; Lorthiois, E.; Martoglio, B. A sensitive ﬂuorescence in-

tensity assay for deubiquitinating proteases using ubiquitin-rhodamine110-glycine as substrate. Anal. Biochem. 2007, 371, 201–207

.

[CrossRef]

55. Boudreaux, D.A.; Maiti, T.K.; Davies, C.W.; Das, C. Ubiquitin vinyl methyl ester binding orients the misaligned active site of the

ubiquitin hydrolase UCHL1 into productive conformation. Proc. Natl. Acad. Sci. USA 2010, 107, 9117–9122. [CrossRef] [PubMed]

56. Koharudin, L.M.I.; Liu, H.; Di Maio, R.; Kodali, R.B.; Graham, S.H.; Gronenborn, A.M. Cyclopentenone prostaglandin-induced

unfolding and aggregation of the Parkinson disease-associated UCH-L1. Proc. Natl. Acad. Sci. USA 2010, 107, 6835–6840.

[CrossRef] [PubMed]

57. Contu, V.R.; Kotake, Y.; Toyama, T.; Okuda, K.; Miyara, M.; Sakamoto, S.; Samizo, S.; Sanoh, S.; Kumagai, Y.; Ohta, S. Endogenous

neurotoxic dopamine derivative covalently binds to Parkinson’s disease-associated ubiquitin C-terminal hydrolase L1 and alters

its structure and function. J. Neurochem. 2014, 130, 826–838. [CrossRef] [PubMed]

58. Nakamura, T.; Oh, C.; Liao, L.; Zhang, X.; Lopez, K.M.; Gibbs, D. Noncanonical transnitrosylation network contributes to synapse

loss in Alzheimer’ s disease. Science (80-.) 2020, 0843, 1–18.

Molecules 2021, 26, 1227

24 of 24

59. Bishop, P.; Rubin, P.; Thomson, A.R.; Rocca, D.; Henley, J.M. The Ubiquitin C-Terminal Hydrolase L1 (UCH-L1) C terminus plays

a key role in protein stability, but its farnesylation is not required for membrane association in primary neurons. J. Biol. Chem.

2014, 289, 36140–36149. [CrossRef]

60. Garcia-Calvo, M.; Peterson, E.P.; Leiting, B.; Ruel, R.; Nicholson, D.W.; Thornberry, N.A. Inhibition of human caspases by

peptide-based and macromolecular inhibitors. J. Biol. Chem. 1998, 273, 32608–32613. [CrossRef] [PubMed]

61. Mons, E.; Jansen, I.D.C.; Loboda, J.; Van Doodewaerd, B.R.; Hermans, J.; Verdoes, M.; Van Boeckel, C.A.A.; Van Veelen, P.A.;

Turk, B.; Ovaa, H. The Alkyne Moiety as a Latent Electrophile in Irreversible Covalent Small Molecule Inhibitors of Cathepsin K.

J. Am. Chem. Soc. 2019, 141, 3507–3514. [CrossRef]

62. Kathman, S.G.; Xu, Z.; Statsyuk, A. V A Fragment-Based Method to Discover Irreversible Covalent Inhibitors of Cysteine

Proteases. J. Med. Chem. 2014, 4–9. [CrossRef]

63. Strelow, J.M. A perspective on the kinetics of covalent and irreversible inhibition. SLAS Discov. 2017, 22, 3–20. [CrossRef]

64. Resnick, E.; Bradley, A.; Gan, J.; Douangamath, A.; Krojer, T.; Sethi, R.; Geurink, P.P.; Aimon, A.; Amitai, G.; Bellini, D.; et al.

Rapid Covalent-Probe Discovery by Electrophile-Fragment Screening. J. Am. Chem. Soc. 2019. [CrossRef]

65. Wilkinson, K.D.; Lee, K.; Deshpande, S.; Duerksen-hughes, P.; Boss, J.M.; Pohl, J. The neuron-speciﬁc protein PGP 9.5 is a ubiquitin

carboxyl-terminal hydrolase. Science (80-.) 1989, 246, 670–673. [CrossRef] [PubMed]

66. Larsen, C.N.; Krantz, B.A.; Wilkinson, K.D. Substrate speciﬁcity of deubiquitinating enzymes: Ubiquitin C-terminal hydrolases.

Biochemistry 1998, 37, 3358–3368. [CrossRef]

67. Nakatsu, F.; Sakuma, M.; Matsuo, Y.; Arase, H.; Yamasaki, S.; Nakamura, N.; Saito, T.; Ohno, H. A Di-leucine signal in the

ubiquitin moiety Possible involvement in ubiquitination-mediated endocytosis. J. Biol. Chem. 2000, 275, 26213–26219. [CrossRef]

[PubMed]

68. Shih, S.C. Monoubiquitin carries a novel internalization signal that is appended to activated receptors. EMBO J. 2000, 19, 187–198.

[CrossRef] [PubMed]

69. Hewitt, C.S.; Krabill, A.D.; Das, C.; Flaherty, D.P. Development of Ubiquitin Variants with Selectivity for Ubiquitin C-Terminal

Hydrolase Deubiquitinases. Biochemistry 2020, 1. [CrossRef]

70. Ekkebus, R.; Flierman, D.; Geurink, P.P.; Ovaa, H. Catching a DUB in the act: Novel ubiquitin-based active site directed probes.

Curr. Opin. Chem. Biol. 2014, 23, 63–70. [CrossRef] [PubMed]

71. Kim, H.J.; Magesh, V.; Lee, J.; Kim, S.; Knaus, U.G.; Lee, K. Ubiquitin C-terminal hydrolase-L1 increases cancer cell invasion by

modulating hydrogen peroxide generated via NADPH oxidase. Oncotarget 2015, 6, 16287. [CrossRef]

Article Doi

Optimization and anti-cancer properties of fluoromethylketones as covalent inhibitors for ubiquitin C-terminal hydrolase L1

DOI: 10.3390/molecules26051227

Source and publish data:

Authors:

Article abstract of DOI:10.3390/molecules26051227

Full text of DOI:10.3390/molecules26051227

Products guided by the article

R&D Labs maybe for 72873-24-6

Relevant to this article

Hot Product